RE: colorsync-users digest, Vol 3 #900 - 11 msgs
RE: colorsync-users digest, Vol 3 #900 - 11 msgs
- Subject: RE: colorsync-users digest, Vol 3 #900 - 11 msgs
- From: "Steve Lawrence" <email@hidden>
- Date: Thu, 17 Jul 2003 18:31:07 +0100
- Thread-topic: colorsync-users digest, Vol 3 #900 - 11 msgs
Graham Gill wrote:
>
> Actually, it's the other way around: even an 8-bit RIP needs more
than
>
> 8 linear bits of precision for deciding whether to print a dot. If
>
> you only had 8 bits, you could not print with less than 1/256 ink
>
> coverage, which is surprisingly dark, particularly at high
resolutions
>
> on most printers.
>
>
It depends on the screening technique chosen. For instance,
>
using a stochastic or ordered fixed cell screen, it's
>
possible to apply the linearisation curve to the screen
>
thresholds themselves, giving a high output resolution (ie.
>
14+ bit precision output levels for a 167x167 screen), while
>
the input threshold values are 8 bit. Given the typically
>
high dot gain of current inkjets at their maximum
>
resolutions, this is an effective technique. Typically a step
>
wedge of 256 input levels reproduces with good subjective
>
linearity, and almost imperceptible stepping (given an
>
effective calibration system to define the curve of course!).
Yes, a good point, well illustrated.
Our halftoning is fractional, but does not require more than 8 bits per
channel input to achieve the good attributes you mention.
Steve Lawrence
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.