PDA

View Full Version : Is digital 6x9cm quality as good as 5x4" film"



wnw
11-Jul-2008, 11:08
Being fed up with the inability of digital lenses to resolve images to anything like the quality I am looking for I am considering a Linhof 679 which will give me access to Schneider/Rodenstock lenses.

I need to decide on digital or film or both but have a question:-

I recognise 5x4" film gives amazing quality but the Linhof matched to a Phase One P45+ back seems to be 6x9cm format (given it's a 49x38cms chip). If not 6x9cms then what format is it?

As this seems to be half the size of 5x4 film can anyone explain how it is now generally recognised 6x9cm will give better or equivalent quality??

I'd really appreciate a simple answer if there is one (and also no generic debate over film v digital). I just need to understand the considerations of the two formats in terms of the end quality of images produced.

Kirk Gittings
11-Jul-2008, 11:15
My two cents based on limited first hand knowledge and allot of reading. The P45 back has much smaller capture area than 6x9, but the quality IMO is somewhere in between drum scanned 6x9 and 4x5.

Given equal quality lenses and film flatness, quality drum scans etc., IME 6x9 film gives about half the quality of 4x5 film.

sanking
11-Jul-2008, 11:27
The Phase One P45 back has a sensor that is 49.1 X 36.8 mm in size, or about 1.9 X 1.4". That is actually less than the size of a 6 X 4.5 cm negative. If you do the match you will find that the resolution limit of the P45 back is about 72 lines/mm. In practice a maximum resolution of 60-64 lines/mm is what the back delivers.

The P45 is a great creative tool and gives enough resolution to satisfy most professional requirements in print sizes up to about 24X30". However, you can actually get a lot more effective detail in a 6X7 cm or 6X9 cm film negative if you use a top quality film camera than you get from the P45. This is really kind of a no-brainer since the film area of a 6X7 cm or 6X9 cm negative is much greater than the sensor area of the P45, and you can get at least as much resolution in lines/mm from top quality MF lenses.

However, to take advantage of the potential of the 6X7 and 6X9 cm negatives you will either have to print optically or scan the negatives with a high quality scanner. You probably need a drum scan to pull everything out of the negatives, but a dedicated film scanner like the Nikon LS-9000 or a high end flatbed will come close.

If you are using a lot of film and work professionally the P45 might be a wise investment. For most of the rest of us MF is a lot less expensive and, when digitized, is capable of higher absolute performance than the P45. I say this with the caveat that it depends a lot on the film because one of the advantages of digital capture is that it can be interpolated up in size better than film that has a lot of grain.

Sandy King




Being fed up with the inability of digital lenses to resolve images to anything like the quality I am looking for I am considering a Linhof 679 which will give me access to Schneider/Rodenstock lenses.

I need to decide on digital or film or both but have a question:-

I recognise 5x4" film gives amazing quality but the Linhof matched to a Phase One P45+ back seems to be 6x9cm format (given it's a 49x38cms chip). If not 6x9cms then what format is it?

As this seems to be half the size of 5x4 film can anyone explain how it is now generally recognised 6x9cm will give better or equivalent quality??

I'd really appreciate a simple answer if there is one (and also no generic debate over film v digital). I just need to understand the considerations of the two formats in terms of the end quality of images produced.

wnw
11-Jul-2008, 11:28
That's great, thanks.

Do we know why the digital back gives this level of quality and how it can compare so highly? I'm just confused because of the small sensor size when compared to a massive piece of film measuring 5x4"

wnw
11-Jul-2008, 11:40
What I seem to be asking is: How can medium format digital compare to 5x4" film in quality - doesn't seem to make sense to me.

Thanks for your patience.

sanking
11-Jul-2008, 11:43
I touched on this in my previous message. A file produced by digital capture is very clean and has no grain, and this compensates for the fact that it has less real detail than in a 6X9 cm negative.

However, this is not an absolute as there are some very fine grain films which when scanned show little or no grain, and even this can be reduced considerably by post-scan processing with a program like Noise Ninja.

In terms of real detail the P45 breaks down at about 24X30", without interpolation, while there is enough detail in a MF 6X7 or 6X9 cm negative to go to about 30X40", assuming a high quality scan.

Sandy King



That's great, thanks.

Do we know why the digital back gives this level of quality and how it can compare so highly? I'm just confused because of the small sensor size when compared to a massive piece of film measuring 5x4"

wnw
11-Jul-2008, 11:52
It's the whole post processing/Noise Ninja/CS3 thing I'm trying to avoid. I really want to try to capture as pure an image as possible at the time of capture and not rely on all the digital gizmos.

I am an accomplished digital photographer in 35mm but when I see 5x4" film images it blows me away and is so far ahead of anything 35mm can capture. Hence the 6x9cm decision. It's really a choice of film or digital but it seems 6x9cms P45+ digital can compete with large format film, given all the financial considerations.

5x4" film or 6x9cm digital is the quandry - the camera would be the same. 6x9cm film doesn't seem the answer I'm probably looking for in terms of quality.

Sorry to sound uninformed but your replies are really helping.

Dave Parker
11-Jul-2008, 12:01
If you don't want to rely on "digital gizmos" then why are you trying to compare a digital back to film?, shoot film and have it high quality scanned, 6x9 film handled properly can result in some very stunning images, as can 4x5 and up..different animals do not result in similar baby's...your going to have to figure out which poison you want to use to get the results you are after..if your trying to get images into a computer, no matter what you do, your going to have to used "digital gizmos" to do it, I have not seen a camera or a back yet, that does not require some processing.

