Did you record using the MediaStreamRecorder?
With the MediaStreamRecorder I got the same 'error', if I use StereoAudioRecorder everything works fine, the inconvenient is the file size

10 seconds:
350 KB with StereoAudioRecorder (numberOfAudioChannels: 1)
67 KB with MediaStreamRecorder

I want to capture in mp3 format to upload it to firebase but I get error in infinite audio

@gdurga can you provide a part of code how you get duration?

@vladpoltorin Sure! Here is the code
First record the audio with the MediaStreamRecorder to reduce the file size, like this:

    recorder = RecordRTC(microphone, {
        recorderType: MediaStreamRecorder,
        type: "audio",
        bufferSize: 4096,
        sampleRate: 44100,
        checkForInactiveTracks: true
    });

Then convert the file using ffmpeg to fix the infinite audio duration
ffmpeg -i <original_file> -c copy -fflags +genpts <fixed_duration_file>

The resulting file are pretty much the same file size of the original and it has a defined duration

Also, when I try to record video with MediaStreamRecorder or MultiStreamRecorder I get infinity duration. StereoAudioRecorder works fine, giving the duration of the audio recording and WhammyRecorder results in a valid duration also.

@gdurga
Is there a way to get the original duration without using ffmpeg?

Hi @nikosgi, unfortunately the only way I found to get the duration using the MediaStreamRecorder is using ffmpeg to convert the file after the record :/
Using the StereoAudioRecorder works fine for me too but the file size will be much larger

Here is a very basic example with Django -> https://github.com/gdurga/Django-RecordRTC-example
Converts the file after the upload using the python os.system

Someone mentioned this as well, however I didn't try it myself yet:

https://github.com/legokichi/ts-ebml

Can someone make it work without type-script compilers?

When I record the screen using throw the recordRTC but video length is infinite. Could you please tell me what should I do?

Did anyone get a solution?

I have integrated below solution. (It's working fine in desktop Chrome Version 100.0.4896.75 (Official Build) (64-bit))
Still need confirmation from different OS and browser.

In order to get the duration inside recorded file we have to add ts-ebml dependency.
I have used webm-duration-fix npm.

By integrating above solution it's fixed my recording duration issue but still need to test in other OS and browsers.

I have integrated below solution. (It's working fine in desktop Chrome Version 100.0.4896.75 (Official Build) (64-bit)) Still need confirmation from different OS and browser.

In order to get the duration inside recorded file we have to add ts-ebml dependency. I have used webm-duration-fix npm.

By integrating above solution it's fixed my recording duration issue but still need to test in other OS and browsers.

I did some test in android and IOS latest version by today date (14/04/2022).

It's working fine.

can this be used for audio recording as well?

This works for me!

useEffect(() => {
    if (!audioRef.current) return
    const audioContext = new window.AudioContext()

    fetch(audioRef.current.src)
      .then(response => response.arrayBuffer())
      .then(arrayBuffer => {
        return new Promise((resolve, reject) => {
          audioContext.decodeAudioData(
            arrayBuffer,
            buffer => resolve(buffer),
            error => reject(error),
          )
        })
      })
      .then(buffer => {
        setDuration((buffer as AudioBuffer).duration)
      })
  }, [])