Wait for the equalizer to be shut down properly by the main thread
before destroying it. Otherwise, the main thread could crash on stop,
due to accessing the equalizer handle while it's being torn down in the
output thread.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Now the API makes both PCM and FFT data optional, and will do nothing if
neither are requested. Also, it now supports a latency offset in seconds
with floating point precision. The two built-in visualizations currently
request zero larency. Increasing the latency asks for even older samples
while specifying a negative count requests samples from the "future"
relative to what the listener is hearing.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Don't bug out on end of playlist, when didBeginStream will receive a nil
track pointer, which should result in unsetting the current track in the
player, and not send a DidBegin notification to everything, including
the visualization views' event handlers.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Fix counts for tracks with subsongs from piling all the counts onto the
first subsong seen, by using the URL fragment in the filename check and
storage.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Remove the sandbox reference, because the user will add folders outside
the sandbox, and we have entitlements to access these folders.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Cuesheets can now expose which URLs they contain, which may help with
sandbox path configuration. That is, if the CUE sheets are already
readable.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
If upsampling the audio by a significant factor, it may be necessary to
process more than one buffer at a time, rather than lose input.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
The visualization buffer now holds up to 45 seconds of loop, and the
latency measurement code now caps this at 30 seconds, and restarts the
output if latency exceeds 30 seconds, such as if a sound output is
reset.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Don't post a metadata event on open, because inputs will relay it to the
player as an early notification bubble, which is unwanted.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
For one thing, the example code I followed was Swift and handled auto
releasing handles in the background, while Objective-C requires manual
handle reference management.
For two, there was no autoreleasepool around the block handling the
input audio chunks, which need to be released as they are pulled out and
disposed of. This also contributed to memory leakage.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Tabulate first seen information when loading the metadata, rather than
when first adding the tracks to the playlist. This should fix first seen
information when metadata is available, as the information will be
useless without track titles.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Cuesheets were invoking a seek operation on open, rather than on first
playback, and this has a heavy toll on FFmpeg audio formats, apparently.
Defer the initial seek to the first readAudio call, and do not invoke it
if a seek was already called on that input session.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
This is essential for chapters, as otherwise, we would be skipping an
awful lot of samples every chapter, or every seek within a chapter.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Determine the length of the file from the container, rather than the
individual audio stream. The former is more likely to be set than the
latter is.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Correctly configure AVFoundation with the channel layouts supported by
WAVEFORMATEXTENSIBLE speaker position flags, which includes varied
formats supported by FFmpeg and Core Audio inputs.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Update this patch to the latest FFmpeg master source, and update to use
fmtconvert instead of a naive for loop.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Stop output when requested, except on natural completion of the last
track in the play queue. Also fix deadlocks with stopping and
restarting.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Stream metadata encoding may not be UTF-8, even though the Vorbis
Comment specification clearly calls for this.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
The crash was because we weren't copying the results array before
iterating over it, and the deadlock was because this was forced to go
through the main thread, rather than going through its calling thread,
which could lock up if the main thread was busy working with the Sandbox
Broker object.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Commit only once the entire batch is loaded and processed. Also commit
using the correct function.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Made the OpenMPT / legacy OpenMPT and mpg123 libraries pre-built.
Changed the OpenMPT and vgmstream plugins to import the libraries as
they are now. Made mpg123 embedded and imported by the main binary,
since it's now shared by two plugins.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
The playlist was being pruned of entries marked for deletion, but they
were not being pruned from the set that was then added to the player.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
The Path Suggester will now automatically open when new files are added
to the playlist and a given path is not in the sandbox settings. It will
also pop for both the File Tree and MIDI SoundFont path configuration
settings being changed.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Synchronize storage access to main thread only, to prevent enumeration
from hitting a case of the main thread writing to the storage.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Added a title to the window, make the table view resize properly, and
remove the font size inheritance from the main window.
Signed-off-by: Kevin López Brante <kevin@kddlb.cl>
The output now uses AVSampleBufferAudioRenderer to play all formats, and
uses that to resample. It also supports Spatial Audio on macOS 12.0 or
newer. Note that there are some outstanding bugs with Spatial Audio
support. Namely that it appears to be limited to only 192 kHz at mono or
stereo, or 352800 Hz at surround configurations. This breaks DSD64
playback at stereo formats, as well as possibly other things. This is
entirely an Apple bug. I have reported it to Apple with reference code
FB10441301 for reference, in case anyone else wants to complain that it
isn't fixed.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
All optional fallback code for older versions has also been removed, and
everything now assumes 10.13.0 or newer. Some cases are still included
for point releases, such as 10.13.2.
Signed-off-by: Christopher Snowhill <kode54@gmail.com>
Christopher Snowhill
Kevin López