HDR High Dynamic Range Examples

[Frank Petronio]

Here's something of interest that I'm swiping third-hand:


A Flickr photo-pool devoted to a Photoshop plugin has accumulated hundreds of beautiful pictures. The "High Dynamic Range" plugin merges several shots of the same scene taken at different exposures, producing pictures that have breathtaking, vivid, hyperreal colors. The HDR pool on Flickr is filled with astonishing examples of the form, and active discussions of the best way to capture great HDR images. link (http://www.flickr.com/photos/bmaas/98086361/in/pool-hdr/)

I dunno. Maybe if it were better executed instead of just showing off what's possible. Does anybody have any links to some good HDR images?

[Donald Qualls]

I don't know that tool, but I've seen some examples where the technique it represents overcomes one of the basic shortcomings of digital -- dynamic range more like older slide film than modern negative stock. I don't know that it'd be that much help for most of what we do, though I can see that something similar could be used to extract a broader range from a B&W negative than my scanner will do with one compromise setting.

[Roger Hein]

To me many of the examples with 'sky' in them look like those badly shot commercials you see on tv or in magazine ads where the conditions weren't photographically 'ideal' but the photographer was under the gun by the art director to shoot anyways. Digital effects, like anything, can be taken too far.

[Walt Calahan]

You don't need to buy a plug-in to do what's built into PhotoShop already.

Blending Mode always image makers to extend the tonal range of all images. A very powerful tool.

Works very well when scanning too. It's not just for digital anymore (HA!).

Take a color neg that is perfect, scan it for the shadows, then scan it for the mid-tones, then scan it for the highlights, and then merge all three images together using the best information from each scan to make a complete photograph. With this technique a photographer can overcome a lot of lost information when translating the negative to an ink-jet print.

[Frank Petronio]

I know, I've done it with architectural interiors for years. But the idea behind HDR is to make it easier and more widespread. I guess what is making me uncomfortable about the image is seeing "too much."

How many stops does an adjusted (not dialated) human eye see? I've always though we could see much more than the ten stop Zone System range, but now I am starting to wonder if we simply ignore a lot of the excess info and only see into the shadows and highlights selectively, when we make a subconcious choice...? Seeing "14-stops" really bothers me.

For that matter, how many people really see the difference between Zone 0 and 1, and 9 and 10? Much less Zone -2 and Zone 12?

[Jim Rhoades]

OK, nice postcard. How come it's a blue sky sunny day, with no directional light? What planet was that taken on?

[David Luttmann]

Donald,

I'm not sure where your dynamic range information comes from. Digital sensors do not have the dynamic range of old slide film. That is a saying that is thrown around simply because highlights can be blown out with the same effect as slide film. In fact, modern day DSLRs have a dynamic range of between 9.5 & 10 stops. This is a pretty close match to color neg film. The idea behind HDR shots is to provide a range of between 14-15 stops. This can't be used for every target as it may not appear natural to the eye.

Walter, HDR is NOT designed for film use from a single color negative for example. Scanning for shadows, midtones & highlights does not give the same effect from a single negative due to the limited dynamic range of the negative....approx 10 stops. Running multiple scans still leaves you with 10 stops, whether or not you are using HDR in CS2. The idea behind HDR is to make multiple exposures (5) with digital RAW files that are underexposed by 2 to 3 stops and over exposed by 2 to 3 stops.....with shots in the middle range as well. This gives you between 14 & 15 stops.

This old idea of saying a digital sensor has a 5 to 7 stop dynamic range like slide film comes from people who don't know how to expose a scence to maintain highlights. This does not hold up under tests though.

[Michael Kadillak]

You hit the nail on the head Jim.

It is possible to extract nominal portions of tonalities beyond what is considered to be a conventional "range" with conventional materials. See Barnbaums book The Art Of Photography.

However, if you were to stand at this precise point and experience the situation for yourself it would still be beautiful. But it would not be bathed in the surreal.

This technology is great if you want to sell vacations or cars, but that is where the buck stops for me. Present this as art and that is when I am heading for the door....

Cheers!

[David Luttmann]

Jim,

You can see the direction of the light in the rock straight ahead. The sun is towards the right. The lack of directional light on the person is probably because they are in the shadow of the trees and a cliff. However, I'll agree that this shot doesn't look that natural....mainly because we aren't used to seeing the compression of such a large dynamic range.

[Kirk Gittings]

"Works very well when scanning too. It's not just for digital anymore (HA!)."

This technique with color negatives does help some with a long dynamic range scene but it is hardly a panacaea. You are still stuck with the dynamic range of a color negative range as Dave says. Also flatbed scanners are generally not accurate enough to get two scans identical in size to precisely overlay, because of temperature differentials, slop in the gearing etc. Silverfast is attempting in its next release to upgrade its registration feature to help more with this.

[Jack Flesher]

FWIW, I was a beta tester for CS2 and the HDR Merge tool got a lot of discussion. Bottom line is it is a tool in the works for Adobe; they simply added it in its current form as a basic functionality in CS2 because they had what you see already developed.

As I understood it, they did this for two reasons, 1) it is a need in certain forms of digital video editing and 2) since they had the basics available they felt it might also whet the appetites of traditional digital imagists -- ostensibly a more functional version of HDR will give them a reason to upgrade to CS3...

Cheers,

[Jack Flesher]

Oh, and to answer Frank's original question: though I have seen impressive representations of incredible DR, I have never seen what I'd term a "really good" image made using the tool.

All FWIW,

[tim atherton]

"This noise
is there due to a lack of bit depth in the shadows hence shooting to the
right on a histogram"

I trust you aren't using the in camera histogram as an accuarate representation of where to place the highlights and shadows when actually shooting a scene?

[David Luttmann]

Matthew,

I'm afraid you are the one who isn't informed. I've already pointed out tests that prove that the dynamic range of DSLRs today attain approx 10 stops....which is equal to that of color negative film. If you can't obtain those 10 stops, it has to do with your processing technique and not limitations of the sensor. The Imatest software package is quite accurate in this respect and is used extensively in bench tests between DSLR & different films. In fact, it's the industry standard in this regard.

I have already produced dynamic range tests using Imatest to show the dynamic range I said a DSLR a capable of. Color neg film tests in between 9 & 10 stops just like a DSLR. I've tested neg film, slide film, and DSLRs. You on the other hand, have produced nothing other than an opinion. Until I see some comparison tests from you, I would appreciate you not blathering on about something you apparently know nothing about.

The fact that the toe extends highlight detail somewhat does not equate to extra stops of dynamic range. You apparently need to learn a little bit about how this range is measured. Produce your test results confirming that color neg film has a greater dynamic range than some modern DSLRs like the 20D or 5D from Canon, or keep to yourself!

[Jack Flesher]

Matthew:

Dave is 100% correct on this point.

Modern DSLR's do regularly achieve between 9 and 10 stops of DR, and some of the higher-end models approach 11. BUT! They have to be shot in their raw mode to deliver that range.

The biggest disservice -- really an obfuscation -- to the DSLR-DR subject was a recent online test done by one of the prominent websites... Unfortunately they chose to do their comparison by shooting their selection of DSLR's in JPEG mode and avoid any RAW file processing! Of course JPEG mode seriously compresses all images, and most DSLRs generate less than 7 stops from their in-camera JPEGs...

[David Luttmann]

First off, I didn't "pluck" from other sites. I don't put down film all the time....I just defend misinformation from people like you who have never done any testing and just parrot back the "film is better because I say it is" nonsense without ever properly comparing. You'll notice that others here are pointing this out to you as well.

