mirror/kissfft
1
0
Fork 0
mirror of https://github.com/mborgerding/kissfft.git synced 2025-05-27 21:20:27 -04:00
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • ca97282da6 does no copy fast conv filtering Mark Borgerding 2004-01-24 00:46:30 +0000
  • 227c021f39 extra cleanup Mark Borgerding 2004-01-04 23:15:44 +0000
  • c80f0cafb7 test code for avoiding buffer copies of output in fast conv filtering Mark Borgerding 2004-01-04 06:32:25 +0000
  • 8ffc08b07b slightly closer, still overflows in some cases Mark Borgerding 2004-01-02 22:28:03 +0000
  • e749a114cb real fast convolution filtering works mostly, sometimes it does not make all the samples ( obeserved with ./fastfir.py -n 1024 -r -l 100000 ) Mark Borgerding 2004-01-01 18:51:17 +0000
  • 0243552944 fast convolution filtering works at least for complex numbers. Mark Borgerding 2003-12-31 22:49:14 +0000
  • 212ec326e0 prototype code for fast convolution filtering Mark Borgerding 2003-12-31 04:06:24 +0000
  • b468fd96d7 Added mucho comments. Mark Borgerding 2003-12-30 15:18:41 +0000
  • ec3b64a62e added test code for kiss_fftnd FFTs can now have an arbitrary # of dimensions Also, buffer copies are eliminated. Mark Borgerding 2003-12-30 05:56:50 +0000
  • 377ac796f0 streamlined Mark Borgerding 2003-12-30 03:08:56 +0000
  • 22ff70b71d n-dimensional FFT works Mark Borgerding 2003-12-30 02:50:45 +0000
  • 5e6199c868 no real change Mark Borgerding 2003-12-30 00:09:50 +0000
  • 62ecf69446 added prototyping for N-dimensional FFT Mark Borgerding 2003-12-29 23:05:45 +0000
  • 10a4a5c46b 2d fft now uses kiss_fft_stride to avoid buffer copies. Mark Borgerding 2003-12-29 21:07:19 +0000
  • 4552a14801 added stride to fft input -- should simplify multi-dimensional FFTs Mark Borgerding 2003-12-29 18:50:13 +0000
  • 604e238834 split kfc.c into source + header files. Added copyright. Mark Borgerding 2003-12-29 17:05:13 +0000
  • 630089ded6 kfc works Mark Borgerding 2003-12-29 16:46:27 +0000
  • cd94cd1ed0 testing kfc (kiss fft cache) Mark Borgerding 2003-12-29 16:38:03 +0000
  • 0bd304429d file to cache fft cfg objects for ease of use. Mark Borgerding 2003-12-26 03:59:46 +0000
  • f4961ed74b fixed a bug with nfft==1 and added utility for more testing Mark Borgerding 2003-12-16 03:43:05 +0000
  • 95a7b856d1 compiles OK under MAC OS X Mark Borgerding 2003-12-15 03:53:14 +0000
  • 6b23ebb5c1 getting ready for v100 Mark Borgerding 2003-12-15 03:23:46 +0000
  • 573536f48f fixed scaling for short real moved fft.py from ./ to sample_code Mark Borgerding 2003-12-14 05:13:11 +0000
  • 559c14b49b changed alloc functions to allow for initialization of user-supplied buffer Mark Borgerding 2003-12-14 03:02:30 +0000
  • ab32979a47 *** empty log message *** Mark Borgerding 2003-12-11 01:56:18 +0000
  • ef7f8ff6c0 getting ready to release version 1.0 Mark Borgerding 2003-12-11 01:54:42 +0000
  • 8a992d0e98 better, reduced mutliply in real fft Mark Borgerding 2003-12-10 05:48:42 +0000
  • b880c599ef optimizing Mark Borgerding 2003-12-10 05:20:21 +0000
  • b737756abb real fwd and rev fft works Mark Borgerding 2003-12-09 04:10:50 +0000
  • 14e9b0dce4 about to start working on real inverse fft optimzation Mark Borgerding 2003-12-09 01:51:29 +0000
