Bitmap sampling rules

Thread View: adobe.postscript.programming

2 messages

[T] Back to Threads [V] Tree View [G] Group Articles

2 total messages Started by Andrew_Garrard Thu, 18 Jan 2007 19:19

#460

Author: Andrew_Garrard
Date: Thu, 18 Jan 2007 19:19

17 lines
2596 bytes

Hi all.

A quick question to the elite, from a PostScript novice:

I have a vague recollection that someone (possibly on this forum many moons ago, in which case my apologies) told me that PostScript RIPs will downsample non-bilevel images to fit the line screen frequency of the device.

If I didn't imagine it, could I ask whether that's standard behaviour, or the behaviour of a particular RIP?

I'm interested because one could have, for example, an image containing both grey scale and purely black-and-white elements, and if the above is true, the bilevel elements would only be rendered at (double) the line frequency of the device, rather than the full RIP resolution - and might also have grey level interpolation. Correspondingly, a sharp-edged grey circle on a white background at high resolution in an image could either look like a sharp circular cut-out of the corresponding grey halftone screen, or like a grey circle downsampled to double the line frequency (which isn't the same thing); it could even be something more complicated if there's some low-level image processing going on to compensate for the sharp edge and printer effects.

I'd be interested in whether the same question applies for non-images - is a PostScript grey circle actually a cookie-cutter slice from a grey screen, or mapped onto a line-screen frequency set of blocks? (In this case, I'd expect it to be a pure cut-out, since the algorithm for per-pixel colour mapping to halftones is well-known, but the image case is less clear to me.)

If I imagined it, the guideline not to provide images at more than twice the target line frequency has a distinct exception in this case (although obviously it's *usually* sensible).

I realise I could find an experimental answer, but I'd really like to know what the official position, if anything, is before I make any sweeping statements on the subject to anyone. I may be missing it, but I couldn't find a detailed description of the sampling mechanism in the Red Book.

I'd be most grateful for any inside information.

Re: Bitmap sampling rules

#464

Author: Andrew_Garrard
Date: Tue, 23 Jan 2007 15:03

14 lines
1354 bytes

Just to clarify, having done the obvious thing and found a
LaserJet with PostScript support to try this on, I'm aware
that there's (apparently) no such filtering happening on
this particular printer. That is, a 600dpi bitmapped grey
circle has partial halftone clusters on the edge, just as
does an outline grey circle; I've not checked a more
contentious case (such as whether a grey line running along
the diagonal between halftone cells comes out white).

I'm just wondering whether HP's implementation is standard
(for all I know it's their own), or indeed whether there
* is* a standard behaviour for this.

Any feedback much appreciated.

Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads