Convert AudioClip to Byte[]

Unity Engine
,
1

Hi, i need to convert an audioclip to a byte array, i know GetData from audioclip doesn’t help because i tied writing a wav file with it and it can’t be reproduced. Someone who can help me?

2

GetData is the way to get audio data from an AudioClip. That doesn’t mean you can just dump that data into a .wav file and expect it to be playable. I’m not sure how WAV files work but there’s probably more to it than that.

I found some code online that might help if you’re wanting to write a WAV file: Unity3D: script to save an AudioClip as a .wav file. · GitHub

3

In fact, i have used that, it creates the wav file first and the write more data over it, it’s good but i need to have the bytes for sending to a server, this server must be able to read it. What i am doing now is using that code you send me, but after create the file, i make unity read all bytes of that file and then upload to the server, something weird :frowning:

4

I needed a straight forward solution for converting audio clip to byte array :

  public static byte[] Convert(AudioClip clip)
    {
        var samples = new float[clip.samples];
        clip.GetData(samples, 0);

        MemoryStream stream = new MemoryStream();
        BinaryWriter writer = new BinaryWriter(stream);

        int length = samples.Length;
        writer.Write(length);

        foreach (var sample in samples)
        {
            writer.Write(sample);
        }

        return stream.ToArray();
    }
1 Like
5

I am having difficulty converting audio clip to a byte array (using darklord’s solution). It seems that the float array is populated by NULL after I use GetData and I have no idea why.

Also, why are you writing the value of the float array length into the memory stream?

6

float can not be NULL … and GetData doesn’t reassign the samples variable … the issue seems to be somewhere else

And you prepend the length of the array so that on receiver side you can recreate the array again by

  • first reading 4 bytes → length
  • create new float[length]
  • now copy the rest into the float array with given length