side by side comparison... large print digital back VS 4x5 color film

[archivue]

Most of people using MD digital back claim that it's superior than 4x5 scan for their pro uses (ofset...).
But is there anybody out there that made a comparison between large prints (inkjet or lambda...) that came from MFDB file and 4X5 scan ? same subject (landscapes, architecture, difficult light..) ?

[Bruce Watson]

Most of people using MD digital back claim that it's superior than 4x5 scan for their pro uses (ofset...).
But is there anybody out there that made a comparison between large prints (inkjet or lambda...) that came from MFDB file and 4X5 scan ? same subject (landscapes, architecture, difficult light..) ?

There are a couple. I seem to remember one from Charles Cramer. Might be on the Luminous Landscape website. IIRC it's a fairly flawed comparison because the scanner used on the film was a Tango, so the scans were somewhat soft. This is due to the Tango's 11 micron fixed aperture which limits the Tango to something around 2200-2400 ppi maximum optical resolution (that is, what it can read from something like the 1951 USAF Resolution Test Chart (http://en.wikipedia.org/wiki/1951_USAF_Resolution_Test_Chart)).

Even at that when I was looking at the images themselves I picked the film images as somewhat better. That was not the conclusion of the article however. If they had found otherwise I doubt they would have published their findings.

If you find any comparisons that use a better drum scanner (something with a 6 micron aperture more or less, or even a 3 micron aperture) I'd like to see it so post a link.

[Stephen Best]

Might be on the Luminous Landscape website. IIRC it's a fairly flawed comparison because the scanner used on the film was a Tango, so the scans were somewhat soft.

The scans were fine, just that the 4x5 shot was clearly out of focus in the area selected for the comparison (or the film had popped). It makes you wonder when people base purchasing decisions on really amateur tests like this.

There's plenty more to consider than which has the highest resolution. For instance, if you do your sums the working depth-of-focus (namely at the sensor plane) is about a 10th of that for 4x5.

[Tyler Boley]

What is the price of a P45 system again? Like that used in the comparisons?
Tyler

[archivue]

prices ? difficult to say...

in the same shop, you have :
P45+ back only for 26000 euros excluding taxes
while the P45+ phase one kit (including body and 80) is at 19000 euros excluding taxes
A simple P45 refurb cost 11000 euros excluding taxes
and 12000 with a phase one camera and 80... 1 year warranty.

While the canon 5DII will be around 2300 euros including VAT... but that's an other story !

[archivue]

P45 refurb kit costs the price of 650 Portra 4x5 (film+C41) and 130 hires scans

[Lenny Eiger]

What is the price of a P45 system again? Like that used in the comparisons?
Tyler

I believe it was $39,000. Basically, $1K per megapixel.

Lenny

[Lenny Eiger]

Most of people using MD digital back claim that it's superior than 4x5 scan for their pro uses (ofset...).

It's 39 megapixels. I get 320 off of a 4x5 piece of film, and 568 megapixels off of 8x10. There's no comparison. The tests are flawed.

That said, one has to get over 360 dpi to get the top quality in color. Almost any digital camera can do that for an 8x10. All you need is a chip where one side is at least 3600 pixels. Digital cameras are great for many commercial uses... And when one goes to the very limited cmyk offset color space, you can't see much difference.

Of course, fine art is another matter....

Lenny

[Rakesh Malik]

That's about the size of it.

I just put up a print in a cafe, a 40x50 inch from a 4x5 slide (Velvia 100). My friend, the one who did the printing (did a good job, btw) also scanned and printed one of his own black and white sheets of 4x5 film, and printed it at 40x50 inches.

The detail in both is incredible; in mine you can walk right up to the print and see the texture in the water, fine detailing in the leafless trees across the bay, and the fine detail in the clouds where the sun's disk isn't overwhelming them. In his, you can see rivets on the bridge, and fine enough detail to separate pairs of cables that look like single cables when you stand back a bit.

Michael Reichmann's now claiming once again that he was right about his assertion that digital outperforms film... but I have yet to see a digital image from anything short of a BetterLight scanning back come even close.

Oh, and my friend doesn't have a drum scanner or anything like that; he has the new Microtek.

[Gordon Moat]

Don't forget that Reichmann is getting older and his eyesight might not be as good as in the past. Many of the comparisons I read on the internet involve a viewed comparison, which always throws into question the eyesight of the writer.

Ciao!

Gordon Moat Photography (http://www.gordonmoat.com)

[David Luttmann]

Don't forget that Reichmann is getting older and his eyesight might not be as good as in the past. Many of the comparisons I read on the internet involve a viewed comparison, which always throws into question the eyesight of the writer.

Ciao!

Gordon Moat Photography (http://www.gordonmoat.com)

Not just that....his impression changes with the weather as well. Before, he was claiming that his old Canon 1Ds surpassed 6x7 MF film. Now in the most recent test, he found that the new 1Ds 2 was "almost as good" as 645. It sure left me scratching my head.

[Tyler Boley]

I believe it was $39,000. Basically, $1K per megapixel.

Lenny

I just wanted the numbers to be part of the discussion. There's a borderline absurd nature to these discussions that taints the concept of "progress".
Tyler (I use both as appropriate, so don't think I'm evangilising)

[DrPablo]

We talk about this stuff all the time, but honestly why aren't there any published, methodologically sound comparisons out there?

First of all, pick an outcome measure for the comparison. If it's resolution, then why are we looking at crops of trees and not looking at quantifiable resolution targets? Why are we not controlling for critical variables like aperture / diffraction, lens characteristics, and focus? It would make these conversations so much easier if we had something useful to reference.

This just feels like a big rhetorical pissing match a lot of the time, where we find a way to justify (or perhaps rationalize) the superiority of our chosen method, and the best we can do is cite internet heresay or personal anecdotes. Fundamentally the decision of a digital or film system usually comes down to things other than pure decisions about resolution.

[Adam Kavalunas]

A good friend of mine, Ron Flickinger, did a lot of testing with his P45 and his Chamonix 4x5. He owns and operates a Screen Cezanne and knows how to scan quite well. He shot the exact same scenes with both cameras, and did prints of each. He sent me 100% samples of each, I think he did an up-res of the digital, and a down res of the film in 2 seperate tests, and the digital won in both on screen, and in prints. The P45 captures 39mp images, but that is compressed, uncompressed i think its around 120mp(don't quote me on that). If you would like more info about the tests, I would suggest to contact him. He is obsessed with quality, and it was enough for him to switch......

Adam Kavalunas
www.plateauvisions.com

[DrPablo]

Thanks, Adam. I myself am not so interested that I'd contact him, but I'd love it if it were either formally published or at least posted on the internet with an open methodology as a resource to us all.

[Lenny Eiger]

A good friend of mine, Ron Flickinger, did a lot of testing with his P45 and his Chamonix 4x5. He owns and operates a Screen Cezanne and knows how to scan quite well. He shot the exact same scenes with both cameras, and did prints of each. He sent me 100% samples of each, I think he did an up-res of the digital, and a down res of the film in 2 seperate tests, and the digital won in both on screen, and in prints. The P45 captures 39mp images, but that is compressed, uncompressed i think its around 120mp(don't quote me on that). If you would like more info about the tests, I would suggest to contact him. He is obsessed with quality, and it was enough for him to switch......

Adam Kavalunas
www.plateauvisions.com

Yes, and the Cezanne is a mediocre scanner. It's blurry in comparison to a top-level drum scanner. Why doesn't he compare something that is reasonable....

Lenny

[Adam Kavalunas]

Lenny,

I'm no scanner expert, but I'll give my experiences. I've had drum scans done by many, including 2 Howtek's which many say is one of the sharpest drum scanners. The cezanne scans are sharper. Also, a few years back I think it was Seybold that did a scanner comparison test and the Cezanne came out on top. So to say that it is an unreasonable scanner to use in a test, may be unreasonable in itself. And besides, I just gave my experience in my first post, I said nothing of it being the end all comparison test. Your response was quite arrogant.

Adam

[Peter De Smidt]

Ted Harris thought the Cezanne was a pretty good scanner. Not the best, certainly, but not mediocre. He seemed to know what he was doing.

Lenny, you could've said: "I wonder what the test would've showed if the scan had been done on a high quality drum scanner?" That would've gotten your point across without the negativity or needless controversy.

[PenGun]

Ted Harris thought the Cezanne was a pretty good scanner. Not the best, certainly, but not mediocre. He seemed to know what he was doing.

Lenny, you could've said: "I wonder what the test would've showed if the scan had been done on a high quality drum scanner?" That would've gotten your point across without the negativity or needless controversy.
You would think of all people who could settle this Lenny has the equipment and the motivation. What about it?

[Peter De Smidt]

But not the needed objectivity.

[sanking]

Just as a point of interest, what is the true optical resolution of the Cezanne Elite when scanning 4X5" film?

Sandy King

[Peter De Smidt]

Hi Sandy,

I own the regular Cezanne, not the elite, and unfortunately I don't have access to the appropriate resolution test slide. Which one do you recommend?

[sanking]

Pete,

No idea. I tried to find the answer to my question by looking at Cezanne Elite specifications on the web and came up dry. All I know is that while the maximum optical resolution of the Cezanne Elite is 5300 dpi it can only scan at this resolution in a relatively small strip, about 1 1/2" or 2" wide I believe. To scan anything wider the Cezanne has to reposition the lens and place it farther from the CCD, which results in less optical resolution. I am guessing that the maximum true otical resolution for 4X5" is 2400 dpi -3200 dpi but don't know for sure.

Sandy





Hi Sandy,

I own the regular Cezanne, not the elite, and unfortunately I don't have access to the appropriate resolution test slide. Which one do you recommend?

[Peter De Smidt]

Hi Sandy,

You're right: 5300 spi is only for originals 1.5" wide. When I scan 4x5 one pass, I scan at 2400 spi. For two pass, which involves combining the two scans in Photoshop, I scan at 4000 spi, as I've not found any benefit from doing 3-pass scans at 5300 spi, with my materials and for my uses.

I scan medium format at 4000 spi one pass, and 35mm at 5300 spi one pass.

-Peter

[PenGun]

But not the needed objectivity.

How 'bout a "cage match" with Lenny Eiger and Michael Reichmann is a knock down drag out fight to the finish. We can judge if there is no knockout. ;)

[Peter De Smidt]

Edmund Scientifics's variable frequency resolution target 5-200 lp/mm, stock # NT43-488, looks like it would be a good test subject, although since it's rigid, it'd be useless on a drum scanner. Oh, yeah, and it costs almost $500!

[Rick Russell]

I'm certainly no expert in the matter, and don't have a dog in this fight, but I understand that the Aztec Premier drum scanner, used by Mr. Eiger and the company I use for my scans, Pixelnation, is one of, if not the best scanner on the market. One of its principal differences is that it captures scan sizes based on the size of the grain of the film, which results in sharper scans.

I too would like to see a side by side comparison of the state of the art original digital source vs. the state of the art analog source. Perhaps Mr. Eiger will assist us, as this discussion is often entertaining, but does not often shed light.

Rick Russell
richardrussell-1@ca.rr.com

[cobalt]

How 'bout a "cage match" with Lenny Eiger and Michael Reichmann is a knock down drag out fight to the finish. We can judge if there is no knockout. ;)


Dude! You almost made me shoot coffee through my nose!

[Bruce Watson]

I'm certainly no expert in the matter, and don't have a dog in this fight, but I understand that the Aztec Premier drum scanner, used by Mr. Eiger and the company I use for my scans, Pixelnation, is one of, if not the best scanner on the market. One of its principal differences is that it captures scan sizes based on the size of the grain of the film, which results in sharper scans.

I've always found this claim quite dubious. The size of the film grain clumps (dye clouds in color films) vary widely in most films used to capture real subjects with realistic subject brightness ranges. IOW, there is not one single size of the grain of the film -- film grain clump size is fairly stochastic (clump sizes increase generally with increasing density). And the scanner doesn't vary its aperture size as it scans - the aperture size is fixed. So the claim that the scanner "captures scan sizes based on the size of the grain of the film" (and yes, I realize it's not you making this claim, it's Aztek's marketing hype) is a non sequitur.

More on what film grain is and what it actually looks like can be found in Tim Vitale's excellent paper (http://aic.stanford.edu/sg/emg/library/pdf/vitale/2006-03-vitale-filmgrain_resolution.pdf) on the subject.

[PenGun]

I sent Michael an email.

I yam a fight promoter today. ;)

Come on Lenny ... you'll crush the poor fool. Think what it would mean for your business.

[Lenny Eiger]

I sent Michael an email.
Come on Lenny ... you'll crush the poor fool. Think what it would mean for your business.

Nothing like a little violence to soothe the soul...
;-)

Lenny

[Lenny Eiger]

I've always found this claim quite dubious. The size of the film grain clumps (dye clouds in color films) vary widely in most films used to capture real subjects with realistic subject brightness ranges. IOW, there is not one single size of the grain of the film -- film grain clump size is fairly stochastic (clump sizes increase generally with increasing density). And the scanner doesn't vary its aperture size as it scans - the aperture size is fixed. So the claim that the scanner "captures scan sizes based on the size of the grain of the film" (and yes, I realize it's not you making this claim, it's Aztek's marketing hype) is a non sequitur.

More on what film grain is and what it actually looks like can be found in Tim Vitale's excellent paper (http://aic.stanford.edu/sg/emg/library/pdf/vitale/2006-03-vitale-filmgrain_resolution.pdf) on the subject.

Bruce,

There are a number of things that work - that don't make apparent sense... I know you have seen the effects of aperture on scans. It may be based on something else entirely, but you know that sample size (aperture) is a huge factor in the quality of the resulting scan. Too small and you get salt and pepper, too large and you get blurry scans.

The Premier has aperture settings every two or three microns, where the 4500 jumps form 6 to 13, to 19, etc. It's just a little less, altho' 13 works for a lot of things in b&w neg or transparency, so it doesn't matter. The 19 works great for color negs. I don't really need the 8 and the 10 - they aren't used that often.

It's better than the Tango, which is set at 11, and can't vary. Some others have small, medium and large settings, which are not particularly specific.

It's a huge difference when you get it right, however. You end up smoothing the grain just a bit but retaining the sharpness. I think the problem is that there are multiple things going on. First of all, you aren't really looking at grain, you're looking at scan samples...

is it clouds we are looking at? I don't really know. I've invited Tim to come over many times, and he almost has... However, the long and the short of it is that the aperture does do a lot...

Lenny

[Bruce Watson]

I don't have a problem with using aperture size as a tool to make a good scan. We agree it's a valuable tool.

What I have a problem with is the idea that one can match the aperture size to the grain clump size. The aperture is fixed for the duration of the scan, while the grain clump size varies (sometimes markedly) over the density range of the film. IOW, there is not a one-t0-one correspondence; the laws of physics prevent it.

What setting the aperture size is then is an art. You and I and any other drum scanner operator has to work with the film and decide which aperture is the best compromise of sharpness and smoothness for that particular film. With my own work the correct aperture setting varies from sheet to sheet of the same film developed the same way. IOW it's not a constant under the best of conditions. It's a compromise, and it's that idea of compromise that they completely leave out of the marketing brochures.

It's just the way their marketing says it. In trying to simplify it into a sound bite they get the physics wrong. I hate it when that happens.

[Lenny Eiger]

It's just the way their marketing says it. In trying to simplify it into a sound bite they get the physics wrong. I hate it when that happens.

I totally agree.

I had many conversations with Phil Lippincott where I would ask a question and get a 2 hour response that had little if anything to do with my question. I think his brain used a language from another planet. However, as I said in a previous post, there are a number of things that don't make sense, yet work a certain way.

I would also agree that setting aperture's and settings is an art, rather than a science, and perhaps that's why the operator is so important. I would add that color management is also an art, way worse than scanning.

Lenny

[Bruce Watson]

I totally agree.

I had many conversations with Phil Lippincott where I would ask a question and get a 2 hour response that had little if anything to do with my question. I think his brain used a language from another planet.

Ah yes. I had some of those oddly frustrating conversations with Phil too. He was certainly a piece of work. Interestingly convoluted thinking. But it seemed to work for him.


However, as I said in a previous post, there are a number of things that don't make sense, yet work a certain way.

This one makes sense to me. I just can't figure out how to reduce a description of what's going on into a sound bite. Good thing I don't have to!


I would also agree that setting aperture's and settings is an art, rather than a science, and perhaps that's why the operator is so important. I would add that color management is also an art, way worse than scanning.

Well, at least color management is not a black art. ;-)

[Lenny Eiger]

Well, at least color management is not a black art. ;-)

Uh, actually, that's what I meant.... ;-)

I bought ColorThink to analyze my profiles and I got all the curves to match up perfectly - then made a horrible print. I called the guy who made it and he said - well, it's just a guide.

A $400 guide... I now make a profile, then print, then make another until the print looks right... and I have iO Table, I know my StudioPrint upwards and downwards and it still doesn't mean its right unless the print works....

Black art? Yes, matey....

Lenny

[Tim Povlick]

I've always found this claim quite dubious. The size of the film grain clumps (dye clouds in color films) vary widely in most films used to capture real subjects with realistic subject brightness ranges. IOW, there is not one single size of the grain of the film -- film grain clump size is fairly stochastic (clump sizes increase generally with increasing density). And the scanner doesn't vary its aperture size as it scans - the aperture size is fixed. So the claim that the scanner "captures scan sizes based on the size of the grain of the film" (and yes, I realize it's not you making this claim, it's Aztek's marketing hype) is a non sequitur.

More on what film grain is and what it actually looks like can be found in Tim Vitale's excellent paper (http://aic.stanford.edu/sg/emg/library/pdf/vitale/2006-03-vitale-filmgrain_resolution.pdf) on the subject.

While I would agree film grain clump size can be viewed as a Stochastic function, the
clump sizes increase generally with increasing density is not the reason why. Can you elaborate on this? I checked "Probability, Random Variables, and Stochastic Processes" by Papoulis.

