WIP: Core Audio Backend #6056
WIP: Core Audio Backend #6056
Conversation
|
Thank you for the work you have done and are continuing to do on this. I feel that lack of CoreAudio support has been a huge drawback up until now. |
|
@kfreezen, are you still working on this? And are you willing to do what it takes to make this pull request mergeable? If so, then I can give you as much or as little help as you need. Otherwise, I would be happy to take over from here. |
|
@mattmcclinch Yes, I'm still working on it. I got to the point where I didn't know what else to do. I thought that I'd add a PR here and get others' input on it. As for specific comments on the code, I will be replying to those throughout the day today. Thanks for your interest in this. It makes me a little happier than I was before I saw this thread this morning. |
|
By the way, if you don't know what else to do, you can work on reformatting your code so that it follows the coding rules. |
|
@mattmcclinch is there an auto-formatter somewhere I can use for that?
…On Fri, May 8, 2020 at 8:58 AM mattmcclinch ***@***.***> wrote:
By the way, if you don't know what else to do, you can work on
reformatting your code so that it follows the coding rules
<https://musescore.org/en/handbook/developers-handbook/finding-your-way-around/musescore-coding-rules>
.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#6056 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AACUWLBKUITHUVW7VUPVVQTRQP6W3ANCNFSM4M3UHY4A>
.
|
|
You will probably have to do it by hand. |
|
There is an xml file suitable for QtCreator in the developerts' handbook |
|
Are you using the MacBook Pro speakers or something else?
…On Fri, May 8, 2020 at 22:44 mattmcclinch ***@***.***> wrote:
***@***.**** commented on this pull request.
------------------------------
In audio/drivers/coreaudio.cpp
<#6056 (comment)>:
> + OSStatus osErr = AudioObjectSetPropertyData (kAudioObjectSystemObject, &theAddress, 0, NULL, sizeof(CFRunLoopRef), &theRunLoop);
+ if (osErr != noErr) {
+ initialized = false;
+ return false;
+ }
+
+ AudioObjectPropertyAddress outputDeviceAddress = {
+ kAudioHardwarePropertyDefaultOutputDevice,
+ kAudioObjectPropertyScopeGlobal,
+ kAudioObjectPropertyElementMaster
+ };
+
+ AudioObjectAddPropertyListener(kAudioObjectSystemObject, &outputDeviceAddress, &onDefaultOutputDeviceChanged, this);
+
+ AudioDeviceID deviceId = getDefaultOutputDevice();
+ return openAudioUnit(&audioUnit, &deviceId, this);
MacBook Pro (15 inch, 2018), macOS Catalina Version 10.15.4.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#6056 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AACUWLBJXHPY2DUUFMAVYG3RQS7SHANCNFSM4M3UHY4A>
.
|
|
The pitch issue occurred when I was just using the MacBook Pro speakers. If I had earphones plugged in or a bluetooth headset connected, it sounded fine. |
|
Please squash all these commits into just one |
kfreezen
force-pushed
the
feature/coreaudio
branch
2 times, most recently
from
461e790 to
38a168e
Compare
|
I'm hoping that covers all the formatting changes. Tried to make sure I got them all. There seems to be a glitch I'm running into intermittently while using the CoreAudio driver. The position indicator froze at position 0 in the score and the sequencer buffer filled to max. I can't get that to reproduce consistently for me though. |
kfreezen
force-pushed
the
feature/coreaudio
branch
3 times, most recently
from
9c828a8 to
c9d5ef6
Compare
kfreezen
force-pushed
the
feature/coreaudio
branch
from
c9d5ef6 to
ae807c9
Compare
|
I'm curious whether there's anything I missed on this one, or if it's waiting on someone else? |
|
It needs a rebase, adjusting to the new coding style. There's a script to help with that, in tools/codestyle. |
kfreezen
force-pushed
the
feature/coreaudio
branch
2 times, most recently
from
4b96e6a to
8b109e9
Compare
|
There are some glitches that a friend and I have run into while testing this branch over the last 3-4 months. The main glitch is what looks like the sequencer freezing after the program has been running in the background for a while. I have not narrowed down what causes the bug, since it's fairly intermittent and shows up when you hit the Play button. It seems like the problem is fixed if I edit the score before playing, however. Maybe someone can give me some ideas on where to look for this one? |
Added option and tabstop for Core Audio to preference dialog
kfreezen
force-pushed
the
feature/coreaudio
branch
from
8b109e9 to
611fc52
Compare
|
I'm curious what architecture changes are being actively made in MuseScore. Is there a source or documentation on those arch changes? Is audio driver code moving to another project? That's what it looks like. |
|
@igorkorsukov should be able to answer that. |
Yes, that's right, the architecture of the project has changed a lot. The new audio driver is here, you can preview it. I am very sorry that we were not able to take advantage of your work. |
|
@igorkorsukov no problem. As long as I can get "instant switching" with my Airpods, I'm happy. Where is a good place to ask questions about the general architecture and MuseScore's long term goals with it? |
You can ask questions in pull requests or in the development chat https://t.me/musescoreeditorchat (if the question is about architecture, then it's better to mention me) |
igorkorsukov
added
the
archived
@psmokotnin FYI |
I added a basic Driver class for a Core Audio backend, and added a RadioGroupBox to the preference dialog for selecting Core Audio.
Why did I want to add this code/What are the benefits?
I wanted MuseScore to switch output devices automatically if I were to plug one in. PortAudio required me to mess with the I/O tab in the Preferences dialog to switch output devices. PortAudio has no API for tracking output device changes.
Using Core Audio directly has another benefit, which is a lower "Energy Impact" score. My not-so-scientific test yielded a difference between 5 and 2.5 when playback was stopped. It's not groundbreaking, but it is a small improvement.