"Osroubek" wrote: > > I have recently started shooting with a Canon 20D from shooting MF > then scanning on the Minolta Multi Pro. I shoot in RAW format and >convert with Adobe Camera Raw in CS2. The > MF shots were on transparency then converted to BW in PS using > usually the channel mixer. >I notice that the scanned film when converted to black and white > results in a better and cleaner looking BW image ... I have also started using the 8 MP Canon chip in the XT and enlarge to 16 x 20. My other cameras are MF. What I've found is that the 8 mp chip can make a 16 x 20 that is very close to the film shot, but it probably takes more work. Here are some of my observations. (The shot I just worked up is on my web page, below.) The limited latitude of the sensor requires bracketing in many shots, also not the usual sun-lit scenic. The noise/grain level of the digital image is less than Tmax 100 in the bright areas, but can be a problem in dark smooth areas. The depth of field advantages of the small sensor are significant. CS2 smart sharpening is very useful, but still not enough (or I'm not good enough with it) to avoid too-bright small highlights in some areas by the time the image is sharp enough. To correct for this, I have to us a curve to pull these down in the affected area. The very high contrast edges may require that the green channel alone be used. The green channel, maybe because it has 2 sensors, seems to give the best image in some areas. I have not found third party raw converters to be better than the CS2 download. Overall, based on very limited experience, I'd say 8 mp can make a good 16 x 20 -- not perfect, but then neither is MF 100 ISO film. Neither can match MF Tech Pan, but that film is history (aside from the pile in my freezer). Finding workflows to get around the limited latitude is probably the area where more work is most needed. Paul www.PaulRoark.com
Message
RE: [Digital BW] Digital vs scan for BW Print
2005-07-05 by Paul Roark
Attachments
- No local attachments were found for this message.