| Tue, 08 Oct 2013 08:18:11 +0100 |
Chris Cannam |
Pad or truncate fft input as appropriate
|
| Tue, 08 Oct 2013 08:18:00 +0100 |
Chris Cannam |
Square wave is a bit cheaper
|
| Tue, 08 Oct 2013 08:17:50 +0100 |
Chris Cannam |
Cache Kaiser windows
|
| Tue, 08 Oct 2013 07:58:09 +0100 |
Chris Cannam |
Some tidying
|
| Tue, 08 Oct 2013 07:57:27 +0100 |
Chris Cannam |
Cache the factorials, makes things slightly quicker
|
| Sat, 05 Oct 2013 11:29:39 +0100 |
Chris Cannam |
Fix silliness
|
| Fri, 04 Oct 2013 21:15:48 +0100 |
Chris Cannam |
Use arraycopy and fill
|
| Fri, 04 Oct 2013 15:52:25 +0100 |
Chris Cannam |
Use unzipped subsets of filter array
|
| Fri, 04 Oct 2013 15:52:02 +0100 |
Chris Cannam |
Tidier output
|
| Fri, 04 Oct 2013 12:01:19 +0100 |
Chris Cannam |
Make matrix filter always return a sparse matrix; add matrix "any" and "all"; use these in tests
|
| Fri, 04 Oct 2013 11:01:18 +0100 |
Chris Cannam |
Get correct number of samples at the end of resample
|
| Tue, 01 Oct 2013 14:55:12 +0100 |
Chris Cannam |
Store resample pending buffer "virtually" without actually representing all the zeroes
|
| Tue, 01 Oct 2013 07:57:20 +0100 |
Chris Cannam |
Checkpointing half-overhauled tests
|
| Mon, 30 Sep 2013 12:55:48 +0100 |
Chris Cannam |
We now have a "direct" resampler, then, though at the moment it's just as slow as the two-stage one
|
| Mon, 30 Sep 2013 12:36:05 +0100 |
Chris Cannam |
Test forward-magnitude
|
