Combine audio file dud to OpenAI characters limitations

Questions & Answers
,
1

I have an automatic system to transform our weekly newsletter into a podcast uploaded to SoundCloud. It works like a charm if the original text file is less than 4096 characters. If it’s more than 4096, the system fails as chatgpt module cannot create the audio file from the text.

What automatic solution I can design around this issue ?

One solution I had in mind would be :

  • create a formula in Airtable to divide the original text into pieces that would not exceed 4096 (I still need to see how I can do this one to not cut in the middle of a word)
  • create x audio files based on the number of sections of text I got
  • combine the audio file together to get one file.

Is there a way to combine 2 audio files into one with make and if yes, how ?

Thanks for your help.

2

You could get OpenAI to summarise your newsletter to less than the character limit first.

This is because you may not be able to find an API that merges two audio files together.

2 Likes
3

This was one of the option, or I just merge them manually which is not ideal and defies the purpose of automation :wink: My newsletter is already summarized in one podcast but we have also a main part that we want to keep as it is in a secondary podcast, this is the one we are struggling with.