grayscale or rgb for b+w negs

[lenshood]

longtime film shooter, just gearing up for hybrid workflow. i shoot b+w film only. i've seen occasional advice to scan b+w negs in 48-bit color (with saturation set to 0) instead of 16-bit grayscale. does that make sense? is there an advantage? (epson v800 / epson scan or silverfast)

 

[chriscrawfordphoto]

There is NO advantage to scanning BW film in color. Scan in 16 bit grayscale, using Gray Gamma 2.2 as the colorspace (if you scanner allows you to select a grayscale profile). Scanning in RGB just makes a file that is 3 times as large on your hard drive, with no additional tonal information.

 

[shutterfinger]

I scan B&W in 48bit RGB as it gives me more adjustment latitude. Others say to use just the blue channel as it gives the best B&W due to the wavelength of blue light.
With an older scanner the light source is a cold cathode fluorescent lamp so the differences between 16 bit greyscale and 48 bit RGB may not be that distinct but the V800 uses RGB Led's so the difference should be more noticeable.
The important factor here is how critical are you? Do you want good or the maximum your scanner can produce?

As for software use EpsoScan. Silverfast SE scans in 48 bit but converts to 24 bit RGB; or 16 bit grayscale to 8 bit grayscale. I think you have to upgrade to Silverfast Ai or Studio to be able to save in 16 bit per channel color (48 bit RGB) or greyscale.

 

[lenshood]

The important factor here is how critical are you? Do you want good or the maximum your scanner can produce?

shutterfinger, that’s a great question - I guess I’d say I haven’t worked with enough scans to be super critical *yet* - for the moment I’m just trying to attain basic competence. But I know that down the line I’ll be more demanding of myself & my results, so if I start off by structurally reducing my options, I know I’ll regret it.

thanks for the response

 

[lenshood]

There is NO advantage to scanning BW film in color. Scan in 16 bit grayscale, using Gray Gamma 2.2 as the colorspace (if you scanner allows you to select a grayscale profile). Scanning in RGB just makes a file that is 3 times as large on your hard drive, with no additional tonal information.

Thanks for the reply - and for the tip re: the colorspace

 

[Alan9940]

If you want the best grayscale scan and conversion, scan a 16-bit linear raw file and convert with the ColorPerfect plugin. My opinion of course, but based on 20 years experience scanning B&W film; mostly using Silverfast software.

 

[Alan Edward Klein]

I scan in 16 bit grayscale, normally 2400 bit and save as tiff. Gamma 2.2. I use Epsonscan. I only set the white and black points to just past the histogram. All other settings are turned off. All adjustments, sharpening, spotting, curves, exposure, etc are done in post processing normally with Lightroom. I've done that with my V600 and presently with my new V850. Check my Flickr page for samples of 4x5 with V850. Good luck.
https://www.flickr.com/photos/alanklein2000/albums/72157714124881023

 

[lenshood]

Thanks, all, for the wisdom. I’ll play around a bit, but I’m leaning toward grayscale.

 

[removed account4]

I only scan bw negatives and prints in color. no idea why I would want to scan anything in bw... even when I am making a digital negative on paper or film, always a color file that I start with.

 

[Andrew O'Neill]

I scan b/w negatives in 48 bit colour. Then in photoshop, I remove the red and blue channels, leaving only the green channel...which is ever so slightly sharper than the red and blue. Finally, the green channel is grayscaled. I learnt this from Ken Lee's scanning tutorial.

 

[chriscrawfordphoto]

I scan b/w negatives in 48 bit colour. Then in photoshop, I remove the red and blue channels, leaving only the green channel...which is ever so slightly sharper than the red and blue. Finally, the green channel is grayscaled. I learnt this from Ken Lee's scanning tutorial.

I don't know what scan software you use, but if you use Vuescan, you can tell the software to make the grayscale file from only one of the color channels and eliminate the need to manually do that in Photoshop.

 

[removed account4]

I scan b/w negatives in 48 bit colour. Then in photoshop, I remove the red and blue channels, leaving only the green channel...which is ever so slightly sharper than the red and blue. Finally, the green channel is grayscaled. I learnt this from Ken Lee's scanning tutorial.

yup !
no need to scan in b/w

 

[Alan Edward Klein]

I scan b/w negatives in 48 bit colour. Then in photoshop, I remove the red and blue channels, leaving only the green channel...which is ever so slightly sharper than the red and blue. Finally, the green channel is grayscaled. I learnt this from Ken Lee's scanning tutorial.

