Yes, size matters. Imagine a smooth gradient, across a cloudless sky for example. It came from a very clean drum scan of tech pan at 16-bit. Now you downsample to 8-bit, _without_ dither, and this gradient goes from 226 to 240 i.e., 15 steps. It s a large print, so this section is 15 long. Thats one 8-bit step per inch. Now, what does the printer dither pattern do at the distinct transition between steps? A very keen eye might see posterization, which can be corrected by adding noise to the image. If you did not downsample to 8-bit, you would not have the need to fix the posterization because the step size is well below human eye detection. The same print is now much smaller, and the gradient is only 0.5 long. There are now 0.5/15 steps per inch, or about 0.033 per step. The printer dither over this short distance will blend in with the adjacent steps, making it much harder, if not impossible to see. This of course depends on your RIP, paper, profile, and image. Tyler has commented in the past that better systems can actually look worse, for cases like this, since the system is finely tuned to put down precisely what is in the file. I hope I have not misstated his point with this example. In any case, it is very subtle, and you need to be looking hard it to find it. Best regards, John Moody -----Original Message----- From: DigitalBlackandWhiteThePrint@yahoogroups.com [mailto:DigitalBlackandWhiteThePrint@yahoogroups.com]On Behalf Of john dean Sent: Monday, November 13, 2006 3:50 PM To: DigitalBlackandWhiteThePrint@yahoogroups.com Subject: [Digital BW] Re: the times, they aren't a-changing-so can we start over again? 16 bit output? Would the size of the print come into play in this? [Non-text portions of this message have been removed]
Message
RE: [Digital BW] Re: the times, they aren't a-changing-so can we start over again?
2006-11-13 by John Moody
Attachments
- No local attachments were found for this message.