Tony Sleep wrote
>
A ratio of (say) 1:1024 *cannot* be mapped linearly to any smaller 
ratio, eg 1:256 of 8bits. 
<

Sorry, Tony, but it looks like yes, you can. Just multiply by 255/1024 
and round to integer each 1:1024 value. You have your data 8-bit, 
linearly coded. Then, when you want to use it, just multiply by 
1024/255. Your range is preserved, you only *lost resolution* by a 
lower bit enconding.

Regards,

Rick Trelles

Reply via email to