My tests were done with the following:

Fuji Reala 100, Fuji NPH 400, Fuji Velvia 50, Fuji Astia 100, Fuji Provia 400 & Kodak Portra160. (Yes, I like Fuji film)

All scans were done on my following scanners:

Imacon 343 in 16 bit, and Minolta Scan Dual 4 in 16 bit , 16x multisampling mode.

I would say that these two scanners are representative of what most of the public can afford (the Minolta) and what most working pros can afford (the Imacon). I'm not concerned with $200 scans from drum scanners as I highly doubt you would have all your work drum scanned. And if you could afford the drum scanner to get an extra stop.....calling it 11 stops.....then you could afford a 16 bit digital back with 11-12 stops of dynamic range.

Really, this topic is a non-issue. The dynamic range measured in stops has been tested to death on various sites and by numerous pros....including myself. Even if there is a 1/2 stop or 1 stop difference between the two, be it in digital's or film's favor, that is not enough to have a real meaningful difference in the quality of photographic imaging. It is really only fodder for measurebators who have nothing better to do than try and put down digital capture.....and will grasp at every straw to do so.

[David Luttmann]

Oh, and by the way....the drum scans I have had done show little better than an extra 1/2 stop.

[David Luttmann]

Matthew,

I'm not going to redo all these tests to prove something to you that others have already acknowledged. This is old news. I did these tests quite a while back and see no reason to do them again. That & the fact that I can't find where I saved the Imatest results.

My suggestion is to go & purchase Imatest and try these tests yourself to see how they work in your workflow. Once you find that the results echo what has already been said, then you too can move on. I have no problem with film as a medium....I still use it a fair bit because I like to.....but I also understand its strengths & limitations. Dynamic range is no longer one of the strengths.

Feel free to share your test samples. Until then.....

[Jack Flesher]

Matthew:

Okay, if I understand you, what you are saying is that a direct digital file does not look at its base like scanned C41. If that is it, no argument here.

And I don't think Dave is bashing film either...

The bottom line is this: IF we have similar DR to begin with, and IF enough color information in the original digital file (say anything greater than 10 bits per channel) then we can almost certainly emulate (though not necessarily duplicate) the "look" (in digital parlance we call it a profile) of the C41 emulsion. This is fact and not conjecture.

Cheers,

[bglick]

Half all the tonal range of a digital SLR is in the top stop, half the remanining tonal ranges are in the "next to top" stop. So, 2 stops uses up 75% of all the available tonal range, 3 stops, we used up 87.5% of the total tonal range available. Now, regardless whether its 12 bit, or 14 bit capture, after maybe 4 stops of digital capture, there is so little tonal range remaining, I wonder how valuable the remaining stops are. I think this point is being overlooked, and its one area neg film still is superior as the tonal ranges remain higher throughout the same image exposure lattitude. How much? I don't know, but I would love to see this analzyed by someone :-)

So the ability to see detail / texture in more then 4 or 5 stops is very possible with a RAW file, as its remarkable how much hidden detail exist, but as for tonal range, well, its just not there. For some images with solid colors in the underexposed stops, this is not an issue, but where there is larger tonal range needed to be seen, well, it often results in everything from posterization to muddy appearance. How bad its displayed is also at the mercy of digital capture, such as if the color is red or blue, the sensor has only half the pixels to record with, since Bayer records in RGGB mode.

I feel this is still one major shortcoming of digital capture. It's my guess, its this shortcoming that Canon and other makers are focussing on in the R&D labs, not more MP. I am curious if the new 39 MP chips are recording tonal range in a fashion different then what digital SLR's record in, as I just described. Scanning backs record much better in this regard, but I am not sure of the method used.

As one poster mentioned, when you can control lighting, such as in studio, digital capture is remarkable...but one area it seems film still shines, specially negative film is recording large amounts of tonal ranges throughout large exposure lattitude. Of course, negative film has much less resolution, so often you have to go up in format size to maintain the sharpness you had in the next smaller format when using chrome films.

So, I think there is some confusion here regarding recording texture / detail, which RAW files come close to that of negative film, vs. recording the same amount of tonal range in each of the stops where texture is recorded. This inheritent weakness in digital capture is what HDR is all about, but as someone mentioned, its a very immature feature of PS CS2, but the premise is good. My guess is, Adoble won't develop it if they feel the weakness will be overcome by hardware technology in the sensor / electronics. I guess we will have to wait till CS3 to find out their positon.

[Emre Yildirim]

This discussion has been really interesting so far.

But serously, do either of you really care that much about which one has more dynamic range?

This argument has turned personal, as much as either of you like to deny it.

All I can say is that digital has gotten much better in this area, and I don't think anyone can deny this.

[David Luttmann]

Sorry Matthew,

Convenient? I have shown you the digital test before. I believe with the Canon 20D. You harped on and on about how a digital source couldn't have 10 stops of dynamic range. I believe I had shown you about 9.9 stops. Now you want me to dig up old Imatest results?

Sorry, I don't have time to waste for people like you. Reason being, when I showed you the results from a DSLR, you denied them to be accurate....in other words, no matter what test results I show you, you will call them to question unless they agree with your bias towards film.

As to the extra 1/2 to 1 stop for neg film from a drum scan....I did say you could go to a digital back to get between 11-12 stops.....but I'm sure you call that into question as well.

I think I'll leave it at that. Go do your own tests and quit trying to find a conspiracy with everyone who provides you with tests that don't agree with your opinion.

[Ken Lee]

You can find a nice discussion of the technique, with sample images and Photoshop info, on Luminous Landscape (http://www.luminous-landscape.com/tutorials/hdr.shtml" target="_blank).



Like so many techniques, it's good to know in case you need it.



Some day soon, they might come up with sensors that accomodate a wider range of brightness - or incorporate some kind of fast multi-sampling. In electronics, a fraction of a second can be a rather long time, and it wouldn't surprise me if it were possible for a camera to actually shoot several images in quick succesion, and overlay them instantly, when the contrast level calls for it.



As for me, TMY + Pyrocat HD + Ansco 130 gives a fairly long range (http://www.kenleegallery.com/images/gallery/hadleymtn236.jpg" target="_blank)already.

[Jack Flesher]

>>Half all the tonal range of a digital SLR is in the top stop, half the remanining tonal ranges are in the "next to top" stop. So, 2 stops uses up 75% of all the available tonal range, 3 stops, we used up 87.5% of the total tonal range available. Now, regardless whether its 12 bit, or 14 bit capture, after maybe 4 stops of digital capture, there is so little tonal range remaining, <<

WG: Where are you getting this -- Ahemmm -- information?

[Frank Petronio]

All I know for sure is that some of you guys are full of it. The problem is I don't know which of you guys are! (But I have some ideas...)

Back to the origianl question - anybody have any web links to HDR - Zone 14 images?

[bglick]

Jack, this "information" is readily available in any RAW tutorial....most notably in Bruce Fraziers book called Camera Raw, it explains RAW files, the digital capture process (Bayer sensors), and Adobes Camera Raw software....it's a good read... and based on your post, I think you would strongly enjoy it. :-)

It seems this little gem of information has somehow eluded most digital fans...something I am sure the makers of Bayer sensors surely do not want to shed light on. They have done an excellent job keeping it under wraps. AFter you read up on it, I am curious as to your comments, as it was s tough pill for me to swallow a few years ago when I first read about it.... so be careful, and absorb it in - small doses! :-)

I explained this in detail in posts awhile back, but don't have the time to dig it up... and yes, even today, I am still bothered by the fact there is 3 stops of data that contain 87% of all the tonal ranges available, that leaves the lower stops, say 5 or 6 which can not record an 8 color step wedge, or even worse for R and B, probably not even a 4 step color wedge....

