Audio Interface Not Working with AI Software/Agents - RESOLVED

So a few days back, I wrote an article about an issue I was having. A quick summary is that when i was using my audio interface (a Focuslite Scarlett Solo 4th Gen), ChatGPT and Superwhisper would not pick up anything from the microphone. They could see the Scarlet just they weren’t picking up any audio from it. Everything else worked, I could record in Logic, Quicktime, Garageband, etc, etc.

At the time I tried various things to get it to work, check out the article if you’re curious, however here are the things I did that resolved the problem for me.

• In Applications -> Utilities - Open the audio midi setup app.

• Select the audio interface and then the Format menu dropdown and change the frequency to 44100hz.

• Next open the software for your audio interface and in the settings, look for an option to combine both inputs and enable it.

This should have fixed it for ChatGPT.

For SuperWhisper one more step is needed;

• Go into Settings, click Sound and switch on “Use New Audio Engine”

• Quit and Restart SuperWhisper.

That should do it for SuperWhisper as well and everything should be working now.

Couple of notes / things worth mentioning

The issue seems to be, as near as I can tell, that the mic on my interface is on the second channel and apps expect it on the first channel.

You may need to switch off/disable the option to combine both inputs for some software, otherwise you will record in stereo with one side/channel empty.

SuperWhisper support was tremendously helpful and responsive

I hope this helps some of you fighting with the same issue.