On Wed, Jun 15, 2005 at 02:26:07PM -0700, Michael Stevens wrote:
>I think you're wrong here and I just tested it to be somewhat sure; be sure
>to let me know if I have my head in my ass.
>
>I don't know what you mean by "100% JPEG group conversion" but your
>statement that the "high quality" setting turns off JPG comression is false,
>I believe.

The JPEG working group provides a sort of benchmark jpeg converter.  It
measures quality from 1 to 100.    At a quality of 100 the quantization
table is set to all 1s.  This cuts quant loss down to the lowest
possible value, but doesn't affect subsampling or other kinds of loss.
Use of these quality settings is discouarged because it creates gigantic
files without really doing much for image quality.

I found the values by comparing Canon created jpegs from the camera and
software to jpegs I encoded myself from the raw data.  High quality is
about 100%, and low quality is about 90%.

>I just tested it on an image I have in TIFF form, LZW compressed. It's 7610k
>on disk. Saved from PhotoShop as a quality 12 JPG it's a 3094k file on disk.
>So, even at the highest setting there is JPG comression going on. It's just
>so small that it's unnoticable to the human eye.
>
>Now, Canon may have some proprietary type of JPG compression that lets the
>camera know that at it's highest level it shouldn't do any JPG compression
>but I kinda doubt it.
>
>Mike

-- void *(*(*schlake(void *))[])(void *);
*
****
*******
***********************************************************
*  For list instructions, including unsubscribe, see:
*    http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************

Reply via email to