[Jack Flesher]

>> Jack, this "information" is readily available in any RAW tutorial.... <<

I thought you might say something like that...

So now please do us all a favor and explain the differences between the linear RAW file and the converted RAW file as it relates to DR -- after all, they are significant.

[Jack Flesher]

>> I knew from experience that I needed far more than 1 digital exposures to get a similar feel to what I was achieving conventionally in 1 shot c41 capture. <<

And because of comments like this, all Dave and I are suggesting is that perhaps some film folks have not yet learned how to process digital files properly to achieve maximum USABLE DR...

The big difference in DR is NOT the total DR available, but how the DR that is there gets distributed. The "automatic" settings in most raw converters do not do the total digital file justice. Hence, one must tweak the converter to achieve the desired result -- and that result can get very close to replicating C41 emulsions in latitude, color and tonal distribution. Heck, we can even replicate the grain once we pull the file into Photoshop ;)

Cheers,

[tim atherton]

"Half all the tonal range of a digital SLR is in the top stop, half the remanining tonal ranges are in the "next to top" stop. So, 2 stops uses up 75% of all the available tonal range, 3 stops, we used up 87.5% of the total tonal range available. Now, regardless whether its 12 bit, or 14 bit capture, after maybe 4 stops of digital capture, there is so little tonal range remaining, I wonder how valuable the remaining stops are. I think this point is being overlooked, and its one area neg film still is superior as the tonal ranges remain higher throughout the same image exposure lattitude. How much? I don't know, but I would love to see this analzyed by someone :-) "

this is the sam for a raw scan from colour negative is it not? Do not the linear files from both follow this?

"It seems this little gem of information has somehow eluded most digital fans...something I am sure the makers of Bayer sensors surely do not want to shed light on. They have done an excellent job keeping it under wraps. "

Conspiracy theory...! - hmm not exactly a big secret. We were discussing this 10 or 12 years ago at least

[Frank Petronio]

I seem to get good images from both color neg and my raw DSLR files, darn it. What am I doing wrong?

I do know that if I need to work with a 2-stop underexposed image, I much rather have a underexposed RAW file than a 2-stop under color neg.

Since just about every form of repro involves compressing the tonal range of our images, even from color chromes, at best we're going to see a little more shadow and highlight detail from a longer toe color neg, raw digital file, or 14-bit HDR. What bothers me about the HDR images I've seen is that the jpg (8-bit) on screen I see seem to place Zone II shadow detail at Zone IV (and Zone VIII highlights at Zone VI). Is this just poor judgement on the image-maker's part or are they simply trying to cheat things so that you can tell they are using HDR? I'd still love to see a practical application that makes a beautiful image.

[Oren Grad]

Frank, this reminds me of Phil Davis's DI#13 developer formula for TMX. It's intended to tame scenes with an exceptionally long brightness range. It accomplishes that by doing weird things to the characteristic curve, flattening the midtones while preserving decent contrast in the shadows and the highlights. For this to be pictorially pleasing, though, the lighting in the original scene has to be just so.

It sounds like the HDR technique poses similar challenges. In principle the digital file ought to allow somewhat more flexibility in deciding where on the scale to place the tonal compression for optimal effect, but in practice it may be that under most circumstances it's very hard to compress such a long scale in a way that looks any good.

[paulr]

"Of course, negative film has much less resolution, so often you have to go up in format size to maintain the sharpness you had in the next smaller format when using chrome films."

this is another piece of common misinformation. today's color negative emulsions easily equal chrome emulsions in resolution of fine detail. chromes have the advantage of less aparent grain (the grainy part of the image ends up in the shadows, where it's ... um, dark). and some chrome films have a sharpness advantage at low resolutions--velvia, for example, produces huge adjacency effects in the 1 to 5 lp/mm range. This makes a velvia chrome look preternaturally sharp to the naked eye or when viewed with a low power loupe. it does not help it resolve fine detail (in fact it hurts it in this regard). A large print from a velvia chrome will not be more detailed or any sharper than one from a similar speed, similar size negative. though it will look less grainy.

check out the mtf curves on kodak's and fuji's websites ... it's all there for the taking.

[Jack Flesher]

>> Jack why are you just totally ignoring what WG has clearly defined in terms of how levels in digital capture work? <<

I am not ignoring it-- If you read what I wrote, I asked WG to explain the difference between a linear RAW and a converted RAW as they are not the same.

So, I am now waiting for his respoonse...

[David Luttmann]

Jack,

I hear Matthew harping on about tonality.....while at the same time he's not grasping the fact that tonality issues with low bit depth in the shadows effect a film scan the same way as a processed digital file. There can be no argument that digital files have a better signal to noise ratio than film....thereby having less noise in the shadows than film.

No one is hiding from tonality issues. The fact is Matthew, you don't seem to grasp that film has those same problems when scanned as well as when printed conventionally. The color in the shadows of film is incredibly inaccurate and riddled with high amounts of noise......just like the digital files. And because of these similarities in terms low quality in shadows, once again, this becomes nothing more than measurebating as it really makes no difference in terms of final output quality. Any differences are those in the way that the raw files are processed with curves, etc, and thus are issues to do with the skill of the operator and not to do with poor quality of the media choice.

Finally, as you have admitted to not having tested differences in a meaningful matter, you aren't in any position to call into question anything to do with conclusions others have made from testing.

[Jack Flesher]

Matthew:

And while we are at it, why don't you or WG explain to us how the linear file from your scanner works?

Clarification: I am asking you or WG to explain the difference to us between the linear RAW scan file, and converted RAW scan file...

[bglick]

Jack

>I thought you might say something like that...

Huh? I quoted a book which you can read it, or do a Google search and read about it? Not good enough? What kind of answer were you expecting?

> So now please do us all a favor and explain the differences between the linear RAW file and the converted RAW file as it relates to DR -- after all, they are significant.

First, my comments were not involving dynamic range, but rather the total tonal values that exist in each stop of of the dynamic range. My point was, how useful is these lower stops of DR you pull from a digital image? Here is a Canon sensor Tonal range recordings....

4096 total bits of tonal range available, 2^12=4096, Canon records in 12 bit....

So the stops of tonal range are as follows.....



1st stop = 2048 Tonal ranges

2nd stop = 1024

3rd stop = 512 ......

4 256

5 128

6 64

7 32

8 16

9 8

10 4

11 2



Hopefully this illustrates my point.... how useful is the later stops? Stop 11 can only define, color or no color.... This only represents the tonal ranges available from exposure, it does not introduce the other huge variable, noise in underexposed pixels, leaving in the end, almost unusable data. How much tonal range is enough? This is dependent on the subject matter in the stop in question. In addition, the magnfication of the image area at that stop as it realtes to the pixel density is another issue, as in fine detail, the Bayer sensor does not do well when exploited at its pixel level.

So to answer your question, how does this relate to the post processed raw file? Well, the ol saying, garbage in, garbage is appropiate. Although the software tries to interpolate these tonal ranges, the softwares success is still dependent on what its fed. For example, do you feel software magic can accurately display an area with large tonal range at stop 11 vs. stop 2?

