Is this in master or 1.2? We very recently checked in a change to rangecompress/rangeexpand that contains different coefficients for the transformation, for this exact reason. The new coefficients are much better at avoiding black pixels from the negative lobes. This was a simple mistake all along: the program from which SPI so kindly donated the formula had two different sets of coefficients, which we didn't realize at the time that we took the *wrong* ones that were not actually used for resizing. The new ones we just put in as replacements are the ones actually used and which work much better.
Give a try to master and see if it helps? On Thu, Oct 10, 2013 at 5:30 AM, Ben De Luca <[email protected]> wrote: > Hey thanks for that! > > Im looking at the various filters to try and determine how we are going to > deal with the ringing artifacts we have on areas that are superbright. I > saw rangecompress function, which helps but at least in my test image the > specular reflection component is so bright I get the black artifacts with > almost any filter with negative lobes. > > > IBA::rangecompress()/rangeexpand(), and oiiotool --rangecompress / > --rangeexpand compress the excess >1 values of HDR images to a log > scale (leaving the <= 1 part linear), and re-expand to the usual > linear scale. This is very helpful to reduce ringing artifacts that > can happen when an HDR image is resized with a good filter with negative > lobes (such as lanczos3), by doing a range compression, then the resize, > then range expansion. It's not mathematically correct and loses energy, > but it often makes a much more pleasing result. > > > On Wed, Oct 9, 2013 at 6:35 PM, Larry Gritz <[email protected]> wrote: > >> I think you want >> >> oiiotool input.ext --resize:filter=blackman-harris 50% -o >> /tmp/test.exr >> >> options are appended to commands with the ":name=val" format. >> >> Filter choices (see end of filter.cpp) are box, triangle, gaussian, >> sharp-gaussian, catmull-rom, blackman-harris, sinc, lanczos3, >> radial-lanczos3, mitchell, bspline, disk. >> >> The default -- which I think pretty good -- is to use blackman-harris >> when magnifying, and lanczos3 when minifying. >> >> >> >> On Oct 9, 2013, at 5:04 AM, Ben De Luca wrote: >> >> > Hi All, >> > I was trying to pass a filter option to the resize command, but I >> cant figure out what the format should look like. >> > >> > I thought it might be some thing like. >> > >> > ./oiiotool -v input.ext --resize 50% filter=blackman-harris -o >> /tmp/test.exr >> > >> > Is that the right sort of format? I looked that the code and I cant >> actually see how I can pass in the filter option. >> > >> > Also, is there a list of implemented filters? >> > >> >> -- >> Larry Gritz >> [email protected] >> >> >> _______________________________________________ >> Oiio-dev mailing list >> [email protected] >> http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org >> > > > _______________________________________________ > Oiio-dev mailing list > [email protected] > http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org > > -- Larry Gritz [email protected]
_______________________________________________ Oiio-dev mailing list [email protected] http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org
