Mercurial > hg > constant-q-cpp

log

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
less more | rev 112: (0) tip
age author description
Wed, 14 May 2014 14:59:45 +0100 Chris Cannam Smaller range, faster chromagram
Wed, 14 May 2014 14:59:38 +0100 Chris Cannam Make kernel a bit more robust to weird parameters
Wed, 14 May 2014 14:04:34 +0100 Chris Cannam First cut at a chromagram plugin as well
Wed, 14 May 2014 12:49:10 +0100 Chris Cannam Make CQVamp implement two plugins, one with MIDI pitch range and one with Hz
Wed, 14 May 2014 11:40:18 +0100 Chris Cannam Slightly neaten
Wed, 14 May 2014 11:23:02 +0100 Chris Cannam Fixes to peak diff display
Wed, 14 May 2014 11:22:51 +0100 Chris Cannam Adjust resampler spec for speed/quality tradeoff
Wed, 14 May 2014 10:41:20 +0100 Chris Cannam Tail padding fix
Wed, 14 May 2014 10:30:49 +0100 Chris Cannam Start making this program adaptable enough to use for testing
Tue, 13 May 2014 13:31:24 +0100 Chris Cannam Merge
Tue, 13 May 2014 12:34:51 +0100 Chris Cannam Latency fixes: avoid rounded fractional per-octave latencies (makes the inverse difficult to handle because we'd need to have access to exact forward resampler latencies); reduce inverse transform latency by one big block
Tue, 13 May 2014 11:45:22 +0100 Chris Cannam Set output format (input might have been not a WAV file at all)
Tue, 13 May 2014 11:46:50 +0100 Chris Cannam Fix buffer underrun
Sat, 10 May 2014 13:44:24 +0100 Chris Cannam Simplify a bit, add some debug out
Sat, 10 May 2014 11:57:59 +0100 Chris Cannam Write diff of tail as well
Sat, 10 May 2014 11:48:21 +0100 Chris Cannam Add ability to write an extra sound file with the difference between input and output
Fri, 09 May 2014 17:43:00 +0100 Chris Cannam Alignment is now right in inverse transform, but the padding at end is still wrong
Fri, 09 May 2014 15:36:23 +0100 Chris Cannam Fix bin ordering within octave
Fri, 09 May 2014 15:16:48 +0100 Chris Cannam Inverse runs; now merely needs to run correctly
Fri, 09 May 2014 11:41:44 +0100 Chris Cannam More work on inverse
Fri, 09 May 2014 11:15:56 +0100 Chris Cannam A bit more on inverse
Fri, 09 May 2014 10:56:23 +0100 Chris Cannam Bit more on inverse
Fri, 09 May 2014 10:05:16 +0100 Chris Cannam Rearrange classes so the basic ConstantQ produces a complex output, and CQSpectrogram (formerly CQInterpolated) is required if you want a real output
Fri, 09 May 2014 08:25:24 +0100 Chris Cannam More toward inverse CQ
Thu, 08 May 2014 19:01:15 +0100 Chris Cannam Start on C++ inverse
Thu, 08 May 2014 09:35:10 +0100 Chris Cannam Fix compiler warning
Wed, 30 Apr 2014 15:06:30 +0100 Chris Cannam Reconstruction now appears to be (roughly!) correct in the yeti code
Wed, 30 Apr 2014 13:32:48 +0100 Chris Cannam Inverse now produces a (wrong) output
Wed, 30 Apr 2014 08:38:40 +0100 Chris Cannam Add small program to run some experiments
Tue, 29 Apr 2014 17:32:52 +0100 Chris Cannam Another step in icqt
Tue, 29 Apr 2014 17:27:04 +0100 Chris Cannam A bit more on icqt
Tue, 29 Apr 2014 16:52:16 +0100 Chris Cannam More on tests and inverse stuff
Tue, 29 Apr 2014 08:36:47 +0100 Chris Cannam Start on inverse code, fix up forward interface a bit
Mon, 07 Apr 2014 16:14:24 +0100 Chris Cannam Backed out changeset 3393a2898a1d
Fri, 04 Apr 2014 13:50:27 +0100 Chris Cannam Try chaining the resamplers
Fri, 04 Apr 2014 13:29:16 +0100 Chris Cannam Avoid compiler warnings
Fri, 04 Apr 2014 12:33:15 +0100 Chris Cannam Linear interpolation fixed
Fri, 04 Apr 2014 11:49:53 +0100 Chris Cannam Interpolation for CQ. But this is wrong, it is interpolating between points in the nearest full-height columns even where there is a nearer point in the not-full-height ones
