|[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]|
In a message dated 10/12/99 9:58:35 PM, firstname.lastname@example.org writes: >I assume that representing the tonal response of these photoprinters of > >which you speak (e.g. LightJet, Fuji, Lambda, etc.) must require greater > >than 8-bits per channel? If not, wouldn't any space larger than 24-bit >True > >Color be wasted? > > > >And since, without adding bits to each channel, you can't pay Paul without > >robbing Peter, more purple (for example) must mean tonal resolution was > >"stolen" from complimentary colors and the available "numbers" were simply > >remapped to mean something different. This must result in posterization, > >even if subtle. My prediction is that these printers (and filmwriters) will be the first devices to break the 8 bit per channel limit and start utilizing high bit data through to output. That is because the huge gamut they are capable of will benefit from having more than 256 levels per channel. But at the moment we still must optimize the data in it's high bit state, then convert to 8 bit at the last minute to get the needed format for certain operations, and for printing. So while having only 256 levels (or even considerably fewer) does not show on press output, and 256 seems sufficient for Epson inkjets, photoprinters are indeed capable of resolving this number of levels and will show posterizations in large areas of very even gradients. The only solution at the moment is to dither such areas selectively by selecting them, feathering the selection, then applying noise or blurring to break up the posterized lines. I suppose a clever software developer could devise a smart filter that would search for areas that meet this criteria and automatically break up the posterizaton, but probably by the time the product was developed and marketed the problem would have been transcended. C. David Tobie Design Cooperative CDTobie@designcoop.com - Turn off HTML mail features. Keep quoted material short. Use accurate subject lines. http://www.leben.com/lists for list instructions.