6c2c2733f9
More Envelope::GetValueRelative; add point for join only as needed
2017-05-25 08:15:45 -04:00
7900aa51ba
Update for commands which interact with clips using the keyboard.
...
Changed the criterion for deciding when two clips are immediately next to each other, and made it into a function: WaveClip::SharesBoundaryWithNextClip.
2017-05-16 14:23:07 +01:00
327e5d8557
Correct envelope update for Join that includes a gap between clips
2017-05-09 09:22:07 -04:00
aba52bc79e
Update envelope properly for TimeTrack and WaveTrack editing...
...
... Formerly this was done correctly only for cut and delete from WaveTrack,
paste into WaveTrack, and sync-lock adjustment of WaveTrack (either lengthening
or shortening).
Now also properly done for TimeTrack cut and paste, and also for:
Split cut
Split delete
Trim
2017-05-09 09:22:06 -04:00
d2acf1f3e5
Bug842: rescale clip offsets, envelope times when setting track rate
2017-05-07 11:12:24 -04:00
4be19128c0
Better constructors for Envelope
2017-05-07 11:00:55 -04:00
55439247ad
All the catches are in place for BlockFile and other errors, now throw.
2017-04-03 23:32:40 -04:00
fc0f093db7
Sequence::Append* give strong guarantee
2017-04-02 21:33:23 -04:00
63cf80d244
Sequence::Delete gives strong guarantee
2017-04-02 21:33:23 -04:00
06f6953c91
Sequence::SetSamples, ::SetSilence give strong guarantee
2017-04-02 21:33:23 -04:00
73e61592aa
Sequence::ConvertToSampleFormat gives strong guarantee
2017-04-02 21:33:23 -04:00
3c65731f38
Sequence::InsertSilence gives strong guarantee
2017-04-02 21:33:22 -04:00
160d846643
Sequence::Paste gives strong guarantee
2017-04-02 21:33:22 -04:00
6b0c5c096b
WaveClip::Flush returns void, throws on error
2017-04-02 21:33:22 -04:00
5963f278f1
void returns from many WaveClip methods, comment safety guarantees
2017-04-02 12:49:47 -04:00
aa83c4cf29
Implement the more than weak guarantees needed for recording...
...
... in WaveClip and WaveTrack, to save as much recording as we can,
assuming the strong guarantees that Sequence will give.
Also comment that some other WaveTrack methods can give strong guarantee,
incidentally to making HandleClear give strong.
2017-04-02 12:46:47 -04:00
608dcc9b3f
Reimplement prohibitions on moving clips by throwing exceptions...
...
... Also comment some strong and weak guarantees
2017-03-31 17:47:19 -04:00
d1cbac4181
more finally, more const, remove unused, fix warnings
2017-03-31 17:47:18 -04:00
38b8e57e4e
Specify whether to throw on bad read in Mixer and WaveTrackCache...
...
... Do throw when exporting or mixing-and-rendering; don't if playing back or
drawing a spectrogram, but then just use zeroes.
2017-03-21 17:23:35 -04:00
39b8d99a56
WaveTrack::Get, WaveClip::GetSamples take a mayThrow=true argument...
...
... and pass non-default in all needed places.
Don't throw, don't put up error dialogs, in drawing or hit-test code.
2017-03-21 17:23:34 -04:00
70d9e4bdc7
GetMinMax, GetRMS functions take a mayThrow argument, return numbers
2017-03-21 14:19:02 -04:00
b81cdee7e3
Comment where xml writing functions may throw
2017-03-18 11:45:05 -04:00
58574f2f78
Redo type HFFT as a smart pointer, remove malloc and free
2017-03-17 17:52:51 -04:00
18be1bdad6
Remove naked new[] in: tracks
2017-03-17 17:52:47 -04:00
bb5dc07c51
Fix one scalar naked new that crept back in
2017-03-17 17:52:41 -04:00
b4734ff790
Duplicate command copies cutlines
2017-03-17 17:52:40 -04:00
934a505e1a
Changes in WaveClip construction...
...