Back to the main point, scanning and film grain, (or whatever one wants to call it), the aperture is being set not to match simply the grain size of film, but the average size of the smallest grain. In Stochastic terms this would be the expected value of grain size. So I agree with what Aztek says, as long as one realizes it's the average grain size that set's aperture. Therefore one should not vary aperture while scanning, If one is set to 10 microns and along comes a dark area that is "clumped" at 55 micron, the 10 micron aperture will read the density properly.

Attached is an intentionally "bad" scan I did with the Premier on Ecke 25, 4K dpi and 8 micron ap. The light and dark areas show the same general "grain size", at least to my eyes.

FWIW I have worked with the folks at Aztek and found them to be straight talkers. This from an EE of 30 years experience that includes designs using PMT's.


Regards,

(West Coast) Tim

[sanking]

I have seen what seemed to me very significant grain reduction with software like Noise Ninja, without any apparent decrease in apparent sharpness. In what fundamental way does the mechanism of adjusting the aperture with a drum scanner differ from the software? Are we not with the software doing the same thing, i.e. adjusting edge rounding to match the average grain clump size?

Sandy King







Back to the main point, scanning and film grain, (or whatever one wants to call it), the aperture is being set not to match simply the grain size of film, but the average size of the smallest grain. In Stochastic terms this would be the expected value of grain size. So I agree with what Aztek says, as long as one realizes it's the average grain size that set's aperture. Therefore one should not vary aperture while scanning, If one is set to 10 microns and along comes a dark area that is "clumped" at 55 micron, the 10 micron aperture will read the density properly.


Regards,

(West Coast) Tim

[Lenny Eiger]

I have seen what seemed to me very significant grain reduction with software like Noise Ninja, without any apparent decrease in apparent sharpness. In what fundamental way does the mechanism of adjusting the aperture with a drum scanner differ from the software? Are we not with the software doing exactly the same thing, i.e. adjusting edge rounding to match the average grain clump size?

Sandy King

This is a little hard... it took me a while to understand this. I think the two processes are quite different from each other. Tim will have his own understanding, deeper in some ways than mine, but here goes. If someone has a better way to describe this, I'm all ears. I may also not be perfectly accurate in this description and feel free to correct me.

I believe Ninja blurs, at least that's what it did on my test. Maybe Sandy or someone else can fill in here if there is more to this... that's all I know and I don't mean to shortchange it.

The aperture is a sample, it isn't like this, but I imagine an opening to the sensor that's like a vertical shade - or shutter perpendicular to the drum. As the drum goes around, the sensor can see thru this shutter. One can widen the shutter or tighten it.

If you sample many times as the drum goes around, thru this opening, you get a series of samples, that make up the pixels of an image. A Premier has a 38,000 step stepper motor, with that many possibilities around a single circumference. It takes 8,000 of those steps and makes samples.

That's vertical. The drum can move over in increments of a16,000 of an inch at a time. It moves over a specific distance based on the ppi/aperture settings. It occurs to me I don't exactly know which one of these it chooses, but it must be the ppi if the aperture can oversample... If the slit that it looks thru is smaller than the RMS Granularity, the scanner will sample the same grain cloud more than once, creating an effect called grain anti-aliasing. If the slit matches the granularity, then you have a sharp scan with no ill effects. If the slit is larger then the image will be blurred.

It isn't a slit, its a circle in a rotating disc, it's just how I visualize it. There are usually two options that are "correct", one that will be grainy and sharp, and another where the grain is smoothed somewhat, but is still sharp. One can make a 1/8in x 1/8in sample and blow it up full size and watch the effects at different micron sizes. Differences are quite visible.

Here's a pic I made that I think illustrates the "slice" and the grain cloud it is looking at, and why I think this effect is occurring. I have a link to the comparison of how different micron settings affect the image here: http://eigerstudios.com/scancompareMicrons.html

I hope this helps...

Lenny

[Oren Grad]

the aperture is being set not to match simply the grain size of film, but the average size of the smallest grain. In Stochastic terms this would be the expected value of grain size. So I agree with what Aztek says, as long as one realizes it's the average grain size that set's aperture.

This makes no sense in purely mathematical terms. By definition, the expected value will never be the smallest grain size unless the grain is uniform in size. And in that case, talking about the "average grain size" is superfluous.

[Tim Povlick]

This makes no sense in purely mathematical terms. By definition, the expected value will never be the smallest grain size unless the grain is uniform in size. And in that case, talking about the "average grain size" is superfluous.

A little clumsy wording on my part. Edit "smallest" to "smaller" and it makes perfect sense to me.

I am trying to say one sets the aperture to match the average size of the grain. I was trying to exclude any large grain clumps from being included in the average.

Thanks,

Tim

[Bruce Watson]

While I would agree film grain clump size can be viewed as a Stochastic function, the is not the reason why. Can you elaborate on this? I checked "Probability, Random Variables, and Stochastic Processes" by Papoulis.

You're not likely to find it in a statistics text. This is about how film works. A stochastic process implies a level of randomness with an overlay of a guiding element. In this case the randomness is the distribution of silver halides in the emulsion and the guiding element is the exposure of image itself.

I'm just talking about metallic silver grain here. In color films dye couplers are used to replace the metallic silver grain clumps with dye clouds. So where I reference grain clumps you can substitute dye clouds and get the same basic meaning for color films.

Areas of low exposure (shadows) tend to capture the few photons that come their way at the surface levels of the emulsion. Areas of high exposure tend to capture photons throughout the depth of the emulsion. After processing, the resulting image show less graininess in the areas of low density and considerably more graininess in areas of higher density.

So far so good, yes? What makes graininess interesting is how it's formed. It's not just individual film grains. It's groups of grains, usually interlocking strands of metallic silver. These groupings form both in 2D and in 3D. That is, in the direction of the plane of the surface of the film, and in the depth of the emulsion. These groupings are commonly referred to as grain clumps. Note that overlapping of grain clumps as you look into the depth of the emulsion just makes them bigger. This is the view the scanner uses.

Finally grain clumps work together to create density in the film. There are at least two ways this happens. First, the size of the grain clumps increases. Second, the interstitial spaces between the grain clumps decreases. The net effect is that it's more difficult for light to pass through.

So... as grain clumps grow and become closer together they create the film's density. I believe I've read somewhere over the years that depending on the film, processing, and the image, that film grain clump size can vary a couple of orders of magnitude. Say from 1 to 100 microns.

To say that you can make any kind of a match between a single aperture and this range of grain clump sizes is, to me at least, quite a stretch.


Back to the main point, scanning and film grain, (or whatever one wants to call it), the aperture is being set not to match simply the grain size of film, but the average size of the smallest grain.

Huh? The average of the smallest grain? I don't know what the means.


In Stochastic terms this would be the expected value of grain size.

I don't follow you. You could use a statistical function to get an average or the mean maybe. Are you implying the RMS average grain size?


So I agree with what Aztek says, as long as one realizes it's the average grain size that set's aperture.

I would at least not argue so much if that's what they said. But as I recall they leave out that pesky word "average."


Therefore one should not vary aperture while scanning,

Should not? How about can not? I've never seen a scanner that would vary aperture during a scan. I wonder what kind of results you'd get if it did. Hmmm...


If one is set to 10 microns and along comes a dark area that is "clumped" at 55 micron, the 10 micron aperture will read the density properly.

I disagree, but that's an argument for a different time and place. This dive into the rat hole is probably driving the OP crazy and isn't helping him much if any.


Attached is an intentionally "bad" scan I did with the Premier on Ecke 25, 4K dpi and 8 micron ap. The light and dark areas show the same general "grain size", at least to my eyes.

What can one say to that? The laws of physics dictate that the grain clump sizes and distances must be different. Just because you can't see it doesn't mean that the scanner can't see it.


FWIW I have worked with the folks at Aztek and found them to be straight talkers. This from an EE of 30 years experience that includes designs using PMT's.

I'm not saying they are bad people. I'm just saying that some of their marketing hype is bad physics. I'm also just pointing out that people shouldn't believe all the marking hype they see -- just because somebody printed it doesn't make it fact. All marketing hype regardless of source should be taken for what it is -- an attempt to persuade you to buy something. Marketing hype is seldom an appeal to logic. It most often is an appeal to your emotions.

So don't believe everything you read or hear. Including what you read and hear from me. Trust, but verify.

[willwilson]

It sounds like there are many parallels to digital audio at work here, but with some optical twists. In essence we are talking about the fundamentals of digital sampling. In the digital audio world an anti aliasing-filter cuts out the garbage that is out of the range of the sampling frequency. This reduces noise that would be introduced by audio signals above the sampling frequency. It seems the aperture of a scanner serves the same purpose, at least it appears so from the scans Lenny posted a link to.


It moves over a specific distance based on the ppi/aperture settings. It occurs to me I don't exactly know which one of these it chooses, but it must be the ppi if the aperture can oversample... If the slit that it looks thru is smaller than the RMS Granularity, the scanner will sample the same grain cloud more than once, creating an effect called grain anti-aliasing. If the slit matches the granularity, then you have a sharp scan with no ill effects. If the slit is larger then the image will be blurred.


Are drum scanners oversampling up to the mechanical limits of the machine (38,000 samples for every 1/16,000th of an inch)? Where the aperture is determining the frequency of oversampling by controlling the physical size of the sample area?

Doesn't the scanner have to be scanning every grain more than once or it would not be able to decipher individual grains. At a minimum, you must be taking samples at twice the frequency/size of the smallest grain you would like to render accurately, correct?

Wouldn't a higher bit depth substantial increase the accuracy of scanning? 8-bits is only 256 possibilities. That is not very many in the digital sampling world. 16-bit is 65,536 possibilities. Shouldn't scanning at a higher bit depth lower the noise within the sample, as well?

[Lenny Eiger]

Are drum scanners oversampling up to the mechanical limits of the machine (38,000 samples for every 1/16,000th of an inch)? Where the aperture is determining the frequency of oversampling by controlling the physical size of the sample area?

You're getting a little over my pay grade here, but in recent conversations about this, I believe the words "choose 8000 of these" were used. In the past, Phil mentioned that he could have supplied a 38,000 ppi scanner, but chose to do 8,000 as that was the limit of his optical resolution and he wanted to deliver "real" data.



Doesn't the scanner have to be scanning every grain more than once or it would not be able to decipher individual grains. At a minimum, you must be taking samples at twice the frequency/size of the smallest grain you would like to render accurately, correct?

I believe this refers to something different - frequency, size, vs number of samples of the same area. Further, I asked Phil what he thought of the Nyquist theorem and whether it applied to his scanning engineering. To paraphrase it politely, he said, "No."



Wouldn't a higher bit depth substantial increase the accuracy of scanning? 8-bits is only 256 possibilities. That is not very many in the digital sampling world. 16-bit is 65,536 possibilities. Shouldn't scanning at a higher bit depth lower the noise within the sample, as well?

I don't know the numbers for every scanner but they do vary. The older ICG's, I believe - don't quote me, did only 12 or 15 bits, etc. The Premier does full 16 bits. Since I am not clear on what causes noise in a drum scanner, I can't really answer your question. I might check in with Aztek on this.

Lenny

[Rakesh Malik]

I believe this refers to something different - frequency, size, vs number of samples of the same area. Further, I asked Phil what he thought of the Nyquist theorem and whether it applied to his scanning engineering. To paraphrase it politely, he said, "No."


First of all, it's not a theorem, it's a mathematical property associated with analog to digital and digital to analog conversion.

If it doesn't apply, then there are a few possibilities:
His equipment is sensitive enough that it's capable of resolving smaller entities than the grains of film, and therefore is already beyond the Nyquist limit.
It's nowhere near the Nyquist limit.
It's magic.
You misunderstood his response.



I don't know the numbers for every scanner but they do vary. The older ICG's, I believe - don't quote me, did only 12 or 15 bits, etc. The Premier does full 16 bits. Since I am not clear on what causes noise in a drum scanner, I can't really answer your question. I might check in with Aztek on this.


More bits will improve the precision of the analog to digital conversion, and make quantization errors smaller. So it should definitely lower the noise introduced through quantization error, if nothing else.

[Rakesh Malik]

Doesn't the scanner have to be scanning every grain more than once or it would not be able to decipher individual grains. At a minimum, you must be taking samples at twice the frequency/size of the smallest grain you would like to render accurately, correct?


I think what you're looking for is that the smallest grains the laser can reliably resolve are 2x its wavelength.

[Bruce Watson]

Doesn't the scanner have to be scanning every grain more than once or it would not be able to decipher individual grains. At a minimum, you must be taking samples at twice the frequency/size of the smallest grain you would like to render accurately, correct?

A drum scanner makes perfectly square pixels that are a single uniform color. A drum scanner can not decipher individual grains. They are far too small, and almost fractal in nature. Once again, see Tim Vitale's excellent paper (http://aic.stanford.edu/sg/emg/library/pdf/vitale/2006-03-vitale-filmgrain_resolution.pdf) on the topic.

Scanners can not render grain accurately. But they do not need to.

[ljb0904]

First of all, it's not a theorem, it's a mathematical property associated with analog to digital and digital to analog conversion.

Actually, it is a theorem. See "Discrete Signal Processing," Oppenheim & Schafer.

[Rakesh Malik]

Oops, I thought it was a property... thanks for the correction!

[Tim Povlick]

You're not likely to find it in a statistics text. This is about how film works. A stochastic process implies a level of randomness with an overlay of a guiding element. In this case the randomness is the distribution of silver halides in the emulsion and the guiding element is the exposure of image itself.

Actually a Stochastic process can be completely deterministic.



To say that you can make any kind of a match between a single aperture and this range of grain clump sizes is, to me at least, quite a stretch.

I don't follow you. You could use a statistical function to get an average or the mean maybe. Are you implying the RMS average grain size?

I would at least not argue so much if that's what they said. But as I recall they leave out that pesky word "average."

Should not? How about can not? I've never seen a scanner that would vary aperture during a scan. I wonder what kind of results you'd get if it did. Hmmm...


I do not see an advantage in varying the aperture during scanning. The down side is now the sampled values represent different sized samples from the film, placing this data into a fixed 2D array of pixel elements doesn't seem correct. In thinking about this I am of the opinion fixed aperture is the way to go. Thanks for your explanation of grain and grain clumping, this is very interesting and film is certainly different from fixed array CMOS / CCD arrays in this regard.




So don't believe everything you read or hear. Including what you read and hear from me. Trust, but verify.

Agreed and thanks for the interesting discussion. I think our biggest challenge is getting the silver exposed to interesting subjects and developed correctly (at least for me).

Best Regards,

Tim

[Tim Povlick]

It sounds like there are many parallels to digital audio at work here, but with some optical twists. In essence we are talking about the fundamentals of digital sampling. In the digital audio world an anti aliasing-filter cuts out the garbage that is out of the range of the sampling frequency. This reduces noise that would be introduced by audio signals above the sampling frequency. It seems the aperture of a scanner serves the same purpose, at least it appears so from the scans Lenny posted a link to.


I think this is a good analogy, only small distinction is an anti-alias filter prevents higher frequency signals from folding into the sampled data and masquerading as an in band signal. If we sample at 20K samples / sec we must cut off all frequencies above 10KHz to prevent them from aliasing down into the in-band signals. That is, a 12KHz tone would end up being sampled as a 2KHz tone if not for the anti-alias filter. I am sure this is what you meant by the above, I simply am amplifying to it.





Are drum scanners oversampling up to the mechanical limits of the machine (38,000 samples for every 1/16,000th of an inch)? Where the aperture is determining the frequency of oversampling by controlling the physical size of the sample area?

Doesn't the scanner have to be scanning every grain more than once or it would not be able to decipher individual grains. At a minimum, you must be taking samples at twice the frequency/size of the smallest grain you would like to render accurately, correct?


To continue the anti-alias filter concept, I suspect the camera lenses we are using act as such a filter. The really high frequency components that would cause alias problems are stopped by our camera lenses. At the scanner one can push the aperture down and down into the single micron range but the film (at least commercial film and not 'spy' film) has nothing more to give up other than the grain or grain clumps (or whatever the term). This is my theory on the subject feel free to take it with a grain of silver. :)



Wouldn't a higher bit depth substantial increase the accuracy of scanning? 8-bits is only 256 possibilities. That is not very many in the digital sampling world. 16-bit is 65,536 possibilities. Shouldn't scanning at a higher bit depth lower the noise within the sample, as well?

It should, that is the signal to noise ratio (SNR), should improve as bit depth increases otherwise the added bits are totally meaningless. PMT's do have some noise associated with their operation. The interested student is encouraged to check the Hamamatsu site for some interesting reading on PMTs. Including photon counting with PMTs.

Best,

Tim

[Bruce Watson]

Actually a Stochastic process can be completely deterministic.

A stochastic process can be completely random. But if a process is deterministic, it is by definition not stochastic. It's been many years and my old statistics books were sold off years ago. But I don't remember any definition of stochastic that didn't include some level of randomness.

That said I suspect that eventually a stochastic process could give the same result as a deterministic process if that's what you mean.


I think our biggest challenge is getting the silver exposed to interesting subjects and developed correctly (at least for me).

Well there's something we can certainly agree on without reservations! We're discussing the technical minutia of the easy part of photography here. The hard part is finding and capturing the interesting subjects. I've been working one scene for at least six years now. I never can seem to show up when everything is ready for a photograph. That scene is beginning to bug me -- it's caused too many sheets of film to hit the trash can. Sigh...

Scanning a poor photograph can't make it a good photograph no matter how high the scan quality is.

[Edwin Beckenbach]

FWIW...The result of a stochastic process in all but trivial cases is a random variable(s). Thus, it is not deterministic. However the parameters of the resulting random variable(s) can be deterministic. For example, the expected value of the stochastic process can be deterministic (but it can also be a random variable itself).

[willwilson]

A comment on the original topic... Based on Tim Vitale's article (which is worth reading) the approximate digital resolution of 4x5 Tmax 100 would be 9,700x12,200 pixels if your lens could resolve 70lpm. That's 118mp. This would equate to a 30"x40" print at 300dpi. Lenny, how do you get 320mp off of 4x5?

I would like to see a valid large print (30x40+) comparison as well; maybe 5DII, Phase One P65, and Top quality Drum Scanned 4x5 Tmax 100 and Velvia.