  • 008d7fd4ea zero case optimized Mark Borgerding 2003-12-09 01:37:48 +0000
  • 548032e305 real optimization Mark Borgerding 2003-12-09 01:25:47 +0000
  • db56a8b1e2 little better, faster Mark Borgerding 2003-12-09 01:19:08 +0000
  • 09dbf36e97 kiss_fftr.c gives correct answers for forward Mark Borgerding 2003-12-07 04:04:27 +0000
  • c2eaab8606 *** empty log message *** Mark Borgerding 2003-12-05 03:51:33 +0000
  • aaa04d9b60 script to make test vectors using Python and the FFT module Mark Borgerding 2003-12-05 03:40:53 +0000
  • 72de907478 minor doc changes Mark Borgerding 2003-12-05 00:35:25 +0000
  • 39d2974fe3 'make snr' now tests 2d ffts Mark Borgerding 2003-12-04 04:08:52 +0000
  • f3c4a9e9ca internal checkpoint split kiss_fft.c into multiple files. It seems the best approach, all things considered. Mark Borgerding 2003-12-04 02:38:50 +0000
  • 164ab47a25 *** empty log message *** Mark Borgerding 2003-11-14 04:20:19 +0000
  • 63e7df14cf fft prototype code for real fft works Mark Borgerding 2003-11-14 02:11:57 +0000
  • 968e277f93 working on real optimization Mark Borgerding 2003-11-14 01:37:31 +0000
  • 673c7bafa4 real_fft works Mark Borgerding 2003-11-12 02:49:04 +0000
  • 206e28f11c made the factorization a separate routine Mark Borgerding 2003-11-12 01:09:35 +0000
  • 140f106743 benchmark utilities now give memory info Mark Borgerding 2003-11-08 02:48:02 +0000
  • bd23fe8d23 the path I was taking would only work for prime numbers (Galois fields) Mark Borgerding 2003-11-08 01:42:15 +0000
  • e98f9ff29a going to bed Mark Borgerding 2003-11-07 03:42:14 +0000
  • ae305ca400 slightly faster Mark Borgerding 2003-11-07 03:31:59 +0000
  • 3a5791f203 slightly faster Mark Borgerding 2003-11-07 03:23:19 +0000
  • 1486d89291 faster Mark Borgerding 2003-11-07 03:04:52 +0000
  • e9095a161c generic butterfly slightly slower -- hmmm Mark Borgerding 2003-11-07 02:39:49 +0000
  • a296b09dbf wrong alloc size Mark Borgerding 2003-11-07 01:06:44 +0000
  • cb5312efdc 2d fft seems to work Mark Borgerding 2003-11-06 03:59:31 +0000
  • 4c458be5e9 checkpoint -- I don't think I've broken anything (yet) adding 2d fft. Mark Borgerding 2003-11-04 23:25:49 +0000
  • ee3094a0e4 benchmark utilities Mark Borgerding 2003-11-04 02:11:00 +0000
  • 4ebf0b5aca aded a CHANGELOG Mark Borgerding 2003-11-04 02:09:53 +0000
  • 2788fba0bd added a CHANGELOG Mark Borgerding 2003-11-04 02:09:48 +0000
  • 8b4e3bacca minor comments and added some primes Mark Borgerding 2003-11-04 02:00:01 +0000
  • 6c8049cc75 slight changes to Makefile Mark Borgerding 2003-11-04 01:01:37 +0000
  • 7b4de0aa11 a little faster Mark Borgerding 2003-11-03 04:30:50 +0000
  • ad4ee571aa faster radix5 Mark Borgerding 2003-11-03 04:04:01 +0000
  • 0403fb3e4a radix 5 a little optimized Mark Borgerding 2003-11-03 03:48:34 +0000
  • 3c0c0431e2 radix 5 works, but is 6x slower than fftw Mark Borgerding 2003-11-03 03:03:16 +0000
  • 85764e6437 radix 5 doesn't work, but I thik it should. Mark Borgerding 2003-11-01 16:48:33 +0000
  • 8ac63adc77 modified time benchmark to repeat same buffer over and over to avoid IO bottlenecks and get more consistent numbers. Mark Borgerding 2003-11-01 04:44:50 +0000