Dave

Aender Brepsom
11-Jul-2008, 12:10
... a Phase One P45+ back seems to be 6x9cm format (given it's a 49x38cms chip). If not 6x9cms then what format is it?

Mathematics may not be my strongest domain, but a P45 chip, with 49x38mm (not cm!!!) is far from 6x9cm. At least that's how I was taught at school.
When you give the size of the chip yourself (regardless of your error with cm vs. mm), how can you claim it is the same as something that clearly is approx. 3 times larger?

I do not know of any 6x9cm digital back. Is there such a thing?

sanking
11-Jul-2008, 12:17
You rather take me off guard with this comment. Whether you shoot digital or scan film there is always some degree of post-processing involved for optimum results. Using digital, be it film or digital capture, inevitablly involves the use of numerous "gizmos" to produce a final image. The use of a noise reducer to minimize grain is as natural to the process as unsharp mask, luminous layer masks, etc.

Sandy King




It's the whole post processing/Noise Ninja/CS3 thing I'm trying to avoid. I really want to try to capture as pure an image as possible at the time of capture and not rely on all the digital gizmos.

I

Ken Lee
11-Jul-2008, 13:02
In the words of Ken Rockwell (http://www.kenrockwell.com/tech/format.htm):

"This $300 used 4x5 is sharper than a new $3,000 Hasselblad and worlds beyond a $5,000 Leica or Contax"

Keep in mind that 4x5 is on the small end of "Large Format".

So if you are blown away by 4x5, have a look at some prints made from 8x10 and larger.

Kirk Gittings
11-Jul-2008, 13:43
WNW, here is a link to a comparison that alot of people refer to. It is not gospel as far as comparisons go, as some would argue that there are better scanners than the Tango and therefore the 4x5 is capable of a bit more than is represented here.

http://www.luminous-landscape.com/essays/Cramer.shtml

sanking
11-Jul-2008, 14:10
WNW, here is a link to a comparison that alot of people refer to. It is not gospel as far as comparisons go, as some would argue that there are better scanners than the Tango and therefore the 4x5 is capable of a bit more than is represented here.

http://www.luminous-landscape.com/essays/Cramer.shtml

Don Hutton and I, together with a friend from Atlanta who owns a P45 back, did some comparison testing a few weeks ago that involved Mamiya 7II, Canon 1ds Mark III, 4X5 and the P45 back on a Contax. We tested both a real scene and a constructed scene with resolution target. The goal is to make prints at a number of different sizes and get together and compare output, and hopefully write the test up for publication. I really believe that this will be a much more valid comparison than the one at LL.

At this point I have seen the film scans (Fuji Acros and Portra 160) of the 6X7 cm and
4X5, which were drum scanned with a Howtek 4500, but not the files from the P45. However, my understanding is that there is no comparison between the 4X5 and the P45, and that in terms of detail the MF scans are much better than the P45. 4X5 clearly beat the 6X7 cm. How this compares with real images on paper remains to be seen.

Sandy King

Kirk Gittings
11-Jul-2008, 14:28
I look forward to seeing the results of that test.

HeinrichVoelkel
11-Jul-2008, 16:48
A p45 has a dynamic range of 11,5 up to 12 stops, more than a color slide, less than a b/w negative.

Mattg
11-Jul-2008, 17:46
Don Hutton and I, together with a friend from Atlanta who owns a P45 back, did some comparison testing a few weeks ago that involved Mamiya 7II, Canon 1ds Mark III, 4X5 and the P45 back on a Contax. We tested both a real scene and a constructed scene with resolution target. The goal is to make prints at a number of different sizes and get together and compare output, and hopefully write the test up for publication. I really believe that this will be a much more valid comparison than the one at LL.

At this point I have seen the film scans (Fuji Acros and Portra 160) of the 6X7 cm and
4X5, which were drum scanned with a Howtek 4500, but not the files from the P45. However, my understanding is that there is no comparison between the 4X5 and the P45, and that in terms of detail the MF scans are much better than the P45. 4X5 clearly beat the 6X7 cm. How this compares with real images on paper remains to be seen.

Sandy King


Sounds very interesting Sandy, especially as are going to compare prints made from the various formats rather than just looking at pixels on a screen.

I understand the reasons for wanting to scan the film, if it's colour, but doesn't that put it at a disadvantage in a test like this? Film has the advantage over a digital file that it can be printed optically; would the film based prints be shown to their full/better potential by enlarging them rather than scanning them?

Eric Leppanen
11-Jul-2008, 17:54
Don Hutton and I, together with a friend from Atlanta who owns a P45 back, did some comparison testing a few weeks ago that involved Mamiya 7II, Canon 1ds Mark III, 4X5 and the P45 back on a Contax.I am very interested in the results of this, too. What film stock(s) did you use for the test?

sanking
11-Jul-2008, 18:04
We discussed the issue of optical printing versus scanning the film. However, since we have access to a good drum scanner (Howtek 4500) that can resolve about 80 lines/mm I think it will be a fair comparison. If you have ever tested film you know that it is pretty difficult to actually put more than 80 lines/mm on film. I have seen tests that show some of the Mamiya lenses resolving over 100 lines/mm on Tmax-100 but I have not been able to get those kind of numbers with my Mamiya 7II lenses.

So I think the Howtek 4500 will pull out about all of the detail in the film negatives, at least from the MF negatives. Don actually tested a Leica with 50mm aspheric lens that day and it looks like the resolution for it is over 150 lines/mm. The results amazed me, but to pull all of the data from the negative he may have to get Lenny Eiger to scan with his 8000 spi Premier.

