If you re-read the first sentence of my original posting you will see
that this was taken from a posting on another group and is not my own
test. I shoot in raw and process all pixel brightness adjustments
(except final print sharpening) in raw.
Chuck Norcutt
Frank van Lindert wrote:
> I am curious to know what the subsequent file sizes of all your
> 'generations' were, Chuck. After all jpeg is only intended for
> compressing the original file without to much quality loss for your
> specific goal. You can always revert to raw, when quality is
> important.
> If the files stop getting smaller a further jpeg save would make no
> sense, or am I wrong here?
>
> Frank van Lindert
> Utrecht NL.
>
>
> On Sun, 08 Mar 2009 17:53:33 -0400, Chuck Norcutt wrote:
>
>> Be sure to note that the highest quality level was used. Photoshop
>> defaults to quality #8 vs. the max of 10. I suspect degradation would
>> be apparent earlier. I posted this because, while I've heard of the
>> problem, I've never previously seen anything in the least bit quantitative.
>>
>> Doug wrote:
>>> On Sunday, March 08, 2009 07:46, Chuck Norcutt wrote:
>>>> Taken from a note on the PWP forum this morning
>>>>
>>>> Chuck Norcutt
>>>> --------------------------------
>>>> I HAVE tested the loss (change) due to successive saves of JPEG files.
>>>> I opened a JPEG, did no editing and saved it under a new name
>>>> (generation #1). Then I opened Generation #1 and saved it as
>>>> Generation #2. Etc. Etc. [all the JPEG saves were at the highest
>>>> possible quality setting]
>>>>
>>>> After up to 3 generations I couldn't visually detect any difference
>>>> from the original JPEG. I used the COMPOSITE transformation with an
>>>> "Absolute Difference" setting to compare pixel-by-pixel any
>>>> differences and could see hardly anything but black in the comparative
>>>> result -- even with "brightness" highly exaggerated.
>>>>
>>>> But when I compared the original with Generation #10 -- WOW what a
>>>> difference. I could readily detect, visually, significant differences
>>>> from the original image. When I did an "Absolute Difference"
>>>> comparison, the differences were so dramatic that you could see an
>>>> actual image in black and white.
>>> I find this very interesting. My thought is that doing a very slight edit
>>> to
>>> the image and then saving it, would be a much better real world test. I
>>> may
>>> try it later today if I have the time using GIMP to edit and ImageMagick to
>>> compare. I don't know if the jpegs algorithms are implimented differently
>>> in
>>> different software.
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|