Chris Cannam
|
afd1b1f86c
|
Ensure we consume when called with final=true and no samples - this was the default behaviour in R2 and also initially in R3 but changed when resample-before modes were added
|
2023-02-20 15:25:27 +00:00 |
|
Chris Cannam
|
0fb8b19ab5
|
Working on further tests
|
2023-02-20 09:53:20 +00:00 |
|
Chris Cannam
|
b317d0e81a
|
Fix handling of final block in speex resampler - there is no guarantee a single pad call will return enough data on its own either
|
2023-01-17 12:34:50 +00:00 |
|
Chris Cannam
|
cc3e265b26
|
Update copyright dates
|
2023-01-10 11:10:06 +00:00 |
|
Chris Cannam
|
549906e033
|
Remove the last occurrences of "using namespace std"
|
2023-01-10 11:04:50 +00:00 |
|
Chris Cannam
|
c871f92b2e
|
Merge
|
2022-11-16 10:13:20 +00:00 |
|
Chris Cannam
|
3960bca0ab
|
Make this exception optional on NO_EXCEPTIONS, as the others are (required for Android)
|
2022-11-16 10:12:41 +00:00 |
|
Chris Cannam
|
50eb332ec8
|
Fixes suitable for mingw32 builds
|
2022-11-08 08:32:05 +00:00 |
|
Chris Cannam
|
52fc576500
|
Remove obsolete file
|
2022-09-30 10:36:35 +01:00 |
|
Chris Cannam
|
1bd266cdb6
|
Defend against NaN/Inf and negative ratios
|
2022-09-30 10:36:29 +01:00 |
|
Chris Cannam
|
50b231ee23
|
Merge
|
2022-09-29 16:26:46 +01:00 |
|
Chris Cannam
|
11d5deff62
|
Reverting these tweaked values to their 3.0 state - this change wasn't an obvious improvement and just meant that the output no longer matched the last round of listening tests, so revert it
|
2022-09-29 15:57:11 +01:00 |
|
Chris Cannam
|
f194becadf
|
Use number of actually-populated bands, not total number
|
2022-09-29 15:45:47 +01:00 |
|
Chris Cannam
|
de56cd114a
|
Update the resampler logic to follow that of R2 more - ignore the pitch hq/hs setting in offline mode entirely. The alternative is too tricky to handle elegantly in the command line tool, which I think means too complicated
|
2022-09-29 11:19:21 +01:00 |
|
Chris Cannam
|
1dc6abdb96
|
Remove libresample support. It hasn't been exercised in years and nobody should be using it in preference to the other options
|
2022-09-27 08:47:33 +01:00 |
|
Chris Cannam
|
9e423cdd8c
|
Handle non-ideal sample rates in R3; document sample rate parameter
|
2022-09-26 16:02:13 +01:00 |
|
Chris Cannam
|
57d59b4c7a
|
Eliminate the mlock logic entirely. It's caused a couple of issues lately and it turns out it wasn't even used!
|
2022-09-26 10:11:16 +01:00 |
|
Chris Cannam
|
020702e7ea
|
Fix include path. Very poor that this wasn't exercised by my build or CI runs
|
2022-09-26 09:37:55 +01:00 |
|
Chris Cannam
|
56058cc288
|
Fix plainly dumb error
|
2022-09-20 10:27:16 +01:00 |
|
Chris Cannam
|
a5e2bfa444
|
Fix some (benign) shadowing warnings
|
2022-09-20 09:30:48 +01:00 |
|
Chris Cannam
|
e90f377600
|
Don't use drand48, it doesn't always exist and since we don't care about the quality of output, it's not worth testing for
|
2022-09-20 09:30:34 +01:00 |
|
Chris Cannam
|
2eb9a20052
|
There's no longer any good reason to do this; remove it
|
2022-09-09 11:22:53 +01:00 |
|
Chris Cannam
|
92362bdba0
|
Resample in chunks as well, to avoid overrunning resample buffer
|
2022-09-02 15:08:29 +01:00 |
|
Chris Cannam
|
021de9d51c
|
Add resample-before mode
|
2022-09-02 14:49:34 +01:00 |
|
Chris Cannam
|
deb84a5f78
|
Update build files
|
2022-09-02 13:11:15 +01:00 |
|
Chris Cannam
|
8c817265d5
|
Pick more rational rational values
|
2022-09-02 12:07:41 +01:00 |
|
Chris Cannam
|
2025c51d64
|