Mon, 24 Mar 2014 16:30:40 +0000 Chris Cannam vector multiply now takes a list
Fri, 21 Mar 2014 12:04:55 +0000 Chris Cannam Ant build file
Thu, 20 Mar 2014 16:15:43 +0000 Chris Cannam Support returning a magnitude spectrum (dense) etc
Wed, 19 Mar 2014 10:30:20 +0000 Chris Cannam Vamp category and RDF files
Wed, 19 Mar 2014 10:30:13 +0000 Chris Cannam minor tidy
Wed, 12 Mar 2014 08:53:45 +0000 Chris Cannam Copyrights, licence
Mon, 03 Mar 2014 08:32:45 +0000 Chris Cannam Allow wrong max in alternate columns where expected max would be in lower octave (which is not present)
Fri, 28 Feb 2014 17:18:54 +0000 Chris Cannam Plot failing test matrices
Fri, 14 Feb 2014 14:24:37 +0000 Chris Cannam Add little program to plot cq of an audio file
Fri, 14 Feb 2014 11:55:03 +0000 Chris Cannam Pull out tests
Fri, 14 Feb 2014 11:19:52 +0000 Chris Cannam Comment out some debug
Fri, 14 Feb 2014 11:19:45 +0000 Chris Cannam Comment out some debug
Thu, 13 Feb 2014 19:21:10 +0000 Chris Cannam Just an import
Mon, 03 Feb 2014 13:01:47 +0000 Chris Cannam Build fixes
Mon, 03 Feb 2014 12:41:43 +0000 Chris Cannam Merge with pulled changes
Fri, 31 Jan 2014 11:40:01 +0000 Chris Cannam Start on automated tests
Thu, 30 Jan 2014 12:12:16 +0000 Chris Cannam Put spectrogram the "right" way up, add bin labels
Thu, 30 Jan 2014 12:11:53 +0000 Chris Cannam Fix min frequency calculation in accessor, add bin frequency accessor
Thu, 30 Jan 2014 10:33:38 +0000 Chris Cannam Single Makefile, build Vamp by default
Thu, 28 Nov 2013 17:03:57 +0000 Chris Cannam Use MIDI pitch values, as in QM Vamp Plugins implementation
Thu, 28 Nov 2013 14:32:42 +0000 Chris Cannam Correct output latency
Thu, 28 Nov 2013 11:46:39 +0000 Chris Cannam Further tidying, compensate for latency in Vamp plugin
Thu, 28 Nov 2013 11:26:53 +0000 Chris Cannam Simplify
Thu, 28 Nov 2013 11:19:55 +0000 Chris Cannam Further latency adjustment
Thu, 28 Nov 2013 11:00:33 +0000 Chris Cannam Remove debug out
Thu, 28 Nov 2013 10:58:02 +0000 Chris Cannam We shouldn't need to compensate for max(latency)+max(drop) but for max(latency+drop); also the padding can be rounded to hop rather than fft size I think?
Thu, 28 Nov 2013 10:24:57 +0000 Chris Cannam Towards handling the hop alignment (which in the MATLAB is handled on output) at the input end
Wed, 27 Nov 2013 14:14:10 +0000 Chris Cannam concat (map ... ) -> concatMap
Fri, 22 Nov 2013 17:56:20 +0000 Chris Cannam Make additional latency at start a multiple of the big blocksize
Fri, 22 Nov 2013 14:47:41 +0000 Chris Cannam Start to address some padding and latency issues
Thu, 21 Nov 2013 17:04:19 +0000 Chris Cannam Some work on alignment
Wed, 20 Nov 2013 21:39:42 +0000 Chris Cannam Update to revised api
Wed, 20 Nov 2013 14:25:54 +0000 Chris Cannam Use revised framer api
Tue, 19 Nov 2013 19:27:45 +0000 Chris Cannam More debug testing
Tue, 19 Nov 2013 14:37:23 +0000 Chris Cannam Return bin frequencies, for reference; some padding
Mon, 18 Nov 2013 17:11:47 +0000 Chris Cannam Use mat.print
Mon, 18 Nov 2013 16:30:38 +0000 Chris Cannam Adjust test case to match matlab example on wiki; print out results
Thu, 07 Nov 2013 11:55:10 +0000 Chris Cannam Pull out test program
Wed, 06 Nov 2013 16:21:28 +0000 Chris Cannam Get remaining blocks from end of processing; fix some compile warnings, etc
Wed, 06 Nov 2013 14:30:42 +0000 Chris Cannam Add Vamp plugin for testing with
Wed, 06 Nov 2013 14:30:19 +0000 Chris Cannam More on reconstructing CQ shape. This is not always correct yet, though
Wed, 06 Nov 2013 09:09:55 +0000 Chris Cannam Delete the kernel!