[sanking]

You can get (roughly) a 320 mp file from 4X5 by scanning at 4000 spi.

How much of that is useful pixels is debatable. My experience is that there is no gain from scanning at more than the effective resolution of the camera/lens/film system. Scanning at 4000 spi suggests that the scanned material has the equivalent of about 80 lines per millimeter of information. 80 llines per mm on 4X5 may possible in certain specific conditions, but in most practical situations one is very lucky to get as much as 50 lines per mm on film, which is about 2400 spi. That gives a file of about 115 mp, which IMO is about as many effective mp one can get from 4X5 format.

Sandy King


A comment on the original topic... Based on Tim Vitale's article (which is worth reading) the approximate digital resolution of 4x5 Tmax 100 would be 9,700x12,200 pixels if your lens could resolve 70lpm. That's 118mp. This would equate to a 30"x40" print at 300dpi. Lenny, how do you get 320mp off of 4x5?

I would like to see a valid large print (30x40+) comparison as well; maybe 5DII, Phase One P65, and Top quality Drum Scanned 4x5 Tmax 100 and Velvia.

[Tyler Boley]

Yes, this is a very misunderstood topic. Scanning for more resolution than the camera/film system can deliver is proving dubious with experience. Beyond that it gets back to grain representation. Then how is that definition delivered? In the case of a drum, despite common thinking, it's the aperture not file size. Up to a point of course.
Then, to get to a big print, how one gets to the printer native ppi is the real question. More physical sampling from the scanner, higher than that native for the chosen aperture? Rezing in editing software? Letting the print driver or RIP do it?
Lot's of interesting stuff here-
http://www.ddisoftware.com/qimage/quality/
I've drawn no conclusions from that yet, but a few tests show the presentation is viable.

Tyler
http://www.custom-digital.com/

[Lenny Eiger]

You can get (roughly) a 320 mp file from 4X5 by scanning at 4000 spi.
How much of that is useful pixels is debatable. My experience is that there is no gain from scanning at more than the effective resolution of the camera/lens/film system. Scanning at 4000 spi suggests that the scanned material has the equivalent of about 80 lines per millimeter of information. 80 llines per mm on 4X5 may possible in certain specific conditions, but in most practical situations one is very lucky to get as much as 50 lines per mm on film, which is about 2400 spi. That gives a file of about 115 mp, which IMO is about as many effective mp one can get from 4X5 format.
Sandy King

Sandy,
I think useful pixels is certainly debatable. However, my experience doesn't match your conclusions. I think the difference may be in the scanner, or more specifically, PMT's vs CCD's. However, if you are correct, then an Epson 750 at 2400 would yield the same file as a top drum scanner (or even a Cezanne). I just don't see this to be the case. Every time I look at a scan from these animals, they don't match up.

Maybe I am wrong. Do you have a test negative that you use to determine this? I would be happy to scan it here at different resolutions and put it to the test.

That said, I am reminded that I am always frustrated by these tests. Resolution is one thing. Sensitivity, pixel by pixel, is quite another.

And just in case you are wondering I'm not trying to piss you off - I sincerely would love to know the answer...

Lenny

[sanking]

"However, if you are correct, then an Epson 750 at 2400 would yield the same file as a top drum scanner (or even a Cezanne). I just don't see this to be the case. Every time I look at a scan from these animals, they don't match up."

Lenny,

Absolutely not. I am not saying that at all, and indeed am surprised at your confusion. I think we agree that some scanners deliver more real, effective resolution than others. Drum scanners, if calibrated, should deliver close to 100% of stated optical resolution. My EverSmart Pro delivers close to 95% of stated optical resolution. Most consumer flatbed scanners, Epson 4990, Epson V750, etc. deliver only about 40% of stated optical resolution. You can test this by scanning a high resolution target: you may have a pixel count of 4800 spi, but the effective resolution would be only about 40% of this, say about 40 lp/mm.

It is not a question of PMT versus CCD technology in the least. It is simply a question of understanding the fact that all pixels are not equal in terms of their detail/resolution. The concept of a "useful" pixel is only debatble if one has no understanding of what "useful" is.

Sandy King




Sandy,
I think useful pixels is certainly debatable. However, my experience doesn't match your conclusions. I think the difference may be in the scanner, or more specifically, PMT's vs CCD's. However, if you are correct, then an Epson 750 at 2400 would yield the same file as a top drum scanner (or even a Cezanne). I just don't see this to be the case. Every time I look at a scan from these animals, they don't match up.

Maybe I am wrong. Do you have a test negative that you use to determine this? I would be happy to scan it here at different resolutions and put it to the test.

That said, I am reminded that I am always frustrated by these tests. Resolution is one thing. Sensitivity, pixel by pixel, is quite another.

And just in case you are wondering I'm not trying to piss you off - I sincerely would love to know the answer...

Lenny

[Bruce Watson]

You can get (roughly) a 320 mp file from 4X5 by scanning at 4000 spi.

How much of that is useful pixels is debatable. My experience is that there is no gain from scanning at more than the effective resolution of the camera/lens/film system. Scanning at 4000 spi suggests that the scanned material has the equivalent of about 80 lines per millimeter of information. 80 llines per mm on 4X5 may possible in certain specific conditions, but in most practical situations one is very lucky to get as much as 50 lines per mm on film, which is about 2400 spi. That gives a file of about 115 mp, which IMO is about as many effective mp one can get from 4X5 format.

My experience suggests that there is a little to be gained from scanning at more than effective resolution of the camera/lens/film/processing system. If, that is, you are printing somewhat bigger.

For example, if you want to make that 50 x 40 inch print from 5x4 film. You could scan at full file size. Or, you could scan at effective resolution which would give you say a 40 x 32 inch print, and then uprez to 50 x 40 inches.

The little bit to be gained here is the elimination of the uprez step. Because every step that requires some shuffling of pixels to some extent looses some information.

Now is that little bit of gain visible in the final print? YMMV. On the other hand, once you've got the film on the scanner, why not? Again, YMMV.

[Ivan J. Eberle]

Anyone care to comment on how the a sampling frequency of between 2600 and 4000 ppi can be the wrong match for the grain structure of specific films? I've run up against grain aliasing/moire with color print films like NC and Portra which seems to amplify the appearance of their grain structure. Wouldn't this make a strong case for sampling at higher ppi--even if there were no gain in line pair or edge resolution to be found?

[Bruce Watson]

Anyone care to comment on how a sampling frequency of between 2600 and 4000 ppi can be the wrong match for the grain structure of specific films?

Can and did. Spent what feels like a whole friggin' day on it. Read the whole thread this time and you'll see it. For extra credit read the Vitale paper. There is no fixed "grain structure" to match. The film is not the problem.

[Lenny Eiger]

Lenny,

Absolutely not. I am not saying that at all, and indeed am surprised at your confusion. I think we agree that some scanners deliver more real, effective resolution than others. Drum scanners, if calibrated, should deliver close to 100% of stated optical resolution. My EverSmart Pro delivers close to 95% of stated optical resolution. Most consumer flatbed scanners, Epson 4990, Epson V750, etc. deliver only about 40% of stated optical resolution. You can test this by scanning a high resolution target: you may have a pixel count of 4800 spi, but the effective resolution would be only about 40% of this, say about 40 lp/mm.

It is not a question of PMT versus CCD technology in the least. It is simply a question of understanding the fact that all pixels are not equal in terms of their detail/resolution. The concept of a "useful" pixel is only debatble if one has no understanding of what "useful" is.

Sandy King

I'll be the first to say that understanding some of this is difficult. For one thing, we all talk about scans as if we are scanning a pixel - or a bit of the image. However, they are just reads, as in a value from 1-255 in lightness darkness for three different channels of color separated light.

I do understand the stated resolution issue. It shows up clearly in the Tango, which can do 11,000 spi, yet only resolve 4000. However, you have often said that your assessment of the real resolution of the 750, for an example, was around 2000 or so (don't remember the exact #). I understand that this is quite different from pixel count, and is not the stated resolution of 5300 or whatever... However, if it is true that it is at a true optical of 2000, the 2400 you stated as the max one could get from a 4x5 would be very close.

What am I missing?

Lenny

[Lenny Eiger]

A comment on the original topic... Based on Tim Vitale's article (which is worth reading) the approximate digital resolution of 4x5 Tmax 100 would be 9,700x12,200 pixels if your lens could resolve 70lpm. That's 118mp. This would equate to a 30"x40" print at 300dpi. Lenny, how do you get 320mp off of 4x5?

I would like to see a valid large print (30x40+) comparison as well; maybe 5DII, Phase One P65, and Top quality Drum Scanned 4x5 Tmax 100 and Velvia.

I would be happy to show you such a thing. I have 32x40 images on my wall, come on by....

I scan 8x10's at 2666 and 4x5's at 4,000. The latter is 16,000 pixels by 20,000 pixels, or 320 mpixels. It think its pretty good, much better than 50 lpm would suggest. The scanner is rated at just under 8,000 ppi optical...

I don't know if that adds anything or not...

Lenny

[Bruce Watson]

It shows up clearly in the Tango, which can do 11,000 spi, yet only resolve 4000.

The Tango's 11 micron aperture would limit it to about 2200-2400 spi real optical resolution wouldn't it?

[sanking]

Lenny,

First, I think you are missing a lot .

As for the Tango resolution, no way the Tango has real optical resolution of 11,000 spi. The aperture of the Tango suggests no more than about 2500 spi. Of course, we know that is not bad since for many years Tango scans of transparencies were considered state of the art.

Sandy





I do understand the stated resolution issue. It shows up clearly in the Tango, which can do 11,000 spi, yet only resolve 4000.
What am I missing?

Lenny

[Lenny Eiger]

The Tango's 11 micron aperture would limit it to about 2200-2400 spi real optical resolution wouldn't it?

Bruce,
I think something is missing here. Maybe its just something I don't understand. The Tango got a 4094 optical dpi at scannerforum.com, numbers I tend to trust. I think there is more than meets the eye. I think it has to do with the optics required to support a 6-micron engine. The Tango likely has a 6 micron capability, but they chose not to allow the aperture to vary, probably a software issue, I'm guessing, but the optics are still there that support it. 13 is better for almost everything, and they let color negs go and got everything else. Probably all about money or time, whatever.

I don't think resolution is closely related to the aperture setting. For instance, when I scan at 4,000 dpi at 13 microns, by rights I should get a lousy optical rez. There are only just 2000 13 micron slices per inch, but I get a much higher optical rez when I scan at the 4000 - from what I see I get at least the full 4000. I get maybe a few percentage points less than if I set it to match the dpi and aperture. But it doesn't cut itself in half. I have high suspicions that the idea that you can only get 8,000 ppi optical by scanning at 3 microns is entirely incorrect. It may be a little better if you do, but the grain anti aliasing will kill you. I doubt they made this error - I mean why bother with a 3 micron capability if there was no film that had a RMS Granularity to match it. It has to be to support the engineering.

Just my 2 cents,

Lenny

[sanking]

Lenny,

Curious, I thought the aperture setting was exactly the method that was used to determine resolution in ppi for drum scanners?


Sandy





I don't think resolution is closely related to the aperture setting.

Lenny

[Lenny Eiger]

Lenny,

First, I think you are missing a lot .

As for the Tango resolution, no way the Tango has real optical resolution of 11,000 spi. The aperture of the Tango suggests no more than about 2500 spi. Of course, we know that is not bad since for many years Tango scans of transparencies were considered state of the art.

Sandy

Sandy,

I stated right there in my post that the Tango can only resolve 4000, not the 11,000.

I am enjoying this conversation, we have a group of very intelligent people here, discussing some very technical issues. It's great. I'm doing my best to listen - and learn. I am happy to admit what I don't know - and wanted to give you an opportunity to fill me in on something - as I don't see it the way you do. Instead, you take the opportunity to take a shot at me and suggest that I don't know jack.

Don't mistake being willing to listen with not knowing anything. I've been studying this stuff for a very long time and I've made a lot of great scans. I run 2 12 color printers and make all my own profiles, do all the stuff many others do. I know a great deal about all of this. I think there are some key things missing - that we are all missing. I don't think anyone here, including myself, has a grasp of it all, not even a full grasp of how scanners work. Even the great Phil Lippincott, who designed a huge amount of the technology we are talking about had limits when I talked to him about practical application. There aren't great books on the subject, that I know of, and there is no long "scanner school" which there used to be.

And now, once again, instead of discussing the issues, I have to take the time to defend myself against personal attack. I have to stand up and tell people all my experience, that I know a lot etc, and it stinks. Last week, you told people to "diss me" if I spoke up about flatbeds or some such thing. We have a conversation and you agree to disagree professionally and then you pull this crap again. You might have noticed that I have not said anything disrespectful to you - I have always disagreed with respect. I think we ought to stick with the issues.

Lenny

[Edwin Beckenbach]

Lenny,

Curious, I thought the aperture setting was exactly the method that was used to determine resolution in ppi for drum scanners?


Sandy

Yes, aperture largely determines resolution. However you can under or oversample so it does not necessarily determine ppi.

[sanking]

Lenny,

Excuse me, but you stated right here,

"I do understand the stated resolution issue. It shows up clearly in the Tango, which can do 11,000 spi, yet only resolve 4000."

Why would you say that the Tango can do 11,000 spi but only resolve 4000 and then complain because someone questions your language?

Get over your raw nerve. I have zero interst in attacking you personally and have not done so, and it is getting tedious having to explain to you otherwise.

Sandy King





Sandy,

I stated right there in my post that the Tango can only resolve 4000, not the 11,000.

I am enjoying this conversation, we have a group of very intelligent people here, discussing some very technical issues. It's great. I'm doing my best to listen - and learn. I am happy to admit what I don't know - and wanted to give you an opportunity to fill me in on something - as I don't see it the way you do. Instead, you take the opportunity to take a shot at me and suggest that I don't know jack.

Don't mistake being willing to listen with not knowing anything. I've been studying this stuff for a very long time and I've made a lot of great scans. I run 2 12 color printers and make all my own profiles, do all the stuff many others do. I know a great deal about all of this. I think there are some key things missing - that we are all missing. I don't think anyone here, including myself, has a grasp of it all, not even a full grasp of how scanners work. Even the great Phil Lippincott, who designed a huge amount of the technology we are talking about had limits when I talked to him about practical application. There aren't great books on the subject, that I know of, and there is no long "scanner school" which there used to be.

And now, once again, instead of discussing the issues, I have to take the time to defend myself against personal attack. I have to stand up and tell people all my experience, that I know a lot etc, and it stinks. Last week, you told people to "diss me" if I spoke up about flatbeds or some such thing. We have a conversation and you agree to disagree professionally and then you pull this crap again. You might have noticed that I have not said anything disrespectful to you - I have always disagreed with respect. I think we ought to stick with the issues.

Lenny

[Lenny Eiger]

Lenny,

Excuse me, but you stated right here,

"I do understand the stated resolution issue. It shows up clearly in the Tango, which can do 11,000 spi, yet only resolve 4000."

Why would you say that the Tango can do 11,000 spi but only resolve 4000 and then complain because someone questions your language?
Sandy King

Sorry,

The Tango can generate up to 11,000 ppi of samples but has an optical resolution of 4094, as reported by Aztek's independent commissioned study, listed at scannerforum.com.

I hope that fixes it.

Lenny

[Peter De Smidt]

The Seybold report scanner test (Volume 30, Number one) looked at a bunch of different scanners, from an ICG drum scanner to an EverSmart Supreme. Part of the tests involved scanning a chromium-plated glass-platen variable frequency resolution target at various angles. The surprising thing to me is that that they found that interpolation leads to higher resolution.

Here's what they say:

"We tested with and without interpolation, which was only necessary in the pixel direction. The reason is that a scanner steps at the same rate for both orientations."

Here's a shortened version of the chart of the results, in line-pairs per millimeter:

Scanner_____Horizontal________Horizontal interp.______Vertical______Vertical Interp.
Agfa t5000 +____45_________________45______________50_____________85
Fuji Lanovia_____90_________________90______________85_____________85
Purup..EskoScan_55_________________85______________40_____________80
Eversmart S.____95_________________95______________95_____________120
Cezanne________60_________________100______________95_____________120
Cezanne Elite____90_________________90_______________95____________120

The drum scanners weren't tested for this, as the test slide is rigid.

So apparently what's happening is that choosing an interpolated resolution in the pixel direction causes the scanner to take smaller steps, but what confuses me is that for some of the scanners interpolation causes an increase in both horizontal and vertical resolution. Any ideas?

[sanking]

Peter,

Choosing an interpolated resolution higher than the stated optiical resolution causes the scanner to take smaller steps in one direction, and that figure is then mixed with the optical resolution in the other direction to determine interpolated resolution.l

However, it is real optical resolution in one direction and this may give greater overall "real" resolution, depending on the subject.

Sandy King




So apparently what's happening is that choosing an interpolated resolution in the pixel direction causes the scanner to take smaller steps, but what confuses me is that for some of the scanners interpolation causes an increase in both horizontal and vertical resolution. Any ideas?[/QUOTE]

[Peter De Smidt]

Ok, so in theory a Cezanne can resolve a maximum of around 120 lp/mm, which would be equivalent to about 6240 spi, right?

[sanking]

Peter,

In theory that could be correct, assuming the lens of the Cezanne can send that information to the CCD.

On the Scan H-End forum some months ago there was some discussion of this very point. I quote from one of the messages there, by a fellow named Mr. Bill, who used to work for Scitex and knows more about scanners than anyone I know.

"I think Jake may have hit the nail on the head. The EverSmart Pro (and
Pro II) are 3175x8200ppi machines. This means that even an the maximum
scanning resolution, all the pixels in the long-bed direction are
captured optically and the interpolation is going on in the Y direction
and is only A 2.58X interpolation.

Interpolation gets a bad name from all those consumer flatbed scanners
that allow you to interpolate to 9,600ppi with a 1,200ppi optical engine
an 8X interpolation. Scitex only allowed much more conservative
interpolation, quite usable. You should never have any fear of pushing
any EverSmart scanner beyond it's minimum optical resolution.

Mr. Bill"

Same would apply to the Cezanne and EverSmart Supreme which use the same Kodak CCD, but take smaller steps and scan in smaller rows so are capable of greater optical and interpolated resolution than the EverSmart Pro.