Sandy King


Sounds very interesting Sandy, especially as are going to compare prints made from the various formats rather than just looking at pixels on a screen.

I understand the reasons for wanting to scan the film, if it's colour, but doesn't that put it at a disadvantage in a test like this? Film has the advantage over a digital file that it can be printed optically; would the film based prints be shown to their full/better potential by enlarging them rather than scanning them?

sanking
11-Jul-2008, 18:06
I am very interested in the results of this, too. What film stock(s) did you use for the test?


We used Fuji Acros, an ASA 100 B&W film, which I find slightly sharper than Tmax-100, and Kodak Portra 160 VC, which we rated at EFS 100.

Sandy King

Mattg
11-Jul-2008, 19:07
We discussed the issue of optical printing versus scanning the film. However, since we have access to a good drum scanner (Howtek 4500) that can resolve about 80 lines/mm I think it will be a fair comparison. If you have ever tested film you know that it is pretty difficult to actually put more than 80 lines/mm on film. I have seen tests that show some of the Mamiya lenses resolving over 100 lines/mm on Tmax-100 but I have not been able to get those kind of numbers with my Mamiya 7II lenses.

So I think the Howtek 4500 will pull out about all of the detail in the film negatives, at least from the MF negatives. Don actually tested a Leica with 50mm aspheric lens that day and it looks like the resolution for it is over 150 lines/mm. The results amazed me, but to pull all of the data from the negative he may have to get Lenny Eiger to scan with his 8000 spi Premier.

Sandy King

I'm not sure I follow your logic Sandy. I can see that using a very good drum scanner might let you scan the film at a resolution approaching the best the lens can deliver but it is still sampling the film in a regular fashion and then printing it in that same regular pattern.

Wouldn't there be qualitative differences if you printed optically, with different sources of losses in resolution? Put it this way, I'd love to see you compare the scanned results to some optically printed results and see them assessed for more than just a numerical measurment of resolution. Easy to suggest when someone else is doing all the work I know but there it is.

sanking
11-Jul-2008, 19:33
I see your point, but giving a really complete answer is very complicated.

To be brief, my opinion is that film (of about same size as digital sensor size) is at a considerable disadvantage vis-a-vis digital capture unless you scan the film and process it as a digital file. This depends highly on output and at smaller print sizes film printed optically will clearly win. However, as print size increases the capability of applying digital processing to film scans more than makes up for the sampling pattern imposed by the scanner.

To say nothing of the fact that optical enlargement and wet processing of color prints is a skill almost gone with the wind.

Sandy




I'm not sure I follow your logic Sandy. I can see that using a very good drum scanner might let you scan the film at a resolution approaching the best the lens can deliver but it is still sampling the film in a regular fashion and then printing it in that same regular pattern.

Wouldn't there be qualitative differences if you printed optically, with different sources of losses in resolution? Put it this way, I'd love to see you compare the scanned results to some optically printed results and see them assessed for more than just a numerical measurment of resolution. Easy to suggest when someone else is doing all the work I know but there it is.

Mattg
11-Jul-2008, 19:54
I see your point, but giving a really complete answer is very complicated.

To be brief, my opinion is that film (of about same size as digital sensor size) is at a considerable disadvantage vis-a-vis digital capture unless you scan the film and process it as a digital file. This depends highly on output and at smaller print sizes film printed optically will clearly win. However, as size increases the capability of applying digital processing to film scans more than makes up for the sampling pattern imposed by the scanner.

To say nothing of the fact that optical enlargement and wet processing of color prints is a skill almost gone with the wind.

Sandy

How about B+W, could you make some enlargements using top notch APO glass to compare with your digitally output prints? My gut feeling is that they would not appear nearly as sharp but might actually retain as much detail. What about unsharp masking and other processing techniques; if we're happy to use them in the computer, why not use them in an optical process and compare the results?

Sorry to harp on this Sandy but it has been a beef of mine with many film vs digital capture comparisons that they don't actually compare the whole process of each type of capture. Rather they compare one type of digital capture with another type of digital capture. As you said, there is little point following through an optical process with colour materials, but a comparison in B+W might be interesting. Especially if all the tools available to the optical printer are used just as those available to the digital printer are.

David A. Goldfarb
11-Jul-2008, 19:59
Seems to be plenty of wet processing of color prints going on, even if the enlargement is digital.

Don Hutton
11-Jul-2008, 20:27
Matt

I've done comparisons of 4x5 negs I've printed optically with a very nice set-up (Devere 5108 in perfect alignment with a Schneider 150 Apo Comp) and the same images drum scanned and printed digitally - I'd agree with Sandy - the benefits start to disappear as you get to big enlargements because of the advantages of extra control etc. with the digital route appear to more than offset any perceived disadvantages of the sampling process through the scanner and printing. I'm pretty convinced at this point that I can make better prints with a hybrid - film to scan and inkjet process now than I have ever been able to manage in the darkroom. A big part of that is the ability to capture a very substantial amount of what is available on the film at the scanning stage.

Don Hutton
11-Jul-2008, 20:31
Seems to be plenty of wet processing of color prints going on...I'd love to know what "plenty" means in a quantitative sense, or at least outside of APUG - I'd bet that the volume of large color prints being printed optically is not 1/500th of what it was 10 years ago.