I don't use Photoshop. Is there a way to break out the green channel with Lightroom or Epsonscan? Why is the green channel sharper especially when you're scanning BW film?

 

[Ko.Fe.]

longtime film shooter, just gearing up for hybrid workflow. i shoot b+w film only. i've seen occasional advice to scan b+w negs in 48-bit color (with saturation set to 0) instead of 16-bit grayscale. does that make sense? is there an advantage? (epson v800 / epson scan or silverfast)

If you scan bw as color and process it in LR, you'll have extra tool for the exposure, contrast adjustment.
Easy to use two sliders, but It makes a lot of difference, if you need it.

I use Epson scan with Epson scanners. I have SilverFast, but it for Plustek. Epson Scan is the best scanning software among several I have used.

 

[Alan Edward Klein]

If you scan bw as color and process it in LR, you'll have extra tool for the exposure, contrast adjustment.
Easy to use two sliders, but It makes a lot of difference, if you need it.

I use Epson scan with Epson scanners. I have SilverFast, but it for Plustek. Epson Scan is the best scanning software among several I have used.

Which extra tools do I have? Where? What panel?

 

[Andrew O'Neill]

I don't use Photoshop. Is there a way to break out the green channel with Lightroom or Epsonscan? Why is the green channel sharper especially when you're scanning BW film?

I just did a quick google search, and it appears that one can eliminate channels using Lighroom. Here's the link to Ken's tutorial: http://www.kennethleegallery.com/html/scanning/scanningGreen.php

 

[Alan Edward Klein]

I just did a quick google search, and it appears that one can eliminate channels using Lighroom. Here's the link to Ken's tutorial: http://www.kennethleegallery.com/html/scanning/scanningGreen.php

I may be missing it. But it seems Lee shows only how to remove the Blue and Red channels in Photoshop. Nothing said about LR.

 

[Andrew O'Neill]

I may be missing it. But it seems Lee shows only how to remove the Blue and Red channels in Photoshop. Nothing said about LR.

You'll have to google it then.

 

[grat]

I think you have to upgrade to Silverfast Ai or Studio to be able to save in 16 bit per channel color (48 bit RGB) or greyscale.

Don't know if people know this, but if you bought a v800/850, you can upgrade to Silverfast AI Studio for about 1/3 the usual price.

 

[Ko.Fe.]

Which extra tools do I have? Where? What panel?

It is called Temp and Tilt in LR I have.

 

[alanrockwood]

Generally there be one of the channels that gives the sharpest image. The preferred channel might depend on the brand of scanner. As I recall, for my canon FS4000us scanner it is the green channel.

On the question of 8 bit vs. 16 bit, there has been a lot of discussion on that topic. The best thing to do would be to test it on a bunch of your images to see if you can tell a difference.

The concern about using 8 bit scanner seems to mostly revolve around the issue of banding in the image, especially if you do some extreme image manipulation after the scan. This is a very legitimate concern for a pure grainless image. If there is any graininess in the image (or other sources of noise in the system, such as detector noise) then the advantage of scanning in 16 bit is less clear cut. Let me state this in a little stronger way. If the noise in the image (i.e. film grain and any other noise that is included in the scanned image, such as sensor noise) is about the same as the ADC step size of the digitizer then 8 bit vs. 16 bit becomes a non-issue. You won't see banding or related artifacts in the final image because they are smoothed over by the noise. This concept is well understood in the field of digital signal processing. In some signal processing systems noise is actually added to a system in order to wash out the effect of ADC step size. If the signal is an image then noise washes out the potential effect of banding. Note that this only works well if noise is added to the system BEFORE the ADC process takes place. This condition is satisfied in a scanner if the noise comes from film grain, sensor noise, and electronic noise, such as thermal noise, prior to the ADC

Let me refine this concept just a bit more. If you scan in 8 bit then whenever you start doing digital image processing you should always convert the image to 16 bit and keep it at 16 bit all the way through the process, including any intermediate saves and the final save. This is to eliminate issues due to roundoff error.

