rkalwaitis
4th November 2009, 09:15
I hope this is not to silly of a question. Ive been trying to read the corefilter descriptions and apply them in a practice. Of course this is aimed at answering my questions and self education, but as you all know it brings about more questions than answers. So please be patient with me.
Okay the question Ive been farting around with generalconvolution3d. Its pretty cool, i can blur and sharpen. Although Im not so sure how they come up with those numbers, but im still working on that part. So Im reading the documentation and doing my little avs files to look at what happens and its going along fine. I got really excited when they started taking about motion for generalconvolution and there was no example.
And if this is possible (generalconvolution to show motion) I would like to try and use the core filters of avisynth (learning purposes) to do something like previous,current,next frame to apply to temporalsoften. I know it would be slow, but this is okay.
I think I have a really good blur for a prefilter. GeneralConvolution(0,"1 2 1 2 4 1 1 2 1").Generalconvolution(0,"2 4 2 4 8 4 2 4 2").generalconvolution(0,"1 2 1 2 4 1 1 2 1"). I borrowed it from Convolution3D. It is the basis for the application of their denoising process. The documentation stated that the 3x3 box is usually more sufficient and I think it is okay to start smaller and work up.
Can some on please point me to another source that will give me an example, just not an algorythm? Or provide one for me. So that I can play with it.:thanks:
Okay the question Ive been farting around with generalconvolution3d. Its pretty cool, i can blur and sharpen. Although Im not so sure how they come up with those numbers, but im still working on that part. So Im reading the documentation and doing my little avs files to look at what happens and its going along fine. I got really excited when they started taking about motion for generalconvolution and there was no example.
And if this is possible (generalconvolution to show motion) I would like to try and use the core filters of avisynth (learning purposes) to do something like previous,current,next frame to apply to temporalsoften. I know it would be slow, but this is okay.
I think I have a really good blur for a prefilter. GeneralConvolution(0,"1 2 1 2 4 1 1 2 1").Generalconvolution(0,"2 4 2 4 8 4 2 4 2").generalconvolution(0,"1 2 1 2 4 1 1 2 1"). I borrowed it from Convolution3D. It is the basis for the application of their denoising process. The documentation stated that the 3x3 box is usually more sufficient and I think it is okay to start smaller and work up.
Can some on please point me to another source that will give me an example, just not an algorythm? Or provide one for me. So that I can play with it.:thanks: