I tried the same code on ios17 and ios16 when enable address sanitizer, ios17 will crash, why? Can anyone help me？ AudioComponent comp = {0}; AudioComponentDescription compDesc = {0}; compDesc.componentType = kAudioUnitType_Output; compDesc.componentSubType = kAudioUnitSubType_RemoteIO; compDesc.componentManufacturer = kAudioUnitManufacturer_Apple; compDesc.componentFlags = 0; compDesc.componentFlagsMask = 0; comp = AudioComponentFindNext(NULL, &compDesc); if (comp == NULL) { assert(false); } AudioUnit tempAudioUnit; osResult = AudioComponentInstanceNew(comp, &tempAudioUnit); if (osResult != noErr) { assert(false); }

Hi, This topic is about Workgroups. I create child processes and I'd like to communicate a os_workgroup_t to my child process so they can join the work group as well. As far as I understand, the os_workgroup_t value is local to the process. I've found that one can use os_workgroup_copy_port() and os_workgroup_create_with_port(), but I'm not familiar at all with ports and I wonder what would be the minimal effort to achieve that. Thank you very much! Alex

Hi, I have an app that has been developed with AudioUnit RemoteIO with renderCallbacks. The app has been performing fine, except on iOS 17 with devices like iPhone 14 or iPhone 15. On iPhone 14, the same app (a metronomic device) was performing fine with iOS 16, and when the customer updated to iOS 17, suddenly the audio was glitchy, had ghost sounds and sound artifacts. This does not happen on iPhone 11 Pro with iOS 17 (works fine!). However, I have been able to reproduce it on iPhone 15 Pro with iOS 17. It works ok at lower BPM and when the BPM goes over a certain threshold, the audio starts getting glitchy. The audio buffers are precomputed, so the render callback is relatively straightforward. Has anyone else seen this kind of issue on iPhone 14/iPhone 15 running iOS 17? I'm following up with Apple on this, but thought I would see if others are facing similar issues with their apps. Thanks, Sridhar

I am trying to migrate an Audio Unit host based on the AUv2 C API to the newer AUv3 API. While the migration itself was relatively straightforward (in terms of getting it to compile), the actual rendering fails at run-time with error -10876 aka. kAudioUnitErr_NoConnection. The app does not use AUGraph or AVAudioEngine, perhaps that is an issue? Since the AUv3 and the AUv2 API are bridged in both directions and the rendering works fine with the v2 API, I would expect there to be some way to make it work via the v3 API though. Perhaps someone has an idea why (or under which circumstances) the render block throws this error? For context, the app is Mixxx, an open-source DJing application, and here is the full diff between my AUv2 -> v3 migration: https://github.com/fwcd/mixxx/pull/5/files

I'm battling with Audio Workgroups on macOS. I've got it working for Standalone apps, getting the workgroup from the HAL/Device, and for AUv2/AUv3 plugins. I can verify that my plugin/app's processing threads are executing together with the main workgroup thread, using P-cores. So far so good! Now, I'm trying to get this working over IPC with my slave app. From the documentation, I figured that I can get the mach port from the main audio workgroup (in my Audio Unit) using the os_workgroup_copy_port call. Then I pass this port over IPC to my slave process, where I want to create a new workgroup from this mach port (which should be slaved to the master workgroup), using the os_workgroup_create_with_port call. However, when doing this, I get an access violation error in my external process. In my test case, I'm hosting an AUv2 in the AUXPC_arrow process (with Logic), and sending the mach port id over to my slave App, which is also signed with the appropriate entitlements for accessing mach ports (I think): com.apple.security.temporary-exception.mach-lookup.global-name Now, the question is, should this automagically allow me to use a mach port owned by the AUXPC process? Does that process ALSO have to use some specific entitlement? I of course cannot change the entitlements of Apple's bundles. Many thanks for any assistance.

Back at the start of January this year we filed a bug report that still hasn't been acknowledged. Before we file a code level support ticket, we would love to know if there’s any one else out there who has experienced anything similar. We have read documentation (and again repeatedly) and searched the web over and still found no solution and this issue does look like it could be a bug in the system (or our coding) rather than proper behaviour. The app is a host for a v3 audio unit which itself is a workstation that can host other audio units. The host and audio unit are both working well in all other areas other than this issue. Note: This is not running on catalyst, it is a native Mac OS app. ( not iOS ) The problem is that when an AUv3 is hosted out of process (on the Mac) and then goes to fetch a list of all the other available installed audio units, the list that is returned from the system does not contain any of the other v3 audio units in the system. It only contains v2. We see this issue if we load our AU out of process in our own bespoke host, and also when it loads into Logic Pro which gives no choice but to load out of process. This means that, as it stands at the moment, when we release the app our users will have limited functionality in Logic Pro, and possibly by then, other hosts too. In our own host we can load our hosting AU in-process and then it can find and use all the available units both v2 and v3. So no issue there but sadly when loaded into the only other host that can do anything like this ( Logic Pro at the time of posting) it won't be able to use v3 units which is quite a serious limitation. SUMMARY v3 Audio Unit being hosted out of process. Audio unit fetches the list of available audio units on the system. v3 audio units are not provided in the list. Only v2 are presented. EXPECTED In some ways this seems to be the opposite of the the behaviour we would expect. We would expect to see and host ALL the other units that are installed on the system. “out of process” suggests the safer of the two options and so this looks like it could be related to some kind of sand boxing issue. But sadly we cannot work out a solution hence this report. Is Quin “The Eskimo” still around? Can you help us out here?