This analysis of course goes out the window if you have a "grainless" negative, i.e. a negative where the variability in the image density due to grain is less than one about one ADC step size. Interestingly, this also depends somewhat on the pixel size of the scanner and any smoothing that can occur due to resolution limits in the scanner. If the pixel size is very small and there is no blurring of the image due (for example) to optical limitations then you can get away with a smaller word size. If the pixel size of the scanner is large and/or there are other significant blurring mechanisms in the system then you need to use a larger word size. If the pixel size of the scanner is small enough then a one-bit ADC would be good enough, but as a practical matter we would never get to that level.

Note: when I say "pixel size" I mean the size of a pixel as projected on the negative, not the actual size of a pixel on the sensor itself, although the two are obviously related through a magnification factor.

I previously dedicated a whole thread to the issue of 8 bit vs. 16 bit scanning, including posting of numerical simulations and some images from another source to demonstrate the concepts. It was, shall I say, not universally accepted here at photrio, but it is nevertheless true.

 

[Alan Edward Klein]

It is called Temp and Tilt in LR I have.

I have Temp and Tint slider in both color film scans and 16 bit grayscale BW scans. ALl they do is change the BW view to colors when sliding the sliders. How do I use them to get better sharpness?

 

[Alan Edward Klein]

Generally there be one of the channels that gives the sharpest image. The preferred channel might depend on the brand of scanner. As I recall, for my canon FS4000us scanner it is the green channel.

On the question of 8 bit vs. 16 bit, there has been a lot of discussion on that topic. The best thing to do would be to test it on a bunch of your images to see if you can tell a difference.

The concern about using 8 bit scanner seems to mostly revolve around the issue of banding in the image, especially if you do some extreme image manipulation after the scan. This is a very legitimate concern for a pure grainless image. If there is any graininess in the image (or other sources of noise in the system, such as detector noise) then the advantage of scanning in 16 bit is less clear cut. Let me state this in a little stronger way. If the noise in the image (i.e. film grain and any other noise that is included in the scanned image, such as sensor noise) is about the same as the ADC step size of the digitizer then 8 bit vs. 16 bit becomes a non-issue. You won't see banding or related artifacts in the final image because they are smoothed over by the noise. This concept is well understood in the field of digital signal processing. In some signal processing systems noise is actually added to a system in order to wash out the effect of ADC step size. If the signal is an image then noise washes out the potential effect of banding. Note that this only works well if noise is added to the system BEFORE the ADC process takes place. This condition is satisfied in a scanner if the noise comes from film grain, sensor noise, and electronic noise, such as thermal noise, prior to the ADC

Let me refine this concept just a bit more. If you scan in 8 bit then whenever you start doing digital image processing you should always convert the image to 16 bit and keep it at 16 bit all the way through the process, including any intermediate saves and the final save. This is to eliminate issues due to roundoff error.

This analysis of course goes out the window if you have a "grainless" negative, i.e. a negative where the variability in the image density due to grain is less than one about one ADC step size. Interestingly, this also depends somewhat on the pixel size of the scanner and any smoothing that can occur due to resolution limits in the scanner. If the pixel size is very small and there is no blurring of the image due (for example) to optical limitations then you can get away with a smaller word size. If the pixel size of the scanner is large and/or there are other significant blurring mechanisms in the system then you need to use a larger word size. If the pixel size of the scanner is small enough then a one-bit ADC would be good enough, but as a practical matter we would never get to that level.

Note: when I say "pixel size" I mean the size of a pixel as projected on the negative, not the actual size of a pixel on the sensor itself, although the two are obviously related through a magnification factor.

I previously dedicated a whole thread to the issue of 8 bit vs. 16 bit scanning, including posting of numerical simulations and some images from another source to demonstrate the concepts. It was, shall I say, not universally accepted here at photrio, but it is nevertheless true.

Alan It would be helpful if you could list which films require 16 bit? For example, I use Tmax 100, Tmax 400, Velvia 50 and Provia 100. These are very fine relatively grain-free. I normally scan 16 bit. I use Lightroom in post. What are your recommendations?

 

[Ko.Fe.]

I have Temp and Tint slider in both color film scans and 16 bit grayscale BW scans. ALl they do is change the BW view to colors when sliding the sliders. How do I use them to get better sharpness?

I never mentioned to use it for better sharpness.

 

[Alan Edward Klein]

I never mentioned to use it for better sharpness.

Sorry. I got confused with statements others have made to scan in color and then use the green portion for better sharpness. I was trying to understand how to do that with Lightroom.