Cory Papenfuss wrote:
A digital camera RAW file is not a picture, it should be nothing more
nor less than the raw data as read from the CCD (possibly with some form
of lossless compression) so it makes no sense comparing it to image
formats
===========
Which, in point of fact, is why most people recommend using it over
any image
format.
For a Pentax RAW file, it *IS* an image format.
Yes. And as I was trying to say earlier, there is no law saying that you
have to represent the pixels as read, green and blue values for the same
location to call the data "a picture" or the format an "image format".
The bayer pattern data is just another way to describe an image. You may
actually say that it's a different image from the RGB data you might
convert it to, because the latter also has some interpolated data - but
*of course* the file data represents a picture!
It's a TIFF. It's not an RGB TIFF like one may be used to seeing. If
you were to load up the image, you'll see a monochrome image of the
picture. The value of a monochrome pixels will be the luminance
component of the image shot, weighted by the tranmissivity of the
pigment in the particular sensor pixel.
And like I said, more or less, if Pentax had bothered, they might have
tried to introduce an extension to the TIFF format allowing the colour
model to be set to, "bayer pattern data" or whatever, and perhaps rules
for transformation to RGB to be stored as TIFF tags...
But we're discussing minor details here, of course, I must have little
to do these days... Yep, I have, come to think of it; I'm on sick leave
as a result of the fact that I tried something called "sports" last week...
- Toralf
