Re: RAW images...

Kennedy McEwen <rkm@xxxxxxxxxxxxxxxxxx> wrote:
Well, I have checked it time and time again with different
images since you raised it and so far I haven't observed it once
at a compression rate that gives a file size about the same as
the original. It does become a problem if you increase the
compression rate to get a file size of half or less of the
original, but I suspect you would get similar problems without
any resizing anyway. So, in practice, it is a non issue.

Your inability to observe the effects does not make it a
non-issue for critical work.

In fact *any* change that causes the boundary for quantization
to change in non-multiples of the original will have the same
effect. If, for example, that is originally done on 8 pixel
boundaries, then shaving 2 pixels off the edge of the image
and saving it will have the indicated effects...

--
Floyd L. Davidson <http://www.apaflo.com/floyd_davidson>
Ukpeagvik (Barrow, Alaska) floyd@xxxxxxxxxx
.

Relevant Pages

  • Re: RAW images...
    ... images since you raised it and so far I haven't observed it once ... at a compression rate that gives a file size about the same as ... Your inability to observe the effects does not make it a ... it is NOT an issue that is restricted to JPEG ...
    (rec.photo.digital)
  • Re: RAW images...
    ... at a compression rate that gives a file size about the same as ... That doesn't mean that JPG cannot be used for useful, even great, images - even after resampling *and* recompression. ... What you are demonstrating is a complete failure to understand the concept of the JPEG algorithm. ...
    (rec.photo.digital)
  • Re: Learning to see
    ... so the ability of a person to "see" things trough the lenses is not a major factor in what can be observed (as several people can look at the resulting images on the computer). ... Setting up things such as kholer illumination is important, even on a straight bright-field scope. ... Add in imaging systems like DIC or phase contrast, and you now add another 5-6 levers & dials which must be properly tuned for good images. ... Over the years I've learned how to set up my experiments to optimize what it is I'm trying to observe. ...
    (sci.techniques.microscopy)
  • Re: Text rendering in FX3
    ... observe, here is an image of the page using FF *2*, which I DL'd from ... Based upon the above bug reports, it looks like the problem is going ... images, although I can't remember any right now. ...
    (Fedora)
  • Re: chemtrails /unwashed satellite images
    ... satellite images i am very interested in learning about subject of ... Not quite sure what chem-trails are or how to observe them on sat images. ...
    (uk.sci.weather)