Jack, I am a huge digital fan, trust me, so no flames please. Raw file processing is one of the 6th wonders of the world. I marvel at how clever the entire process is, i.e. only save x MB in the field, then run it through the grinder and come out with a file size of 4x MB. So for starters, a processed raw file is already 75% "guess work". (actually more, as the tonal range is also interploated up from 12 bit to 14 bit) And in my opinion, its truly remarkable how accurate the "guessing" is. However, the "accurate guessing" is done in the stops where the software has the most data to work with. I don't care how sophisticated an algorythym is, its hard to produce beatiful tonal ranges in an area that recorded but two ranges to begin with... make sense?

So how does this relate to HDR? It's just my opinion, if your going to record several exposures to grab all the tonal ranges in 15 stop scene, i would suggest taking 6 shots of the scene, not 2 or 3. As if you take too few images, if much of the scene exist at the lower end of the exposure, you can't expect accurate tonal range, and most likely will receive posterization or muddy tonal ranges.

> The big difference in DR is NOT the total DR available, but how the DR that is there gets distributed. The "automatic" settings in most raw converters do not do the total digital file justice. Hence, one must tweak the converter to achieve the desired result -- and that result can get very close to replicating C41 emulsions in latitude, color and tonal distribution. Heck, we can even replicate the grain once we pull the file into Photoshop ;)

All converters are only as good as what you feed them... this is not an issue about converter, its a shortcoming in how a digital image is captured, it has nothing to do with the converter. I hope my point is now much clearer..... the method Bayer sensors use to record tonal ranges is surely the achillies heel of digital capture, unless, the scene is a 3 stop scene, whereas digital capture is just brillaint. (assuming the pixel density was sufficient for its intenended end use)

> this is the sam for a raw scan from colour negative is it not? Do not the linear files from both follow this?

Tim, this is a great question. I don't think its true. I have found virtually NO information regarding this, as makers do not publish such. I am curious if anyone does have additional information on this. From my experience, I do not beleive this is true. A scanner has more time to record the information it sees, and I beleive the hardware / software behind the chips reflects this huge advantage.

> Conspiracy theory...! - hmm not exactly a big secret. We were discussing this 10 or 12 years ago at least

My only point was, I see digital gurus who swear by digital, still have no clue of this concept.... and lets face it, short of pixel density (which is dependent on final size of image) the inability to record large tonal ranges in all stops, must be overcome at the same time the exposure lattitude is improved. If and when, digital technology passes these hurdles, well, I don't want to say it.... but..... (from what I understand, this is not so easy though)

> Since just about every form of repro involves compressing the tonal range of our images, even from color chromes, at best we're going to see a little more shadow and highlight detail from a longer toe color neg, raw digital file,

Frank, I agree..... both image capture methods suffer from this.... but neg film sure does a better job then chrome film, but at a higher price in both rez. and grain.

> this is another piece of common misinformation. today's color negative emulsions easily equal chrome emulsions in resolution of fine detail.

Paul, this is a bit of a mystery to me. I have reviewed all the Fuji MTF film data. This is one of the rare cases where I disagree with the MTF data. My theory is, the MTF data is theoretical and does not consider the grain of image film, which ultimately effects the resolution. I don't have a sound theory on this, but I do see it on a regular basis..

> it does not help it resolve fine detail (in fact it hurts it in this regard). A large print from a velvia chrome will not be more detailed or any sharper than one from a similar speed, similar size negative. though it will look less grainy.

This has never been my experience. I think its interesting to note, that Fuji's uses an entirely different methodology for displaying data for neg color film vs. chrome color film. Of course they don't describe such in their manuals. But considering Kodak never even offered this information, one can not complain. This difference is clearly evident in the grain readings it offers for neg film vs. chrome film. To bring them to "apples to apples" you must divide chrome grain values by 2.3. The resolution values are also not based on the same set of circusmstances....but its been so long since I disucsssed this with Fuji film engineers, I can't remember some of the differences..... as I have relied on my own testing to give me a feel for the information I need. However, even at the high end, i.e. max. enlargements, I have found Velvia to outperform color negative film by at least 30%, which is almost equal to one jump up in format size, such as 4x5 to 5x7. That was my point.... does your experience differ Paul?

[bglick]

Dave, we had some cross posting going on here....

I am interested in learning how you conclude that scanners record tonal ranges the same as Bayer sensors do? I never could find any documentation on this subject, but am very interested in learning more.....

[Jack Flesher]

Dave:

What I hear Matthew harping on is that a scanned C41 files has FAR MORE DR than a direct digital file does.

So the points I see Matthew and WG missing are 1) the fact that scanned and direct linear raw files have the same VERY low tonal distribution and must be properly "converted" to extract their full potentails and 2) proper conversion of either of these will deliver an incredible dynamic range.

Next, certain forms of direct digital capture generate more DR than native C41, slightly surpassing the 11-stop mark. Though admittedly, scanner software can actually expand DR to a degree and may thus be able to take the native stops available in C41 up to around 10.

(It should be pointed out that the scanner software can also do this with E6, so that a film like say Astia can generate a very long tonal final file also. Finally, there are viable arguments that the orange base of most C41 films actually LIMIT the real DR to no more than that of E6 emulsions -- but let's not go there...)

I will add that while I do use a DSLR for direct digital capture, I also still use film, though only 4x5 and only for scanning, and my film of chice for that purpose is Astia. So to be perfectly clear, I am not bashing film or suggesting there are even better alternatives than scanning C41 -- All I am saying is that incredible tonal range can be had from a multitude of sources and is not limited to one specific type of capture.

Cheers,

[David Luttmann]

wg,

CCD pickup devices are similar in terms of tonal issues & bit depth to a DSLR....just minus the Bayer processing. This still leaves you with 12, 14, or 16 bit data to work with. As well, the quality in those bits are only as good as the feed from the AD convertor and is subject to limits such as noise from gain controls, etc.

Because of this, you end up with the same issues as a DSLR or digital back. If you take a look at 11 stops down on a neg scan, you'll see little to no color information as the film simply can't record it in a meaningful way. Realize I'm not saying digital does this better, I'm just saying that film is by no means perfect in this respect.

It is quite often noticed in testing of DSLRs that they are able to dig deeper into the shadows. This is commented on regularly by camera reviewers...even those that want to believe film can do better.

I can sum up this way....based on the tests from my Minolta & Imacon, todays top DSLRs beat out color negative film in terms of dynamic range....but just barely. If this wasn't the case, I would be using more film than I presently do. I also wouldn't bother telling you something that was the opposite of what I found in testing. If I was going to lie about something, I can find better things to lie about.....like explaining to my wife why I couldn't get the reservations at the restaurant she wanted for Valentines Day (because that happens when you remember the morning of ;-)

Cheers,

[Jack Flesher]

Yes, lots of cross-posting -- sorry.

[Jack Flesher]

WG:

No argument here, just getting at full disclosure and clarification ;)

My point is the figures you are quoting above are off the sensor for the LINEAR raw file -- and those are NOT the same as the values generated for the working file AFTER raw conversion...

[Jack Flesher]

Matthew:

Yes -- What I am referring to is the fact that the normal C41 (or E6) distribution gets compressed during the scan just like a direct digital capture does -- with a very low heel and high toes -- and must then be "converted" to restructure the data back to a useable form.

