"We won't provide any conclusions because the results in the given tests
context are self explanatory."
This is another way to say: "Trying to put things in context so that the
results are relevant to real life would expose us to criticism, so we'll
say it's obvious so that other people get the du
Ravi:
The other thing you should know is that sigmoidization
http://www.imagemagick.org/Usage/resize/#resize_sigmoidal should be ripped
out of the LoHalo code: One can do better than to enlarge through linear
light when the filter has negative lobes, but I ascertained last Summer
that sigmoidizati
Note: The current nohalo sampler is interpolatory. So are nearest
neighbour, bilinear, and Lanczos. What this means is that if such samplers
are combined with fast paths, AND the arithmetic is spotless, there would
not be a "jump".
lohalo, which basically is an improved Mitchell-Netravali bicubic,
I stand by Comments 6 and 7 of [1].
Summary: The only way I can see fast paths working well without very very
intricate programming (including round off error tracking etc) and without
making the image "jump" in some situations is to make absolutely sure that
the branching to the fast path occurs
Proof of concept attempt to have VIPS used as a backend for GEGL:
https://github.com/jcupitt/gegl-vips.
Discussion of the topic at
http://gimp.1065349.n5.nabble.com/gegl-vips-td29248.html.
My gut feeling is that although there is a lot of good code and
architecture to be "poached" from the VIPS l
GIMP :)
___
gegl-developer-list mailing list
gegl-developer-list@gnome.org
https://mail.gnome.org/mailman/listinfo/gegl-developer-list
http://www.gimpusers.com/forums/gimp-developer/15360-ways-to-improve-gimp-2-9-performance
Nicolas Robidoux
Senior Research Scientist
Phase One phaseone.cmo
___
gegl-developer-list mailing list
gegl-developer-list@gnome.org
https://mail.gnome.org/mailman
TTBOMK, http://run-edge-laplace-broken.xml is kept there as a reminder that
something needs to be looked into. This is what the "broken" suffix means.
___
gegl-developer-list mailing list
gegl-developer-list@gnome.org
https://mail.gnome.org/mailman/listin
>../haus.png
>
>
>
>
>
> Thanks for any hints,
> Torsten.
>
> ___
> gegl-developer-list mailing list
> gegl-developer-list@gnome.org
> https://mail.gno
> Simply put, after resizing with either of them the final image has a 1px
wide transparent border. That doesn't look right to me :)
Isn't this, basically, step 6) of
http://www.freedomsponsors.org/core/issue/78/add-other-samplers-that-properly-reduce-downsample-and-warp-images
?
_
Dear Jon:
The very last commit to the branch ("Added nohalo sampler (VSQBS).")
contains incomplete code.
Don't merge it. It's to be kept on ice.
I and/or Adam will have a look at what came before in the branch ASAP.
--
Nicolas Robidoux
Image processing and applied mathemat
11 matches
Mail list logo
