Photo Forum / Film Photography / Darkroom / January 2005
Recalibrating film to scan?
|
|
Thread rating:  |
Josh - 09 Jan 2005 18:27 GMT Hello all again,
I figured this was still apropriate to post to R.P.Darkroom as it is about processing film. I am already familiar with the calibration of film development specifically for printing using the Zone System. I currently have mapped out TMX and TMY to get a zone VIII density of 1.35 over FB+F for darkroom printing with various exposures. Now that I have started to dabble in scanning I have realized that the total density range of my negatives is quite narrow compared to the capabilities of the scanner. I was wondering if anyone had specifically dealt with this issue and possibly come out with some target densities that take advantage of the range of the scanner without coming close to the edges of the films capabilities in Dmax.
I would assume right of the bat that since I am interested in the same shadow detail that my film speed would not change (currently 64 for TMX and 200 for TMY for me specifically). However I would think that to expand the overall density range I would have to significantly increase my development times. I would assume that TMX and TMY would remain quite linear to a Dmax of 3.0 or so, so would anyone think an appropriate calibration value for Zone VIII would be on the range of 2.4 - 2.6. This would give me difficult to print conventionally contrasty negatives, but would effectively separate out the values and allow for better information capture in a good film scanner.
Any thoughts?
(whether or not this fits in with your philosophy of "printing", please keep flames to a minimum - I decided I am going to TRY this anyway, I am not throwing out my darkroom and if I am not happy I can always go back. )
-Josh
winddancing - 09 Jan 2005 20:19 GMT Wouldn't it be useful to know which scanner, and software; and then the output intentions?
Gregory Blank - 09 Jan 2005 22:33 GMT > Wouldn't it be useful to know which scanner, and software; and then the > output intentions? It would be for us :-)
 Signature LF Website @ http://members.verizon.net/~gregoryblank
"To announce that there must be no criticism of the President, or that we are to stand by the President, right or wrong, is not only unpatriotic and servile, but is morally treasonable to the American public."--Theodore Roosevelt, May 7, 1918
Gregory Blank - 09 Jan 2005 20:38 GMT > Hello all again, > [quoted text clipped - 29 lines] > > -Josh IMOP nothing is going to be gained by lots of over exposure,....or over development. Once you have the information on the film (even for scanning) there really is no advantage to more density. Basically your just lengthening the time your scan takes.
 Signature LF Website @ http://members.verizon.net/~gregoryblank
"To announce that there must be no criticism of the President, or that we are to stand by the President, right or wrong, is not only unpatriotic and servile, but is morally treasonable to the American public."--Theodore Roosevelt, May 7, 1918
Josh - 10 Jan 2005 01:11 GMT Thanks for your reply.
I was figuring that if a scanner maps all the gray values to lets say 0 to 2047 throughout its Dmin to Dmax range then it would be wise to try to fill all of that space with data. If my whole negative ranges from 0 to 1.5 in density on a film scanner that can see through up to 3.0 density then I'm only occupying half the dynamic range. I do understand that the scanner software will just expand the data to fill the space based on some curve - but that's just mathematical interpolation, it seems the data would be better data if the dynamic range has to be "streched" less.
I hope this makes sense.
-Josh
Nicholas O. Lindan - 10 Jan 2005 18:24 GMT > ... dynamic range. I do understand > that the scanner software will just expand the data to fill the space > based on some curve - but that's just mathematical interpolation The scanner has a programmable gain amplifier to expand/compress the photodiode output so that it fits the input range of the A/D converter. That's why there is no picket-fencing as you get in Photoshop, the calculation is analog, is done in hardware and acts on a signal that has not yet been quantized.
 Signature Nicholas O. Lindan, Cleveland, Ohio Consulting Engineer: Electronics; Informatics; Photonics. Remove spaces etc. to reply: n o lindan at net com dot com psst.. want to buy an f-stop timer? nolindan.com/da/fstop/
Mark in Maine - 10 Jan 2005 18:50 GMT >The scanner has a programmable gain amplifier to expand/compress the >photodiode output so that it fits the input range of the A/D converter. >That's why there is no picket-fencing as you get in Photoshop, the >calculation is analog, is done in hardware and acts on a signal that has >not yet been quantized. In the photography world, I am almost 100% wet photography, but in my professional life, I am an engineer with a background in the design and use of A/D converters.
It seems to me that there is a big technical difference between colour and black and white material when scanning - and that is because with the way the dyes work in transparancy material they behave much more like a continuous image, and you would need an A/D converter with resolution similar to what the transparancy supports. If you look at a b&w negative under a microscope (the way the scanner sees it when scanning), you don't see levels of gray, you see grains which are black, on a white background, and the overall density in an area is determined by the percentage of that area which is covered in grains. To digitize this you have two options - one would be to use a sensor in the scanner which is significantly larger than the grain pattern, and use the A/D to average this out - the result would be an image in which you lose spatial resolution (edges would not be as sharp), the other alternative, and the one which I would chose if I were serious about this is to use a very small sensor (very high number of DPI) and a low resolution A/D converter.
To achieve a scan with no loss of data there is a theory that tells how small the sensor spot has to be (about half the size of the grain). Most scanners that boast very high resolution (DPI) are actually interpolating between actual samples, their ability to resolve black lines is significantly lower. The specification which is really needed is not one of the things that scanner manufacturers publish. In this case the resolution of the A/D converter is almost irrelevant.
If you were to do a scan in this way, and then generated a digital print from it, enlarging it a lot, what you would see would be the same sort of graininess that we get in the darkroom.