[bglick]

Dave interesting post. First, by no means am I considering you telling lies or altered information, I am very interested in your findings, and I am not disputing them, this is how we all learn. It's so often that we the final 10% of information we absorb on a subject is the most decisive portion. That seems to be the case here, as with all these posts, we are narrowing down to the nitty gritty differences between the two in regards to DR. Now as for dinner reservations, McDonalds is always open and requires no reservations? Just a thought.. :-)

A question... I fully understand your post, but again, where is the information derived from regarding how a scanner records bit depth? Leaving out films ability to record tonal ranges (which is another topic) I am curious as to where you got this "scanner" information. If you are correct, then of course I fully agree both digital capture and film scanning have the exact same achillies heel. The link that mathew provided sheds some light on this subject, something I always wanted to see.... but, I need to study the link a bit more, as I am not sure exactly what to conclude, but I am intereted in the comments from others.

Jack, althought I don't disagree with your comment of.....

My point is the figures you are quoting above are off the sensor for the LINEAR raw file -- and those are NOT the same as the values generated for the working file AFTER raw conversion...

I do feel this is misleading, as, post processed files are always dependent on the amount of real data they are fed, as feeding a junk file to the worlds greatest coverter will produce a junk file. So all converters are fully dependent on what you feed them, so regardless of the interpolation done by a RAW file, there is NO single bigger factor in final image quality then the actual "capture data". And that is what we are talking about here, what is captured, as the converter can only improve upon this by a certain %, no miracles....

> Next, certain forms of direct digital capture generate more DR than native C41, slightly surpassing the 11-stop mark

Jack what type of digital capture are you referring to here? Scanning backs? If so, i agree, but that is an entirely new can of worms :-)

[David Luttmann]

Mcdonalds? I guess I'd have to dine in rather than drive-thru! Actually, a fellow at one of our favorites took pity on me and booked us in at 7:30 tonight. I'm a lucky man!

wg,

The best I've seen from a DSLR is RAW shots from a 1Ds MK2 that show approx 11.5 stops of dynamic range. More than enough to be considered equal to or better than any color film. Some digital backs (not scan backs) can show even more than 12 stops. This is more than enough range to work with in a photo. I don't recall ever seeing anyone claim 12 stops for color neg film....that's for sure!

Cheers,

[David Luttmann]

I'm sorry, what exactly is funny.? I gave you the figures for the Canon 20D. Now that I mention tests done on a different camera you accuse me of changing my story? I haven't gone from one figure to another....I quoted a test done on the 1Ds Mk2 done with Imatest at the Imaging Resource website. They found 11.5 stops. Are you saying that they are telling a lie as part of a conspiracy againt film? Or could it be that maybe you have never tested this equipment and have no knowledge as to what the real figures are.....but have lots of opinion of what you'd like them to be.

As to 15 stops with Reala, not with a Nikon Coolscan he didn't. The LS 2000 is a 12 bit scanner and is not capable of a 15 stop range. Even with 16x multisampling, you won't get 15 stops. I know because I used to own one! For my low cost scanner, the Minolta, I used 16 bit 16x sampling and still won't get 15+ stops. That range exceeds what the scanner is capable of and doesn't agree with tests I've done on an Imacon nor a drum scan. Looks like flawed method and wishful thinking from where I'm sitting.

As well, the broad shoulder characteristics of Fuji Reala simply mean that it is possible to record highlight detail and keep it maintained well past where one would expect highlight blowouts....especially in specular areas. The problem though is that this is not true useable dynamic range, but rather a compressed area of highlight maintainence because of the broad shoulder of the film. This is what we have been referring to as "usable dynamic range." This compressed are does not equate to a few extra stops of dynamic range as some people like to believe.

Sorry Matthew, but you're running out of straws.

[bglick]

The link mathew provided is very interesting..... and it sure demonstrates just how huge the dynamic range of negative color film is. I am not sure how the Nikon LS2000 measures up to the best of film scanners of today. Does anyone?

Anyway, this test did NOT compare the same vs. digital capture which would be an interesting comparison, as the authors testing methods were very sound in my opinion.

Now, looking at the test results on this link,

http://www.path.unimelb.edu.au/~bernardk/tutorials/360/technical/hdri/results.html

I draw the following conclusions.... on film, Reala can record 10 - 11 stops visual on a light box.... (this is fits well within industry talk) But what these results demonstrate to me is, even film needs its equivalent of "Raw Processing", and although I once assumed digital capture benefited more from "processing"........ well, I now fell differently, as these test results demonstrate just how huge the differences are for the "processed" results. After seeing these results, I can't see how digital capture processed images can compare to negative film processed images. Of course, this only is an issue when such large dynamic range is required.

Of course, there is many caveats here. For example, the author treated each image as an individual scan. Therefore, software handled the interpolation well. These amazing processed images would not have faired so well if they all existed in the same piece of film, i.e. if the one piece of film was shot with 21 targets that had reflected values that varied by one stop, certainly not a practical set-up. My point is, his test results are from scanning 21 pieces of film and the paticulars of software were maxmized for each scan. If the author had all these images on one piece of film, he would have to scan the same piece of film 21 times to acheive the same results. How much would the results would differ?

Also, based on these tests, it does seem that the CCD scanner is recording information very similar to the linear fashion of a digital camera. However, the fact the film is holding more lattitude is helpful, even if the scanner records the same, so its my opinion, the overall effect of more lattitude in color neg film combined with linear recording still yields more tonal ranges, with software manipulation of course.

The most interesting part at the bottom of the link I posted above, is the "mean pixel value" chart.... I think this is the most relevant data as it relates to "usable" stops. From this test, it seems EV 3 through 11 are clearly the most useful, which means for black and white subjects, neg film can record good detail in 8 stops. I buy this...... For subject with less contrast, white-grey, the film recordered well 3 EV through 8 EV, or 5 stops, again, I will buy this. I consider these usable stops are a bit more then double what you can expect from chrome film with low contrast subjects, (equal to contrast of white-grey) you can expect 2.5 stops at most. (maybe a tad more with Astia) Now, just how does digital compare using this same test criteria? My guess is, it will compare equal to chrome film, but not color neg film, i.e. when using the same criteria for "usable" stops. However, digital capture will surely record more stops then chrome film, just not more usable stops based on the same criteria. Just my opinion from experience and reading these tests.

The other issue regarding this test, and possible digital test is, it relates only to black / white / grey subjects. A digital capture test using the same targets would be operating at optium results. However, if this test was done using color targets using the same contrast (not in the B&W, but rather the white-grey, grey - black contrast range) such as light green and dark green, the digital capture would do much worse, due to the RGGB nature of digital capture. This is an area that benefits film, as EVERY film pixel can record R, G or B. I have tested Bayer sensors with color targets, and it does extremely poor with R and B, no surprise, but the results are dramatic, specially when the test pattern is not square to the sensors grid.

So I guess the answer many of us are trying to get at - is not clearly presented in a sound test using both digital and film. Although maybe Dave did such? Anwyway, with all the variables at play, contrast of target, color of targets, number of scans made of the same film, type of digital processing used, type and generation of digital sensor used, type of scanner used, scanner software used, its nearly impossible to draw solid conclusions, although I can't wait till someone takes the time to do a more thorough comparison.

Dave, 11 stops for 1DsMII....what contrast was the sujbect, black and white, I assume?

