Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 

   


   


   















      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[filmscanners] Re: JPEG Lossless mirror?




--- Julian Robinson <jrobinso@pcug.org.au> wrote:
> FWIW the following is from http://www.jpg.com/products/wizard.html
> It
> implies that normally you would introduce artifacts when doing a
> mirror and
> re-saving, but I think is claiming that with this technology you
> won't
> degrade the image at all.

That is very well possible. The reason is that for mirroring you do not
have to decompress and compress again. Instead you rearrange the DCT
coefficients. That is why this method only work for multiple of 8
pixels, I believe. If you would decompress, mirror the image and then
compress no such limitation would exist.

> My guess is that it does have to clip to nearest 8 x 8 pixel block to
> do
> this because the boundaries of these blocks would have to change
> following
> a mirror, but this could be wrong.

Correct.

> RECOMPRESSION WITHOUT LOSS!* Recompress your JPEG files again and
> again
> without introducing generational image loss normally associated with
> recompressing JPEG files! Because of the underlying Pegasus
> technology, you
> can recompress, rotate or mirror JPEG images without introducing
> recompression artifacts!

It is not clear what the pegasus technology is and when it applies. If
it is only for mirroring and rotation by n*90 degress then you do not
need to decompress and recompress which avoids any image loss. But this
is a very limited functionality.
Another additional technique could be that only the blocks that have
been changed are recompressed. All blocks that have not been changed
are only decompressed for display purposes but kept as they were
originaly when the file is saved. That assumes the image was not scaled
during editing.
Finally, I could think of another technique. When you decompress the
image you keep all the data as floiting point data in memory. You only
truncate it for display but keep the floating point data in memory.
Then when you compress again without having edited the image no
additional (or at least much less) loss will occur. Not sure about
that, though.

Rob

__________________________________________________
Do You Yahoo!?
Send FREE Valentine eCards with Yahoo! Greetings!
http://greetings.yahoo.com

----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body




 




Copyright © Lexa Software, 1996-2009.