Mattg
11-Jul-2008, 20:32
Matt

I've done comparisons of 4x5 negs I've printed optically with a very nice set-up (Devere 5108 in perfect alignment with a Schneider 150 Apo Comp) and the same images drum scanned and printed digitally - I'd agree with Sandy - the benefits start to disappear as you get to big enlargements because of the advantages of extra control etc. with the digital route appear to more than offset any perceived disadvantages of the sampling process through the scanner and printing. I'm pretty convinced at this point that I can make better prints with a hybrid - film to scan and inkjet process now than I have ever been able to manage in the darkroom. A big part of that is the ability to capture a very substantial amount of what is available on the film at the scanning stage.

Thanks Don,
what are the beneits at smaller sizes with optical printing and at roughly what sort of reproduction ratio do the digitally processed prints start to look superior?

Don Hutton
11-Jul-2008, 20:44
Matt

These comparisons are highly subjective - what looks better to one might not to the next. At smaller reproduction ratios, I'd suggest that the subjectivity is bigger. I don't imagine I'm an awesome silver print maker - merely competent by my own standards and those likely differ from the experience of others.

Despite your investment of skills and equipment etc. in your chosen process, I'd strongly encourage anyone who has any doubts or questions of the merits of achieveing the same goals through a different to process to fully investigate it personally. A couple of weeks ago I was considering whether 8x10 vs 4x5 had any merit for prints under 40x50. I did some tests and discovered that, for my purposes, it did. We are all going to draw the line at a different place. Make the best possible print you can from a negative you understand and like optically and then send it off to be done digitally at the same size and see what you think.

Mattg
11-Jul-2008, 20:55
Matt

These comparisons are highly subjective - what looks better to one might not to the next. At smaller reproduction ratios, I'd suggest that the subjectivity is bigger. I don't imagine I'm an awesome silver print maker - merely competent by my own standards and those likely differ from the experience of others.

Despite your investment of skills and equipment etc. in your chosen process, I'd strongly encourage anyone who has any doubts or questions of the merits of achieveing the same goals through a different to process to fully investigate it personally. A couple of weeks ago I was considering whether 8x10 vs 4x5 had any merit for prints under 40x50. I did some tests and discovered that, for my purposes, it did. We are all going to draw the line at a different place. Make the best possible print you can from a negative you understand and like optically and then send it off to be done digitally at the same size and see what you think.

Thanks Don, that's good advice.

Oren Grad
11-Jul-2008, 21:15
I'd love to know what "plenty" means in a quantitative sense, or at least outside of APUG - I'd bet that the volume of large color prints being printed optically is not 1/500th of what it was 10 years ago.

Don, I think David's point was that in this brave new hybrid world, a lot of traditional chromogenic print materials are being used for digital enlargements, e.g. via laser printing in a Durst Lambda or Theta.

David A. Goldfarb
11-Jul-2008, 21:38
Don, I think David's point was that in this brave new hybrid world, a lot of traditional chromogenic print materials are being used for digital enlargements, e.g. via laser printing in a Durst Lambda or Theta.

Yes, that was my point--Lambda, Theta, LightJet, Chromira, and then all the minilabs that still print wet with LED printers to RA-4 paper, because for volume printing, RA-4 is way faster and less costly than inkjet.

Don Hutton
11-Jul-2008, 22:11
Yes, that was my point--Lambda, Theta, LightJet, Chromira, and then all the minilabs that still print wet with LED printers to RA-4 paper, because for volume printing, RA-4 is way faster and less costly than inkjet.
I'm sure for volume printing of smaller prints for it is. The topic of this post is really concerned with quality as opposed to volume, which is why it hadn't honestly crossed my mind that the automated RA4 stuff was at all relevant. I think (but could stand corrected) that Sandy was alluding to the fact that the considerable skill and application involved in producing a high quality - i.e. exhibition quality - color wet darkroom print was a disappearing art. I'm certain that there will be a never ending stream of pimpled minilab operators who will be producing plenty RA4 prints for some time to come, despite their lack of knowledge or skill surrounding the production of a fine print.

Gene McCluney
11-Jul-2008, 22:18
Every WalMart supercenter has a Fuji Frontier printing laser-exposed RA-4 paper. While the quality of these prints is not "pro" quality, the volume of paper is high enough that I think we can say that "so far" the RA-4 Color paper manufacturing is secure for the forseeable future. The bulk of mini-lab printed digital files is onto laser-exposed RA-4 paper.

Don Hutton
11-Jul-2008, 22:24
Gene

I really don't think anyone is disputing that. I just don't understand why it gets brought up in a post about ultimate large print quality.

sanking
11-Jul-2008, 22:24
Every WalMart supercenter has a Fuji Frontier printing laser-exposed RA-4 paper. While the quality of these prints is not "pro" quality, the volume of paper is high enough that I think we can say that "so far" the RA-4 Color paper manufacturing is secure for the forseeable future. The bulk of mini-lab printed digital files is onto laser-exposed RA-4 paper.


Yes, that is true.

But my original point was that very few labs are capable of high quality *optical* printing of wet processed color.

If you want to print at WalMart or Costco you start with a digital file. Or maybe they process your film and give you digital files on CD.

There are exceptions, of course. Elevator Digital in Toronto still has the capability to print color optically from color anaogue material, and very high quality at that. But there are not many Elevator Digitals in North America.

Sandy King

Eric Leppanen
11-Jul-2008, 22:46
Make the best possible print you can from a negative you understand and like optically and then send it off to be done digitally at the same size and see what you think.That's exactly what I did a year or so ago, to compare the merits of 4x5 versus 8x10 film, analog versus digital prints, at a 16x20 inch print size. I picked 16x20" because it was the largest size that I could get my B&W negs printed commercially using analog processes, and typically I don't require B&W prints larger than that size anyway.

