- From: Frédéric Kayser <f.kayser@free.fr>
- Date: Wed, 15 Jan 2014 01:27:58 +0100
- To: public-respimg@w3.org
- Cc: Eric Portis <lists@ericportis.com>
- Message-Id: <D73D1B3A-A05E-4C48-B77E-0EFF833F0382@free.fr>
Hi, Setting quality to 50 in Photoshop activated 4:2:0 chroma sub-sampling, but you didn't use sub-sampling for the JPEG2000 file (-s 2,2). To visually see the effects of chroma-subsampling this test pattern can be used: https://bug856375.bugzilla.mozilla.org/attachment.cgi?id=813475 Notice the impact on color when quality is set to 50 and below compared to 51 and more. Eric Portis wrote: > On Tuesday, October 22, 2013 at 12:38 AM, Yoav Weiss wrote: >> I think it would be extremely helpful if someone could run some JP2K tests and prove or disprove Henri's claims. >> The following comment https://groups.google.com/d/msg/mozilla.dev.platform/9NKc7OeEFLM/kH35YH62uMUJ suggests that it's a matter of encoder settings. > > It took a few months, but I finally got around to running some tests: > > http://ericportis.com/posts/2014/resolution-progressive-jpeg2000-performance/ > > The results for JPEG2000 were not good. > > See the encoder settings used in footnote #3. From my reading of the docs, the key one seems to be setting RLCP progression order, vs the default LRCP (which is progressive quality, ala Progressive JPEGs). > > > —eric
Attachments
- text/html attachment: stored
- image/png attachment: quality50vs51.png
Received on Wednesday, 15 January 2014 00:28:30 UTC