Can anyone help me with this issue? I'm spending a day to figure out this problem. Those are all base64 encoded audio buffer I prepared to send to Open AI real time API. For

base64-encoded3.txt

, I get the correct answer from OpenAI. However, for

base64-encoded1.txt

, I get answer from OpenAI which says he can't give answer for audio clip or voice sample. https://cdn.discordapp.com/attachments/1243025299361370176/1316469253829951560/base64-encoded1.txt?ex=677ec188&is=677d7008&hm=adba460e5c5487b9e00bcb663472530cb4e425e155d2386128ce4afa3c00ee64& https://cdn.discordapp.com/attachments/1243025299361370176/1316469254249386044/base64-encoded3.txt?ex=677ec189&is=677d7009&hm=95fea18802a5ef964ecf28b4500b268d3cfb1cd0469faa0ba8e3c9a475b38b6d&