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: TIFF's, PNG's, JPEG's and quality



I apologise profusely if this message has already been seen.

I didn't see it get to the list after a subscription change, so am trying
again.



To: filmscanners@halftone.co.uk
From: markthom@camtech.net.au
Subject: filmscanners: JPEGs, TIFF's and suchlike

OK, Ed, I'll bite!

I've just tried to do some sort of meaningful comparison of originals
versus JPG's at 100% quality, and the results are here - no apologies for a
hastily knocked up, ugly, web-page ;( :

http://www.adelaide.net.au/markthom/compare.htm

Note that 1024 x 768 resolution would be useful for viewing these images.
(But don't panic, I'm only using tiny bits of a scan, and the total web
page is only about 40K - I have set it up so that your browser does the
magnification..)

In summary, I think that although the differences CAN be subtle, if using
any form of lossy compression, keep your eyes VERY wide open, and your
magnifier ready!


If my methodology is suspect, fire away..

Regards, Mark T.

>Ed Hamrick writes:
>
>> I've just had my 6x7 trannies scanned by a lab specialising in digital. I
>>  paid for a 50MB scan and got a 7MB jpeg back. I took the CD back to the
>>  manger, who told me that it was a 50MB scan, compressed to 7MB and that
all
>>  the information would be there when I opened it up!
>
>I think the manager is correct.  This is only 7:1 compression, and I
>suspect that if you did a bit by bit subtraction of the jpeg and tiff
>files you'd find at most some errors in the least significant bit.
>
>There's certainly no way you'd see a difference with the naked
>eye or by printing the images, even after manipulating them
>digitally (assuming the images are within 10% of their final
>intensities).  I'm sure lots of people will jump in and claim this is
>incorrect, but it would be most useful if these people would post
>web pages with images that show a counterexample to this.
>
>In this case, a picture truly is worth a thousand words, and
>I hope this doesn't degenerate into a thread with a thousand
>messages <smile>.



==========================================
Mark Thomas   markthom@camtech.net.au
http://www.adelaide.net.au/~markthom




 




Copyright © Lexa Software, 1996-2009.