swmcl
24-Jan-2015, 19:17
Hello all,
I have recently done a film test and in the interests of open-sourcing the data and checking my thinking against that of the luminaries it is attached in a pdf.
What I've done is shoot a roll of film and I've also sampled a Stouffer step wedge. The results are in the one-page pdf. The actual shoot data is presented alongside all my conclusions regarding gradients etc.
These are my suppositions that I would like discussed:
1. The film speed is in actuality around 64 ISO whereas I shot it for an anticipated 100 ISO.
2. Not only would I shoot at 64 ISO, I would be well advised to keep the lowest textures at no lower than -2.5 zones below the mid-tone zone of Zone 5. The film is recording tones at 3 stops below the mid tone of Zone 5 but textures are a different story.
3. This film and this developer are producing a pretty linear result to at least 9 stops above Zone 5 - a density of just above 2.0.
4. The scanned step wedge has a useful optical density range from around 0.4 to not much more than 2.0. This corresponds to around -2.5 stops through to +9 stops on the film.
5. If this film were used to optically create a print I am able to print from say -2.5 stops through say +4 or +5 stops on a condenser enlarger but a little more on a diffusion enlarger. If I were to scan the film and print digitally there is more information on the film from say +4 or +5 stops through to +9 stops that could be digitally processed and printed. The scanning could give me another 4 stops worth of information.
6. If the developer I used could linearly process a film at a lower gradient out to the same final densities then I could jam even more into a scan. This data is showing a clear 9 + 2.5 = 11.5 stops but perhaps it could go another 3 stops which would make an SBR of 13 to 14 stops recorded on the film to be able to be scanned on the scanner. To my mind, a digital print will contain the more information. Whether or not this is important is another matter altogether. Aesthetics aside, the hybrid workflow will produce the most information in a final print.
The scanner was an Epson v700 with Vuescan and all settings neutralised I believe. I hope it is a straight scan.
I'm most interested in the facts here, not whether or not the final print would be good to look at. As I understand it, if one were to shoot in the anticipation of an optical process and for some reason one were to use the film in a hybrid workflow, one could be assured that the decision to process for the optical workflow is not one that sufficiently degrades this hybrid workflow. It would be a case of scanner to the rescue to some extent.
I'm not sure why the pixel values are being expressed as 8-bit values in Photoshop. I tried to set a full-blooded 16-bit scan in Vuescan.
Whether or not a printer can make the fine changes of densities that I might be able to read in a scan is another matter altogether. It may be that the printer can't do any better than the optical process anyway which means the film retains the full image data although only a part of that data can ever be seen.
I would appreciate any comments to address my misunderstandings.
Cheers,
Steve
I have recently done a film test and in the interests of open-sourcing the data and checking my thinking against that of the luminaries it is attached in a pdf.
What I've done is shoot a roll of film and I've also sampled a Stouffer step wedge. The results are in the one-page pdf. The actual shoot data is presented alongside all my conclusions regarding gradients etc.
These are my suppositions that I would like discussed:
1. The film speed is in actuality around 64 ISO whereas I shot it for an anticipated 100 ISO.
2. Not only would I shoot at 64 ISO, I would be well advised to keep the lowest textures at no lower than -2.5 zones below the mid-tone zone of Zone 5. The film is recording tones at 3 stops below the mid tone of Zone 5 but textures are a different story.
3. This film and this developer are producing a pretty linear result to at least 9 stops above Zone 5 - a density of just above 2.0.
4. The scanned step wedge has a useful optical density range from around 0.4 to not much more than 2.0. This corresponds to around -2.5 stops through to +9 stops on the film.
5. If this film were used to optically create a print I am able to print from say -2.5 stops through say +4 or +5 stops on a condenser enlarger but a little more on a diffusion enlarger. If I were to scan the film and print digitally there is more information on the film from say +4 or +5 stops through to +9 stops that could be digitally processed and printed. The scanning could give me another 4 stops worth of information.
6. If the developer I used could linearly process a film at a lower gradient out to the same final densities then I could jam even more into a scan. This data is showing a clear 9 + 2.5 = 11.5 stops but perhaps it could go another 3 stops which would make an SBR of 13 to 14 stops recorded on the film to be able to be scanned on the scanner. To my mind, a digital print will contain the more information. Whether or not this is important is another matter altogether. Aesthetics aside, the hybrid workflow will produce the most information in a final print.
The scanner was an Epson v700 with Vuescan and all settings neutralised I believe. I hope it is a straight scan.
I'm most interested in the facts here, not whether or not the final print would be good to look at. As I understand it, if one were to shoot in the anticipation of an optical process and for some reason one were to use the film in a hybrid workflow, one could be assured that the decision to process for the optical workflow is not one that sufficiently degrades this hybrid workflow. It would be a case of scanner to the rescue to some extent.
I'm not sure why the pixel values are being expressed as 8-bit values in Photoshop. I tried to set a full-blooded 16-bit scan in Vuescan.
Whether or not a printer can make the fine changes of densities that I might be able to read in a scan is another matter altogether. It may be that the printer can't do any better than the optical process anyway which means the film retains the full image data although only a part of that data can ever be seen.
I would appreciate any comments to address my misunderstandings.
Cheers,
Steve