On Dec 6, 2004, at 3:22 PM, B. Campbell wrote: > >> If you scan at a reduced resolution eg. 1200ppi instead of 4800ppi > (assuming >> that is the optical resolution), your scanner simply throws away >> three-quarters of the detail in your film, leaving you with just > one-quarter >> of the original detail. > > I thought that ppi in the scan was relevant in determining the amount > of > "detail" in the print but that detail in the scan was a function of the > scanner's optics and mechanics, not the ppi setting. I would expect to > lose > detail in the print if for example I scan a 4x5 negative at 300 ppi > and then > make a 16x20 print but I never thought that if the scanner is capable > of say > 4800 ppi and I scan at 1200 ppi I've necessarily lost three quarters > of the > "original detail" in the negative regardless of the size at which I > print. > Or stated another way, I would have thought I could scan the 4x5 > negative at > say 720 ppi with a 4800 ppi scanner and make a 4x5 print that has as > much > detail as the scanner, printer, and paper are capable of producing, > i.e. > that I wouldn't lose most of the detail in the negative by doing that. I think in this context you should interpret the previous posters use of the term "detail" as really to mean "data". Your analysis is correct on the whole, when taking a less mathematical definition of detail - if you simply mean the visible details of a particular image, then of course it varies from image to image, and you may well not lose 3/4's of the detail by sacrificing 3/4's of the scanning resolution - you will loose 3/4's of the scanned data however. This distinction between significant visual detail and the much more literal "data" needs to be kept pretty clear... scanning at twice the resolution will often not give you four the detail, just four times the data. - R
Message
Re: [Digital BW] Computing power
2004-12-06 by Roger Howard
Attachments
- No local attachments were found for this message.