I was wondering if it is possible to export the audio we hear on playback. I just noticed that there is an .mp4 export option but I don’t hear any audio coming out… So I guess the answer is ‘No’ but thought I’d check here to confirm :)
Hi! Thanks for the question and sorry for the late response on this!
Currently our video capture exporter (capture to .mp4) does not support capturing audio. There are plans to improve the audio capture support in future versions, but for the moment your options are unfortunately limited. That said however, there are still a couple of ways you can capture the audio output from your stage now:
There is a debugging feature for audio that allows all the output from playback to be captured to a .wav file on a local disk. Some documentation on it can be found here. Effectively, once a filename has been set, toggling the “Enable Stream Dump” checkbox on will start a capture of all audio output and toggling it off will stop. Be warned though, toggling it back on will overwrite any previous capture to that same file. This was only meant as a debugging tool really so it’s not the friendliest interface.
Using the “Streamers” API in python. This is available in Omniverse Machinima now, but may not be available in Omniverse Create yet. Its documentation has also unfortunately not been published yet, but will eventually appear here. This API allows an arbitrary number of ‘streamer’ objects to be created (audio = omni.usd.audio.get_stage_audio_interface() followed by audio.create_capture_streamer()). Each object is capable of capturing the stage’s audio output to a file (filename is set with audio.set_capture_filename(filename)) using audio.start_capture().
Note that both of these capture methods use the same functionality under the hood. Technically they can write to any supported output format including WAV, FLAC, Ogg/Vorbis, and Opus. However at this time only WAV has been thoroughly tested.
Audio support in Omniverse Create is an ongoing work in progress. Feedback and feature requests are very welcome!
It’s great to know there is work in progress! While I can see how I can use the options above to get out of the line, I’m going to wait until a fully featured implementation is in place as I’m in no rush.
But it would be pretty cool to be able to export said audio, especially if we want to add Audio Samples to play when via triggers, it would be nice to be able to capture that.
We have the ability inside of our Machinima app to create a parallel audio track to the movie capture, and then you can take that file and add it to the file generated by movie capture and they should line up well.
We still are working on the ULTIMATE solution, which would be to just record audio and video in one action. It will still require that you press record, then we create all the rendering frames, and then we run through it again to generate the audio. This allows the video to record at its fastest, non-real-time rate, and then audio to do the same, even if those rates are different.
Hope that helps,. let me know if you’d like more detail (or less!)