The previous poster mentioned the use of a variable gain amplifier (VGA) prior to the A/D to expand/compress the range of the scanner. This is done, but a more common use for the VGA is to use it to compensate for non-uniformity in the illumination intensity. During a calibration cycle, the scanner measures the bulb intensity and saves it, then during the scan it alters the A/D sensitivity to compensate for bright and dark areas. - A dandy technique, but really only valuable when scanning colour materials.
Josh - 11 Jan 2005 14:19 GMT Wow, thanks for the extremely informative response.
It seems then the best solution for the B&W film for scanning issue is to use C41 film. And in that case I may as well use color film and use the separate color channel levels to "grayscale" the image. The grain concept makes sense - grains may as well essentially be small opaque areas in clear film - if you can capture their size and location you could probably accurately represent a black and white negative with one bit (given enough resolution) theoretically. I would think that scanning even at 4000 DPI (Nikon 9000) would not be enough resolution to accurately depict all the grains of a 6x7 negative. This answers a lot. Thank you.
-Josh
Tom Phillips - 13 Jan 2005 14:17 GMT > Wow, thanks for the extremely informative response. > > It seems then the best solution for the B&W film for scanning issue is > to use C41 film. I don't think that's a correct conclusion to draw at all, unless you're scanning transparency films. Color negative films have a fairly low Dmax similar to B&W. Should make no difference.
> And in that case I may as well use color film and use > the separate color channel levels to "grayscale" the image. The grain > concept makes sense - grains may as well essentially be small opaque > areas in clear film - if you can capture their size and location you > could probably accurately represent a black and white negative with > one bit (given enough resolution) theoretically. Grain patterns for all practical purposes represent continuous tone. You are not seeing individual grains, but pragmatically connected clumps of grains. Also, tT the typical scanning resolution used for digital print output is no where near that of a microscope view.
> I would think that > scanning even at 4000 DPI (Nikon 9000) would not be enough resolution > to accurately depict all the grains of a 6x7 negative. Granularity depends on the film, speed, development, and type (as in t-grain.) Also negative size regarding either enlargement or scanning resolution. 6x7 is a fairly large negative.
I've typically had my 4x5s scanned at resolutions up 6000x4000. There is no visible grain. The idea is not to scan at microscopic resolutions, but a resolution matched to the intended output...
> This answers a lot. Thank you. > > -Josh Tom Phillips - 13 Jan 2005 14:18 GMT > >The scanner has a programmable gain amplifier to expand/compress the > >photodiode output so that it fits the input range of the A/D converter. [quoted text clipped - 10 lines] > because with the way the dyes work in transparancy material they > behave much more like a continuous image, Actually dyes (dye clouds) in transparency films will mirror the grain patterns (clumps) of the original RGB silver layers. One will not appear/ behave more continuous than the other.
uraniumcommittee@yahoo.com - 09 Jan 2005 23:15 GMT This is insane.
Josh - 10 Jan 2005 01:05 GMT Tom Phillips - 10 Jan 2005 05:41 GMT > Hello all again, > [quoted text clipped - 9 lines] > that take advantage of the range of the scanner without coming close to > the edges of the films capabilities in Dmax. The advantage of scanning negative films is it's low D-max threshold. If you start bumping up the negative's contrast and Dmax, you're going to start losing detail in the scan. As it is, the dmax of negative films is well within the signal to noise ratio of even cheap scanners (s/n is what counts.) All you have to do is make contrast adjustments (curves or histogram) to the scan either when scanning or in photoshop after scanning.
> I would assume right of the bat that since I am interested in the same > shadow detail that my film speed would not change (currently 64 for TMX [quoted text clipped - 15 lines] > > -Josh Han Schutten - 10 Jan 2005 10:24 GMT > The advantage of scanning negative films is it's low > D-max threshold. If you start bumping up the negative's [quoted text clipped - 4 lines] > make contrast adjustments (curves or histogram) to the > scan either when scanning or in photoshop after scanning. I can back this. When you expand the density range of your negatives, you'll get the same problems as with scanning slides. It will be difficult to get all the detail from the dense regions of the negative and the contrast range will be more than your display or paper can handle. If your scanner gives you 2047 gray values, as you posted, there is enough head space to adjust the histogram and go back to 8 bits in the end.
Robert Feinman - 10 Jan 2005 14:20 GMT > Hello all again, > [quoted text clipped - 29 lines] > > -Josh I have some tips on scanning B&W film in the tips section of my web site. In general it's less demanding that transparencies. Most modern scanners will not have any technical limitations on density, but some of the provided scanning software clips values and causes posterization so third-part software like Vuescan or Silverfast may be worth a try. I think if your negative prints well it will scan well.
 Signature Robert D Feinman Landscapes, Cityscapes and Panoramic Photographs http://robertdfeinman.com mail: robertdfeinman@netscape.net
Alan Smithee - 11 Jan 2005 20:57 GMT You're on the right track. What you really need to establish is, can your scanner live up to it published specs. My scanner says it can go to 3.4 DMax but when actually put to the test it's more like 2.8 and more noisey at the deep dense end of the neg. The second thing to consider is that scenes will vary. Are you shooting on a 4X5? If so varying the dev time is easy to adjust per frame, 35mm not so easy. Another consideration is the scanning software itself. Vuescan can allow you to pump more light through the neg using the exposure control so a dense neg can be allowed for. Hell, give it a try and let us know how it worked!
> Hello all again, > [quoted text clipped - 29 lines] > > -Josh
|
|
|