TOPAZ STUDIO 2 FOR SCANNED SLIDES MAC
It seems to still work up to Mojave on a Mac but I'm sure that's the end of the line. The best noise reduction for film in my experience was always Noise Ninja, which unfortunately has been discontinued. It would be nice if there was a version trained on film images (not likely to happen). I have contemplated purchase of DeNoise AI but unless there is some indication that it works better I will hang on to my money. Sometimes it work OK but never seems to make much of an improvement. I have tried Topaz Sharpen AI but find that it often confuses grain for image detail and sharpens it which is an epic fail. If anyone has profile recommendations let me know.
I need to explore the film profiles for Neat Image to see what I can find. I used the Nik De-Fine and really liked the results - and even printed with great results Therefore tagging this thread for future helpful tips. I am still working on all this though, no two images are alike. At this point, I have not used more conventional sharpening programs or even different Capture 1 settings, but I would suspect that I would need to use masking more extensively for local control. Have also used Topaz Denoise AI with some limited success. In these areas I apply a mask locally, reducing these by up to 30% prior to sharpening in Topaz. It can help to reduce grain locally with the structure and clarity sliders in Capture 1. When it works, the results can be astonishing in the amount of detail revealed. I have even been able to eliminate nearly all grain from smooth surfaces/sky and still get amazing fine detail and structure which looks amazing. Sometimes it can take a little noise reduction, sometimes up to 90%, depending on the subject and exposure of the film. I start off with low sharpening 15-25% (or even less) and zero noise reduction which I then dial in gradually. Often but not all the time, the Focus module works best.
TOPAZ STUDIO 2 FOR SCANNED SLIDES MANUAL
With a really fast computer I built earlier this year, I can trial and error multiple manual settings quickly. The high MP count of your camera scans may therefore be a factor. Using an old Nikon Coolscan 9000ED, scanning old Kodachromes and negatives, and Topaz Sharpen AI, I have found that scans at 8000DPI are handled much worse than at the standard 4000DPI.
Back then I was concerned about the luminance noise (grain). In the 90s and early 2000s I did not notice that as in the digital era. One of the strange parts about scans from the high end scanners is that you can see the color grains and scans may benefit from some chroma noise reduction. The later versions addded some features, but mainly added multi-core and GPU support, not that I found GPU was usually needed for single images. I know the program still exists but I am out of touch with recent versions. It worked incredibly well at nailing the grain but leaving image details. At the time you could train it yourself to create different noise profiles for different films, and there was also a library of profiles that other users had made. Many years ago I used Neat Image for film grain. I assume it's now been optimized for digital so it may not be as useful as it once was for film. Old versions of Denoise are not available from Topaz but Noiseware is still out there. I used early versions of Topaz Denoise for film (v5, maybe) and also Noiseware. The Topaz AI applications seem to have been trained using digital images and the different noise characteristics of film captured images cause them to fail.ĭoes anyone have a recommendation for applications or methods for grain noise reduction in scanned film images? I find the applications that work well on digital camera noise are ineffective and often counter productive when used to reduce grain noise in scanned images. The results are excellent but the grain in some of the images is pretty objectionable. I have been digitizing a lot of old film images using my Sony A7R4. They work well, with the Topaz applications being my favorites. I have been using several noise reduction applications for digital camera images.
0 kommentar(er)
