On 9/25/05, Moose <olymoose@xxxxxxxxxxxxxxx> wrote:
> I believe the JPEG spec only defines how an image is represented in the
> file format. I don't believe it says anything about the mapping of
> higher brightness resolution images into its 8 bit space. How the data
> is created is the business of the creators, not the file format. No
> matter what algorithm/curve is used, mapping 12 to 8 bits loses data.
Actually, jpeg is the compression scheme and is fairly well defined
(has to be or else you wouldn't be able to decompress the file), the
file format is usually a jfif file but it doesn't have to be (how's
that for useless trivia?). In terms of how it handles brightness, I'm
not sure. I seem to recall that in one point of the process the image
is converted to the LAB colour space so the luminance (or brightness)
is separated out from the colour. I believe that jpeg only handles
24-bit colour images so I'm not sure if the standard defines what it's
supposed to do if it gets more colour data than that or if it's left
to the software that's creating the jpeg to convert the image to
24-bit colour before compression. If that's the case then it could
really vary on the software or digital camera in terms of how the
brightness is converted.
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|