Legacy ImageMagick Discussions Archive

NicolasRobidoux wrote:

henrywho wrote: Yes, definitely better "moire fighting".... if you are talking about "-define filter:window=Spline -define filter:lobes=3 -distort resize"

Yes, I was.
It looks like "thin whiskers" do give something, at least in EWA.

Your 'thin whiskers' was puzzling me, until I realised just now you mean a tappering windowing function, which gives a very slight contribution from the lobes further away. Yes for Morie handling I would think that would be the case, with Hamming being a very special class of this type of thing.

Anthony:
I've not double checked, but I think that the default number of lobes may be 3 across the board when -distort Resize.
Drop the "lobes" commands from your defines...
(I'm actually happy with that.)

NicolasRobidoux wrote:Anthony:
I've not double checked, but I think that the default number of lobes may be 3 across the board when -distort Resize.
Drop the "lobes" commands from your defines...
(I'm actually happy with that.)

Hmmm Yes. When Sinc function gets substituted it looks like all EWA Windowed Jinc is given a 3 lobe version.

The problem is the support comes (as is logical) from the weighting function. So when Jinc weight was substituted, the later support looking also gets changed.

I'll have to have a look at how to fix this.

If you fix things so that the "default with Jinc is always 3-lobes" is no more, please let me know: I'll have to put some "-define filters:lobes=3" in "The Recommendations".

If it appears it is likely in 3 weeks, after my return from China and I re-sync my sources with the offical ones. I get a lot of work done in china, at least in previous trips. My not in this one as it is shorter than others.