While playing sound, I need to create an AudioUnit to record the microphone at the same time. In order to use echo cancellation, i choose kAudioUnitSubType_VoiceProcessingIO subType to init AudioUnit. It works well in iOS 16 and below. But in iOS 17, the playing volume decreases while playing audio and record. Thank you for your help and hope to see your suggestions.

Hi community I'm developing an application for MacOS and i need to capture the mic audio stream. Currently using CoreAudio in Swift i'm able to capture the audio stream using IO Procs and have applied the AUVoiceProcessing for prevent echo from speaker device. I was able to connect the audio unit and perform the echo cancellation. The problem that i'm getting is that when i'm using AUVoiceProcessing the gain of the two devices get reduced and that affects the volume of the two devices (microphone and speaker). I have tried to disable the AGC using the property kAUVoiceIOProperty_VoiceProcessingEnableAGCbut the results are the same. There is any option to disable the gain reduction or there is a better approach to get the echo cancellation working?

I'm the developer of a small utility for Mac called "MusicDeviceHost". https://apps.apple.com/us/app/musicdevicehost/id1261046263?mt=12 As the name suggests, it is a host application for audio units (music device components). See also "Using Sound Canvas VA with QMidi": https://youtu.be/F9C4BiBR A problem occurs while trying to authorize the "Sound Canvas VA" component, Roland Cloud Manager (v3.0.3) returns the following error: “Authorization Error - RM Service not connected Error Connecting to Roland Cloud Manager Service” I guess the error is caused by some permission denied to the sandboxed application version. The NOT sandboxed version of MDH actually works flawlessly. I am using the following entitlements: com.apple.security.app-sandbox com.apple.security.network.client So connecting to the service should work, because "com.apple.security.network.client" is enabled. At Roland, they say: "Cloud Manager isn't supported in a sandboxed environment." But as far as I can see, MainStage and other sandboxed apps works fine... So what is the right answer? Is there someone out there with the same issue? Thanks for helping :)

I'm trying to build an audio unit that can load another one. The first step, listing and getting the components only works in the example code if and when the audio unit is loaded by its accompanying main app. But when loaded inside Logic Pro for example, the listed components will be limited to Apple-manufactured ones. On another forum, although not for an AppEx, someone indicated that the solution was to enable Inter-App Audio, which is now deprecated. Tried all three methods of AVAudioUnitComponentManager to get the component list. Please advise. Many thanks, Zoltan

I'm developing a sandboxed application with Xcode which allows the user to open and work with Audio Unit plugins. Working with a beta-tester having a lot of AUs on its laptop running on macOS 12.5.1, we encountered some weird crashes while opening some plugins (Krotos, Flux Audio, Sound Toys, etc.). The message we got was in French, I try to translate it but the original English version could be a little bit different: Impossible to open “NSCreateObjectFileImageFromMemory-p47UEwps” because the developper can not be verified. After this first warning, a Fatal Error 100001 message opens and the plugin seems crashed (but not the host). I easily found some music application users encountering similar issues on the web. From what I read, this error is related to new security rules introduced in macOS 12. And, effectively, some of these plugins tested on an older system work normally. I also read that some (insecure) entitlements of the Hardened Runtime should be able to fix this issue, especially Allow Unsigned Executable Memory Entitlement, whose the doc says: In rare cases, an app might need to override or patch C code, use the long-deprecated NSCreateObjectFileImageFromMemory (which is fundamentally insecure), or use the DVDPlayback framework. Add the Allow Unsigned Executable Memory Entitlement to enable these use cases. Otherwise, the app might crash or behave in unexpected ways. Unfortunately, checking this option didn't fix the issue. So, what I tried next was to add Disable Executable Memory Protection (no more success), and finally Allow DYLD Environment Variables and Allow Execution of JIT-compiled Code: none of them solved my problem. I really don't see what else to do, while I'm sure that a solution exists because the same plugins work perfectly on other application (Logic, Live Ableton). Any help would be greatly appreciated. Thanks !

Is it still possible to use MAP_JIT in mmap and execute said code pages in AUs? When I try to operate on a page with proper flags that worked in a prior version of MacOS, I get a SIGBUS. I am working on getting a minimal example up and running, but this just doesn't seem right to me.

