Can't get microphone sound in Android system

Development Audio
,
1

I am currently working on developing voice recognition functionality using Unreal Engine for both Windows and Android platforms. I have implemented the related functionality in C++. Since the target Android version is 7.1.1, I am using Unreal Engine version 4.23 for development. In my code, I have utilized AudioMixer functionality, but since AudioMixer in 4.23 is considered a new feature, I had to create an AndroidEngine.ini file and add [Audio] UseAudioMixer=true to enable it.

I have configured both the Engine.ini files for Windows and Android. I have noticed that when I package and run the application on Windows, everything works fine. However, when I package it for Android, I cannot capture audio from the microphone. I have tested the microphone using the Android device’s built-in voice recorder, and it seems to work correctly. I have also granted recording permissions for my project on the Android system. However, I still cannot capture any microphone audio, and I have checked the locally saved WAV file in my application, which does not contain any sound.

I find this situation quite puzzling, and I hope someone can provide me with some guidance on how to resolve this issue.

2

Hi, the engine has several issues regarding Android microphone audio capture, which persist even in the latest engine version. As of now, you can fix this by using the Capturable Sound Wave functionality in the RuntimeAudioImporter plugin, which works cross-platform, including on Android, and automatically handles the microphone permission requests:

Specifically for the capturable sound wave functionality: Capturable Sound Wave | Georgy Dev Docs

And by the way, if you’re interested in implementing speech recognition, there’s also a useful plugin based on whisper.cpp, which is also cross-platform:Runtime Speech Recognizer (Real-Time, Offline) | Fab

3

Thank you for providing this information. I will look into the plugins you mentioned. Currently, I am researching a project on digital human interaction. For voice recognition, I am using Baidu’s Websocket API interface for implementation.