As a subject I used the Methodist church at the ghost town of Bodie, CA. This is a two-story, unpainted, highly weathered wooden structure with lots of fine detail (wood grain, nail heads, etc.). I initially photographed this structure with 4x5 (150mm Sironar-S lens, Fuji Acros processed in Xtol) and later with 8x10 (300mm Sironar-S lens, HP5+ processed in Xtol). I then had 16x20" selenium-toned, fibre-based analog prints made from each sized negative by the proprietor of my local B&W custom print lab (who has been in the business for many years and is considered by the southern california photography community to be an excellent B&W printer). I then sent the negatives to West Coast Imaging, where they were scanned by their Tango drum scanner (300MB file size, 8-bit) and made into 16x20" prints using an Epson 9800 printer with K3 ink set and Museo Silver Rag paper.

I then took all four prints back to my local B&W lab, spread them out on a table without any identification, and asked the various photographers passing by to rank the prints by sharpness, tonality, etc.

By a broad consensus, the results were as follows:

- The 8x10 analog print was clearly superior to the 4x5 analog print, as expected.
- The 8x10 digital print upon close examination showed a slight improvement in resolution and tonality versus the 4x5 digital print, but the difference was subtle and would probably not be noticed when placed behind glass.
- The 8x10 analog print was perhaps a smidge softer than the 8x10 digital print, but it had slightly better tonality, a better "continuous tone" look, and ever so slightly deeper blacks. On balance the 8x10 analog print was considered the best of the bunch. However, these differences were sufficiently subtle that few folks would have noted the difference when mounted behind glass.

Our conclusion at the time was that, while 8x10-based analog prints yielded the absolute best results by a small degree, 4x5-based digital prints were the most practical solution for a 16x20" print size.

These results were consistent with previous tests I had done with color film stock. Basically a 16x20" print size was the inflection point beyond which 4x5 and 8x10-based digital prints started to diverge in terms of noticeable quality differences.

The Mamiya 7 is an interesting case in that, in my experience, 16x20" digital prints made from Mamiya 7 chromes typically had comparable perceived resolution and sharpness versus 4x5-based digital prints. Where 4x5 often won out by a small margin was in tonality; some 4x5-based digital prints seemed to have slightly richer, deeper tones and a more three-dimensional appearance versus their M7-based counterparts. But these differences were often subtle and heavily influenced by subject matter. However, at larger digital print sizes, the Mamiya 7 runs out of gas, at least with color subjects (I never really used my Mamiya 7 with B&W). A 24x30" Mamiya 7-based color digital print would still have excellent perceived sharpness, but its tonality was clearly not on par with LF.

This is why I am curious about the MF digital backs. Can I achieve top notch 24x30" color digital prints (my personal benchmark) with an MF-based platform? Some folks such as photographer Charles Cramer feels that a P45+ yields results comparable to 4x5 up to a 24x30" digital print size using his Mamiya 645-based platform, but other folks are not as optimistic.

I used to think that the next-generation of digital backs (50+ megapixels) would close whatever remaining quality gap there is with 4x5 film. But now I am concerned that, for this additional resolution to be useful, the required f-stop settings would be so restrictive (to avoid lens resolution loss due to diffraction) as to severely limit landscape photography applications (I can't get everything in focus af f/8!). So either the MF digital back must use camera movements to keep the f-stops reasonable (losing its main convenience advantage versus LF film), or be limited to applications where extensive DOF is not critical.

If this analysis is correct, then the P45+ and its ilk may be as good as it gets as far as digital capture is concerned for landscape applications, at least for the immediate future. I am very curious to see how well it performs.

wnw
11-Jul-2008, 23:38
The replies are enlightening (despite a few going off message) so thanks to everyone. Sorry to Mr Brepsom for my small typo. I admit to being less than perfect.

With regard to 'digital gizmos', and my original point, I should explain that I see 99% of 35mm digital images enhanced by clever and ever increasing numbers of little Photoshop actions (I do not refer to digitisation in general as should have been fairly obvious hopefully), without which the original RAW file as captured in the field remains flat, has unbelievably low dynamic range and needs every bit of software processing to bring it up to even a basically acceptable standard. The proliferation of PS seminars and courses seems to prove this point, the focus of which completely centers around the 'Wow, you should see this new PS action I've got, want to trade it?" philosophy to photography. Forums everywhere dealing with 35mm digital are loaded with cries of despair from photographers the world over about the time they need to spend on post processing. This is the 'gizmo' factor to which I refer.

Hasselblad, even in MF, seem to follow this route for software corrections which are necessary arising from lenses which cannot resolve the power of their digital chips: color aberration, distortion, vignetting and moiré and a Hassy lens isn't exactly a cheap piece of glass.

It is to look for an alternative to 35mm and MF lenses and the necessity of making corrections in software during post processing to cover those lens problems that my enquiry into view cameras arose.

Having said that I seem to glean from the posts that some MF lenses such as some from Mamiya can indeed resolve the power available from these large chips. Is this so anyone? Anyone have any info. If so, which lenses in MF are up to the standard from Schneider and Rodenstock?

I have no burning desire to lug around a field camera if MF digital will give me the top quality I am after. I take on board the very helpful comments with thanks to Sandy and the others especially regarding the quality of the scan being important to the final result.

Many thanks and also your your patience.

Aender Brepsom
12-Jul-2008, 00:45
Don Hutton and I, together with a friend from Atlanta who owns a P45 back, did some comparison testing a few weeks ago that involved Mamiya 7II, Canon 1ds Mark III, 4X5 and the P45 back on a Contax. We tested both a real scene and a constructed scene with resolution target. The goal is to make prints at a number of different sizes and get together and compare output, and hopefully write the test up for publication. I really believe that this will be a much more valid comparison than the one at LL.