Sandy King




Ok, so in theory a Cezanne can resolve a maximum of around 120 lp/mm, which would be equivalent to about 6240 spi, right?

[DrPablo]

For what it's worth, I've had a number of 4x5 chromes professionally scanned on a Creo Eversmart Supreme, and since getting my Howtek 4500 I've re-scanned them. And the color accuracy and shadow detail are noticibly better on the Howtek thus far. I realize that there may be many non-hardware variables for color, but I just wanted to reinforce that spi resolution is certainly not the only issue and probably not the most important.

[sanking]

This thread has sure gotten off topic. It was, "large print digital back VS 4x5 color film."

Maybe time for the moderators to kill it and allow folks to start another more focused thread.

Sandy King

[Tim Povlick]

Well there's something we can certainly agree on without reservations! We're discussing the technical minutia of the easy part of photography here. The hard part is finding and capturing the interesting subjects. I've been working one scene for at least six years now. I never can seem to show up when everything is ready for a photograph. That scene is beginning to bug me -- it's caused too many sheets of film to hit the trash can. Sigh...

Scanning a poor photograph can't make it a good photograph no matter how high the scan quality is.

That's interesting, certainly a long time pursuing it. I hope you can capture the scene to your expectations and will post a thumbnail. I'd love to see it!

I would like to capture a moon rise in the background with the lighthouse:

http://www.nps.gov/cabr/

in the foreground. Many problems include the park closes early so only winter months are possible, there are few days the moon is in right position per month, lack of coastal fog etc. are making it a challenge. A couple weekends ago I was totally asleep at the switch and missed a golden oppurtunity with everything in place, except me.

Good luck with the pursuit!

Best,
Tim

[sanking]

Lenny,

I took your question, what am I missing? as sarcasm and responded accordingly. Perhaps I misunderstood your intention.

Yes, the figure of about 2000-2200 effective pixels, or a bit more than 40 lpm, is what my tests have shown for the V750. And while that is close to the maximum detail in a 4X5 negative I think you would agree that resolution is only one of the components of image clarity, micro and macro contrast being the others, and in the case of a scan, shadow depth as well. So while the resolution of the V750 comes close to pulling all of the detail out of the typical 4X5 piece of film I certainly agree with you that image clarity will not be nearly so good as wiht a drum scanner or high end flatbed.

Another point. Earlier I stated that there was no need to scan beyond the information level of the film, which is determined by the camera/lens/film system primarily, and of course also by subject and conditions. I should amend that to say there is no reason to print at a resolution beyond the system because there may be several reasons to over-sample and then down size. In fact, I regulary scan at twice the optical resolution of the scanner and then downsize to reduce noise.

So for the record I don't think there is anything extravagant at scanning 4X5 film at 4000 spi, even though the film may not contain that much information. The problem, and this takes us back to the original purpose of this thread, is that a large percentage of the 320 mp count that you get is useless pixels. In my expereince the best 4X5 film shot in optimum conditions will not have more than about 125-50 mp of useful infirmation, regardless of the resolution in spi at which it is scanned. That is still a lot more information than in a P45 back so if there is a shoot out between the P45 and a 4X5 scan with your Premier my money will be on you. In fact, I think even a Howtek 4500 or my EverSmart Pro easly wins in this comparison with P45 versus 4X5 film.

One way that people can determine if the scanner is pulling all of the information out of the film is to scan at various resolutions and then look at the film through a microscope at about 20X. In most cases I think you will find that the limit of a piece of 4X5 film is around 50 lpm, or 2500 spi assuming that is real resolution.

Sandy King




I do understand the stated resolution issue. It shows up clearly in the Tango, which can do 11,000 spi, yet only resolve 4000. However, you have often said that your assessment of the real resolution of the 750, for an example, was around 2000 or so (don't remember the exact #). I understand that this is quite different from pixel count, and is not the stated resolution of 5300 or whatever... However, if it is true that it is at a true optical of 2000, the 2400 you stated as the max one could get from a 4x5 would be very close.

What am I missing?

Lenny

[Lenny Eiger]

Lenny,

I took your question, what am I missing? as sarcasm and responded accordingly. Perhaps I misunderstood your intention.


FWIW, I meant no sarcasm. In the future, if you think I am being rude in any way, please feel free to check with me privately before you respond. It is truly not my intention to do battle with you, or engage in foolish, useless conversations of any sort. I respect your knowledge and abilities. I do realize some of the conclusions we have come to are different, but that's all.



Yes, the figure of about 2000-2200 effective pixels, or a bit more than 40 lpm, is what my tests have shown for the V750. And while that is close to the maximum detail in a 4X5 negative I think you would agree that resolution is only one of the components of image clarity, micro and macro contrast being the others, and in the case of a scan, shadow depth as well. So while the resolution of the V750 comes close to pulling all of the detail out of the typical 4X5 piece of film I certainly agree with you that image clarity will not be nearly so good as with a drum scanner or high end flatbed.

So for the record I don't think there is anything extravagant at scanning 4X5 film at 4000 spi, even though the film may not contain that much information. The problem, and this takes us back to the original purpose of this thread, is that a large percentage of the 320 mp count that you get is useless pixels. In my expereince the best 4X5 film shot in optimum conditions will not have more than about 125-50 mp of useful infirmation, regardless of the resolution in spi at which it is scanned. That is still a lot more information than in a P45 back so if there is a shoot out between the P45 and a 4X5 scan with your Premier my money will be on you. In fact, I think even a Howtek 4500 or my EverSmart Pro easly wins in this comparison with P45 versus 4X5 film.


I think the claim that there is "no more there" is what many people are using to justify the purchase (or recommendation thereof) of a consumer level scanner (for professionals!), or even a 22 mpixel camera. I have even heard digital camera people say they have better pixels than me, pixel for pixel, which with such a low dynamic range, doesn't make any sense to me.

I suspect you are incorrect in the max of 150 mp of info, however I have no data to support such a claim. It doesn't look like that to me on my scans - but that's not particularly scientific. I think what would be useful - perhaps for the entire group here - is to develop some testing methods to figure this stuff out. I think there remain some questions as to what is possible with scanners at the low, mid and high end. I think incorrect info abounds.... much of it no doubt put out by people marketing digital cameras, or the Reichman's of the world.

If a 20x microscope is the way to figure this out, by all means let's find one and put it to the test. I think we ought to develop a list of questions and hunt them down.... Of course, I have to get back to work now rather than start on this, but I think it would be great...

Lenny

[PenGun]

Just to wander a bit further. ;) It's interesting that Ken Rockwell has now decided that film is better even at the 35mm level:

http://www.kenrockwell.com/tech/00-new-today.htm

It's also interesting to me as a sometime audiophile that a drum scanner is an analog device until we take the output of the PMTs.

Digital sound is ugly in general and the comparison with CCDs and tubes, which populate my pre and main amps, is kinda cool.

[mcfactor]

I think the only truely fair comparison would be between a 4x5 optical print made in the darkroom, i.e., not scanned, and a print of the same size made by a lightjet or digital optical device, or even an inkjet print. scanning negtive film brings in all of the above scanner issues and necessarily degrades quality.

[Lenny Eiger]

I think the only truely fair comparison would be between a 4x5 optical print made in the darkroom, i.e., not scanned, and a print of the same size made by a lightjet or digital optical device, or even an inkjet print. scanning negtive film brings in all of the above scanner issues and necessarily degrades quality.

I'd rather have a scanned image and inkjet than a darkroom print any day. I think the quality is far higher than what you see in the darkroom. There are a lot of factors, not the least of which is dynamic range.

As to a personal preference for the kind of surface one likes, that's for everyone to decide for themselves. However, you can't suggest that inkjet prints don't have at least the quality of a darkroom print - it's simply not supportable.

Lenny

[Bruce Watson]

...comparison with CCDs and tubes, which populate my pre and main amps, is kinda cool.

You do realize that a PMT is a vacuum tube, yes? You could think of a drum scanner as a big single stage vacuum tube amplifier.

[sanking]

I agree with Lenny. Assuming that you have a scan that pulls all of the detail out of the negative or transparency the quality of digital printing is much higher than darkroom printing, especially with magnifications at or over about 3X-4X. Once the negative is scanned there is so much one can do to enhance it both in terms of tonal values and apparent sharpness that simply can not be done in the darkroom without an extraordinary amount of work.

There are exceptions, however. Working with a late model Leica camera and aspheric lenses and a low speed film with the camera on a tripod one can expect to put as much as 150-200 lpm on the film. Not even a a Howtek 4500 can pull all of the detail out of a negative like so so unless you have a HR-8000 or Premier drums scanner you will get higher quality in the darkroom with wet processing.

Sandy King



I'd rather have a scanned image and inkjet than a darkroom print any day. I think the quality is far higher than what you see in the darkroom. There are a lot of factors, not the least of which is dynamic range.

As to a personal preference for the kind of surface one likes, that's for everyone to decide for themselves. However, you can't suggest that inkjet prints don't have at least the quality of a darkroom print - it's simply not supportable.

Lenny

[PenGun]

You do realize that a PMT is a vacuum tube, yes? You could think of a drum scanner as a big single stage vacuum tube amplifier.

Indeed, my very point. Made better by you though ;).

[Lenny Eiger]

Just to wander a bit further. ;) It's interesting that Ken Rockwell has now decided that film is better even at the 35mm level:

http://www.kenrockwell.com/tech/00-new-today.htm


I just called them. They apparently use a film scanner. They were very defensive - at first they didn't want to tell me the brand of their scanner - I finally just asked them what type they had. ("We don't gie out the name of our scanner".) I don't know anything about Mr. Rockwell or why we ought to listen to him - or not, either way. Some day he might like to get a truly great scan...

I say that as their prices were so low - on everything, especially printing. I can't imagine they have any time to do things right... maybe they are good at film developing...

Lenny

[mcfactor]

my point was that it adds another layer to the process. If you optical print an image there is only one other layer, the enlarging lens. when you scan, there is the scanner, the processing of the computer, and the printer.

I did not mean to suggest that the quality of the prints were worse, just that there was another layer added to the process that could (as the above discussion demonstrates) alter the quality.

Also, the discussion was not about dynamic range as much as it was about sharpness and resolution, which are affected by scanning.

[sanking]

Over the past year or so I have had the same 6X7 color negative scanned by several people with different professional flatbed and drum scanners, including Howtek 4500, Howtek 6500 and Screen 1030ai. I have also scanned the same negative with an EverSmart Pro and with a Leafscan 45.

I do have a microscope, and in looking at the negative with a 30X eyepiece it is clear that for all practical purposes the scan made with the EverSmart Pro at 3175 spi is as good as those made with the Howtek 6500 and Screen 1030 ai at 5000 spi. In other words, all of the scans have as much information as can be seen in looking at the negative with a microscope with a 30X eyepiece.

From this I must conclude that the limiting factor was the film, not the scanners. And this color negative was made with a high quality Mamiya 7II camera and a 65mm lens at f/8, with the camera on a tripod, which is about as good as it gets in the analog world of film.

BTW, for more details have a look at the thread I started on the Hybrid forum about this. http://www.hybridphoto.com/forums/showthread.php?t=743

Sandy King









I suspect you are incorrect in the max of 150 mp of info, however I have no data to support such a claim. It doesn't look like that to me on my scans - but that's not particularly scientific. I think what would be useful - perhaps for the entire group here - is to develop some testing methods to figure this stuff out. I think there remain some questions as to what is possible with scanners at the low, mid and high end. I think incorrect info abounds.... much of it no doubt put out by people marketing digital cameras, or the Reichman's of the world.

If a 20x microscope is the way to figure this out, by all means let's find one and put it to the test. I think we ought to develop a list of questions and hunt them down.... Of course, I have to get back to work now rather than start on this, but I think it would be great...

Lenny

[Tyler Boley]

As Sandy mentioned earlier, this thread has wandered so much I'm not sure what the real focus is any more. But when it comes to in depth comparisons of scanned film, capture, darkroom vs ink, etc. etc. the conversation becomes much more complex than some of this indicates.
Given that, I have to stress something Bruce often states here, all that matters to me is what happens on paper, the final object. The in between steps, ccd, pmt, whatever, can only be judged by how they impact the final piece. Looking at comparisons of file magnifications on the monitor only gets us so far. Also, if we are going to judge, what are the criteria? What constitutes higher quality? What does better mean? The opinion that the quality of scanning/inkjet surpasses silver is just that, an opinion. I ask this as someone committed to drum scanning and the highest quality inkjet systems and that aesthetic, I love it- how come few, if any, ink prints can sit next to one of Caponigro's (just as an example) finest? A Tice, whatever...
I ask this against my own interests, I offer drum scanning, inkjet printing, etc etc for a living, my own personal work is committed to these processes and I depend to a smaller degree on sales of that too. But in the interests of real artistic pursuit, real scientific knowledge, and advancement of the craft we love, absolute honesty is required, difficult, but required.
Tyler
http://www.custom-digital.com/

[PenGun]

I've looked at very many of the available comparisons found on the net. From that rather poor perspective what I have noticed is that while the actual resolution and detail is similar on good scans from good machines the shadows are owned by the PMTs. I've seen very nice scans with actual antialiasing in the darker parts of CCD scans that were otherwise very impressive.

[sanking]

Tyler,

To put this in perspective, I print with a rather arcane, but extraordinarily beautiful in my opinion, 19th century process called carbon transfer. I would wager that less than 1% of the people on this forum have even heard of it, and probably only one in a thousand has ever seen a real carbon transfer print. That in itself demonstrates a fair amount of commitment to the craft of photography. And I consider myself first and foremost a shooter and a printmaker, with everything in between the path.

That said, I love the path, or the process to the final print. The path is how most of us spend our time in the pursuit of our artistic objectives, and in the end the path is more meaningful than the end result, which is when all is said and done a thing, or product. It takes me a fraction of a second to see and make a negative, and once the final print is in hand the mystery is over, but all between is a magical path, and time I treasure.

Sandy King



As Sandy mentioned earlier, this thread has wandered so much I'm not sure what the real focus is any more. But when it comes to in depth comparisons of scanned film, capture, darkroom vs ink, etc. etc. the conversation becomes much more complex than some of this indicates.
Given that, I have to stress something Bruce often states here, all that matters to me is what happens on paper, the final object. The in between steps, ccd, pmt, whatever, can only be judged by how they impact the final piece. Looking at comparisons of file magnifications on the monitor only gets us so far. Also, if we are going to judge, what are the criteria? What constitutes higher quality? What does better mean? The opinion that the quality of scanning/inkjet surpasses silver is just that, an opinion. I ask this as someone committed to drum scanning and the highest quality inkjet systems and that aesthetic, I love it- how come few, if any, ink prints can sit next to one of Caponigro's (just as an example) finest? A Tice, whatever...
I ask this against my own interests, I offer drum scanning, inkjet printing, etc etc for a living, my own personal work is committed to these processes and I depend to a smaller degree on sales of that too. But in the interests of real artistic pursuit, real scientific knowledge, and advancement of the craft we love, absolute honesty is required, difficult, but required.
Tyler
http://www.custom-digital.com/

[Lenny Eiger]

Over the past year or so I have had the same 6X7 color negative scanned by several people with different professional flatbed and drum scanners, including Howtek 4500, Howtek 6500 and Screen 1030ai. I have also scanned the same negative with an EverSmart Pro and with a Leafscan 45.
From this I must conclude that the limiting factor was the film, not the scanners. And this color negative was made with a high quality Mamiya 7II camera and a 65mm lens at f/8, with the camera on a tripod, which is about as good as it gets in the analog world of film.
Sandy King

I had seen this before. Perhaps another read is in order. If you want to send the neg over here, I'd be happy to scan it on my machine... just to see...

Lenny

[Lenny Eiger]

Given that, I have to stress something Bruce often states here, all that matters to me is what happens on paper, the final object. The in between steps, ccd, pmt, whatever, can only be judged by how they impact the final piece. Looking at comparisons of file magnifications on the monitor only gets us so far. Also, if we are going to judge, what are the criteria? What constitutes higher quality? What does better mean? The opinion that the quality of scanning/inkjet surpasses silver is just that, an opinion. I ask this as someone committed to drum scanning and the highest quality inkjet systems and that aesthetic, I love it- how come few, if any, ink prints can sit next to one of Caponigro's (just as an example) finest? A Tice, whatever...
But in the interests of real artistic pursuit, real scientific knowledge, and advancement of the craft we love, absolute honesty is required, difficult, but required.
Tyler
[]

Tyler,

Of course you're right. Great prints are just that, great prints. I love them, in any media. My favorite is a gravure I have, a Clarence White, from Camerawork, called Morning. Ultimately, its about photography.

I would say two things. The first that on a site like this many participants are asking for advice on what to purchase. Is it worth it to buy an expensive scanner, for example? They have to make a buying decision and they can be costly. As you have I am sure, I have many software programs, plugins and the like that I just threw away, as they were not as useful as we had hoped. There are lots of mistakes to be made. When one is buying a camera, we can't tell them its about artistic pursuit, we have to ask what they want to accomplish and help them get what they need to accomplish it. When it comes to scanning, aesthetics come into play, as some people aren't after what you or I might call a great print. The criteria gets muddled. Manufacturers stretch the truth. Marketers... it would be nice to have clear criteria outlined and results tabulated. Maybe I'm just dreaming... but this topic does come up a lot...

The other thing is from your lips to God's ears. I wish more of photography had a primary component of honesty. There is way too much post-modernism these days for my taste.

Lenny

[Tyler Boley]

Sandy, I'm in the 1%, and even the one in a thousand, and am also aware of what you do and your commitment. I absolutely agree with everything you said. I don't think the point I wanted to make is at odds with your post. I also live in the details of the process, and love every educational second of it. I hope I never arrive at any dogma about it, only dogmatic commitment. I'm only saying that beyond pure technical prowess evaluated about each small step, placing judgment on one step's methodology must be in the context of it's contribution to the end piece. I am disturbingly committed to the demands of each step, I didn't mean to suggest otherwise, as demonstrated by this article-
http://www.custom-digital.com/2008/09/bw-print-quality/

Thanks for your post, nice to read, I am in agreement entirely and hope my point was clear. I'll do anything that I stumble on to get to the print I want, some things I'd never admit amongst the demanding technicians here, other things at the bleeding edge. I hope I'm making sense... There are factual differences between processes that must be acknowledged, but they may or may not be relevant to the desired end.

