Re: [Announce] WinSANE 0.1.0.0 Release

Nick Lamb (njl98r@ecs.soton.ac.uk)
Tue, 11 May 1999 01:37:42 +0100 (GMT)

Nick Lamb (njl98r@ecs.soton.ac.uk) wrote:
> A series of three 10-bit samples (e.g. an RGB pixel captured by a one-pass
> 30-bit parallel port scanner) will arive through SANE as...
>
> 98765432 10------ 98765432 10------ 98765432 10------

On Mon, 10 May 1999, Ewald R. de Wit wrote:
> This seems to be the mostly used format, but the HP 10 bit packing is
>
> ------98 76543210
>
> I would like to know if there are any other brands using LSB alignment for
> their >8 bit packing.

Are you suggesting that the HP backend exports this bizarre format?
If so, the HP backend is non-compliant wrt the SANE standard. The SANE
standard indicates two possible ways of encoding 8 < bits <= 16
This is neither of those two ways, so it's not possible for a properly
written frontend to correctly understand that format.
(The two possible ways are distinguished by ENDIANESS, and shown in my
original mail to the list, quoted above)

Otherwise, I pity the poor HP backend, having to transform this bizarre
format into a usable SANE format, and presumably fix up every single
sample as it passes through the backend. I hope there aren't any more
scanners in production which choose this non-optimal solution.

Nick.

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