At this point I have seen the film scans (Fuji Acros and Portra 160) of the 6X7 cm and
4X5, which were drum scanned with a Howtek 4500, but not the files from the P45. However, my understanding is that there is no comparison between the 4X5 and the P45, and that in terms of detail the MF scans are much better than the P45. 4X5 clearly beat the 6X7 cm. How this compares with real images on paper remains to be seen.

Sandy King

How good did the results from the 1Ds Mark III appear in your test?

Emmanuel BIGLER
12-Jul-2008, 01:54
I'm coming late to this discussion and to answer some initial questions, if we look at the ultimate difference between direct digital capture and film capture, it happens that the detection quantum efficiency of an amateur-grade colour-bayer silicon sensor is already in the range of 20%, whereas the best colour films do not exceed a few percent. If we look at the progress in colour film between, say, Kodachrome 25 and the last generation of ISO 100 fine-grain slide films, we have a gain in efficiency of about 4x at the same level of grain and resolution. But silicon is already much better in terms of quantum efficiency and professional silicon detectors used for military and scientific applications already exceed 80% of quantum efficiency. It is unlikely that futher developments with films will increase the present performances significantly, so we have to live with the already superb films we have available.

In terms of ultimate noise, it means that you need a much bigger surface of film with big pixels capturing many, many more photons in order to get something equivalent to a medium format digital sensor in terms of signal to noise ratio and total number of pixels in the image. The actual ratio in size between a silicon chip and the required surface of film that will, eventually, deliver the same quality in terms of noise and resolution is probably between 4x and 16x in terms of surface, i.e. about 2x to 4x in terms of diagonal size. This is fuzzy since it depends as mentioned on the efficiency of the sensor, the quality of the scanner and all processing tricks that cheat with edge sharpness and pixel averaging to reduce noise.

However I do not believe that small-sized digital sensors smaller than 18x24 mm will ever be able to compete with LF film simply because when you scale down an optical design, the F-number of the lens has to increase in order to get the same effect of diffraction. Lens designers have made a superb effort to bring to the market digital view camera lenses with a max aperture of 4.5 and a best f-stop around F/8 for use with 39x47mm sensors. Those lenses are close to the diffraction limits at the cenrte of the field and cannot really be much improved. If you extrapolate to a centimeter-size sensor you'd have to provide a lens with max apêrture at f/1 (simply apply the ratio of the diagonals in order to keep the absolute iris diameter constant for keeping diffraction effects constant) and best aperture at f/2 or something which sounds more difficult technically to obtain that a f/8 lens covering a larger format with bigger pixels.

This for the ultimate physical limits.

Now for practical purposes, there are as I can see it as an amateur, many advantages tu use medium or large format film if you do not have to produce hundreds of top-quality images per day.
The quality of modern colour negative films is incredible, their ability to record and compress a range of about 11-12 f-stops in the initial scene is unique.
And with film you do not have to care for maintaining a digital image archive. Just keep your originals in a safe place and eventually re-scan them when your CDs self-erase of when your redundant arrays of inexpensive hard disks crash all at the same time some day !! ;)
Again, using film can be "viable" from an economic point of view only if you aim at a small number of outstanding images produced per year. Whether this is enough to make a living for a professional photographer in 2008 is another issue. Very few fine-art photographes can make a living selling only images recorded with LF cameras on LF film.
For regular day-to-day production, delivered to regular customers formatted by a 625-line TV image quality ;), my understanding is that various small-format digital cameras do such a good job that many professional photographers stopped using their view camera !!

But as an amateur, I am perfectly happy with modern film, since I do not have to serve any customer, I do not care for any return-on-investment, since I just spend my money on leisure activities ;)
So for me, film is OK, the choice & assortment is still comfortable. I can have access to a professionnal scanner if I need it about once a year I want to scan the image-of-the-century. And I can still print them myself with my enarger or with an amateur-grade scanner+printer in the interim..

audioexcels
12-Jul-2008, 02:34
As a subject I used the Methodist church at the ghost town of Bodie, CA. This is a two-story, unpainted, highly weathered wooden structure with lots of fine detail (wood grain, nail heads, etc.). I initially photographed this structure with 4x5 (150mm Sironar-S lens, Fuji Acros processed in Xtol) and later with 8x10 (300mm Sironar-S lens, HP5+ processed in Xtol). I then had 16x20" selenium-toned, fibre-based analog prints made from each sized negative by the proprietor of my local B&W custom print lab (who has been in the business for many years and is considered by the southern california photography community to be an excellent B&W printer). I then sent the negatives to West Coast Imaging, where they were scanned by their Tango drum scanner (300MB file size, 8-bit) and made into 16x20" prints using an Epson 9800 printer with K3 ink set and Museo Silver Rag paper.

I then took all four prints back to my local B&W lab, spread them out on a table without any identification, and asked the various photographers passing by to rank the prints by sharpness, tonality, etc.

By a broad consensus, the results were as follows:

- The 8x10 analog print was clearly superior to the 4x5 analog print, as expected.
- The 8x10 digital print upon close examination showed a slight improvement in resolution and tonality versus the 4x5 digital print, but the difference was subtle and would probably not be noticed when placed behind glass.
- The 8x10 analog print was perhaps a smidge softer than the 8x10 digital print, but it had slightly better tonality, a better "continuous tone" look, and ever so slightly deeper blacks. On balance the 8x10 analog print was considered the best of the bunch. However, these differences were sufficiently subtle that few folks would have noted the difference when mounted behind glass.