[David Luttmann]

Matthew,

I've shown that modern day DSLRs have a range between 9.5 & nearly 12 stops. You have shown absolutely NOTHING to prove these "conclusions" incorrect. I have pointed out numerous tests that show these results. As usual though, any result that conflicts with your established opinion of what you want is judged by you to be incorrect. I won't waste any further time with you as you need to get out and actually compare before you'll have anything to discuss. As you appear to refuse to do this, seemely because you feel you already KNOW, I don't expect to hear back from you on this matter anytime soon.

wg,

The tests are done with step wedges. You can check out the method from the Imatest web site. I purchased the pro version and have found it to be a great tool to evaluate lenses prior to purchase, as well as comparing films and digital bodies. The biggest surprise I had was evaluating a friends Fuji Finepix S3. Using ACR from RAW, I got 11.98 stops recorded. This pretty much matches the 12.1 stops recorded from the imaging resource web site. I am hoping that Fuji can come out with a higher rez model with the same DR. This would be perfect for outdoor portrait work that I do on a regular basis.....giving me about 2 stops better than what I presently have.

Best regards,

[bglick]

Dave, I am not disputing the results you got, or the technique you used. I have seen the Imatest at work, and I agree its very fundamentally sound. Just so we are clear.... the issue that seems un resolved is, just how many "usable" recorded detail exist when using the same criteria when comparing....

Digital caputure
Chrome scanned
Negative scanned

This is what I refer to as, a missing link in this quandry.

Matt, what you have done is the same as the author of the test has done, you used a single reference on the film, and allowed the DR of the scanning system to use that as a reference point, which works very well specially with neg film as the author of the test demonstrated. However, the same can not be done when only one single scan is made of an entire image whereas there is tonal ranges throughout many captured stops. Much of this is dependent on how many stops the scanner can record and where you set the range to begin recording these stops. Not all scanners / software is capable of this shifting technique.

[David Luttmann]

"You are a very selective reader and seem to avoid any tests you can provide with regards to neg."

Matthew,

you haven't provided ANY tests....just opinion. I've already pointed out that the Nikon LS2000 is incapable of the 15 stops you claim from Reala. Having owned this scanner, I'm well versed in what it can & can't do. While is was a good (albeit slow) scanner, it was not capable of the range claimed from the test site. Yes, you can go further....you could even claim 16 or 18 stops, but the signal to noise floor would be so bad that discerning signal from noise would be an exercise in futility. That is why threseholds are established with Imatest to prevent going far off the true data stream....otherwise you'd have people claiming 14 or 15 stops of DR, even though the signal would be pretty much lost within the noise floor.....much like what the web site you mentioned is doing.

Once again though, these aren't tests you've done, but, to quote you, "stuff you've picked from the web." I wonder why people who haven't actually tested something are so sure of themselves with regards to outcome. Poor science background I guess.

[David Luttmann]

Good day.

[Jack Flesher]

>>Yes Jack ,agreed . But if there is more information there to start with ,say in the highlights this will also be compressed . But at least it is there , compressed or not . Which is going back to what I origianlly said about dyanmic range. : ) <<

LOL! Not going to conceed this point that easily -- we're now getting into metaphysics ;)

Fact: ALL of the original DR only exists with 100% accuracy in the ORIGNIAL scene...

Sooooo.... If I capture that scene with film, then scan the film, regardless of what emulsion I choose I am introducing TWO iterations of the original scene before the final print output. So regardless of total DR captured, the overall tonality has been altered twice -- and who knows what damage has been wreaked to the original scene? If OTOH, I do a direct digital capture, I am only introducing a single iteration of the original before I print it out. (:D) Hence the original DR is only being altered once and I can thus make an argument that the direct digital capture has to be better since it is closer to the original scene than the scanned film. *HUGE* Grin

Obviously, I am not actually arguing this ridiculous point, but rather making a point about how this debate has morphed -- The bottom line is we are getting really close to arguing about how many angels can dance on the head of a pin.

I fi understood the above comment, you are saying that film captures more DR before it is scanned than a digital sensor -- and I do not argue that point. (Surprise!) But at the same time I am saying the LINEAR raw file is not the end of the equation for direct digital capture and one needs to consider the CONVERTED raw file as the end result for direct digital capture! AND I add that same properly converted raw file can have an equally broad DR to that original piece of film.

Notwithstanding, I happen to agree with Frank's premise that an excess of DR does NOT automatically make an image good ;)

Cheers,

[bglick]

>I am saying the LINEAR raw file is not the end of the equation for direct digital capture and one needs to consider the CONVERTED raw file as the end result for direct digital capture!

Jack, not trying to beat a dead horse here...... The only difference between a raw file and converted raw file is..... the RAW file is 100% image captured data, a converted file contains the same captured data, but adds 300% bits of "guessed" data. So even though a converted file has about 4x the information of a raw file, it's the raw "captured data" the converting software makes its educated guesses from. So captured data (raw data) is the basis of the entire file. So not sure why you put all your emphasis on the converted file, as if the data does not exist in the capture, then coverting software can make no educated guesses upon it? It does not add data from thin air... Raw converstion does nothing more then "fill in the blanks" of the two missing colors for each pixel and converts from 12 bits to 14 bits (or whatever the paticular camera / converting software does)



And of course I agree with Frank, and mentioned several times above, if you are shooting a 2 stop scene, none of this matters!

[tim atherton]

"I if understood the above comment, you are saying that film captures more DR before it is scanned than a digital sensor -- and I do not argue that point."

and if not scanned, you begin to encounter limitations of enlarger and paper

[Jack Flesher]

WG: Not relevant, because the same thing happens in a scan -- the only difference is the scanning software converts it wothout giving you a linear raw option...

:D,

[Jack Flesher]

Oh yeah...

and if not scanned, you begin to encounter limitations of enlarger and paper

Thanks Timothy!

[tim atherton]

"the only difference is the scanning software converts it without giving you a linear raw option..."

Which I think (?) is what the vuescan RAW option gives you - a linear raw file - which, for example, lets you then do ALL the conversion in Photoshop with something like NegPos for B&W and Colour neg

[David Luttmann]

"the RAW file is 100% image captured data, a converted file contains the same captured data, but adds 300% bits of "guessed" data. So even though a converted file has about 4x the information of a raw file, it's the raw "captured data" the converting software makes its educated guesses from."

wg,

Where in the world did you get this "guessing" from? If you are referring to the fact that RAW files are smaller than their tif counterparts, that has only to do with lossless RAW data compression. There is NO guessing on the part of the RAW converter that corresponds to differences in file size.

[bglick]

> Not relevant, because the same thing happens in a scan -- the only difference is the scanning software converts it wothout giving you a linear raw option...

jack, I would still like to see one bit of documentation on this fact.... But regardless, even if film scanners are subjected to the same linear bit depth capture, you must compare the bit depth per channel of capture of each. Most good film scanners capture either 14 or 16 bits, and from what I can tell, digital sensors capture much less.

It appears the Canon 20d and 1DsII record only 12 bit per color. Is this correct? If so, a scanner that records 16 bits, is at a big advantage.

2^12 = 4,096 (from previous post)
2^16 = 65,536

If you replace the 4096 I used in the chart above, and start with 65,536, I think you will see there is a huge advantage in the usable stop range, so I don't think your premise of, the same linear losses happen scan or digital capture is accurate. For example at each stop, the scanner has 16x more tonal ranges, for example, at stop 6, digital capture has 64 tonal ranges vs. 1,024 tonal ranges for the scanenr. This 16x difference is much greater then one would think when hearing digital capture at 12 bit vs. 16 bit for scans. On the surface, it appears a scan is only on 33% increase in tonal range, but in practice, its actually 1,600% increase in tonal range.