Tyler

[sanking]

Tyler,

Thanks for your comments.

If you ever get to my area of the country bring some prints and plan to stop by and stay for a while. I think we would have lots to talk about.

Sandy







Sandy, I'm in the 1%, and even the one in a thousand, and am also aware of what you do and your commitment. I absolutely agree with everything you said. I don't think the point I wanted to make is at odds with your post. I also live in the details of the process, and love every educational second of it. I hope I never arrive at any dogma about it, only dogmatic commitment. I'm only saying that beyond pure technical prowess evaluated about each small step, placing judgment on one step's methodology must be in the context of it's contribution to the end piece. I am disturbingly committed to the demands of each step, I didn't mean to suggest otherwise, as demonstrated by this article-
http://www.custom-digital.com/2008/09/bw-print-quality/

Thanks for your post, nice to read, I am in agreement entirely and hope my point was clear. I'll do anything that I stumble on to get to the print I want, some things I'd never admit amongst the demanding technicians here, other things at the bleeding edge. I hope I'm making sense... There are factual differences between processes that must be acknowledged, but they may or may not be relevant to the desired end.

Tyler

[Tyler Boley]

Sandy, Likewise. I've got Todd Gangler right up the street so have seen good tri-color. Have seen mono as well but no contemporary. Speaking of alt processes, anyone seen a woodburytype? Lovely.
OK, I'm way OT now...
Back to the OP, I saw side by side, tango scanned 4x5 Velvia and P45 of exact same scene, printed large on Premium Luster at Bill Atkinson's place. He thought it was evidence that capture had arrived, I liked the scanned film better. They were admittedly close. Again, I get lost at the cost comparisons..
Tyler

[bglick]

> All I know is that while the maximum optical resolution of the Cezanne Elite is 5300 dpi it can only scan at this resolution in a relatively small strip, about 1 1/2" or 2" wide I believe. To scan anything wider the Cezanne has to reposition the lens and place it farther from the CCD, which results in less optical resolution. I am guessing that the maximum true otical resolution for 4X5" is 2400 dpi -3200 dpi but don't know for sure.


From Screen Cezanne Elite pdf spec sheet:

The advanced CCD array and XY zoom mechanism help the Cézanne Elite attain its remarkable 589 to 5,300 dpi range of optical resolutions across the entire scanning bed. This allows operators to use the most appropriate resolutions according to the sizes of the originals and the desired enlargement/reduction ratios. Output is possible at up to 20,000 dpi with the aid of the Cézanne Elite’s precision interpolation function, and the user can choose RGB, CMYK, Grayscale, or lineart output.


I am curious, maybe this was one of the improvements the Elite had over the original Cezanne? I have not noticed the shortcoming Sandy mentioned above, then again, I never tested for it either...


> The P45 captures 39mp images, but that is compressed, uncompressed i think its around 120mp(don't quote me on that).


A pixel is a recording site, period. The MP of the camera is the number of sites in which the sensor can sense light. From there, everything is interpolation. You could make a similar argument with film.... and say its recording capacity is its grain size.



> Yes, and the Cezanne is a mediocre scanner.

This is a $45k scanner, maybe today down to nearly half this amount. As Adam correctly stated, in the last Seibold 01 head-to-head test, judged by a panel of experts, the Cezzanne beat every scanner tested, except the ICG drum, which just barely beat it. I have used many drum scanners and the Cezzane, I consider the Cezzane to be at least equal to the best drum scanners, except maybe the ICG, which I have not used.


As mentioned previously, this thread wandered all over the map, but nonetheless, some great contributions... I would like to add a few comments....

As others have noted, most comparisons between film and digital are absurd.... in order to compare the two, there must be specific set of categories for comparison, such as type of capture (long DOF shots, shooting flat 2d subjects, subject color, subject orientation), lenses used for each, film type (low rez neg color, higher rez trannie, ulra high B&W, etc) , camera quality, scanner type used, how well tuned is the scanner and operator, etc. How well was the film processed, how accurate was the focus....you could go on forever. The point being, often times the generalizations don't nail down the variables, so the conclusion might be accurate, but not applicable when the variables change...


The scanners themselves are mystery boxes. No one knows the exact workings of the scanner themselves. For example... when a scanner scans at 2000 dpi.... does this mean it offers a final output at 2000 dpi, or does it take in 6000 dpi, interpolate an output 2000 dpi. I think this is "one" of the reasons we see low end consumer grade scanner suggest 4000 dpi, but deliver 2000 dpi, as Nyquist is at play here... so you need a great deal of over sampling to actually resolve 2000 dpi.... so a mix of "whats in the box" combined with marketing BS, leaves all this mysterious in the end. As mentioned, what counts is the finished scans....and how they compare to other scans...or to digital capture.

[Lenny Eiger]

> Yes, and the Cezanne is a mediocre scanner.
This is a $45k scanner, maybe today down to nearly half this amount. As Adam correctly stated, in the last Seibold 01 head-to-head test, judged by a panel of experts, the Cezzanne beat every scanner tested, except the ICG drum, which just barely beat it. I have used many drum scanners and the Cezzane, I consider the Cezzane to be at least equal to the best drum scanners, except maybe the ICG, which I have not used.


I don't care how much things cost. The fist scanner I used was a Hell 299, it was in the hundreds of thousands, if memory serves. The Seybold test is not considered authoritative. You probably never tested a Premier, it's better than the older ICG's, for certain. The Cezanne is not close. It don't think it would be equal to any drum scanner. I don't know what you are trying to convince yourself of, but there is physics to be taken into account.



The scanners themselves are mystery boxes. No one knows the exact workings of the scanner themselves. For example... when a scanner scans at 2000 dpi.... does this mean it offers a final output at 2000 dpi, or does it take in 6000 dpi, interpolate an output 2000 dpi. I think this is "one" of the reasons we see low end consumer grade scanner suggest 4000 dpi, but deliver 2000 dpi, as Nyquist is at play here... so you need a great deal of over sampling to actually resolve 2000 dpi.... so a mix of "whats in the box" combined with marketing BS, leaves all this mysterious in the end. As mentioned, what counts is the finished scans....and how they compare to other scans...or to digital capture.

It has nothing to do with Nyquist. It has to do with the different between how many pixels one can make vs the actual optical resolution of the sampling mechanism (how many bars it can separate). In a drum that would be related to the stepper motor and the choices of steps where the samples are taken.

if you are making scans you are happy with - with a Cezanne, that's great.

Lenny

[sanking]

This has come up before. The Screen Cezanne Elite uses XY zooming, not XY stitching. So it can scan at 5300 spi anywhere on the bed, but at that zoom magnification the strip is limited to 1.5" wide. For anything wider the lens zooms out to cover the larger area. This is quite unlike the XY stitching used by EverSmart and IQSmart scanners, which mow the lawn and then stitch it together. Unlike the Cezanne and Cezanne Elite the EverSmart and IQSmart scanners can scan at the full platen size at the maximum optical resolution.

The optical resolution of flatbed scanners is determined by how many times they sample at 90 degrees to the direrection of travel of the CCD. My EverSmart Pro, for example, samples at 8020 spi in the direction of the run of the CCD as determined by the stepper motor, but only 3175 spi in the other direction. The stated optical resolution is therefore 3175 spi, or about 62 lpm. However, if you actually test a resolution target with this scanner it will show effective resolution of about 78 lpm. I assume that the Cezanne will do likewise as Peter alluded to in a message a couple of days back.

I don't own a Cezanne and can not comment on its quality. However, in actual comparison tests I have had done with the EverSmart Pro and severla drum scanners, all with higher optical resolution, the EverSmart Pro scans were very comparable in quality. Obviously operator competence has to be taken into consideraton here as in the comparison my friend Don Hutton made a much better scan with his Howtek 4500 than another fellow did with a 6500.


Sandy King


"The advanced CCD array and XY zoom mechanism help the Cézanne Elite attain its remarkable 589 to 5,300 dpi range of optical resolutions across the entire scanning bed. This allows operators to use the most appropriate resolutions according to the sizes of the originals and the desired enlargement/reduction ratios. Output is possible at up to 20,000 dpi with the aid of the Cézanne Elite’s precision interpolation function, and the user can choose RGB, CMYK, Grayscale, or lineart output.

The scanners themselves are mystery boxes. No one knows the exact workings of the scanner themselves. For example... when a scanner scans at 2000 dpi.... does this mean it offers a final output at 2000 dpi, or does it take in 6000 dpi, interpolate an output 2000 dpi. I think this is "one" of the reasons we see low end consumer grade scanner suggest 4000 dpi, but deliver 2000 dpi, as Nyquist is at play here... so you need a great deal of over sampling to actually resolve 2000 dpi.... so a mix of "whats in the box" combined with marketing BS, leaves all this mysterious in the end. As mentioned, what counts is the finished scans....and how they compare to other scans...or to digital capture." [/QUOTE]

[bglick]

> It has nothing to do with Nyquist.

You are mistaken, it has everything to do with Nyquist.... 1/R is the basis of recordable resolution, period...




> It has to do with the different between how many pixels one can make vs the actual optical resolution of the sampling mechanism (how many bars it can separate).

This is Nyquist and 1/R, you are contradicting yourself.....




> In a drum that would be related to the stepper motor and the choices of steps where the samples are taken.


That is ONE variable of the drum scan, its equivalent to one axis of resolution on a XY scan. At 8000 dpi, this equates to 157 pp/mm. A diffraction limited lens viewing film at f2.8 can deliver 1500/2.8 = 536 lp/mm, almost 4x the resolution of a 8000 lines / inch on the drums rotation. A lens has more potential than revolving drum. A drum scanners native image can be thought of as threads of a screw. I think you are overlooking the potential of a sharp optics.


> You probably never tested a Premier, it's better than the older ICG's, for certain. The Cezanne is not close. It don't think it would be equal to any drum scanner.


I have compared scans from a Howtek 8000 I owned, and still use vs. the Premier, I can see little or no differences. As you know. the basis of the premier was the Howtek 8000. A friend of mine a few years back, went to visit Phil when considering an upgrade from his 8000 to the Premier, he brought film he had scanned on his 8000, and Phil could barely match it after tweaking it for quite some time. Bottom line, the last generation of Howtek scanners, mainly the 4500 and 8000 scanners were great scanners, period. (of course, anyone can get a bad sample) After comparing Howtek scans to Cezanne scans, for color film, I would consider them nearly the same, hence why I finally left the drum world and went flat bed. However, if I was scanning B&W, I would still own a drum.


As for the price of scanners, lets get real, we all know scanners in the early 90's were in the 200k price range.... I was comparing prices of todays scanners, not early 90's like you converted my reference to... no company can sell a $30k scanner if it performs like a consumer level "soft" scanner as you have implied many times for the Cezanne. Others on this list have mentioned their Cezanne scans have equaled or beat drum scans, yet somehow, you continue to ignore what they write. You accuse everyone of defending their scanners, instead, it appears from your posts, it's you who continually defends your Aztek....



> The Seybold test is not considered authoritative.

It was the last non biased, independently juried scanner test I have read about....do you have any other tests to share?

[bglick]

> The optical resolution of flatbed scanners is determined by how many times they sample at 90 degrees to the direrection of travel of the CCD. My EverSmart Pro, for example, samples at 8020 spi in the direction of the run of the CCD as determined by the stepper motor, but only 3175 spi in the other direction. The stated optical resolution is therefore 3175 spi, or about 62 lpm. However, if you actually test a resolution target with this scanner it will show effective resolution of about 78 lpm. I assume that the Cezanne will do likewise as Peter alluded to in a message a couple of days back.


Fully agreed with above.... but as I mentioned previously, much of this is premised on some of the variables we don't know. Such as, at each step, what size image is captured? how is that image spread out over the pixels on the CCD chip? How much over sampling? Depth of focus at the CCD chip? etc, etc. So my point is, many of these unknown variables still leaves some mystery...but in the end, it all comes out in the wash when you make a scan, so IMO, it makes no sense beating up all the "possible" internal hardware / software / optics that each manufacturer utilizes to achieve their desired scan resolutions. This is whats nice about a head-to-head comparison like Seibold did in 2001....

[Lenny Eiger]

> It has nothing to do with Nyquist.
You are mistaken, it has everything to do with Nyquist.... 1/R is the basis of recordable resolution, period...
> It has to do with the different between how many pixels one can make vs the actual optical resolution of the sampling mechanism (how many bars it can separate).
This is Nyquist and 1/R, you are contradicting yourself.....


First of all, Phil Lippincott told me that Nyquist wasn't a factor in what he did. We already went over this, some possible explanations for his stance were offered. That was good enough for me. I don't have to rehash it.



> In a drum that would be related to the stepper motor and the choices of steps where the samples are taken.
That is ONE variable of the drum scan, its equivalent to one axis of resolution on a XY scan. At 8000 dpi, this equates to 157 pp/mm. A diffraction limited lens viewing film at f2.8 can deliver 1500/2.8 = 536 lp/mm, almost 4x the resolution of a 8000 lines / inch on the drums rotation. A lens has more potential than revolving drum. A drum scanners native image can be thought of as threads of a screw. I think you are overlooking the potential of a sharp optics.


There are a lot of other factors here. CCD's vs PMT's and everything else. Accuracy of sampling area, the machine knowing where it is. We've been doing this thread for days now.



> You probably never tested a Premier, it's better than the older ICG's, for certain. The Cezanne is not close. It don't think it would be equal to any drum scanner.
I have compared scans from a Howtek 8000 I owned, and still use vs. the Premier, I can see little or no differences. As you know. the basis of the premier was the Howtek 8000. A friend of mine a few years back, went to visit Phil when considering an upgrade from his 8000 to the Premier, he brought film he had scanned on his 8000, and Phil could barely match it after tweaking it for quite some time. Bottom line, the last generation of Howtek scanners, mainly the 4500 and 8000 scanners were great scanners, period. (of course, anyone can get a bad sample) After comparing Howtek scans to Cezanne scans, for color film, I would consider them nearly the same, hence why I finally left the drum world and went flat bed. However, if I was scanning B&W, I would still own a drum.


Gee, the BG in BGLick wouldn't be short for Bill Gillooley (SP?), would it?

What you are saying simply does not jive with my experience. I have a client that did a lot of scans on the Cezanne - some of which he brought to me and were analyzed by the both of us - and came here only because this machine could outdo his by such a wide margin. That was last year, he moved back to Australia... and I checked, don't have any of those older originals. I saw no reason to keep them.


As for the price of scanners, lets get real, we all know scanners in the early 90's were in the 200k price range.... I was comparing prices of todays scanners, not early 90's like you converted my reference to... no company can sell a $30k scanner if it performs like a consumer level "soft" scanner as you have implied many times for the Cezanne. Others on this list have mentioned their Cezanne scans have equaled or beat drum scans, yet somehow, you continue to ignore what they write. You accuse everyone of defending their scanners, instead, it appears from your posts, it's you who continually defends your Aztek....


Everyone makes claims about everything. Luminous Landscape will have you believing that a 4x5 piece of film is less than the latest 35mm digital camera. I simply respond to how I see things. I don't think a Cezanne is worth 30K.



> The Seybold test is not considered authoritative.
It was the last non biased, independently juried scanner test I have read about....do you have any other tests to share?

I refer to the tests at Aztek, with my discussions with them, etc., and their independent scannerforum site. According to them they made a real effort to be fair... and I have no reason to distrust them. They aren't the kind of folks who would stretch it just to further their product, certainly not with me...

[sanking]

What goes on inside a scanner is pretty much a mystery to me, but I do understandhow to test my equipment to evaluate end results.

What I did want to emphasize about the Cezanne Elite, which perhaps you understand, is that if you are scanninig anything over 1.5" wide at 5300 spi there is some interpolation going on since the lens will have to zoom out to cover a larger area, and this results in decreased optical resolution. This may or may not make much difference in the actual end result sharpness -- that is for Cezanne users to evaluate and judge.

Sandy King





> The optical resolution of flatbed scanners is determined by how many times they sample at 90 degrees to the direrection of travel of the CCD. My EverSmart Pro, for example, samples at 8020 spi in the direction of the run of the CCD as determined by the stepper motor, but only 3175 spi in the other direction. The stated optical resolution is therefore 3175 spi, or about 62 lpm. However, if you actually test a resolution target with this scanner it will show effective resolution of about 78 lpm. I assume that the Cezanne will do likewise as Peter alluded to in a message a couple of days back.


Fully agreed with above.... but as I mentioned previously, much of this is premised on some of the variables we don't know. Such as, at each step, what size image is captured? how is that image spread out over the pixels on the CCD chip? How much over sampling? Depth of focus at the CCD chip? etc, etc. So my point is, many of these unknown variables still leaves some mystery...but in the end, it all comes out in the wash when you make a scan, so IMO, it makes no sense beating up all the "possible" internal hardware / software / optics that each manufacturer utilizes to achieve their desired scan resolutions. This is whats nice about a head-to-head comparison like Seibold did in 2001....

[bglick]

> Gee, the BG in BGLick wouldn't be short for Bill Gillooley (SP?), would it?


Is this a joke I am not getting? Who is Gillooley?




> First of all, Phil Lippincott told me that Nyquist wasn't a factor in what he did.


Phil was excited to sell his products. You seem to think anything Phil once said is Gospel, and all other knowledgeable people should be dismissed. Your Aztek allegiance is loud and clear. There is nothing wrong with that. Anyone that understands image capture can not escape the Nyquist premise. All scanners are digital cameras, they all have a subject plane and an image plane.



> Everyone makes claims about everything. Luminous Landscape will have you believing that a 4x5 piece of film is less than the latest 35mm digital camera.

I think you are exaggerating to make a point..... Mike R never claimed this, but he did compare 4x5 to 39MP MF back, which i addressed in another post, which should be moved here, as I mistakenly posted it to the wrong thread.