... Eliminate CreateFromCopy, add new one-step constructor instead.
It was wasteful to create a copy only to re-create the Sequence at once.
Sequence::Copy is a factory returning a unique_ptr.
Some error checks are removed, but there will be exceptions instead later.
2017-03-17 17:52:39 -04:00
65c722bfe5
Comment: resample, reformat of WaveClip needs no recursion for cutlines
2017-03-17 17:52:39 -04:00
87775d35e0
null pointer checks on all uses of WaveTrackCache::Get
2017-03-17 17:52:35 -04:00
8ddbc1d3d3
Assume number of spectrogram bins = 1/2 FFT window in fewer places...
...
... and that they have equal width on the linear scale
... also simplify the logic involving zero padding factor
This is more preparation for constant-q.
2017-03-17 17:52:28 -04:00
aa0d55ac83
Use enum class ProgressResult, don't interconvert with int or bool
2017-03-17 17:52:24 -04:00
81285ee0c1
More const and override
2017-03-17 17:52:20 -04:00
a9d33f87c4
Comments and minor off by one fixes for drawing wave clips...
...
... and rename two shadowing variables
2016-10-01 13:01:19 -04:00
921b769b52
Fix the previous fix! ... in copying of wave clip display caches.
2016-10-01 10:24:50 -04:00
a41eb4422c
Fix mistake in partial reuse of caches for drawing wave & spectrum...
...
... But this mistake probably had no visible consequences, if total width
in pixes was unchanged.
2016-10-01 10:07:29 -04:00
f993f1eadf
Regularize casts as (int), so they are easier to find and review
2016-09-20 08:42:11 -04:00
ad04187a41
Change sampleCount arguments, variables, return values to size_t...
...
... whenever they really describe the size of a buffer that fits in memory, or
of a block file (which is never now more than a megabyte and so could be fit in
memory all at once), or a part thereof.
2016-09-15 21:03:17 -04:00
078ff056e2
Make many conversions sampleCount->size_t are explicit and checked...
...
... with run-time assertions.
I examined each place and reasoned that the narrowing was safe, and commented
why so.
Again, there are places where the sampleCount variable will later be changed
to have a different type, and they are not changed here.
2016-09-15 21:02:32 -04:00
78be459fa1
Convert sampleCount <-> floating or -> long long explicitly ...
...
... A non-narrowing conversion out to long long is a necessity, but the
conversions to float and double are simply conveniences.
Conversion from floating is explicit, to avoid unintended consequences with
arithmetic operators, when later sampleCount ceases to be an alias for an
integral type.
Some conversions are not made explicit, where I expect to change the type of
the variable later to have mere size_t width.
2016-09-15 21:02:31 -04:00
ed21545c80
pixel column counts and sample window sizes use unsigned types
2016-09-11 20:28:13 -04:00
43b4d3cc99
travis
2016-09-07 13:38:39 -04:00
301fa337d5
Fix bug when expanding a cutline that contains another cutline...
...
... rare case, indeed.
2016-09-07 12:50:53 -04:00
919d77d176
More uses of min and max
2016-09-06 20:46:25 -04:00
c8e7372886
Rewrite Resample::Process to take and return size_t values
2016-09-06 12:39:58 -04:00
180da5a769
change reserve to resize, clarify what omp atomic is used
2016-08-31 11:25:32 -04:00
ce92dce856
improved OpenMP SpecCache::Populate
2016-08-28 13:00:44 -04:00
127b2a6acf
Fix more narrowings of sampleCount
2016-08-26 12:42:22 -04:00
1e0fcbf872
Profiling of spectrogram is opt-in for normal builds.
2016-08-26 12:32:54 +01:00
b2d1470062
OpenMP-ized SpecCache::Populate
2016-08-25 13:44:12 -04:00
52417c444d
Remove unnecessary casts to sampleCount
2016-08-24 14:50:46 -04:00
fd2b050d6f
Type agnosticism for some other variables that were not sampleCount...
...
... in some cases, this fixes narrowings.
2016-08-24 14:50:45 -04:00
79c79f9cd3
Remove many mentions of sampleCount with auto and decltype...