Dave, I have only read a few books on digital so far, and have only worked with it for a few years, so I am no expert. But here is what I recall from the authors. Each pixel can only record ONE color value, it sees no other colors. So a blue pixel will record only a tonal value for the amount of blue it sees during exposure. Now, in the processed file, each pixel mysteriously has a distinct value for R, G and B. So, if each pixel only records the value for one color, the remaining two colors must be "guessed", "interpolated", "converted", so the final processed file now reflects all pixels with a distinct value for R, one for G, and one for B. Right?

I am not doubting there is some compression going on, but thats irrelevant, as just this color interpolation triples the data, then there is other interpolation and probably some compression that makes up the rest of the file ballooning. Did you understand RAW processing differently?

[Jack Flesher]

WG:

If you already know it is the darkest color in your file, how many bits of electronic data do you need to confirm that fact?

[bglick]

Jack, you switched gears on me, you left the discussion, I can only assume why....that's fine.....

It seemed we were making progress.... but the learning and contributions usually stops when information is presented contrary to someones previous position.... Hopefully someone with no previous position can comment.

[David Luttmann]

Actually wg,

my contributions stopped as I got tired of trying to change Matthews opinions with facts....to no avail. Oh well.

[bglick]

Dave, I am sure you realize I was referring to jacks post.

But you are more then welcome to comment on my post, no flames here....

Also, you sort of attacked my explanation of raw conversion, or interpolation..... you asked "where in the world" I came up with such. I explained myself and am curious if you disagree with me, or just failed to comment. I am open minded and just want to learn.... I did not take your post personal, but I really am curious if I am missing something here as you seem very knowledgeable.

To me, this is a shortcoming in many forums, a thread can only go so far, then, once clumination starts to set in, opposers drop out, but never sure why, did they go on vactation, can they not admit to being wrong, did they never come back to read the thread, etc.??.

The best analogy I can think of..... many threads are like court trials, the prosecutor makes a strong case, but then, the defense disproves much of the case with strong evidence and common sense.....but then, instead of cross examining witnesses, exploring other possibilities, etc., the prosecuting team just leaves the court room.... so did the defense really win? Would the prosecuting team won if they did not leave the court room? Who knows? The same is true with many of these threads. If one searches the archives, you find most threads end in more confusion then they started with, even though there is some strong cases made throughout......

It's my guess from reading a lot of threads, NIH syndrome (Not Invented Here) is the leading cause of much of the inconclusive threads. I am not referring to threads which are based on opinion and subjectivity such as, does ink jet look better then darkroom prints.

[Jack Flesher]

>>Jack, you switched gears on me, you left the discussion, I can only assume why....that's fine..... <<

Uhhh I never left it WG. I asked you a simple question that related to your 12-bit comment o nt he Canon's directly right above it. All it required of you was an answer...

Am I to assume by your more recent comment that you now are not wanting to participate?

Cheers,

[bglick]

Jack, why do I need more information to confirm my darkest point? That is not the issue here....

What I described above was the fact 16 bit capture offers 16x the amount of tonal ranges, in any given stop, for scaned file vs. digital capture. Regardless whether its in the darkest or lightest area of the image. But most importantly it applies to the darker areas where the 12 bit capture has few tonal range choices to pick from as a result of the linear degradation. Make sense? I am interested in your comments, truly, I know these are issues close to your heart :-)

[Jack Flesher]

WG: I did not say you needed any more information, nor did I imply it -- I only asked a simple question.

I'll repeat my question in a slightly different form: How many bits of data do I need to CONFIRM pure black in one channel if I already know it is my darkest color?

We'll get to your 16-bit query when we get over this hump.

[bglick]

Pure black probably needs very few bits to select from, as its the top of the color heep..... But I don't see the relevance to the question? I thought all this was about tonal ranges within usable stops, right? So where ya headin with this Jack?

[Jack Flesher]

PS: And FTR, to your comment that I left the discussion: I have a real job during the day and attend to it. So contrary to popular belief, I do not live connected to the web 24/7/365 and am not always able to respond to internet forum threads in real time.

And while we're at it, let's face the fact that the way this site's forum software works, this thread will die tomorrow afternoon when it hits page 2.

[bglick]

Jack, when I mentioned you "left the discussion", it was in regards to your post, NOT your lack of a post. I certainly understand not everyone is available 24/7, so lets be fair.

My comments regarding people dropping out of threads when evidence moves against them, was not regarding you, it was a general comment explaining why most threads never reach solid conclusions.

[Jack Flesher]

Okay, fine. So now I am only waiting for you to answer the simple question I posed...

[Jack Flesher]

Sorry -- I meant to say please be more specific than very few -- how many? It's okay to guess.

I promise, you will see where this is going in short order

:D,

[bglick]

My guess is 2, black and white, OK?

One minute you don't have time to respond, then next you insist on these guessing games??? Why not just state your position so we can benefit from what you are concealing?

[Jack Flesher]

Good guess. It is actually 1. 1 bit gives us two choices -- on or off; sensitized or not. If it is off, it is black, if it is on it is something else, hence we can confirm it.

Now move one level up on our DR scale. With 10 stops of DR total, this is still a very, very dark color.

Can you estimate how many bits of data we need to determine its relative tonal values since we already know it is one level lighter than the preceding full black?

[bglick]

I meant two tonal ranges, which you have correctly stated is one bit, but,

in the method used above, which is the premise of this discussion, the tonal ranges have already been broken down to tonal steps per stop, not bits per stop... so again, this is all mute.....

If you want to make a point and educate us all, I would be glad to read it and comment, but I don't have time for a Q&A session with you.....

[Jack Flesher]

No WG... the purpose behind my Q&A method was to take it one step at a time, presenting the math clearly and logically so that YOU ultimately would be able to understand a complex subject in simple enough fashoin that you could explain it to yourself...

Since you don't want to do it that way, I am not going to waste any more energy explaining the mathematics of deconvolution to you and instead leave you to go figure it out for yourself...

Good day,

[bglick]

Jack, your Q&A methodology would be fine during a phone conversation or face to face, but to conduct a conversation in a thread, with constant Q&A in a forum like this, IMO, is a bit odd.

Anyway, the purpose of the forum is to benefit everyone, not just me. It's possible if you gave a thorough post explaining it, I would not be intelligent enough to understand it, but I am sure others would. I won't waste my time figuring out what your talking about, as I have no idea what you are suggesting. Again, a thread that ends with with no real conclusion, the plaintiff leaves the court room halfway through the trial....

[Kirk Gittings]

Guys this is like a flame war on qualudes......zzzzzzz

[Saulius]

How about HDTR? (http://mkweb.bcgsc.ca/fun/hdtr/) I know it's not what you were talking about but I thought this was an interesting image. (http://mkweb.bcgsc.ca/fun/hdtr/hdtr-14.jpg)
Sorry if this might have been posted already but I couldn't get myself to wade through all of those posts.

[David Luttmann]

Of course, it has already been explained that film has little to no tonal accuracy at these same shadow areas....but that will continue to be ignored by some I'm sure.

Jack, just give up. It appears that it is not that some people don't understand.....it's that they don't WANT to. Afterall, some people don't want their sacred cow (film) to be gored! That's fine. It just seems silly to be blathering on about lack of color tonality with digital sensors in these deep shadows when film has exactly the same problem.

[Frank Petronio]

I just like how people can do 100 posts while never getting back to the original question ;-)

[Tom Westbrook]

>Dave sed: film has little to no tonal accuracy at these same shadow areas

Sorry if this has been covered, but isn't that a problem of human perception and not so much one of method of capture? People don't perceive colors well in dark areas. I think you'd have to talk on more of a technical level using densitometer data or RGB values to see whether one was better/worse than the other in that area. So who's going to care if one does better than the other in that area?

BTW, are there data plots around for various sensors like the sensitometric data they give you in film data sheets? Curves showing sensitivities at various exposure levels, spectral sensitivities, that kind of thing?

[David Luttmann]

"As I have now shown this is not the case. "

Sorry Matthew,

I must have missed your example of where you were not able to dig as deep in a digital source. Please post your link of the same shot as you have said you have shown "this is not the case."

I'll wait for your post proving your point.....unless of course, you haven't done this with a digital source and as a result have once again not proven your point at all.

Regards,

[tim atherton]

"Guys this is like a flame war on qualudes......zzzzzzz"

It's the slow, contemplative Large Format approach to flaming... :-)

[David Luttmann]

Matthew,

You have summed up the ENTIRE problem that a lot a film users like you have....they are unwilling to do a test to compare side by side. You rely on what you think to be true...what you want to be true.....but you refuse to compare.

I've been through this many times. I remember comparing 4x5 Astia scans blown up to 16x20 along side a 16x24 enlargement from a Canon 1Ds MK2. There were a few people from a local camera club who maintained that it was absolutely impossible for the digital source to even compare. I asked how they did their comparisons? The answer I got back is typical....they didn't need to do comparisons because they KNEW it was impossible. Well with logic like that, a discussion is a waste of time. However, when they viewed the print comparisons, they were astonished to find that they viewed the prints as being similar in quality. Each had a slight edge over the other in different areas. The film with an ever-so-slight advantage in resolution....only visible by pressing your nose to the print. The digital won out in terms of color accuracy, sharpness & accutance, low noise, and ability to dig into the shadows.

A lot of people will try to hide behind numbers, etc. I have found that when you do tests on your own, and read tests & comparisons of well known photographers & printers, there is a common thread virtually all of them....and that is the they are quite often better able to dig into the shadows from a digital source. Even Charlie Cramer recently commented on this when he was testing a digital back.

I can understand where you're coming from because I was the same. I refused to believe it until people started showing me comparison prints and tests. I then did the same to prove them wrong....only to find that they were correct. That is the main reason why an overwhelming majority of my work is now digitally based. I chose it because for me, and for the print sizes I do (mainly 16x24 & less) I get quality that is similar to film, but without any grain.

So, as I said, film exhibits the same problem in scans as digital files do with regards to tonality. You conveniently ignore that. So, unless you are printing color with an enlarger, you will experience the EXACT same tonality issues.....which when you compare side by side, are actually not issues at all.

Unfortunately Matthew, you don't understand this because you have never compared and thus, you continue to blather on about bit depth, etc, when they are of no issue.

Sorry, but until you compare & test, you have nothing to offer here other than opinion. Until you compare & test, I will ignore further posts from you as they are nothing more than uneducated opinion.

If this sounds harsh, I apologise. I am however tired of people who claim something that they have no comparisons, tests, or data to back up there opinions with. Telling me something is so, just because you think it is, does not hold any water with me.

Good day.

[bglick]

> Jack, just give up. It appears that it is not that some people don't understand.....it's that they don't WANT to.

Nothing could be further from the truth...how can we understand something when the prosecution left the court room before making their case? Ok, so blame the jury for not understanding, I follow the logic here?

> Of course, it has already been explained that film has little to no tonal accuracy at these same shadow areas....but that will continue to be ignored by some I'm sure.

Just so its clear, as this thread has gone more towards right/wrong, vs. the sharing of information..... The issue here is NOT tonal accuracy, we all agree film and digital have spectral response curves that are poor in underexposed areas. However, much of these areas can be isolated in PS and color corrected. So thanks to digital processing, this is not the issue....the problem is, what happens when there is not sufficient tonal ranges and we get posterization type effects from the huge gaps in tonal ranges available in the given stop of exposure. Somehow, the prosecutors bailed out on this concept as the thread developed in this area??

As for many of us being anti-digital, well, I can only speak for myself, but I find that absurb, I am the biggest digital fan out there, I love its potential and use it often, but that does not mean I want to ignore the differences.

> You have summed up the ENTIRE problem that a lot a film users like you have....they are unwilling to do a test to compare side by side. You rely on what you think to be true...what you want to be true.....but you refuse to compare.

Dave, if there was perfect test done on these topics, by qualified technicians, with detailed white papers explaining the results, we would not be having this discussion. I too wish these tests were more readily available, but even then, the results would be challenged as pro digital, anti digital, etc. So the best we have is the occasional sharing of information on forums like this, but unfortunately the human elelment is the limiting factor as this thread has demonstrated so well. I have never fully understood how quickly a thread can turn sour, once someones information is challenged.... such as Jacks sudden insistence of engaging me into a Q&A session, to entice him to share his point? If I don't comply with his up/back, up/back, tutorial, he punishes us by staying silent? On a public forum?? Which adults use? I admit, much of this is just beyond me.

It's too bad, as there is a lot of good minds on this forum. But I will continue to offer what I can to help those who can benefit from my experiences, and I hope some contributors also continue with thie same theme of "offering what they can, when they can".

[David Luttmann]

"So thanks to digital processing, this is not the issue....the problem is, what happens when there is not sufficient tonal ranges and we get posterization type effects from the huge gaps in tonal ranges available in the given stop of exposure. Somehow, the prosecutors bailed out on this concept as the thread developed in this area?? "

wg,

I have pointed out that these posterization effects are present from both film & digital, albeit in slightly different ways. However, instead of acknowledging this, it has somehow been focussed upon by some as a digital only problem....in spite of many attempts to clearify the fact that this problem is evident in both film and digital capture.

Because this problem is evident in both capture methods, there is nothing to be gained by silly comments about how this effects only digital capture....as it doesn't. Breath is simply being wasted here. This would be similar to complaining about digital noise at high iso, while avoiding the fact that film has the SAME problem.

As to qualified technicans, I think those of us using industry standard software tools to measure different media have already posted enough evidence.....the problem is, for some, it will never be enough.

[bglick]

> I have pointed out that these posterization effects are present from both film & digital

Dave, no one is disputing this?



> it has somehow been focussed upon by some as a digital only problem

Well, if you have been reading my posts, nothing could be further from the truth. The purpose of advancing this thread was to determine the differences in linear bit depth recording using the two methods, i.e. scanned film vs. digital capture. Of course they both suffer the same consequences, but the goal was to define which method can record more steps of tonality at a given under exposed value, right? To me, thats when the shi%$ hit the fan.....

> there is nothing to be gained by silly comments about how this effects only digital capture

I can only assume you are not reading my posts. I am not trying to upset you, but your comment makes no sense to me. This thread is comparing the two, it's not a conspiracy of one vs. the other. From your tone, it seems you really think many of us are anti-digital, which is absurd, as I constantly search for more justifications to be fully digital. Anyway........

[David Luttmann]

wg,

I'm commenting on the statements made by Matthew as well as others in the past on this very topic. You have been one of the few to actually wish to participate in a logical discussion. My apologies if you thought some of my comments were directed to you.

Cheers,

[bglick]

Dave, you did not preclude me, so I assumed I was in the masses.

Is it possible you can direct your posts to me? :-)

It's just gettin good.....