> I refer to the tests at Aztek, with my discussions with them, etc., and their independent scannerforum site. According to them they made a real effort to be fair...

hmmmm.... a manufacturer running a test, with their scanner being one of the participants. I bet I know the outcome before even reading the results. And you feel this would represent a less bias test than the Seibold, which you were quick to dismiss? I do not agree with your logic here....

[sanking]

This is kind of funny. Bill is a former Scitex operator/technician who knows a lot about EverSmart scanners, and scanners in general including drums. He is known as Mr. Bill on teh Scan Hi-End forum and has given us EverSmart scanners a lot of technical support.

It is a hoot that Lenny would think you were BG. Bill has many times made the point that EverSmart scanners are sharper than Fuji Lavonia and Cezanne because of a higher quality lens and XY stitching.

Sandy






> Gee, the BG in BGLick wouldn't be short for Bill Gillooley (SP?), would it?


Is this a joke I am not getting? Who is Gillooley?




> First of all, Phil Lippincott told me that Nyquist wasn't a factor in what he did.


Phil was excited to sell his products. You seem to think anything Phil once said is Gospel, and all other knowledgeable people should be dismissed. Your Aztek allegiance is loud and clear. There is nothing wrong with that. Anyone that understands image capture can not escape the Nyquist premise. All scanners are digital cameras, they all have a subject plane and an image plane.



> Everyone makes claims about everything. Luminous Landscape will have you believing that a 4x5 piece of film is less than the latest 35mm digital camera.

I think you are exaggerating to make a point..... Mike R never claimed this, but he did compare 4x5 to 39MP MF back, which i addressed in another post, which should be moved here, as I mistakenly posted it to the wrong thread.



> I refer to the tests at Aztek, with my discussions with them, etc., and their independent scannerforum site. According to them they made a real effort to be fair...

hmmmm.... a manufacturer running a test, with their scanner being one of the participants. I bet I know the outcome before even reading the results. And you feel this would represent a less bias test than the Seibold, which you were quick to dismiss? I do not agree with your logic here....

[Lenny Eiger]

> Gee, the BG in BGLick wouldn't be short for Bill Gillooley (SP?), would it?
Is this a joke I am not getting? Who is Gillooley?

Bill's a poster on another list.



> First of all, Phil Lippincott told me that Nyquist wasn't a factor in what he did.
Phil was excited to sell his products. You seem to think anything Phil once said is Gospel, and all other knowledgeable people should be dismissed. Your Aztek allegiance is loud and clear. There is nothing wrong with that. Anyone that understands image capture can not escape the Nyquist premise. All scanners are digital cameras, they all have a subject plane and an image plane.


I have been very up front that the Aztek folks are friends of mine. They've been good to me. They aren't perfect, by any means. I don't want to dismiss anyone who knows something. I do give Phil's opinion a lot of weight. Of course, I don't know who you are, what your experience is - or anything about you. For all I know you could be a great genius. Maybe I should take your word over Phil's. I'm not being sarcastic. But I don't know your name, there is no link to a web site, there is no way for me to judge how I should weigh your comments.



> Everyone makes claims about everything. Luminous Landscape will have you believing that a 4x5 piece of film is less than the latest 35mm digital camera.
I think you are exaggerating to make a point..... Mike R never claimed this, but he did compare 4x5 to 39MP MF back.


He didn't compare his P45 to a high end drum scanner. I think he should have. If you think the Cezanne is just as good, maybe it doesn't matter to you. I think if you compare the top product in one category you should use the top product in the other.

[dwhistance]

Hi, I posted this in response to Bglick's comment on the other thread but as it really relates to this one will post it here as well:

Bglick, many thanks for the well reasoned comments. Can I ask one question which bothers me - how does the Bayer filter (if I've got the right term!) affect your resolution figures? Surely with a one shot back you need to reduce the effective number of pixels to take account of the distribution of the colour pixels? Also presumbaly the interpolation itself loses some resolution? If you do adjust for this how would the revised numbers look?

David Whistance

[Peter De Smidt]

What you are saying simply does not jive with my experience. I have a client that did a lot of scans on the Cezanne - some of which he brought to me and were analyzed by the both of us - and came here only because this machine could outdo his by such a wide margin. That was last year, he moved back to Australia... and I checked, don't have any of those older originals. I saw no reason to keep them.



So your evaluation of the Cezanne amounts to results from one scanner/user? That seems a rather small sample to support your rather sweeping generalizations.

[Gordon Moat]

If you consider a simple approach of how the CCD in a MFDB works through a Bayer pattern, that is one factor, though more in colour interpolation than resolution (not file sizes). There are twice as many green filtered areas as blue or red. There is also a factor of how much light passes through the colour filtering, and gets collected in the well. Quite simply, the physical size of the pixel, and the spacing from one pixel to another, can have a greater affect on optical capability than the Bayer pattern.

I have yet to see any test of any MFDB or D-SLR that can resolve a true clean optical capability greater than about 70% of it's file size suggested theoretical limits (excluding scanning backs). My terminology might not be as technical as other's here, but hopefully that makes sense. Addition of an anti-alias filter or anti-moiré layer can further reduce resolution, though can be useful to avoid other possible errors.

Remember that white is a full signal capture while black is no signal capture (or think of it as full charge and no charge respectively). Due to this factor, and Bayer filtration, and interpolation between adjacent pixels, darker areas of a scene could be more prone to errors or reduced accuracy than brighter areas. Sometimes this shows up as noise, while other times it is false colour at a near pixel level.

Some of the current MFDBs are quite good, and can allow very impressive and compelling results. I don't see a need to state or claim that this newer technology is better nor worse than scanned film. In reality, a well shot image on film, translated nicely into a compelling printed image, can potentially be quite good, and impress many people. The biggest arguments in favor of MFDBs are largely speed and convenience.

Ciao!

Gordon Moat Photography (http://www.gordonmoat.com)

[Peter De Smidt]

What I did want to emphasize about the Cezanne Elite, which perhaps you understand, is that if you are scanninig anything over 1.5" wide at 5300 spi there is some interpolation going on since the lens will have to zoom out to cover a larger area, and this results in decreased optical resolution. This may or may not make much difference in the actual end result sharpness -- that is for Cezanne users to evaluate and judge.

Sandy King

I own a Cezanne, and Sandy is absolutely right. That said, it isn't that hard to stitch multiple scans together in software, if needed.

[Jim Graves]

Amazing thread ... WAY over my head, but amazing. Has given me a lot to think about. I am convinced that I need to learn this area, leave the wet darkroom (as much as I hate that thought) and enjoy the fruits of what the digital revolution has done for film photography.

With the worry of going farther afield in this thread: Two things:

1) Lenny had a booth at the Fort Collins FOTO3 conference and spent at least 45 minutes with me doing a free scan of 4x5 negative ... with a running monologue of sampling ... etc., etc. ... very gracious and generous with his time and expertise ... and patient ... I thank him for that ... I learned a lot.

2) I was recently introduced to Carbon Printing (see Sandy's earlier post) and loved the end result ... I have signed up for the April 2009 Ansel Adams Gallery workshop at Yosemite for carbon printing ... it is an amazing process with enormous returns ... can't wait!

[PenGun]

Yes indeed. Some heat but a lot of light. I've learned a great deal in this thread.

[bglick]

Peter, have you ever put test targets in the center and the edges of the scan area, then scanned at 5300 dpi? what was the results?


Do you have the Elite? I am just curious if Screens spec sheet is not accurate, or poorly worded. My scanner is not hooked up right now, otherwise I would try this...

[Peter De Smidt]

No, I haven't done any testing with appropriate resolution-determining test slides, mainly because one of those slides costs half of what I paid for my scanner.

I have the original Cezanne. I don't have my manual handy, but here's a quote from the Seybold report:

"The Cezanne scans reflective and transparency copy up to a maximum input format of 348x530mm. Although the top resolution is nominally 5,300 ppi, the Cezanne's resolution actually varies according to the type and size of the image to be scanned." p12 of the scanner test issue.

The resolution wouldn't vary if the Cezanne could combine rows of max-res scans to cover it's entire bed. As Sandy said, placement on the bed does not change resolution, but size of original does, as the lens must zoom out to get the whole original imaged on the 8000 element ccd array.

The xy aspect of the scanner means that maximum resolution is available anywhere on the bed but not _over_ the whole bed. This is in contrast to scanners such as Agfa's T2500 in which maximum resolution is only available over a 2" strip down the center of the glass.

[Btw., the late, great Ted Harris, who also owned the original Cezanne, confirmed this view in a private email.]

The original Cezanne doesn't stitch with grayscale or continuous color scans, but it might with copydot scanning, as I seem to remember the manual stating this.

If you look at the brochure pdf for the Elite, available at:

http://www.screen.co.jp/ga_dtp/en/product/pdf/204-144E.pdf

You'll note that the specified optical resolution for continuous toned originals is listed as "589 to 5300 dpi". Doing the math confirms that the 589 dpi figure corresponds to what an 8000 element ccd array would produce when focused over the entire front to back depth of the Elite's scanner bed.

[aluncrockford]

As somebody who uses the P45 every day and used to use 10/8 and 5/4 both scanned by drum and Imacon scanners ,I would say that from a commercial point of view the P45 wins hands down , When we did the comparison tests we were working on a job shooting on 10/8 we scanned the trannie with a high res drum scan at 140mgs and compared it with the P45 file ,the digital was sharper cleaner and faster , the agency went with the digital file and that was the last time I shot film on a job . the files from the digital capture are a lot cleaner and can be rezed up without too much loss of detail with film there is a maximum scan size that beyond that point all you are doing is enlarging the grain and that has a tendency to give the image a much softer look . Again from a commercial point of view when we are working on massive prints ( the size of a house) agencies will insist on P45 digital files over film every time.

[bglick]

> the files from the digital capture are a lot cleaner and can be rezed up without too much loss of detail with film there is a maximum scan size that beyond that point all you are doing is enlarging the grain and that has a tendency to give the image a much softer look .


A very valid point which I failed to mention.... all digital capture files have better up rez potential vs. scanned film. When you scan film at its max. you are truly at the max. Its remarkable how well digital capture files up rez... specially the best backs with large pixel sites, low noise and 16 bit A/D.


I am curious though, in this example you cite, what f stop did you shoot the P45 at?

[Gordon Moat]

It is quite often that the largest printed size becomes the limiting factor, more so than the capture methods. As commercially printed sizes go up, the printing requirements per unit area decrease, to keep file sizes down, better match the gear, and to ensure reasonable printing times. A billboard is not a high resolution image, just because you are unlikely to stand next to it, just to use a (very) obvious example.

I have yet to ever see any example that was more than a convenience comparison. Sharpness is not resolution, but is often mistaken as improvement, and that can be applied to film-to-film comparison too.

Q. Can you get a very nice large print from 4x5 or 8x10?
A. Yes

Q. Can you use one of the latest MFDBs to get a nice large print?
A. Yes

See how simple that is . . . no need for one to be better than the other. Why doesn't someone throw a 6µm pixel square grid over a piece of film, and then maybe a Bayer pattern, and improve the scanning capability of film? That way we can create discrete blocks of capture that can up-rez into more blocks.

I would not argue against the convenience of a MFDB, and I have enjoyed using one. Does that mean I think film sucks, scanning sucks, or the combination is somehow inferior . . . No. I don't think anyone should buy, lease, nor rent a MFDB based upon anything other than meeting your needs; don't make the commitment based upon claims, in favor or against. Up-rezzing never increases the amount of detail captured; all it does is spread it out more.

Ciao!

Gordon Moat Photography (http://www.gordonmoat.com)

[SimonV]

Hello, awhile ago there was post by Victoria Perelet on this forum with link on her site to side by comparison of drum vs flat bed scans of 8x10 neg, contact/digi print and few other comparisons. I googled and searched this site but can not find URL. If anybody bookmarked it - it may add to this thread.

[Paul Kierstead]

Q. Can you get a very nice large print from 4x5 or 8x10?
A. Yes

Q. Can you use one of the latest MFDBs to get a nice large print?
A. Yes


The single best, and most relevant, analysis so far. Academic analysis is interesting, but that is the bottom line.

[bglick]

> I don't think anyone should buy, lease, nor rent a MFDB based upon anything other than meeting your needs;

> Academic analysis is interesting, but .....


It seems all these threads through the years follow a similar pattern. They go something like this.....


1) A question about resolution between film formats, digital, scanners, etc.

2) some math to explain the differences

3) Disputing the math from real world results (of course all the relative details are rarely provided)

4) The math naysayers (artist) jump in..... Photography is not about math, "just go take pictures" !!!

5) Justification of all the options often by personal experience..

Then, usually a final...

6) None of this matters, you have to test everything to see what meets your needs.


To the newbie, and to even some pros who don't obsess over the numbers, these threads have to generate a lot of "head shaking" , i.e. total confusion, and rightfully so.... to that end, I would like to offer the following....

1) Many people use this information to attempt to make "buy" decisions for their photography. It's unfortunate, but most of this equipment is not available to be tested side by side. This specially applies to equipment such as scanners which are no longer made.


2) While math will often NOT give you ever final detail about a given comparison, in certain areas it can define the physical limits of what's possible. In most cases, this will get you in the ballpark, vs. without the math, you will be in "the wrong continent." This is significant, but is often completely dismissed because there is so many variables at play within a comparison, analysis can never be 100% accurate, in which case, many people assume it's better to "throw the baby out with the bath water". IMO, this is a mistake. I rather get "close to" an answer vs. "have no clue".


3) Unfortunately, in these comparisons, "the devil is in the details". And often these details are not available in forum threads. Gordon touched on this above, but I would like to expand on this... by simply altering 3 "user controlled variables" - changing the apt. of a taking lens, film type / format, and the scanner used to scan the film, I can demonstrate:

a) Large Format Film will out resolve a MF back

b) MF back can out resolve LF film.


Both can be true under a given set of variables. Unfortunately, you must dig into the details to get to these variables. Many of these variables surfaced throughout this thread, size of format, film type, lenses used, DOF, f stop, color of subject, MP's, up rez potential, scanner quality, etc.

[Lenny Eiger]

Q. Can you get a very nice large print from 4x5 or 8x10?
A. Yes

Q. Can you use one of the latest MFDBs to get a nice large print?
A. Yes


Maybe, in color. I am not that interested in "very nice". I am interested in very specific qualities.

I think that in a decent size, say up to 40 inches, in black and white, that I just haven't seen the capability from any digital capture. I'm talking about a museum quality print, with tons of richness, that a gravure printer, say a Paul Strand, would be proud of. I'm more interested in tonality than resolution and certainly depth of field vs critical sharpness. Just my thing...

In response to aluncrockford, all due respect and all, I wouldn't trust what agencies want - they are mostly non-technical, non-photographers, and have no idea of what they are talking about. It's been a very long time since those people were craftspeople.

For commercial work, it's there with all everyone needs. If I was doing commercial work I would use whatever digital fit the bill for the work I was asked to do. I don't do commercial work and don't really want to and my priorities are therefore different...

Lenny

[Joerg Krusche]

hi,

there is an easy and inexpensive way for you to test our own scanner .. just buy a USAF 1951 test chart on film (will cost you about 50 euros) which goes up to group 7 .. did that with the lanovia .. max resolution for me is in the range of 50 to 55 lps/mm .. this means that you can with your nose close up still will have your 5 line pairs per mm .. the picture/print will be critically sharp at all viewing distances .. and 4x5 will give you a 40x50 print.... greater viewing distances do not even require those about 2500 ppi .. such a simple test will tell you what your specific equipment can deliver ... and actual 2500 ppi is quite a lot .. btw Fuji with reference to the lanovia and max optical resolution speaks of 2500 ppi in case of halftone and 5300 if line art.

just my 2 ..


joerg

[Gordon Moat]

Mostly I agree with Lenny, despite that I do commercial work. My art endeavors are often low resolution (oil paintings and Polaroid manipulations). However, I have the benefit of being near the Museum of Photographic Arts most of the year, and I get to see many examples of fine art photography. The presence of MOPA also drives a small community of art photographers, largely under two groups representing many people both digitally and film based. The digital group is largely composed of people in their 50s or older, while some younger individuals make up the other group. So I get to see a ton of prints from all different types of gear, all year round.

The best prints I have ever seen were Edward Burtynsky's images. Many of those were from 4x5 colour negative, a few from 8x10, and some were not printed from an enlarger. These were RA-4 process prints, and not inkjet. I have never seen any inkjet prints come close. These have been (so far) the absolute best colour prints I have ever seen, though a few Andreas Gursky prints come very close.

In very large B/W prints, I saw some David Fokos (http://www.davidfokos.net/) prints that were quite amazing. I was surprised that they were C-Prints, and not silver gelatin. They were vastly better than the large prints in the Annie Leibowitz exhibit I saw at the San Diego Museum of Fine Art.

When you see the best out there, that draws comparison to every other printed image you will see later. I fail to be impressed by the quality of inkjet prints, though I have seen some amazing and compelling images printed that way. Mount inkjet prints in a sandwich under glass (or lucite/plexi), and it gets tougher to tell it apart from C-Prints, but not impossible.

One trend I have noticed is sharpening, and often to the point of over-sharpening. This makes sense when you consider that many doing these prints are in their 50s or older, and their eyesight is not as good as it was in the past. Sharpened prints will look better to older people, even if the resolution is not as good. This is more of a trend in post processing and printing, though unfortunately too often used to compare prints and images.

On a given day, with a given set of gear, one photographer can do better than another. The difference is less down to the gear, and more down to the ability to work within the limits of the gear. This is all that gear matters, in that is does not get in your way, nor hinder your creative vision. Give a hack the most expensive gear on the planet, and they will still be a hack.

We have far better gear choices now than the revered photographers of the past, yet we struggle to create more compelling image than they accomplished. Buying another piece of gear will not elevate any of us above their level.

Ciao!

Gordon Moat Photography (http://www.gordonmoat.com)

[bglick]

If soft image is the photographers intent, then these techno threads are senseless to those type of artist. Softness in an image is easy to achieve, specially within PS. The basis of these threads are often to examine how to maximize sharpness through gear and technique...this is when techno mumbo jumbo has a home and often where gear plays a VERY significant role to achieve such.


