Silverfast 9 ghost tif issue

[jad3675]

Well, not sure this is 100% the correct forum, so apologies in advance.

I've been working with a PrimeFilm 7200 35mm scanner and SilverFast 9. I was never really happy with how much tweaking I had to do within SF to get decent images. I had the genius idea to scan as negative and then use Fred's 'negative2positive' script to convert - only using isrd within SF. No other adjustments.

I scanned an an entire roll this way as .tif files, and noticed this when I started going through them. A faint ghost image in the upper left corner.

Converted .tif

OG Negative Image, scanned and saved as a .tif. I don't see the ghost image, at all.

Re-scanned as a jpg and converted as a jpg - no ghost image.

And the original negative, scanned and saved as a jpg.

I re-scanned the roll using vuescan (which takes 2x as long as SF), saved the scans as tifs and those conversions were great - no ghost images at all.

So....Does Silverfast save tifs with something extra in the format that imagemagick stumbles on?

 

[MattKing]

Interesting. I have no idea what might be causing this, but have moved the thread to the Scanning and Scanners sub-forum.

 

[DWThomas]

Were the good scans and bad scans under the same room lighting conditions? One almost wonders if extraneous room light was getting in there somewhere -- maybe try a black cloth over the scanner?

Just a couple of thoughts 'cause the pattern reminded me a little of some lighting I occasionally see with the mini-blinds here in the computer room!

 

[jad3675]

Were the good scans and bad scans under the same room lighting conditions? One almost wonders if extraneous room light was getting in there somewhere -- maybe try a black cloth over the scanner?

Just a couple of thoughts 'cause the pattern reminded me a little of some lighting I occasionally see with the mini-blinds here in the computer room!

Yep same location, same lighting conditions. I even put tape over the window on the top of the scanner, thinking the same thing. Otherwise, not much for light to get in. Positive scans with SF and saved as tif do not exhibit this behavior, either. It is....odd.

 

[_T_]

Have you tried rescanning the frame with the settings that originally had the problem to see if it is reproducible? Also I see that you scanned an entire roll this way, did you see the problem on any other frames?

 

[MattKing]

Try re-scanning with the film emulsion same way up, but turned 180 degrees. See if the ghost image moves.

 

[brbo]

I can't see a ghost image like that happening on a line ccd scanner.

More likely the script for inverting the negative is not handling the IR channel in the Silverfast raw file correctly.

 

[Romanko]

More likely the script for inverting the negative is not handling the IR channel in the Silverfast raw file correctly.

Can you upload the raw file somewhere and share a link?

 

[jad3675]

I can't see a ghost image like that happening on a line ccd scanner.

More likely the script for inverting the negative is not handling the IR channel in the Silverfast raw file correctly.

That makes a lot of sense, actually. Guess it's off to google to see how I can pull out the channels from the tif.

I'm using this script which relies on imagemagick.

Fred's ImageMagick Scripts: NEGATIVE2POSITIVE

Fred's ImageMagick Scripts - negative2positive - converts from a (scanned film) negative image to a positive image

www.fmwconcepts.com

Can you upload the raw file somewhere and share a link?

Have you tried rescanning the frame with the settings that originally had the problem to see if it is reproducible? Also I see that you scanned an entire roll this way, did you see the problem on any other frames?

About 1/3rds of the scans came in this way. I've scanned the negative multiple times all with the same settings and the same results. When I get some time today, I may try it with isrd turned off and see if that helps.

Try re-scanning with the film emulsion same way up, but turned 180 degrees. See if the ghost image moves.

I'll give that a try.

Thanks again everyone!

 

[jad3675]

So, that's interesting. I used imagemagick to pull the channels out of the Vuescan negative tif. Three channels.

Silverfast has 6 channels in the tif. I think the conversion script is choking on the extra channels.

I also had a few errors on the convert process:
convert-im6.q16: Unknown field with tag 35243 (0x89ab) encountered. `TIFFReadDirectory' @ warning/tiff.c/TIFFWarnings/949.
convert-im6.q16: Wrong data type 3 for "PixelXDimension"; tag ignored. `TIFFReadCustomDirectory' @ warning/tiff.c/TIFFWarnings/949.
convert-im6.q16: Wrong data type 3 for "PixelYDimension"; tag ignored. `TIFFReadCustomDirectory' @ warning/tiff.c/TIFFWarnings/949.

 

[jad3675]

I combined the first three channels from silverfast (_0, _1, _2) and got this image:

Attempting to combine all 6 of the channels into an image resulted in a 'images are not the same size.' I guess that explains the slightly smaller ghost image in all the images.

 

[jad3675]

And, I found out what the issue is - thanks to everyone for the help.

https://forum.silverfast.com/imaging-in-general-f2/tiff-image-appears-to-contain-2-layers-t13937.html

Turning off 'Include HiRepp' in the preferences dialogue of Silverfast 9 eliminates this issue. The resulting tif only has three channels and no ghosting.

Thanks to everyone for the help with this. Hopefully this can help someone else out in the future.