| commit | 818efc820d723addb9852928532a401817baf681 | |
| author | martin <martin@cyberpower.kcl.ac.uk> | |
| Fri, 27 Aug 2010 20:43:01 +0000 (27 21:43 +0100) | ||
| committer | martin <martin@cyberpower.kcl.ac.uk> | |
| Fri, 27 Aug 2010 20:43:01 +0000 (27 21:43 +0100) | ||
| tree | 4f9140f6b6916903d60e67584824a0e07f85f726 | treesnapshot (tar.gz zip) |
| parent | 312c6aeed310949325ea7272f0bf47e305de9f15 | commitdiff |
played with convolution
convolve-circ gives an inverted result (with cuda-fft)
i'll have to check if fftw works. there was a bug in
convolve-circ - the kernel should be fftshifted.
in run.lisp there is convolve-nocirc that calls
convolve-nocrop and crops the result to the original
array size
convolve-circ gives an inverted result (with cuda-fft)
i'll have to check if fftw works. there was a bug in
convolve-circ - the kernel should be fftshifted.
in run.lisp there is convolve-nocirc that calls
convolve-nocrop and crops the result to the original
array size
| frontend/angular-psf.lisp | diffblobblamehistory | |
| frontend/draw-model.lisp | diffblobblamehistory | |
| run.lisp | diffblobblamehistory | |
| vol.asd | diffblobblamehistory | |
| vol/convolve.lisp | diffblobblamehistory |