On 18 Nov 2004, at 22:56, Martin Orpen wrote:
on 17/11/04 11:59 am, matthew ward wrote:
I doubt if it is photoshop, its maths, when you go from 65000 levels ofEven Photoshop lies about bit depths.
precision to 256 levels of precision you are going to get a rounding
error.
It uses a rather odd 32,769 levels!
Maybe (for legal purposes) they are claiming that 15-bit +1 is the same as
16-bit?
It is.
Now I am getting very confused about this thread, but here goes nothing:
if you want to express something to between 32,769 and 65536 levels on precision, you need a 16 bit environment to do it in. If you want to express something to 32,768 levels of precision you could get away with a 15 bit environment.
Computers like/need either 8 bit numbers or 16 bit numbers.
An 8 bit number requires 8 'spaces' to be reserved for it, a sixteen bit number requires 16 'spaces'.
If you give them a number in between they will add noughts to fill in all 16 spaces.
Your example gave a grey defined in 8 bit ie it was exactly one of the 256 greys available to you.
When you converted it to 16 bit you basically added a load of noughts to fill in the extra 8 spaces in the number, when you converted it back you removed them.
Had you converted to 16 bit and then changed something, some of the noughts would have become numbers, when you converted back to 8 bit, photoshop would have rounded these extra numbers up or down to the nearest whole number.
Hope this helps
Matthew Ward
=============================================================== GO TO http://www.prodig.org for ~ GUIDELINES ~ un/SUBSCRIBING ~ ITEMS for SALE
