There is no reason to go back the other way around toward JPEG lossless and/or JPEG-LS since those are 'compression' only codec ('winzip' on steroid for images). By now we should all be doing -at least- progressive JPEG for sharing our 4K family pictures that will be shown on a smartphone. Sorry again if this totally out of scope on this newsgroup, but the argument has remained the same since 1992, when hierarchical JPEG mode (ITU T.81) had been introduced. In this case, I would really appreciate that you used 'Jasper', 'OpenJPEG' (put a name.), instead of making people believe that (as a standard) 'JPEG 2000' is a poor standard: that can easily out-performed by something like JPEG-LS. Your argument(s) seems to always be on the quality implementation side. I've already mentioned on Twitter my point of view, since I believed discussing J2K implementation details were totally out of scope on c.p.dicom. I have also found claims of 'better' lossless compression (in the post CALIC/JPEG-LS/J2K era) to often be false, falsified, impractical, not generalizable across modalities, or of such minuscule improvement as to not justify the interoperability hit.įor me, interoperability is a much higher priority than squeezing out the last drop of compression performance ( ) I have always found JPEG-LS to be much simpler to implement, faster, less resource (memory and CPU) intensive and generally compresses as well or better than J2K reversible, BUT it has neither the lossy (irreversible) nor other features of J2K, so if one wants one scheme that does (almost) everything, I can see why some folks gravitate to J2K. What are these 'new techniques have been discovered that can produce higher lossless compression ratios than JPEG2000', and just how much better do they compress? > I just posted a new blog entry on JPEG 2000 that members of this group might find interesting:

On Thursday, Septemat 7:05:50 AM UTC-4, Chris Hafey wrote:

J2k compression