Our conclusion at the time was that, while 8x10-based analog prints yielded the absolute best results by a small degree, 4x5-based digital prints were the most practical solution for a 16x20" print size.

These results were consistent with previous tests I had done with color film stock. Basically a 16x20" print size was the inflection point beyond which 4x5 and 8x10-based digital prints started to diverge in terms of noticeable quality differences.



I'm trying to get this straight. The 8X10 analog print was clearly superior to the 4X5 analog print, but once you digitized both, the 8X10 print's virtues became nearly indistinguishable from the 4X5 prints that it made little sense to digitize the 8X10 film since 4X5 gives you 4 more shots than 8X10 does...practically and economically, shooting 4X5 gave you the better results over 8X10 unless printing larger than the size you would typically print to. Is this correct or did I miss something along the way?

sanking
12-Jul-2008, 08:31
The Mamiya 7 is an interesting case in that, in my experience, 16x20" digital prints made from Mamiya 7 chromes typically had comparable perceived resolution and sharpness versus 4x5-based digital prints. Where 4x5 often won out by a small margin was in tonality; some 4x5-based digital prints seemed to have slightly richer, deeper tones and a more three-dimensional appearance versus their M7-based counterparts. But these differences were often subtle and heavily influenced by subject matter. However, at larger digital print sizes, the Mamiya 7 runs out of gas, at least with color subjects (I never really used my Mamiya 7 with B&W). A 24x30" Mamiya 7-based color digital print would still have excellent perceived sharpness, but its tonality was clearly not on par with LF.



One way to even things out between 6X7 cm and 4X5" is to use a slower speed and finer grain film with the smaller format. In prints up to 17 X 22" I don't perceive much if any difference in tonal qualities between 6X7 cm with Fuji Acros and 4X5" with TMAX-400.

In contrast to your work, I use the Mamiya 7II primarily with B&W.

Sandy King

sanking
12-Jul-2008, 08:33
How good did the results from the 1Ds Mark III appear in your test?

I have not yet seen the results of that test yet. But the results with the 1Ds Mark III should be about as good as possible since the lens used was a 50mm Leitz aspheric glass.

Sandy King

Eric Leppanen
12-Jul-2008, 09:36
I'm trying to get this straight. The 8X10 analog print was clearly superior to the 4X5 analog print, but once you digitized both, the 8X10 print's virtues became nearly indistinguishable from the 4X5 prints that it made little sense to digitize the 8X10 film since 4X5 gives you 4 more shots than 8X10 does...practically and economically, shooting 4X5 gave you the better results over 8X10 unless printing larger than the size you would typically print to. Is this correct or did I miss something along the way?You didn't miss anything. Typically digital printing allows one to print at a larger print size than would otherwise be possible with analog prints, without a corresponding loss of quality. Some folks (like David Muench, among others) have said that digital printing allows one to "gain a film format" as far as enlargeability is concerned.

At a 16x20" print size, the analog prints were already showing significant differences between 4x5 and 8x10, while the digital prints were largely indistinguishable. If the print size were increased beyond 16x20", the 4x5 and 8x10 digital prints would also start to diverge. Please note that the point of divergence is influenced by a variety of variables, including subject type, film stock, etc. The subject I chose was very "detail intensive"; other B&W subjects would show little noticeable difference between 4x5 and 8x10, even between 16x20" analog prints.

I cannot speak to the B&W performance differences between 6x7 and 4x5. In 6x7 I shot almost entirely color (mostly Provia), which cannot resolve to the level of Acros or TMX.

gari beet
13-Jul-2008, 00:50
Regarding the weight issue, it is worth bearing in mind the amount of ancilliary gear you may need to carry such as leads and batteries/chargers etc. This would of course depend how and what you are shooting, wether you are out for extended periods or close to the car etc.
As an example, I went to Norway a few years ago shortly after buying my first LF. I had at the time a Nikon D70 and 3 lenses, when packed I compared the weights of the digital camera with chargers and image banks etc and it was not that short of the 5x4 with a couple of lenses, meter, filters etc. I took the 5x4 in the end. My current kit of Tachi and 4 lenses, meter, filters and holders is maybe 20/30lb with the tripod.

I have no direct experience but have read that there is also an issue of colour fringing with digital backs when movements such as swing and tilt are used, depending on what you are shooting this may have a bearing also, it seems that it is an issue with LF and MF sensors but not APS or FULL FRAME.

So what is it you will be shooting?

Gari

bob carnie
13-Jul-2008, 06:22
Hi Sandy

I will do output on Lambda at 400ppi 8bit and Cannon Inkjet 16bit for your tests on high gloss stock at 30 x40 and larger magnification.
give me a call if you like.

Bob

Don Hutton and I, together with a friend from Atlanta who owns a P45 back, did some comparison testing a few weeks ago that involved Mamiya 7II, Canon 1ds Mark III, 4X5 and the P45 back on a Contax. We tested both a real scene and a constructed scene with resolution target. The goal is to make prints at a number of different sizes and get together and compare output, and hopefully write the test up for publication. I really believe that this will be a much more valid comparison than the one at LL.

At this point I have seen the film scans (Fuji Acros and Portra 160) of the 6X7 cm and
4X5, which were drum scanned with a Howtek 4500, but not the files from the P45. However, my understanding is that there is no comparison between the 4X5 and the P45, and that in terms of detail the MF scans are much better than the P45. 4X5 clearly beat the 6X7 cm. How this compares with real images on paper remains to be seen.

