* Revised PNG palette optimization
* Added some debug logging in coders/png.c.
* More precise blur values for Lanczos2Sharp and LanczosSharp.
* Added location of first Mitchell crossing (=8/7) to the filters data
structure.
* Added Lanczos2D* filters now named Lanczos2*
* Reorganization of AcquireFilter() to make it work better
* Clearer EWA filters (LanczosSharp etc) comments.
* Added LanczosSharp (3-lobe Lanczos with sharpening)
* Filter sharpening factors are also always applied regardless of usage.
* CubicBC filter formulas simplified by constant folding. In
particular, P1 coefficient (always zero) removed from coeff.
* Revert the Robidoux filter to a Keys cubic with C=(108 sqrt 2-29)/398
(as already specified in the documentation).
* Ignore PS bounding box offsets if -page is set.
* Add support for -evaluate exp.
OBS-URL: https://build.opensuse.org/package/show/graphics/ImageMagick?expand=0&rev=47
* Added "filter:sigma" expert setting defining the 'sigma' for the Gaussian
filter only. This is similar in action to 'blur' but only for Gaussians,
and does not modify the filters support, allowing you to set a very small
sigma, without the function 'missing' all pixels.
* Patch for DrawableRotation() and DrawableTranslation()
* The webp format requires the webpconv delegate program (experimental).
* Replaced "Robidoux" with Cubic 'Keys' filter that is near equivelent to
the previous sharped "Lanczos2D" filter. (C=0.31089212245300069)
This also is very similer to a Mitchell filter but specifically designed
for EWA use and is the new default filter for Distorting Images.
* Added new filter 'Lanczos2D' a 2-lobe Lanczos as defined by
Andreas Gustafsson in his thesis "Interactive Image Warping" (page 24)
http://www.gson.org/thesis/warping-thesis.pdf
* Added filter "Robidoux" which is a slightly sharpened version of the
"Lanczos2D" filter (blur=0.958033808) specifically designed to be less
'blurry' for horizontal and vertical lines in no-op distortions.
* Add support for "pattern:vertical2" and "pattern:horizontal2".
* Add support for "pattern:vertical3" and "pattern:horizontal3".
* Properly handle PSD layers with negative offsets.
* Added sqrt(2) bluring default for Gaussian Filter if used as
a Cylindrical EWA filter. This resulted removing the last aliasing
issue that was present in tests for Gaussian EWA resampling. Of course
it is still a very blury filter for default use in EWA.
* Adjusted Variable Mapping Blur Composition so user arguments actual
relate properly to the sigma of the blur for a maximum mapping value.
* Fix horizon anti-alising for output-scaled perspective distortions.
* 'Bessel' filter is now offically and more accuritally named 'Jinc'
however 'Bessel' while not visible as a filter option can still be used
OBS-URL: https://build.opensuse.org/package/show/graphics/ImageMagick?expand=0&rev=45
* Automatically set the quantum depth to 16 for HDRI.
* IPTC profile not always wrapped properly inside an 8BIM profile.
* TIFF tile geometry must be divisible by 16.
* Rename of SincPolynomial to SincFast for easier user understanding.
Ditto for LanczosChebyshev to LanzcosFast.
* Switch default resize filters to using the faster SincPolynomial
filter by default internally. However 'Sinc' will still use the
Trigonometric function, and can be used to assign the trig version
of Sinc() to filters using the filter expert options.
* The default filter for 'distort' was found to be a very blurry inaccurate
filter function. It was removed and replaced with a correct Gaussian
filter (as used by resize)
* Added a switch so that "-interpolate filter" will force the use of
a cylindrical filter for ALL pixels in distorted images. That is you can
use that switch to use a cylindrical filter even for images that are
being enlarged by the distortion. However EWA is still currently using
a fixed 2.0 sampling radius. This switch complements the use of "-filter
point" which turns off EWA filters in favor of interpolation for all
pixels in a distorted image. BOTH switches should not be used together.
* A bug in the support radius of the EWA resampling function was found,
now that correctly defined resize filters are being used. Suddenly Normal
Gaussian distortions are not so blurry, and tests with distortions of
the 'Rings' image show extremely good and clear results, with only minimal
blurring. The filter 'blur' expert option can be used to adjust this
further.
* Don't negate the geometry offset for the -extent option.
* The RGBO format is now listed as a supported format.
* Added the Nicolas Robidoux and Chantal Racette Lanczos resize filter
OBS-URL: https://build.opensuse.org/package/show/graphics/ImageMagick?expand=0&rev=43
* Eliminate useless message about assuming zero delay when writing
a single-frame MNG, and changed it from Error to Warning when
writing a multiple-frame MNG.
* Only use the first alpha channel in PSD image.
* Only use XPM complying colors for XPM images (e.g. green is rgb(0,255,0)).
* Eliminate bogus "invalid colormap index" when pinging ICO images.
* Support -set density / units.
* Properly map PNG intent to image->rendering_intent
* The orient option sometimes improperly set "undefined" (reference
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=16795).
* Only list orientation options for the -list orientation option (reference
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=16801).
* Return proper standard deviation for combined channels (reference
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=16786).
* Handle transparency properly for the PSD image format.
* Emit a warning if the PNG encoder can't satisfy the color type and
bit depth requested with a "-define" directive.
* The -fx 2e+6/1e+5 argument no longer returns the wrong results (reference
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=16833).
* Add -subimage-search option to the compare utility.
* Throw exception if image size differs for the compare program but the
-subimage-search option is not specified.
OBS-URL: https://build.opensuse.org/package/show/graphics/ImageMagick?expand=0&rev=38