Wed, 06 Nov 2013 08:10:33 +0000 Chris Cannam Much of the actual CQ processing logic
Tue, 05 Nov 2013 16:58:18 +0000 Chris Cannam Don't forget top octave in buffer list
Tue, 05 Nov 2013 16:44:12 +0000 Chris Cannam Start on CQ process implementation
Tue, 05 Nov 2013 08:41:00 +0000 Chris Cannam Stub class
Mon, 04 Nov 2013 18:56:23 +0000 Chris Cannam Ah, the code was right, the test was wrong
Mon, 04 Nov 2013 18:17:26 +0000 Chris Cannam Almost finish normalising kernel and storing it sparsely
Mon, 04 Nov 2013 17:51:27 +0000 Chris Cannam Minor updates
Mon, 04 Nov 2013 14:20:29 +0000 Chris Cannam bf -> vec, matrix concat rename
Mon, 04 Nov 2013 10:54:49 +0000 Chris Cannam Keep the kernel dense for the moment, and use std::complex
Fri, 01 Nov 2013 17:58:39 +0000 Chris Cannam Fix frequency calculation in cpp, and some diagnostics
Fri, 01 Nov 2013 16:13:22 +0000 Chris Cannam Calculate (so far unnormalised) CQ kernel
Wed, 30 Oct 2013 18:26:04 +0000 Chris Cannam Restore reshape into atom columns; drop atoms for alignment (not working, we may be compensating in the wrong direction)
Wed, 30 Oct 2013 17:29:14 +0000 Chris Cannam Return only the actual results (i.e. space with zeros rather than duplicates)
Wed, 30 Oct 2013 16:31:25 +0000 Chris Cannam Build assembled blocks
Wed, 30 Oct 2013 10:51:23 +0000 Chris Cannam More on assembling blocks
Wed, 30 Oct 2013 10:20:55 +0000 Chris Cannam Start trying to assemble blocks as they arrive
Wed, 30 Oct 2013 09:06:52 +0000 Chris Cannam Print out results (to go to file to load into e.g. SV)
Tue, 29 Oct 2013 16:55:55 +0000 Chris Cannam Towards shuffling the cq cells into the right arrangement for return
Mon, 28 Oct 2013 18:23:43 +0000 Chris Cannam Reshape result vectors
Mon, 28 Oct 2013 17:16:55 +0000 Chris Cannam Comments
Fri, 25 Oct 2013 17:39:35 +0100 Chris Cannam Return the atom count & spacing as well
Fri, 25 Oct 2013 17:39:28 +0100 Chris Cannam Comments
Fri, 25 Oct 2013 16:57:04 +0100 Chris Cannam Start on code to apply the transform
Fri, 25 Oct 2013 16:56:46 +0100 Chris Cannam Make into a function!
Fri, 25 Oct 2013 14:38:17 +0100 Chris Cannam Normalisation -- now it *really* matches the toolbox
Fri, 25 Oct 2013 12:49:04 +0100 Chris Cannam Use symmetric window -- kernel now matches the MATLAB one
Fri, 25 Oct 2013 11:53:28 +0100 Chris Cannam Sparse etc
Thu, 24 Oct 2013 18:53:41 +0100 Chris Cannam Return sparse complex matrix
Thu, 24 Oct 2013 15:09:38 +0100 Chris Cannam Minor fixes
Thu, 24 Oct 2013 14:43:36 +0100 Chris Cannam Generate the actual kernels
Thu, 24 Oct 2013 11:18:18 +0100 Chris Cannam Another minor few sums...
Wed, 23 Oct 2013 17:31:04 +0100 Chris Cannam Initial kernel parameters
Wed, 23 Oct 2013 14:36:52 +0100 Chris Cannam Directory structure, paper, matlab etc
less more | rev 112: (0) tip