Sandy King

bob carnie
13-Jul-2008, 06:25
Sandy and Don

You both can come to Toronto, and do optical and digital prints for your test, we have room and would love to see you again Sandy,, Don I am not so sure about, (does he drink Errol Flynns)

Bob

I'm not sure I follow your logic Sandy. I can see that using a very good drum scanner might let you scan the film at a resolution approaching the best the lens can deliver but it is still sampling the film in a regular fashion and then printing it in that same regular pattern.

Wouldn't there be qualitative differences if you printed optically, with different sources of losses in resolution? Put it this way, I'd love to see you compare the scanned results to some optically printed results and see them assessed for more than just a numerical measurment of resolution. Easy to suggest when someone else is doing all the work I know but there it is.

bob carnie
13-Jul-2008, 06:39
After reading this whole thread, my offer stands,

Sandy , Don, and for that matter anyone interested can come to TO for a weekend.
We have a studio with lights, electronic flash and hotlights.
I will make available a wet sink to process large prints wet from analogue and digital.
I have 4x5 - 11x14 enlargers capable of 30x40 prints , also a horizontal to do 8ft prints. Enlargers are Deveere and Durst .
As well 60inch Cannnon 9000 to test large prints as well axcess to the new Epson.
The Lambda can output to 30inch paper flex or gloss or matt to any magnification we would want.

What I do not have is all the cameras mentioned but those wanting to test can bring their own.
I have a Imocan Scanner which is being replaced this summer season so any scans beyond Imocan quality should be brought as well.

Sandy is quite familiar with our facilities and most welcome to do these tests here, I totally trust his judgement .

sanking
13-Jul-2008, 08:07
Hi Sandy

I will do output on Lambda at 400ppi 8bit and Cannon Inkjet 16bit for your tests on high gloss stock at 30 x40 and larger magnification.
give me a call if you like.

Bob

Hi Bob,

Thanks for the offer. I will have to discuss this with the other two participants in the test but I am definitely interested. One of the things I wanted to include in the test was output by an independent third party, and your offer appears to cover that.

Sandy King

Don Hutton
13-Jul-2008, 08:28
Sounds like an excellent excuse to visit Toronto! I'd also be very interested to see your set-up Bob - I've only heard fabulous reports about it and the work done in it.

bob carnie
13-Jul-2008, 08:38
Lets discuss this and make it happen then .

Elevator would happy to be a third party on this as a lot of the questions , suggestions , have been floating around in my head, I have tried to do some of the tests myself but it really is a monster task to cover all angles and as well be non biased.
I know I can get some support on materials from the manufactures of paper , chems and inks *digital and analogue*. I have no pull on capture devices, but since we print 7days a week I know I can get some consumables for these tests.

Ted Harris would have loved to be involved with this as we discussed a lot of these issues when he was here in the spring.

August is not a good month , but maybe October is good as Sandy will be in Ontario on other issues.
I will leave it up to Sandy and Don to decide.

Clay Turtle
13-Jul-2008, 14:18
Was at the local (actually I had to go out of town) photo shop to pick up some Dektol & pick up a pack of paper too. On the way in I noticed an ad on the door about the new digital coming soon. As I was checking out, the patron before me started a conversation as I mentioned 35 mm film & he was still shooting film too. The salesman behind the cash register asked us about 'going digital', which we both responded with not interested. Then he asked why was it because we wouldn't consider digital?
I keep hearing these stories about how digital is better than or just as good as film so being that I really don't pay much attention to either digital backs or digital cameras, I would have to ask just how fine are the receptors in these new generation of digitals? I mean the smallest film grain is to be found with C41 processed black & white dye film (Ilford) so if we were to set it as the zero point & scale receptor to grain size would it fall in the catagory of 1) fine grain slow speed film 2) medium speed 200 ASA 3) high speed 400 ASA 4) 1600 ASA or 5) 3200 ASA film?

Mattg
13-Jul-2008, 14:40
Good on you Bob, any plans to open an Elevator Sydney?

Gordon Moat
13-Jul-2008, 15:00
. . . . . . . . . .

I have no direct experience but have read that there is also an issue of colour fringing with digital backs when movements such as swing and tilt are used, depending on what you are shooting this may have a bearing also, it seems that it is an issue with LF and MF sensors but not APS or FULL FRAME.

Microlenses or diffusion layers simply are not efficient enough guiding light beyond a certain angle (somewhat dependent upon design). There was an excellent write-up about these technology issues in LFI a few issues ago, though mostly relating to the Kodak work on the Leica M8. Each photosite is more like a well or pit, and light needs to fall (or be guided) down that well. Fringing can occur as a result of guidance issues combined with Bayer pattern interpolation errors.

Why you see this (generally) more with larger sensors is that the photosites tend to be physically larger, though this is not always a good generalization. As photosites become physically smaller, the potential optical resolution improves, but there are more likely to be colour issues from Bayer interpolation, or noise intruding upon accumulated charge at each photosite.

Possible solutions are in the works, through a variety of different approaches. In video cameras, interpolation issues can be solved by having three separate sensors, and no Bayer pattern; unfortunately cost is a huge barrier to doing that in stills cameras. Three chips also creates a separate issue in that lenses need to be more optimized to focus on each chip the same. Blue filtration can still cause the most problems, since the chips are often receiving the least light under blue filtration (much worse than green, though only slightly worse than red).

Ciao!

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

bob carnie
14-Jul-2008, 05:08
Not in the forseable future, but a road trip around Austrailia is definately in the plans.

Good on you Bob, any plans to open an Elevator Sydney?