When I started processing the scans of these old glass slides I found that about half of them had blown-out highlights.
I'd been adjusting the exposure in the scan preview to maximise but preserve the highlights so I was a bit curious about what was going on. It turns out that many of the slides required a negative exposure offset to bring the highlights within range but this is not applied correctly to the prescan. I think the scanner does the prescan at its default exposure and the software compensates for the preview. Each time you change a slider you can press a button to update the preview without redoing the prescan. Even if you do a prescan with the updated values it applies the offset in the same way. It's not until you do the full scan when the exposure values are actually sent to the scanner. That all works well until you have a bright slide where the raw prescan values are clipped at 255/255/255 (I can't be bothered typing it out for 16bit;). When you turn the slider down the preview adjusts from 255/255/255 but the original is actually brighter than that so the numbers you see are lower than reality. Now I'm getting around it by doing the prescan as before then adjusting down a bit, then selecting a small section of the brightest highlights and scanning that. Then I open the file in Photoshop at 100% and check the RGB values. Bloody fiddly but I don't have a sheet of ND filter material handy! Not that I want any more potential for Newton rings. I could speed up the process by reducing the scan resolution but I'm only scanning at 1200ppi anyway so the extra scans only take a few seconds. Anything higher I'd just be creating larger files with no extra detail. I'm not happy about discovering this after scanning all 70+ slides... Oh yeah, the scanner is a Minolta Multi Pro. IIRC they use the same software for many other models. Other than this little quirk it's excellent. Cheers, Dave -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
