Re: [Image-SIG] Radius bug with GaussianBlur

2010-04-25 Thread Fredrik Lundh
On Mon, Mar 1, 2010 at 10:49 PM, Stani wrote: > Strange indeed. The same is true for UnsharpMask: More silly than strange, I'd say -- looks like a cut-and-paste-propagated typo that weren't caught by the (rather shallow) filter tests. Trivial patch here: http://hg.effbot.org/pil-2009-raclet

Re: [Image-SIG] Radius bug with GaussianBlur

2010-03-01 Thread Stani
Strange indeed. The same is true for UnsharpMask: class UnsharpMask(Filter): name = "UnsharpMask" def __init__(self, radius=2, percent=150, threshold=3): self.radius = 2 self.percent = percent self.threshold = threshold def filter(self, image): return i

[Image-SIG] Radius bug with GaussianBlur

2010-02-22 Thread Tiberio Uricchio
Hi, i just wanted to report a simple bug in ImageFilter.GaussianBlur: class GaussianBlur(Filter): name = "GaussianBlur" def __init__(self, radius=2): self.radius = 2 def filter(self, image): return image.gaussian_blur(self.radius) As you can see, radius is always fix