Re: 16-bit frames and interleaving

David Mosberger-Tang (David.Mosberger@acm.org)
Sun, 28 Feb 1999 11:35:30 -0800

>>>>> On Tue, 9 Feb 1999 18:21:54 -0500 (EST), Tripp Lilley <tlilley@perspex.com> said:

Tripp> The docs say that 16-bit deep frames contain the samples in
Tripp> the machine's native byte order. Can I assume that they will
Tripp> always come in pairs, and that they are sample-interleaved,
Tripp> not byte-interleaved?

Tripp> ie: on a little endian machine, would data look like this
Tripp> (where the first character is R, G, or B, for the color of
Tripp> the sample and the second character is 'L' or 'M' for the
Tripp> 'Least' or 'Most' significant byte):

Tripp> RL RM GL GM BL BM ...

Tripp> (which is what I am assuming)

Tripp> or can I expect some sort of nuttiness like

Tripp> RL GL BL RM GM BM

Tripp> Of course, I /realize/ that this would be a totally asinine
Tripp> way of rendering the samples :-) But the spec leaves a
Tripp> /teensie/ bit of room for asinine interpretation :-)

The intention of the spec is to allow the former format only (assuming
a little-endian machine; on a big-endian, you'd get RM RL GM GL BM BL
...).

--david

--
Source code, list archive, and docs: http://www.mostang.com/sane/
To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com