No you don't. I actually this question had occured to me also...
> ...a 4 bit sample could be scaled by a factor of _seventeen_...
>
> or maybe
>
> e.g., a 4 bit sample could be shifted left by 4 bit to represent a
> sample value of depth 8 or even better multiplied by a factor of
> 0x11 (17) so that white pixels still map to white pixels (15*17=255)
>
Err?
white pixel xxxx 0000 <<4 --> 0000 0000 (0)
White pixels (0) are still white pixels, while black pixels have the
value of (240) since shifting left by 4 is like multiplying by 16. The
only reason to multiply by 17 is to use the whole dynamical range
(0-255) and is not related with the whiteness of pixels AFAIK.
MF
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com