| Mon, 25 Oct 2010 18:34:44 +0900 |
matthiasm |
added chord change value, some other tweaks
matthiasm-plugin
|
| Mon, 25 Oct 2010 11:26:38 +0200 |
Chris Cannam |
* nnls_chroma -> nnls-chroma in cat file; don't track the .o file!
matthiasm-plugin
|
| Mon, 25 Oct 2010 16:58:32 +0900 |
matthiasm |
removed subtraction in chroma dictionary, added to-the-power-of-1.5 in chordino
matthiasm-plugin
|
| Mon, 25 Oct 2010 02:33:47 +0900 |
matthiasm |
* tiny update of README, more tomorrow
matthiasm-plugin
|
| Mon, 25 Oct 2010 02:32:32 +0900 |
matthiasm |
* wondering whether I can make bullet lists ...
matthiasm-plugin
|
| Mon, 25 Oct 2010 02:30:05 +0900 |
matthiasm |
* matthiasm -> nnls-chroma; * nnls_chroma -> nnls-chroma
matthiasm-plugin
|
| Mon, 25 Oct 2010 01:58:37 +0900 |
matthiasm |
regarding the chord estimation:\n * tweaked chord templates\n * that means that the original method also changed
matthiasm-plugin
|
| Mon, 25 Oct 2010 00:52:39 +0900 |
matthiasm |
* HMM implementation for chord recognition
matthiasm-plugin
|
| Sun, 24 Oct 2010 20:43:11 +0900 |
matthiasm |
* added new parameter that toggles NNLS
matthiasm-plugin
|
| Fri, 22 Oct 2010 21:43:57 +0900 |
Matthias Mauch |
README and new parameters, not tested yet
matthiasm-plugin
|
| Fri, 22 Oct 2010 11:58:16 +0100 |
Chris Cannam |
* Add skeleton plugin description RDF
matthiasm-plugin
|
| Fri, 22 Oct 2010 11:44:08 +0100 |
Chris Cannam |
* Add an approximate category file
matthiasm-plugin
|
| Fri, 22 Oct 2010 11:37:35 +0100 |
Chris Cannam |
* Fix empty spectrograms
matthiasm-plugin
|
| Fri, 22 Oct 2010 11:32:47 +0100 |
Chris Cannam |
* tuning -> Tuning
matthiasm-plugin
|
| Fri, 22 Oct 2010 11:31:54 +0100 |
Chris Cannam |
* Rename outputs (so alphabetical ordering works better!)
matthiasm-plugin
|
| Fri, 22 Oct 2010 11:30:21 +0100 |
Chris Cannam |
* Split out NNLSChroma plugin into three plugins (chroma, chordino, tuning) with a common base class.
matthiasm-plugin
|
| Fri, 22 Oct 2010 16:19:40 +0900 |
Matthias Mauch |
changed Makefile: added chromamethods to the plugin code objecs, minimum OSX set to 10.5 (and no ppc)
matthiasm-plugin
|
| Thu, 21 Oct 2010 21:51:36 +0100 |
Chris Cannam |
Merge
matthiasm-plugin
|
| Thu, 21 Oct 2010 20:56:27 +0100 |
Chris Cannam |
Merge
matthiasm-plugin
|
| Thu, 21 Oct 2010 19:23:02 +0100 |
Chris Cannam |
* More tidying
matthiasm-plugin
|
| Thu, 21 Oct 2010 21:48:04 +0100 |
Chris Cannam |
* Use proper Vamp path to look up chord.dict
matthiasm-plugin
|
| Thu, 21 Oct 2010 20:50:22 +0100 |
Chris Cannam |
* Revert to the previous version of nnls.c[c], which used single-precision floats (better for us).
matthiasm-plugin
|
| Thu, 21 Oct 2010 20:19:29 +0100 |
Chris Cannam |
* minor tidy
matthiasm-plugin
|
| Thu, 21 Oct 2010 16:34:58 +0100 |
Chris Cannam |
* Split out some common code into chromamethods.cpp from NNLSChroma.cpp
matthiasm-plugin
|
| Thu, 21 Oct 2010 14:50:24 +0100 |
Chris Cannam |
* Update Makefile
matthiasm-plugin
|
| Thu, 21 Oct 2010 14:38:02 +0100 |
Chris Cannam |
* OK, let's revert to a C version (after discovering that the FORTRAN compiler isn't standard on the Mac)
matthiasm-plugin
|
| Thu, 21 Oct 2010 12:21:41 +0100 |
Chris Cannam |
* Remove some obsolete files
matthiasm-plugin
|
| Thu, 21 Oct 2010 12:12:23 +0100 |
Chris Cannam |
* Consistent indentation (spaces only)
matthiasm-plugin
|
| Thu, 21 Oct 2010 11:58:28 +0100 |
Chris Cannam |
* Rather than worry about provenance of C versions, why not just use the FORTRAN? (We can back this out if it doesn't go well with the build scripts)
matthiasm-plugin
|
| Mon, 18 Oct 2010 09:35:26 +0000 |
matthiasm |
currentstate
matthiasm-plugin
|