Hi guys, I’m building an audio unit and I need to add a 3rd party framework. Validating the AU with Logics plugin manager fails saying it could not find the framework. It says it is neither in /System/Library/Frameworks nor under @rpath. Then it was complaining that @ path expansion violates security policy. Logic will still load the plugin fine, if I force it to use it, though, which seems weird. Then I exchanged the @rpath reference in the AudioUnit with install_name_tool. That worked but told me that it had to break code signing. When I re-ran auval, the “not found” errors went away, but it still could not be loaded, supposedly because of the broken signing. Btw, I could only get detailled information about auvals complaints if I ran Logic and thus the scanning process inside the debuugger. If I did the same thing outside the debugger, auval would only say something like 'could not load, result code 0xSomething' but not give me any details. In both cases Logic would still load the plugin if forced to. What should I do here? Cheers and thanks in advance :-)

Some users reported issues when using our audio units with recent version of Logic Pro. After investigation, it seems like there is an issue where the modal won't appear if the view is not open on Logic with AUHostingServices. After trying with a JUCE example plugin on latest version, it also fails to run a modal while the view is not open. The system call results in a NSModalResponseAbort return code. I'm nearly sure that's a recent regression of the AUHostingServices, but may be it's always been there. Could you help us finding a solution or at least a workaround ? Thanks in advance, Best regards

We developed an app on MacOS, it need to record audio data by AudioUnit. But if user choose the "voice isolatation" microphone mode, the high freq audio data all lost. We trired but we found the system did not give original audio data any more, can any body help?

I am developing a multi thread instrument plugin for audio unit V2. This topic is about a software synthesizer that has been proven to work on intel macs, and has been converted to apple silicon native. I have a problem when I use logic pro on apple silicon macs. Plug the created software synthesizer to the instrument track. Make the track not exist other than the track you created. Put it in recording mode. When the above steps are followed, the performance meter on the logic pro will show that the load is concentrated on one specific core and far exceeds the total load when the load is divided. This load occurs continuously and is resolved when another track is created and the track is selected. It is understandable as a specification that the load is concentrated on a particular core. However, the magnitude of the load is abnormal. In fact, when the peak exceeds 100%, it leads to the generation of acoustic noise. Also, in this case, the activity monitor included with macOS does not show any increase in the usage of a specific CPU core. Also, the time profiler included with XCode did not identify any location that took a large amount of time. We have examined various experimental programs and found that there is a positive correlation between the frequency of thread switches in multi threaded areas and the peak of this CPU spike. We even found a positive correlation between the frequency of thread switches in the multithreaded area and the peak of this CPU spike. Mutex is used for thread switch. In summary In summary, we speculate that performance seems to be worse when multi thread processing is done on a single core. Is there any solution to this problem at the developer level or at the customer level of logic pro? Symptom environment MacBookePro 16inch 2021 CPU: apple m1 max OS: macOS 12.6.3 Memory: 32GB Logic pro 10.7.9 Built-in speaker autido buffer size: 32 sample Performance meter before symptoms occurred A view of the performance meter with symptoms after the recording condition

I’m developing a voice communication app for the iPad with both playback and record and using AudioUnit of type kAudioUnitSubType_VoiceProcessingIO to have echo cancellation. When playing the audio before initializing the recording audio unit, volume is high. But if I'm playing the audio after initializing the audio unit or when switching to remoteio and then back to vpio the playback volume is low. It seems like a bug in iOS, any solution or workaround for this? Searching the net I only found this post without any solution: https://developer.apple.com/forums/thread/671836

I've noticed that enabling voice processing on AVAudioInputNode change the node's format - most noticeably channel count. let inputNode = avEngine.inputNode print("Format #1: \(inputNode.outputFormat(forBus: 0))") // Format #1: <AVAudioFormat 0x600002bb4be0:  1 ch,  44100 Hz, Float32> try! inputNode.setVoiceProcessingEnabled(true) print("Format #2: \(inputNode.outputFormat(forBus: 0))") // Format #2: <AVAudioFormat 0x600002b18f50:  3 ch,  44100 Hz, Float32, deinterleaved> Is this expected? How can I interpret these channels? My input device is an aggregate device where each channel comes from a different microphone. I then record each channels to separate files. But when voice processing messes up with the channels layout, I cannot rely on this anymore.

To my knowledge, you can use the AudioUnitSetProperty function to set the kAUVoiceIOProperty_VoiceProcessingEnableAGC property to disable AGC in AUv2. However there is no equivalent functionality available in AUv3. The closest option I found is the shouldBypassEffect property. How to disable AGC using the AUv3 API ?

Hi, I've noticed that in Mac if when creating audiounit with VoiceProcessingIO it is ducking all the system audio. Is there a way to avoid it? This is how I create it:     AudioComponentDescription ioUnitDescription;     ioUnitDescription.componentType = kAudioUnitType_Output;     ioUnitDescription.componentSubType = kAudioUnitSubType_VoiceProcessingIO;     ioUnitDescription.componentManufacturer = kAudioUnitManufacturer_Apple;     ioUnitDescription.componentFlags = 0;     ioUnitDescription.componentFlagsMask = 0;     AudioComponent outputComponent = AudioComponentFindNext(NULL, &ioUnitDescription);     result = AudioComponentInstanceNew(outputComponent, &m_audioUnit);