| Thu, 25 Jun 2020 09:32:01 +0100 |
Chris Cannam |
Make querying and setting the external alignment program or transform separate from selecting the alignment type - we need it to work that way for a clearer UI
pitch-align
|
| Tue, 16 Jun 2020 15:17:50 +0100 |
Chris Cannam |
Use ModelId for recording model; add mix-to-mono option
smoother-recording
|
| Wed, 03 Jun 2020 13:58:29 +0100 |
Chris Cannam |
Debug tweak only
|
| Mon, 01 Jun 2020 17:13:09 +0100 |
Chris Cannam |
Fixes to aligner destruction sequence when re-aligning during alignment; hide debug output behind a flag
pitch-align
|
| Mon, 01 Jun 2020 17:12:12 +0100 |
Chris Cannam |
Prefer diagonal if the cost is the same - this makes the results the same as our earlier test aligner
pitch-align
|
| Fri, 29 May 2020 17:39:02 +0100 |
Chris Cannam |
Allow use of proper sparse models (i.e. retaining event time info) in alignment; use this to switch to note alignment, which is what we have most recently been doing in the external program. Not currently producing correct results, though
pitch-align
|
| Thu, 28 May 2020 17:52:19 +0100 |
Chris Cannam |
Implement trimmed linear aligner
pitch-align
|
| Thu, 28 May 2020 17:04:36 +0100 |
Chris Cannam |
Fixes to notification and completion in aligners
pitch-align
|
| Fri, 22 May 2020 17:17:44 +0100 |
Chris Cannam |
Provide callback for output preprocessing before DTW, use it for freq-pitch conversion; use direct setting of completion on alignment models instead of creating fake outputs for completion only
pitch-align
|
| Thu, 21 May 2020 16:21:57 +0100 |
Chris Cannam |
Make a start on further alignment methods
pitch-align
|
| Thu, 14 May 2020 16:38:48 +0100 |
Chris Cannam |
Profiling points
|
| Thu, 14 May 2020 16:38:37 +0100 |
Chris Cannam |
Profiling points and comment
|