Use extracted pickNearestRational; ensure the max rational is actually the max for both num and denom
|
2022-09-02 12:06:45 +01:00 |
|
Chris Cannam
|
a525654d6a
|
Pull out nearest-rational into mathmisc so it can be used in other places than BQResampler
|
2022-09-02 12:06:11 +01:00 |
|
Chris Cannam
|
ee3091ab75
|
Merge from branch speexdsp
|
2022-09-02 11:12:35 +01:00 |
|
Chris Cannam
|
e0a6fc686d
|
Add support for external libspeexdsp
|
2022-09-02 11:10:44 +01:00 |
|
Chris Cannam
|
4336d47a02
|
With longer process sizes we can end up with insufficient space in the inbuf, but enough in the outbuf following the previous retrieve call for a consume to be possible before writing to the inbuf
|
2022-09-01 10:41:25 +01:00 |
|
Chris Cannam
|
20d22f76d6
|
Don't pass debug level through to FFT - it doesn't respect the log target
|
2022-08-30 18:55:33 +01:00 |
|
Chris Cannam
|
53eca719fb
|
Apply gradual phase-reset on unity in the R2 stretcher (R3 already does this)
|
2022-08-18 16:41:43 +01:00 |
|
Chris Cannam
|
e58c6018c5
|
This appears benign and should be slightly quicker
|
2022-08-18 15:24:24 +01:00 |
|
Chris Cannam
|
c9565266f2
|
Adjust the limits, with performance in mind - to be auditioned
|
2022-08-18 14:27:57 +01:00 |
|
Chris Cannam
|
8fee46b704
|
Add SLEEF FFT support
|
2022-08-09 15:50:02 +01:00 |
|
Chris Cannam
|
f81598c166
|
Backed out changeset c03ab2acf74e
No, that worked very badly in practice. Revisit
|
2022-08-08 17:02:08 +01:00 |
|
Chris Cannam
|
f027a59fa4
|
I don't like adjusting the rational approximation - it changes (even if only nominally) the effective ratio based on quality. I think this is more appropriate
|
2022-08-08 16:55:53 +01:00 |
|
Chris Cannam
|
6ecc973d41
|
Keep a structure with recommended limits for inhop/outhop in current configuration; skip readahead entirely when inhop is greater than a certain limit - allowing us to reduce start delay further in short-window mode
|
2022-08-05 14:58:12 +01:00 |
|
Chris Cannam
|
4aea160ec3
|
Indentation only
|
2022-08-04 16:58:00 +01:00 |
|
Chris Cannam
|
3eae190a85
|
This is not actually usually a serious problem, unlike the <1 case which can be genuinely troublesome
|
2022-08-04 16:31:23 +01:00 |
|
Chris Cannam
|
63ebc45107
|
... I said, in RatioOftenChanging mode
|
2022-08-04 13:34:31 +01:00 |
|
Chris Cannam
|
76ee5998f1
|
Permit rational precision to vary by quality setting in RatioOftenChanging mode
|
2022-08-04 13:31:33 +01:00 |
|
Chris Cannam
|
7833e1bff1
|
Avoid kick analysis in single-window mode; comment update
|
2022-08-04 12:08:15 +01:00 |
|
Chris Cannam
|
5b26e90a58
|
Fix to windowing offsets
|
2022-08-04 11:19:13 +01:00 |
|
Chris Cannam
|
fe9e86bc3c
|
Per-channel window source buffer, & connect it up
|
2022-08-04 10:31:36 +01:00 |
|
Chris Cannam
|
d1386b0a0c
|
Provide a separate window source buffer to simplify single/multi window logic
|
2022-08-04 10:02:09 +01:00 |
|
Chris Cannam
|
2fa0e1162e
|
Toward properly expressing the window sizes in Guide when in single-window mode, so as to separate the time-domain frame length from the longest FFT within R3Stretcher and allow us to use a shorter time-domain frame. Not working correctly in single-window mode yet.
|
2022-08-03 14:16:17 +01:00 |
|
Chris Cannam
|
9fff2836c6
|
Avoid RatioMostlyFixed in RealTime mode, it can lead to unpleasant surprises
|
2022-08-02 16:22:24 +01:00 |
|
Chris Cannam
|
d211121701
|
Ah, this was still in use in Scavenger. It can stay for now
|
2022-07-14 17:49:22 +01:00 |
|