Interesting take on all the prints you have viewed, what a nice treat to see so much fine photographic art. As you know though, all this discussion was not about printing, but rather getting "to the point of printing", a HUGE distinction. As you point out, printing is the final expression of the image, and the offerings do vary tremendously today, vs. yesteryear.


> This is all that gear matters, in that is does not get in your way, nor hinder your creative vision.


I would disagree with generalizations like this. If someones goal is to make a tack sharp 70" print of a non static scene.... I would suggest a photog with a 8x10" view camera, or a high end MF back would have a much better chance of meeting their objectives, vs. if they had a mini digicam. The end product should drive the gear choice. I realize I exaggerated a situation to make a point... but... Since most hobbiest do not have the ability to own "all" the gear, it makes sense to analyze what type of final images you want to achieve, and work backwards from there.


But of course, more gear or better gear won't radically change the skill of the artist to capture a scene, this is a different animal. But with the advent of digital, and the sense of, "every capture is free", I have seen many elevate the quality of their output.... why? Because where they would once take a few hundred shots when they paid for film, processing and scanning, now they fire many thousands of shots, shoot longer, take more chances, etc. In many forms of photography, specially birding, sports, action, candids, the more "stuff" you throw against the wall, the better chance you have to increase the "keepers". Also, in FAST action photography, the high end DSLR's have superseeded 35mm film.... so the over all image quality I see now in this arena, has also elevated quite substantially. As the digital makers next improve lower noise higher ISO's (Nikon seems to have mastered ISO 1600 already), and next pump up the DR, we will see another huge jump in IQ...

[Lenny Eiger]

I fail to be impressed by the quality of inkjet prints, though I have seen some amazing and compelling images printed that way.

Inkjet printing has a very long tonal range in black and white, and rich, delicious colors ini color. It's a very capable medium. If you haven't seen truly impressive inkjet prints, you just haven't had the opportunity to see some from someone who can really print.

Of course, there are a lot of differing ideas about what makes a great print. Some favor a great, dark black, others don't care about black at all. Some like it shiny, some don't. Etc., etc.

As to shooting millions of pictures to get one good one.... (from a different post), I suppose there are lots of folks who like to do that and for them digital is fabulous. I think folks with large format cameras are the other type....

I do sincerely agree that one should look at what they are trying to accomplish, preferably educate themselves with a little History, and work back to the right tools for the intended goal.

Lenny

[sanking]

Over the years I have tested all of my scanners, and a few belonging to friends, with high resolution targets. I use the chrome on glass targtets that are capable of discrimination up to more than 225 lpm. The information is valuable and has taught me a lot about the capabilities (or lack of it) of some of the scanners tested.

The ability to resolve fine detail is an essential quality of every high quality scanner. But it is not the only important quality, as the ability to pull out information from high density areas, along with contrast, are also very important. A top quality scanner should be able to do all of these things.

Sandy King


hi,

there is an easy and inexpensive way for you to test our own scanner .. just buy a USAF 1951 test chart on film (will cost you about 50 euros) which goes up to group 7 .. did that with the lanovia .. max resolution for me is in the range of 50 to 55 lps/mm .. this means that you can with your nose close up still will have your 5 line pairs per mm .. the picture/print will be critically sharp at all viewing distances .. and 4x5 will give you a 40x50 print.... greater viewing distances do not even require those about 2500 ppi .. such a simple test will tell you what your specific equipment can deliver ... and actual 2500 ppi is quite a lot .. btw Fuji with reference to the lanovia and max optical resolution speaks of 2500 ppi in case of halftone and 5300 if line art.

just my 2 ..


joerg

[Gordon Moat]

. . . . .

"This is all that gear matters, in that is does not get in your way, nor hinder your creative vision."


I would disagree with generalizations like this. If someones goal is to make a tack sharp 70" print of a non static scene.... I would suggest a photog with a 8x10" view camera, or a high end MF back would have a much better chance of meeting their objectives, vs. if they had a mini digicam.

You disagree, but you exactly made my point. If your goal is to take pictures of dangerous animals at a distance, or a dangerous enemy, then a short lens would hinder your creative vision. I felt that point so obvious to not be worth stating, though perhaps I should have been more precise . . . after all this is the internet, and not an in-person discussion.




The end product should drive the gear choice. I realize I exaggerated a situation to make a point... but... Since most hobbiest do not have the ability to own "all" the gear, it makes sense to analyze what type of final images you want to achieve, and work backwards from there.

Isn't this obvious? Let me try another way to state this: if two photographers stand together in front of a scene, quite likely they will each have a different choice of gear, and different approach, despite that their results might appear to be very similar views of the same scene. Obviously the results could also differ greatly. If you want to take that to another extreme, give one photographer a D-SLR and a 600mm lens, and give the other photographer the same D-SLR and a 50mm lens . . . in such a comparison the results are likely to be different, yet despite the great cost of the 600mm there is no guarantee the resulting images will be any more compelling . . . regardless of subject and scene. Anyway, if I typed every situation, I could write a book . . . if all you want to do is take long distance pictures of girls at the beach, without them knowing, then a wide angle lens is going to hinder your creative vision . . . See how this could get?

I don't care what equipment you purchase, or how expensive it is, or is not; but you will not get any guarantee of improving your results. Buying gear is not the path to becoming a better photographer. Overcoming technical problems does not make you a better photographer. Anyone who thinks "if only I had such and such a lens/camera/computer/scanner/printer I would become a better photographer" is delusional and has drank too much of the Koolaid . . . and this illustrates a problem of the internet discussions in that you nearly need to hit someone over the head with a virtual piece of lumber in order to make a point . . . just kidding.
:D :cool:




But of course, more gear or better gear won't radically change the skill of the artist to capture a scene, this is a different animal.

Exactly! I suppose I didn't state this well enough, but hopefully what you stated right there is more likely to be understood than my generalization.



But with the advent of digital, and the sense of, "every capture is free", I have seen many elevate the quality of their output.... why?

Practice and editing. I have seen people do the same thing with film cameras. Of course these people ignore that shutters do wear out, and cameras do fail. Quite often many of them miss certain shots from not taking a breath and slowing down a little. High volume is not a path to greatness; only a different approach.




Because where they would once take a few hundred shots when they paid for film, processing and scanning, now they fire many thousands of shots, shoot longer, take more chances, etc.

I know some photographers who have a vastly greater frame count than many other photographers, yet their images still suck. Why is that?




In many forms of photography, specially birding, sports, action, candids, the more "stuff" you throw against the wall, the better chance you have to increase the "keepers". Also, in FAST action photography, the high end DSLR's have superseeded 35mm film.... so the over all image quality I see now in this arena, has also elevated quite substantially. As the digital makers next improve lower noise higher ISO's (Nikon seems to have mastered ISO 1600 already), and next pump up the DR, we will see another huge jump in IQ...

Ever look at the set-up at a major sporting event? Many of the cameras are remote controlled and simply triggered or fire off when action comes into focus . . . no photographer needed, other than clamping it down. The fans don't care, as long as they recognize the players. That boils down photography into a commodity. The functional difference in the work I do in photography is that I am controlling the image, and not waiting for it to happen.

I get this feeling that you are wowed by the technology. Nothing wrong with that, as long as one does not put too much faith into it. I have shot with much of the latest from time to time, and I have also worked on some future gear for a major manufacturer. Obviously none of this stuff is crap, but it is also not the next best thing to holes in Swiss cheese. The fact remains that many great images were made in the past on far worse (technically) gear than is available now. If you cannot pick up and use a camera somewhat like one of the past masters, and create images as compelling as their images, then the latest and greatest will never elevate your images past their results.

Ciao!

Gordon Moat Photography (http://www.gordonmoat.com)

[bglick]

> I get this feeling that you are wowed by the technology.

No, not really the case.... If anything, I might be wowed by the process of getting from point A to point B. The understanding of the fundamentals of imaging (optics, MTF, Nyquist, printing, etc) all play a role in the chain of meeting your final objectives. Of course, I have made a lot of 20 ft prints, which makes you plan real carefully.


When I shoot images that will be 11x14" or smaller, the freedom I have today amazes me.... anything is possible, no math required, no wow'd of technology, just focus on the scene layout, capture, technique, etc. In this case, everything is a slam dunk. Similar to making a trip to the grocery store 2 miles away, I grab my keys and go, no thinking, no planning, no strategy, no thoughts about cost, cause the technology is overkill for the application. But when I travel cross country in my motorhome.... ah yeah, now its all about thinking, planning, costing, check lists, etc. etc. Photography is a vast field, we all do different things, have different goals, etc.


I don't want to beat a dead horse regarding the "gear" issue.....

http://i.pbase.com/o2/25/583725/1/106242356.zeMh4EiE.beatingadeadhorse.gif


as I think we sort of agree.... but the right gear is sometimes critical to meet the final objectives...whereas you made it sound as if, give a great photographer junk gear, he can can turn chicken chit into chicken salad... I don't always agree with this.

[Sal Santamaura]

I'll admit to only having read every word up through around post 70; after that I scanned (pardon the pun) carefully. Despite Sandy's likely disappointment that this strays a bit from the topic...

My benchmark is an 8x10 black-and-white silver-gelatin contact print. Many observations lead me to accept Ctein's criterion of 30 lp/mm as necessary for "perfect sharpness" in a print. The frequently quoted 5 lp/mm falls far short of contact-print quality in my experience. Therefore, I'm sticking with a completely wet approach until:


A reasonably priced printer (that performs well in all other respects) becomes available with 1500 dpi outputAn affordable digital instantaneous capture device, easily transportable and usable in a variety of outdoor environments, is offered with a 186 Mp sensor.

As an amateur dilettante, I don't need to deal with the commercial factors many of you do. That's satisfying. :)

[bglick]

>Many observations lead me to accept Ctein's criterion of 30 lp/mm as necessary for "perfect sharpness" in a print.


Sal, Ctein's test, it must have been 10 years ago, really rocked me. You are the first person who ever referenced it! Of course, other than contact prints, there is no chance of getting this level of detail into a print. I applaud your high standards!!

[Sheldon N]

I was curious about the reference to Ctein's 30 lp/mm figure as necessary for ultimate print sharpness. I did some quick searching and found this recent article by him...

http://theonlinephotographer.typepad.com/the_online_photographer/2008/07/its-still-a-vis.html

He states that under normal print viewing conditions, the resolution limits of human vision are around 5 lp/mm. However, he states that we are far more adept at discerning acutance even if we can't see the resolution. He also states that "acutance trumps resolution and detail".

Given this, wouldn't the proper control of acutance be the determining factor in the appearance of a print, rather than going up to 30 lp/mm?

[sanking]

That study has come up before as I recall discussing it on a thread here a year or two ago.

It should be noted that 30 lp/mm is obtainable only with contact printing, with a negative that has at least that much information, and only on some silver gelatin papers. It would not be possible in contact printing with platinum and palladium because the texture of the art papers that are used with these processes limit resolution to around 10 lp/mm, or less.

It would be almost impossible to get 30 lp/mm on silver paper when printing by projection, unless printing same size as the negative or at least by no more than about 2X.

Sandy


>Many observations lead me to accept Ctein's criterion of 30 lp/mm as necessary for "perfect sharpness" in a print.


Sal, Ctein's test, it must have been 10 years ago, really rocked me. You are the first person who ever referenced it! Of course, other than contact prints, there is no chance of getting this level of detail into a print. I applaud your high standards!!

[Sal Santamaura]

...wouldn't the proper control of acutance be the determining factor in the appearance of a print, rather than going up to 30 lp/mm?Ctein's results indicated that 30 lp/mm was necessary to provide an appropriate level of acutance. My observations agree.

[bglick]

> Given this, wouldn't the proper control of acutance be the determining factor in the appearance of a print, rather than going up to 30 lp/mm?


Sheldon, in general, I fully agree. Of course, both would be nice, but we must be realistic. The largest factor of sharpness for humans vision seems to be extreme edge sharpness and contrast in the 3 - 5 lp/mm range. This is not in conflict with Cteins findings that human vision can discern resolution up to 30 lp/mm. They are two different findings. Sort of, two ways to skin a cat....

What would be interesting, a test with two images, one a contact print at 30 lp/mm, and a 5 lp/mm print with extreme edge sharpening, and see what the consensus is... My guess is, there would be no consensus, it would be hard to determine which is sharper. Before digital, only one option has been available to us...

I think the radial target comparison on the link provided demonstrates how powerful edge sharpening can be. Specially when you view from 10ft back. In addition to edge sharpening, digital has the ability to add contrast, which can sometimes be as powerful as edge sharpening. There is many who still believe you can never add resolution to the initial capture.... and maybe by the truest definition of resolution, this is true, but who cares about the terminology...in the end, digital manipulation can truly add a great deal to "perceived sharpness".....which IMO, is all that matters - assuming sharpness is the goal, pinhole photographers do not apply.

[bglick]

> Ctein's results indicated that 30 lp/mm was necessary to provide an appropriate level of acutance.


Sal, IIRC, Cteins test was done long before digital manipulation / printing was good enough to test edge sharpening and contrast vs. pure on-film-resolution (contact print). I never performed side by side scientific type comparisons, but it sure would be great experiment. My guess is, digital manipulation would be more powerful than most think.

[Oren Grad]

The article describing Ctein's test results is "Is Your Print Paper Sharp Enough", in the March-April 2002 issue of Photo Techniques magazine. He tested nine different B&W papers, as well as different grades for both graded and VC papers, and found that every one could record well above 30 lp/mm.

We've had this discussion before. Regardless of what techniques one employs in digital processing, the available printers - inkjet, laser on traditional papers, whatever - can't put anywhere near that kind of detail on paper.

A digital print at 5 lp/mm to which extreme edge sharpening has been applied will look like a digital print at 5 lp/mm to which extreme edge sharpening has been applied. It's akin to what some users of traditional B&W film like to do with development techniques that exaggerate edge-and-adjacency effects. The prints will certainly capture one's attention from ten feet back.

I'm with Sal on this.

[sanking]

Development procedures such as stand and other minimal agitation techniques that enhance adjacency effects are not as flexible as digital sharpening but they can still be very powerful. I have a 12X17 carbon transfer print that was made from a digital negative by contact printing. It is one of the sharpest photographs I have ever seen, for two reasons. One, the scene has some features that show high surface relief, which accentuates the impression of detail, and two, the film was developed with extreme minimal agitation to enhance adjacency effects. The sharpness is startling and people who have seen it comment that I must have used unsharp mask in preparing the digital file, but in fact I used no sharpening at all and the line effects are due entirely to surface relief and to the method of development.

If I did the numbers this print really does not have nearly as much detail as one would think. It was made with a 6X9 cm HP5+ negative, scanned at only 2540 spi, and magnified about 6X. This limits detail to a maximum of about 10 lp/mm, and the negative, which was printed on an Epson 3800, would reduce resolution even more. But when I look at this print from a distance of about 12-15 inches it looks sharper than any contact print I have ever made from a LF or ULF negative.

However, in general I agree with Sal and Oren in that silver prints made directly from LF and ULF cameras look a lot sharper to me than the best inkjet prints I have seen.

Sandy King





T
A digital print at 5 lp/mm to which extreme edge sharpening has been applied will look like a digital print at 5 lp/mm to which extreme edge sharpening has been applied. It's akin to what some users of traditional B&W film like to do with development techniques that exaggerate edge-and-adjacency effects. The prints will certainly capture one's attention from ten feet back.

[Lenny Eiger]

However, in general I agree with Sal and Oren in that silver prints made directly from LF and ULF cameras look a lot sharper to me than the best inkjet prints I have seen.
Sandy King

I'll have to go look and see. I'm not sure I would agree. There are a lot of factors. However, I will add that when I was printing primarily in platinum I used a very thin paper, a version of Bienfang Graphics without the OBA's. In the drying process, after washing, it shrank quite a bit, and made for some very sharp looking prints...

Lenny

[Tyler Boley]

Sandy, your eyes are not lying to you. The two processes have different capabilities. Silver (depending on source) has the capability to resolve more and also carry more micro levels of gray, even than the best current inkjet system.
Just toggle between the silver contact and K7 2880 on this page-

http://www.custom-digital.com/info/B&WPrintQuality/index.html

It's quite clear. Inkjet is still a halftone process, though complex and very high quality.
This is not meant to be a judgement, these differences show to the eye very differently under different circumstances, sometimes irrelevant.
Sometimes inkjet can look sharper simply because the dots themselves are sharp forcing a hard edge. But the fact remains that more actual photographic information from the source is getting to paper with silver than with inkjet... and I love inkjet for other reasons...
Complicated.
Tyler
http://www.custom-digital.com/

[sanking]

Tyler,

Great illustrations. Thanks for posting the link.

Sandy




Sandy, your eyes are not lying to you. The two processes have different capabilities. Silver (depending on source) has the capability to resolve more and also carry more micro levels of gray, even than the best current inkjet system.
Just toggle between the silver contact and K7 2880 on this page-

http://www.custom-digital.com/info/B&WPrintQuality/index.html

It's quite clear. Inkjet is still a halftone process, though complex and very high quality.
This is not meant to be a judgement, these differences show to the eye very differently under different circumstances, sometimes irrelevant.
Sometimes inkjet can look sharper simply because the dots themselves are sharp forcing a hard edge. But the fact remains that more actual photographic information from the source is getting to paper with silver than with inkjet... and I love inkjet for other reasons...
Complicated.
Tyler
http://www.custom-digital.com/

[Ken Lee]

Hollywood - which has deep pockets for technology - uses analog capture and digital correction.

I presume they feel that the combination currently gives us the best of both worlds.

I don't mind sacrificing a few line pairs per millimeter, or even a few levels of grey, to gain nuanced (http://www.kenleegallery.com/images/forum/img228C.jpg) corrections and adjustments.