...
... This makes much code agnostic about how other things (functions and
arguments) are typed.
Many of these neeed to become size_t instead of sampleCount.
2016-08-24 14:50:45 -04:00
1281f1b14b
Common function limits buffer size to sampleCount known to be small
2016-08-23 12:46:10 -04:00
46f38708ed
Dir manager uses std::shared_ptr
2016-08-13 12:23:05 -04:00
2917fe5087
One more make_unique for WaveClip, and comment changes
2016-08-13 10:37:52 -04:00
dcceaca13c
Envelope bug
2016-08-12 20:53:38 -04:00
cb48ca2b91
Use arrays of smart pointers to WaveClip
2016-08-12 20:53:38 -04:00
a400a7408c
Distinguish whether clip ptr arrays own; vector; const-correctness
2016-08-12 20:53:38 -04:00
4cbca5d75c
Use std::vector to hold WaveClips
2016-08-12 20:53:38 -04:00
bd57f6392f
Simplify iterations over cutlines in WaveClip.cpp; stl idioms
2016-08-12 20:53:37 -04:00
6e4e710d3b
Remove seeming naked new and deletes in comments and uncompiled code
2016-08-10 11:05:51 -04:00
20bee00115
Remove some naked new amd delete in: tracks and clips
2016-08-08 07:30:44 -04:00
6ac68db5be
Code cleanup: removed the old real FFT code not used since at least 2009.
...
I confirmed that the currently used real FFT code in RealFFTf.cpp is faster
than the old one with a quick benchmark that calls PowerSpectrum() on 4-minute
audio file, with different sizes of computation windows:
Window_size: 256 method: new FFT time_s: 0.393
Window_size: 256 method: old FFT time_s: 1.065
Window_size: 1024 method: new FFT time_s: 0.38
Window_size: 1024 method: old FFT time_s: 0.958
Window_size: 4096 method: new FFT time_s: 0.413
Window_size: 4096 method: old FFT time_s: 1.084
Window_size: 16384 method: new FFT time_s: 0.518
Window_size: 16384 method: old FFT time_s: 1.338
Window_size: 65536 method: new FFT time_s: 0.655
Window_size: 65536 method: old FFT time_s: 1.524
Window_size: 262144 method: new FFT time_s: 0.735
Window_size: 262144 method: old FFT time_s: 1.873
2016-06-25 20:40:16 +02:00
94cf94718e
Redefine ODLocker as movable, and it may try-lock only.
2016-04-17 13:58:21 -04:00
f94b3b3afa
Add some const qualifiers to BlockFile methods
2016-04-10 21:05:20 -04:00
daf92c43f8
Sequence::Copy is a factory returning smart pointer. WaveClip stores it.
2016-04-10 20:47:45 -04:00
0c9deb398c
Fix mac build. No standard headers in Audacity.h.
2016-03-01 12:22:37 -05:00
48a5f55179
More const qualifiers, for copying of tracks, and replacing in lists of tracks
...
... (the tracks may be const, not the list, when replacing)
2016-02-27 17:30:30 -05:00
70c1d57591
More const qualifications, and mutables, for functions that draw tracks...
...
... The display related members that had to become mutable are probably
ones that don't belong in the track classes, ultimately.
2016-02-27 12:35:17 -05:00
9bf098c7d9
Sweep unnecessary wxString copies: rest
2016-02-23 02:15:56 -05:00
dbaa811577
Stack-allocate where possible! ...
...
... Removed many unnecessary naked news and deletes.
