Problems with autocorrections and gamma gradation
PostPosted: Thu Feb 14, 2013 7:34 pm
Hi,
I am looking for a way to scan film positives preferably without autocorrection of any kind (sharpening, dust removal, gamma gradation, etc.). This is what an entire film looks like:
and we scan each of the small slides you see in the image to analyze with Image J individually using tools like area and optical density (lightness and darkness). Unfortunately we have been having an issue where the images which clearly appear the same lightness/darkness in the complete film image are not the same when they are scanned individually (example: 
. This does not seem like an issue with variation in our slides, but an issue with the software. In the second example I linked, the image on the left contains a black spot from preparing our original slides. The software has decided to adjust the lightness of the entire rest of the image to account for it, leaving me with a too-light image. The right side of the 2nd linked example shows a slide without any very dark spots, so the software has decided to adjust for that and make everything in the image darker.
I do not see cropping out the dark spot as a solution because some of our regions of interest we are measuring in tissue are almost black and we don't want them corrected for. If they are dark in the original film positive we are scanning, they should be just as dark in the resulting scan. This tells us that more of our gene of interest is expressed, and that's what we're trying to measure and compare to other slides.
My end goal is to have these all come out with the same or extremely similar background (background meaning where "white" space where tissue we have put on slides is not present) but the gamma gradation seems to pick and choose how to adjust the histogram on its own and this is unacceptable. If we could find a way to either only scan pure, raw, unadultered images that would be great. If not, we could try adjusting for background only with empty space and set this all to the same value for a given film. I do like the darkness of background to be between the two in the side-by-side example as it is easy to see.
Is there a way to do what I want with Silverfast? I would like to turn off the gamma gradation completely to see if that helps but nothing I have tried suggests that this is even possible. We are currently using Silverfast SE version 6.6.2 on an Epson Perfection V700 scanner and using the scan type 16 bit HDR Grayscale. Our images need to stay in grayscale. I have read past advice about using 48bit HDR to get "raw" images but this option does not appear in my software as far as I can tell in grayscale, only color. I tried it anyway to see if these were truly raw images and after manually setting gamma gradation to 0.0 which should be none, the software automatically reset it to .4.
I will not be editing the lightness/darkness in other software. Our post-processing consists of changing images to a smaller 8 bit version for analysis and adjusting the scale.
Others in my lab have found some success with calling Tech Support overseas, but our usual contact (a man named Boris) is not helpful and I hope we can find better help here online. Thanks in advance!
I am looking for a way to scan film positives preferably without autocorrection of any kind (sharpening, dust removal, gamma gradation, etc.). This is what an entire film looks like:
and we scan each of the small slides you see in the image to analyze with Image J individually using tools like area and optical density (lightness and darkness). Unfortunately we have been having an issue where the images which clearly appear the same lightness/darkness in the complete film image are not the same when they are scanned individually (example: . This does not seem like an issue with variation in our slides, but an issue with the software. In the second example I linked, the image on the left contains a black spot from preparing our original slides. The software has decided to adjust the lightness of the entire rest of the image to account for it, leaving me with a too-light image. The right side of the 2nd linked example shows a slide without any very dark spots, so the software has decided to adjust for that and make everything in the image darker. I do not see cropping out the dark spot as a solution because some of our regions of interest we are measuring in tissue are almost black and we don't want them corrected for. If they are dark in the original film positive we are scanning, they should be just as dark in the resulting scan. This tells us that more of our gene of interest is expressed, and that's what we're trying to measure and compare to other slides.
My end goal is to have these all come out with the same or extremely similar background (background meaning where "white" space where tissue we have put on slides is not present) but the gamma gradation seems to pick and choose how to adjust the histogram on its own and this is unacceptable. If we could find a way to either only scan pure, raw, unadultered images that would be great. If not, we could try adjusting for background only with empty space and set this all to the same value for a given film. I do like the darkness of background to be between the two in the side-by-side example as it is easy to see.
Is there a way to do what I want with Silverfast? I would like to turn off the gamma gradation completely to see if that helps but nothing I have tried suggests that this is even possible. We are currently using Silverfast SE version 6.6.2 on an Epson Perfection V700 scanner and using the scan type 16 bit HDR Grayscale. Our images need to stay in grayscale. I have read past advice about using 48bit HDR to get "raw" images but this option does not appear in my software as far as I can tell in grayscale, only color. I tried it anyway to see if these were truly raw images and after manually setting gamma gradation to 0.0 which should be none, the software automatically reset it to .4.
I will not be editing the lightness/darkness in other software. Our post-processing consists of changing images to a smaller 8 bit version for analysis and adjusting the scale.
Others in my lab have found some success with calling Tech Support overseas, but our usual contact (a man named Boris) is not helpful and I hope we can find better help here online. Thanks in advance!