  • 471803ca08 removed unused macro Mark Borgerding 2003-11-01 04:26:02 +0000
  • 7b7aefe7c4 moved scratch buffer to stack variable Mark Borgerding 2003-11-01 03:59:43 +0000
  • 28551899e2 radix 4 faster Mark Borgerding 2003-11-01 03:49:53 +0000
  • d1df249536 radix3 fixed point now works Mark Borgerding 2003-10-31 04:01:09 +0000
  • b1969544a6 radix 3 still doesn't work for fixed Mark Borgerding 2003-10-30 03:00:49 +0000
  • d4f87befda re-added radix 3 butterfly Mark Borgerding 2003-10-30 02:02:29 +0000
  • ca4c74e07c Woops, one should not test with input of all zeros Mark Borgerding 2003-10-29 04:29:01 +0000
  • 97b18f3fef comments Mark Borgerding 2003-10-27 04:02:11 +0000
  • d9fcda04b6 version 0.2 upload to sf Mark Borgerding 2003-10-26 19:29:36 +0000
  • ecb1a76974 added zip creation to tarball make target Mark Borgerding 2003-10-26 04:25:18 +0000
  • 1db3d91ee5 getting ready for next release Mark Borgerding 2003-10-26 04:07:32 +0000
  • 52b4b9ab5c *** empty log message *** Mark Borgerding 2003-10-18 01:45:26 +0000
  • c239ba2c1c slight code cleanup, comments Mark Borgerding 2003-10-18 01:39:36 +0000
  • bca7fd5151 compiles with -ansi -pedantic Mark Borgerding 2003-10-18 01:23:34 +0000
  • e2470b3a03 *** empty log message *** Mark Borgerding 2003-10-18 00:33:38 +0000
  • a3d3217ae6 *** empty log message *** Mark Borgerding 2003-10-18 00:32:54 +0000
  • 6f8bcedc24 radix 3 fixed point still broken Mark Borgerding 2003-10-17 02:59:32 +0000
  • 31d4214f44 radix 3 seems to be pretty fast Mark Borgerding 2003-10-17 02:34:22 +0000
  • 73744b908c check point Mark Borgerding 2003-10-17 01:26:14 +0000
  • 317f11e66e starting point for radix 3 Mark Borgerding 2003-10-17 00:11:19 +0000
  • d6ae498630 took the bitwise and out of the switch case -- may have prevented optimization Mark Borgerding 2003-10-15 03:45:24 +0000
  • 5f0efe8f17 pretty happy with radix 2 and radix 4 next up is radix 3, or maybe 5 Mark Borgerding 2003-10-15 03:38:05 +0000
  • 9504aa79c1 Fixed generic mixed radix butterfly Mark Borgerding 2003-10-15 02:52:34 +0000
  • 0424734e9d radix 4 now about as fast as original version Mark Borgerding 2003-10-15 01:52:13 +0000
  • f609401471 about to make some changes -- just wanted a checkpoint Mark Borgerding 2003-10-15 00:05:50 +0000
  • 2ae7e0f1f2 radix 4 works but slow Mark Borgerding 2003-10-14 02:47:25 +0000
  • 6b76490456 Fixed point works Mark Borgerding 2003-10-14 01:09:33 +0000
  • 8460f1f8f5 added optimization for radix 2 Mark Borgerding 2003-10-14 00:38:58 +0000
  • 0d6d61cfce reduced calling parameters Mark Borgerding 2003-10-11 23:07:16 +0000
  • 0d44569b3b made one single malloc for all buffers Mark Borgerding 2003-10-11 23:00:12 +0000
  • f93a0258df Simplified some inner loop calcs Mark Borgerding 2003-10-11 22:45:35 +0000
  • 911d29d139 changed from static function that wasn't inlining very well to a macro Mark Borgerding 2003-10-11 22:41:17 +0000
  • 11983e5056 used += on complex components Mark Borgerding 2003-10-11 22:39:40 +0000
  • 043da3b65d avoid last recursive call Mark Borgerding 2003-10-11 14:43:13 +0000
  • 7ec9402d5b Fixed point works (in the loosest sense of the word "works") Mark Borgerding 2003-10-11 14:34:01 +0000