2016-02-17 18:15:57 -05:00
7c4c45a0d5
"new"->"NEW" in comments, easier to find remaining naked operator new
2016-02-14 18:52:41 -05:00
56e7653343
"delete"->"DELETE" in comments, easier to find remaining naked operator delete
2016-02-14 18:50:45 -05:00
daa7617e88
auto_ptr (deprecated) -> unique_ptr (preferred in C++11)
2016-02-13 12:00:28 -05:00
ca56876aaf
Various minor fixes in Sequence
2016-02-04 01:06:51 -05:00
51fbc697cf
One less indirection accessing Sequence::mBlock
2016-02-03 22:17:08 -05:00
5de27ac36f
One less indirection accessing InvalidRegion
2016-02-02 11:49:54 -05:00
321d5259a2
More uses of SampleBuffer, eliminating explicit DeleteSamples calls
2016-02-01 10:16:00 -05:00
9c483e2e86
Add (redundant) #include "Experimental.h" in all .cpp files using the symbols
2015-08-31 17:56:10 -04:00
7544a35a6e
Improve calculation of reassigned spectrogram at screen boundaries
2015-08-17 10:03:45 -04:00
9505278c44
Disable caching for reassigned spectrogram
2015-08-17 10:03:21 -04:00
84ee685ee9
reassignment, internals, implementing time correction, but -- ...
...
... it does not yet interact correctly with caching, so results may be
slightly wrong if you scroll the view by less than a screen.
2015-08-17 09:17:46 -04:00
7113d533fb
reassignment, internals, implementing frequency correction only
2015-08-17 09:17:45 -04:00
e4fe449021
more simplification of arguments
2015-08-16 21:14:41 -04:00
5aba06a8a2
simplify argument passing for drawing pitch views
2015-08-16 19:45:09 -04:00
f81231efae
Remove Sequence.h from other headers
2015-07-28 10:02:08 -04:00
d39eaa4e65
Remove WaveTrack.h from other headers
2015-07-28 10:02:05 -04:00
adb4a534e8
Pitch is no longer a special view type... && bug fixes 1040, 1041
...
... It is a choice of algorithm in spectrogram settings, and any frequency
scale may be chosen with it. Spectral selection works in it.
2015-07-27 23:01:20 -04:00
1dffeace93
Access SpectrogramSettings via WaveTrack, which may be nondefault (no UI yet)...
...
... and add accessors to SpectrogramSettings, and remove TrackArtist functions
for getting and setting the globals.
2015-07-26 21:10:16 -04:00
786c1da645
Move class SpectrogramSettings to its own file
2015-07-26 21:10:13 -04:00
8664c877ba
WaveDisplay may or may not manage its own memory.
2015-07-19 11:44:16 -04:00
c14b326913
Publicize class SpecCache. TrackArtist will re-use it.
2015-07-19 11:44:15 -04:00
35e0897bf7
Avoid needless mix/max/rms calculation when showing individual samples.
2015-07-19 11:43:55 -04:00
b25994a82d
Throw away EXPERIMENTAL_FFT_SKIP_POINTS...
...
... with James' consent.
2015-07-18 23:21:27 -04:00
97f06ec09c
Less shimmer and maybe less CPU in display updates for scroll-scrub, because...
...
... tolerant comparison of pps values allows more cache reuse and avoids the
changing alignment of pixels to samples that causes the "shimmery" update.
2015-06-11 10:26:38 -04:00
e8d676c961
fix compilation warning
2015-06-09 14:22:13 -04:00
bb42f6e7c0
More minor spectrogram drawing changes...
...
... Removed an unnecessary zero-filling, added comments, removed a duplication
with a loop.
2015-06-07 17:37:52 -04:00
976eb6164f
More reorganization of wave and spectrum caches, and performance...
...
Simplify the partial copying of wave and spectrum caches. Use memcpy for speed.
Don't memcpy out of the caches into temporary buffers for TrackArtist,
just pass
pointers.
More vectors in the cache classes, fewer deletes.
Pulled big loop, and its body, out of the spectrogram routine into functions.
2015-06-06 11:43:10 -04:00
1916bc3142
Fix compilation of some disabled experimental branches for spectrograms ...
...
... Not promising that they work, just compilation.
2015-06-04 11:30:08 -04:00
afa2fe9fb4
Keep FFT windows for Spectrograms in one place in SpectrogramSettings...
...
... not redundantly in each WaveClip.
2015-06-04 11:29:18 -04:00
Paul Licameli
David Bailes
Paul Licameli
Darrell Walisser
James Crook
Raphaël Marinier