When MovieRecorder's Configuration is set up with hasAudio: true, the internal recorded waits for both a sample of video and audio to actually do the processing
guard self.videoInputs.count == self.configuration.numberOfVideoTracks && self.audioInputs.count == self.configuration.numberOfAudioTracks else {
return
}
So any buffer sent before both one image and one audio have been provided is discarded.
Maybe it shouldn't?
Also, there is no explicit information on that situation, so if one doesn't provide any audio sample (while testing stuff for example), the asset writer ends up in an "unknown" state when trying to complete the recording, and it's not clear why/what happened.
When MovieRecorder's Configuration is set up with
hasAudio: true, the internal recorded waits for both a sample of video and audio to actually do the processingSo any buffer sent before both one image and one audio have been provided is discarded.
Maybe it shouldn't?
Also, there is no explicit information on that situation, so if one doesn't provide any audio sample (while testing stuff for example), the asset writer ends up in an "unknown" state when trying to complete the recording, and it's not clear why/what happened.