[lawrence beck]

See issue nbr 79 (Nov-Dec 2008) of LensWork. Tests performed with the Epson 9880. Best results achieved with Harmon Glossy Fiber AL paper. Quoting Brooks Jensen from his article:

"Testing these four papers, there was a very clear winner: the Harman Glossy Fiber Base AL. It was not only the deepest, darkest density of all — a density that exceeded what I was ever able to get in the darkroom with gelatin silver fiber
papers. Looking up close, from across the room, under various light sources, it was simply better to my eye. I showed these papers to other people, visitors to LensWork, and members of our staff . Almost everyone agreed it was a marvelous paper, the best of the test subjects.

So, with the ability to use Harman papers in combination with Epson printers and K3 inks we can produce prints that, in my opinion, exceed the quality of the Lens Work Special Editions we used to produce in the wet darkroom. In fact, one of the tests we did was to print selected images from the former LensWork Special Editions program on Harman Glossy Fiber Base AL and compare it to the original gelatin silver prints that we made on Ilford Multigrade Fiber Base, selenium toned. In every single case, the Harmon print looked better than, or at least as good as, the toned gelatin silver version. I would never have predicted this. I’ve been a gelatin silver guy since my earliest days in photography. As a matter of fact, in the earlier edition of the LensWork Special Editions we made a big deal out of the fact that we were not producing inkjet prints. We advertised that and used that specific language; no inkjet prints.

Well, times have changed. The new papers and the new inks produce work that I could never produce in the darkroom, with a look and a feel that is absolutely spectacular. I am unapologetic about this. I have no qualms whatsoever in offering these Epson K3 pigment-on-paper prints on Harman paper and claiming with confidence and assurance that they are every bit as good, if not better, than the original LensWork Special Editions gelatin silver prints. Back then, we claimed our Special Editions were visually indistinguishable from the photographer's original work. I stand by that claim."

Maybe it's time to reexamine this "wet darkroom vs inkjet" debate, taking into consideration the current state of the art inkjet technology...

[Tyler Boley]

I certainly hope it doesn't appear that I am interested in, or contributing to, that debate. I don't think it's very helpful. My interest is that during this complete, sudden, and in some sense unrequested overhaul of our process, the standards set by our history remain at the highest level. I wound up looking hard at this stuff, and making information available, not judging anyone's choices. It directly speaks to the thread, which as should be obvious by now, has no clear answers.
Tyler

[lawrence beck]

I'm confused, Tyler.
You state in your latest post "I certainly hope it doesn't appear that I am interested in, or contributing to, that debate. I don't think it's very helpful."

Yet five posts earlier you provide the following link: http://www.custom-digital.com/info/B...ity/index.html to a test that you illustrate on your personal website. You state, reaching a conclusion from that test "It’s clear to me that we have not yet met the level of technical photographic quality our previous methods achieved. Does this matter? Can you see these differences with normal viewing? I think you can, given certain images, and viewing conditions, and eyesight. But I care less about that than keeping the bar high, often we can’t put our finger on a technical reason one print is “better” then another, but adherence to the highest standard at each step of the process clearly yields a higher percentage of stunning prints, every little tweak contributes."

How is there "adherence to the highest standard at each step" when, as you later state under the Technical Notes of the test: "All ink prints were done at 100% full 4000ppi on Hahnemuhle PhotoRag 308 and the silver prints were 1:1 contact, so all were the same size. The Epson ABW prints were made with a 9800, the quad prints with a 9600, and the K7 prints with a 7800."?

You're comparing a silver contact print to five year old technology in the Epson 9800... not to mention even older technology in the 9600, a paper (Hahnemuhle PhotoRag 308) that is a far cry from the best inkjet printing can achieve (in terms of resolution). Hardly "adherence to the highest standard at each step". How can you even make such a comparison?

It appears to my uninitiated eye that this is a rather strong contribution to that very debate that you don't find very helpful.

What am I missing here?

[dwhistance]

Lawrence

No doubt Tyler will respond to your post, however I would like to comment on your statement about the technology - "five year old technology in the Epson 9800... not to mention even older technology in the 9600, a paper (Hahnemuhle PhotoRag 308) that is a far cry from the best inkjet printing can achieve (in terms of resolution). Hardly "adherence to the highest standard at each step". "

Whilst the Epson 9800 has indeed been available for 5 years I think you will find that there is very little difference between it and the "latest" 9880 (or even 11880 or 9900), particularly in terms of B&W output.

Similarly although the 9600 is an even older printer the underlying technology is very similar to the later printers (they added extra inks and ways to stop you using third party products and very little else!). Also in Tyler's case he drives it with a well set up, state-of-the-art, RIP and uses a dual quad set-up. Whilst this is probably no longer state-of-the-art for B&W inkjet printing (I think that is represented by the K7 inks at the moment although Lenny may disagree in favour of his own) it is nevertheless very close to state-of-the-art.

I am inclined to agree with your final comment regarding Photorag - there are papers with higher resolution available now, even matte ones. However, in my opinion, Photorag does have sufficient resolution for a test of this kind (as I think is clear from Tylers scans) and is a well known product which is presumably why Tyler used it.

David Whistance

[lawrence beck]

Thanks for your comment David.
I should have beter qualified my statement by indicating that I was speaking of the 11880 model which has some noticeable improvements and changes to the dither pattern which result in a noticeable improvement in output.

The single most significant improvement would be noticed with a state of the art glossy paper as opposed to a fine art substrate that exhibits far more dot gain. The fact that LensWork has admitted better quality with the 9880 and Harmon Gloss Fiber Base AL makes me wonder how much better the prints would have looked if printed with the 11880.

[Tyler Boley]

Lawrence, your post deserves a serious reply, sorry for the delay... family stuff. More soon, honest. I feel like my enthusiasm for these issues has hijacked the thread with stuff of little interest, and the answers about almost every point you make are complex and lengthy. Anyway, more as soon as possible.
Tyler

[sanking]

What I understood is that Tyler was simply pointing out the different capabilities of printing systems, not trying to debate the merits of one or the other. I understand that there are people who want to push the concept of superiority of one printing system over another, but my impression is that the majority of photographers are reasonable folks who recognize that we are dealing with different media.

It is entirely possible, as you suggest, that the newer line of printers, using smooth papers, will be capable of greater resolution than the previous generation. However, even if that is a fact I doubt very much that the improvement will be enough to bridge the rather huge different that now exists between an inkjet print, on whatever surface, and a silver gelatin print made by contact printing as the difference is on the order of 10 lp/mm versus 40 lp/mm.

However, as others have pointed out, resolution in itself is not the determining factor in print quality, or for that matter, even in sharpness. I previously made the point that one of my carbon transfer prints, made with an inkjet digital negative, is one of the sharpest photographs I have ever seen, yet it has the same limitations in terms of detail as an inkjet print.

I personally never liked silver gelatin prints all that much, which is the major reason I took up alternative printing. Inkjet prints on fine art papers are much more attractive to me than inkjet prints on baryta. So if it were up to me I would say, let inkjet me the best media it can be, not try to imitate the glossy look of silver gelatin. Pt/Pd prints also are very limited in terms of detail compared to silver gelatin prints but few complain that they are not sharp enough as there are other qualities that speak to us.


Sandy King






It appears to my uninitiated eye that this is a rather strong contribution to that very debate that you don't find very helpful.

What am I missing here?

[Lenny Eiger]

The single most significant improvement would be noticed with a state of the art glossy paper as opposed to a fine art substrate that exhibits far more dot gain. The fact that LensWork has admitted better quality with the 9880 and Harmon Gloss Fiber Base AL makes me wonder how much better the prints would have looked if printed with the 11880.

I suppose that I ought to respond since I was mentioned. I cancelled, or let die, my subscription to Lenswork last year. I did this for a number of reasons. First of all there was an article that suggested we all sell our work for $20 a print. Those of us that sell our work often work quite hard at it for many years and I found it simply offensive. Second, he put out another article suggested that his printing technology had gotten to the point where it was the best thing going, better than darkroom, better than inkjet. (We should all give up our inkjets and get presses.) Third, for the longest time he resisted showing photographers whose printing was done on inkjet, suggesting it was lesser somehow. Finally, when I looked at this printing style - I didn't like it. It's very dark, gloomy and depressing. There's a style that gets applied to every photographer's work. On occasion there's some very interesting photography. But all in all, I didn't see it as something that fed me, so I let it go. Given his disdain for the new medium over the years and his predilection for a very specific type of printing, I question whether he is the right person for the "ultimate test."

I have just printed some images on a PhotoRag Baryta paper for someone. Very difficult print, lots of masking and twiddling with the colors to get them right. I did another print right after on PhotoRag. I like the PhotoRag print much better. The matte surface has many advantages over a shiny one. You can't knock PhotoRag and suggest that Harmon's whatever is "better." They are very different mediums. They are as different as black and white is from color. I have done as much profiling of PhotoRag as anyone - probably about 40-50 different profiles of this same paper over the last few years. It is a superb paper. It has exceptional richness. I just went and looked at my profile test print. I have just made a new profile for PhotoRag Baryta. My PhotoRag print of the exact same image (actually set of images) is at least as sharp as the Baryta coated paper. In some cases it looks sharper, but this is likely a minor difference in contrast. The matte surfaces allow for more richness and deeper tonality.

Many of the top papers perform excellently. Hahnemuhle's line is exception, as is Crane's. There are others that I haven't played with as much. A lot of the excellence of these papers (and any paper) comes with working with them over the years, just like it was in the darkroom, just like it is with film.

I would also suggest that great prints are made by great printers, as in humans, not machines. The machines are just there as support. I think a focus on sharpness is misplaced, as is the focus I often hear, of dmax. I think a rich print is made with the midtones, and how they interact with each other. It is also about the entire balance of tones, and not one side of the spectrum or the other.

Of course, these are my opinions, not intended to be a statement of fact. Everyone is welcome to disagree.


Lenny

[sanking]

With the exception of the comments about LensWork, to which I still subscribe and find very interesting, I am going to take the unusual stance and express my total agreement with Lenny about matte papers for inkjet prints and the misplaced focus on sharpness and Dmax. I think he is exactly right in that a rich print is made in the mid-tones, and sometimes with just a bit of highlight or shadow detail.

Sandy King




You can't knock PhotoRag and suggest that Harmon's whatever is "better." They are very different mediums. They are as different as black and white is from color. I have done as much profiling of PhotoRag as anyone - probably about 40-50 different profiles of this same paper over the last few years. It is a superb paper. It has exceptional richness. I just went and looked at my profile test print. I have just made a new profile for PhotoRag Baryta. My PhotoRag print of the exact same image (actually set of images) is at least as sharp as the Baryta coated paper. In some cases it looks sharper, but this is likely a minor difference in contrast. The matte surfaces allow for more richness and deeper tonality.

I would also suggest that great prints are made by great printers, as in humans, not machines. The machines are just there as support. I think a focus on sharpness is misplaced, as is the focus I often hear, of dmax. I think a rich print is made with the midtones, and how they interact with each other. It is also about the entire balance of tones, and not one side of the spectrum or the other.

Of course, these are my opinions, not intended to be a statement of fact. Everyone is welcome to disagree.


Lenny

[Gordon Moat]

I also find myself agreeing with Lenny and Sandy on this, in that there is an over-emphasis on sharpness and dmax values. Unfortunately there are many involved/invested in newer technologies that feel a need to quantify differences.

On a different note, the absolute best non-silver B/W printed images I have ever seen came off an unusual Heidelberg two colour press, that was about the size of a refrigerator. I still have one of the samples, a duotone of Hagia Sophia. Unfortunately Heidelberg sold very few of these prior to discontinuing that press.

There is also a move to solvent based inks, eco-solvent, and UV cured inks. In Epson's commercial line, and in other companies, there are some newer technologies that potentially show another direction, and better control. Eventually some of that might trickle down to smaller desktop inkjet systems, either from Epson, HP, or another company.

The main thing is that printing technologies that are not continuous tone will only approach emulating continuous tone. I prefer true continuous tone, despite that I have seen nice images from dot techniques.

Ciao!

Gordon Moat Photography (http://www.gordonmoat.com)

[Tyler Boley]

I'm afraid my writing is less than clear. How something appears, and how it factually is, are two different issues. As Sandy and Lenny and others have noted, a print can have the distinct impression of sharpness, and turn out to be not so much technically. But for the vast majority of artists, how something appears is surely all that matters after all. Very few, including my friend Brooks, have any need to peer at high resolution scans of dots of ink like I do.
I've tried very hard not to state any judgments on how things appear, even though my preferences may show through. But when it comes to a very specific technical criteria, say file (image) detail written to paper, these things can be judged and compared objectively. Photography has certain unique qualities, those are what I chose to look at, continuous tone and optical clarity.
Now let me make it clear- for a lot of complicated reasons, one print made on one system may give the impression of more sharpness and continuous tone and thus be "better", than another, and that's fine. However, technical scrutiny may show the opposite, but that does not make it any less "better". I'm digging a dangerous hole here, anyway, not "meeting the standards" has to do with scrutiny of very specific and defined issues, not impression or impact.

Regarding your comments about the state of hardware technology, I disagree, and the problem is I have no high res scans to back up my statements so you can take them for what they are worth.

The 9600 on up through the newest 9900 all have the same dot size. The improvements come from how dots can be manipulated with screening and dither pattern etc.. Now we're talking B&W here, since I posted in reply to Sandy's comment about silver print resolution. Though 9900 ABW output may be an improvement over 9800 ABW output, it still can not come up to the performance of even the variable dot 9600 quad output with regard to the SPECIFIC criteria in the tests. Additionally, the 2880 K7 7800 output shown can't be bettered with any large format B&W output to date, again with regard strictly to the same criteria. It can be bettered, believe it or not, with the newest desktop small dot models like the 1400 with a K6 setup, and the 1900 with a K7 setup.
Again I must stress, this does not mean that by your, or even my standards, it's not possible on a given day with a given file, well made by a great printer (as Lenny well said) that a system that did not test as well can not make a "better" print than from one of these high performing systems. I have no doubt Brooks is making wonderful prints, he knows what a great print is, or isn't, better than most.

I'm sorry this is getting long, but the next part is not quick and easy. A paper is "sharper" than another based on how precisely it can hold a dot. With regard to matte, only two papers I have tried hold a sharper dot the HPR, and that would be Museo Portfolio Rag, and Premier Fine Art. However, neither can take the ink load HPR can, so it may be a trade off. If I have to limit each light ink to avoid mottle and bleed, and retain linearity, then I can't let light inks at very high dot frequency describe file detail, I must transition to wider spaced dots of the next darker ink sooner.
I'd say higher density of light inks with slightly softer dots, vs harder dots of slightly darker inks, is not a huge deal and the tests would have been similar. Limiting each of these individual systems to suit Portfolio Rag's needs would have thrown another variable into the mix. A good batch of HPR takes ink from all these systems as they are intended without special needs, therefore revealing the system. I do like Portfolio Rag though, for some work.

Now the "photo" papers like Harman. These papers perform well with ABW, which uses wider spaced dots of darker inks than the K4/6/7 systems. I have done a lot of testing on those papers with custom B&W inks systems trying to maximize a gloss monochrome process. The experimental combo setup I have in the 7800 right now is nuts to say the least. These papers can not take high levels of light inks, at all, even for color. In order to keep dot definition precise without mottle/bleed, light inks must be limited quite a bit, if a GO is also present, that's more fluid and more limiting has to occur. I've had these papers come out of the printer with ink literally running down off them. So again, we are back to having to introduce darker inks sooner, so each ink is nowhere near 100% coverage. Some on this list may have gotten some of my test 2880 K7 gloss samples, on Ilford GFS and Innova USG (Epson Exhibition Fiber). I assure you, those would not have performed as well as HPR did in the tests due to the required over limiting. They have to give up dot positions that could have been used for information, to spacing for accurate "lightness". That's not to say they were not pretty, and sharp to the eye, and nicely continuous tone.
ABW on those papers, including the Harman, would have resolved no more detail than HPR because of the way ABW works. The resolution boast on the Piezography site is no lie, and the ABW part of that test could have been on glossy and showed much the same result. It's because of the entire system, not an individual element's performance, like the paper. Of course though, the system must be idealized for each element. So a K7 system on HPR will look very different than one optimized for Harman. These are not simple black or white issues.

Again, a well done print from the right file on Harman may give the impression of the sharpest inkjet print we have ever seen, that's not what I'm talking about. These impressions are the most important thing to people here, I realize that. In fact they are the most important to me too. My personal printmaking is all about a personal response to the print object, this other stuff goes back in the drawer then. I created a custom ink setup for Arches Cold Press watercolor paper for a special project. It has crap resolution, crap gamut and dmax, but is gorgeous for these particular images. I would call them good prints, though they would look horrible on those tests.

I'm sorry for the length, these are not simple issues or ones that should concern many, but I did not want to blow off the questions. Some of this gets down to novel length discussions of how multiple density inks in series can be used to increase information description from the source file, ink printing process in general, etc etc. For anyone who has little concern for this stuff I REALLY hope you didn't read this far...
Man, this sounds like a lecture. OK, fire away...
Tyler
http://www.custom-digital.com/

[Doug Dolde]

It's 39 megapixels. I get 320 off of a 4x5 piece of film, and 568 megapixels off of 8x10. There's no comparison. The tests are flawed.


One word: BULLSHIT ! Grain is NOT good data !

[Stephen Best]

Whilst the Epson 9800 has indeed been available for 5 years I think you will find that there is very little difference between it and the "latest" 9880 (or even 11880 or 9900), particularly in terms of B&W output.

I have a 4800 and 7900, and before that a 7800. There's just no comparison between the 7900 B&W output and that from the earlier x800 generation. The differences stem from a number of factors: improved dithering (introduced with the 3800 and x880 series and further refined for the x900), more precise dot shape & placement (introduced with the 11880), tonal colour mixing and improved linearization (introduced with the x900 series). I don't pretend that the prints look like traditional media or those from K7 etc inksets, but it would be wrong to assume that Epson's B&W output has stood still over the years.

You can read about some of the changes here (these are real, not just marketing):

http://www.epson.com/cgi-bin/Store/jsp/ProImaging/EpsonInnovations.do?invMoreInfo=EpsonInvAccuPhotoHDRScreening&BV
http://www.epson.com/cgi-bin/Store/jsp/ProImaging/EpsonInnovations.do?invMoreInfo=EpsonInvMicroPiezoPrintHead&BV