Record both input (mic) and output (browser window) on macos

I want to record, simultaneously, the audio from a browser window and the audio from my mic.

Relevant specs:

  • 2015 MBP with 3.1GHz dualcore i7 chipset
  • macOS 12.7.2
  • Audacity 3.4
  • microphone is a Blue Yeti USB mic (which also supports sound output for monitoring)
  • headphones connect via bluetooth but can also support a wired connection (which I could plug into the Yeti mic)
  • “IShowU Audio Capture” audio-routing software installed (it’s basically equivalent to what Soundflower used to do; more recent versions require a more advanced chipset & OS than I have on this machine)
  • Brave Browser Version 1.64.122 Chromium: 123.0.6312.122 (Official Build) (x86_64)

Is this a solved problem, given these specs and resources? My searching of these forums and the internet more broadly hasn’t turned up a solution yet. I have a lot of ideas about ways to solve it, but figured I’d check explicitly first to see if that work had already been done.

You might be able to do it by creating an “aggregate” device in Audio MIDI Setup (look in your Utilities folder). I have no experience doing this so I don’t know if it is even possible.

That’s 3.4.2. The early 3 products had problems.

Use the wired. For no other good reason it will take one layer of complexity out of the experiment.

Make absolutely sure nothing else is running on the machine while you test. Skype, Zoom, Meetings, etc, mess with sound routing.


Why? It’s good to know the goal. If your goal is to record both sides of a Zoom meeting, then I can save you a lot of work. You can’t do that. Zoom (and I think Skype before that) will be happy to record your work for you.

If you’re trying to catch your partner in compromising positions, Audacity can’t be used for surveillance, law enforcement, or conflict resolution.

I did something similar, but I did it with two Macs and a small sound mixer.


There’s something messy about Aggregate Device and I forget what it is.

You can always cheat in a quiet, echo-free room. Lay your phone on the table between you and the computer speakers. No headphones or microphone. Record the room. That’s Voice Memo.

Make sure you can get the work off the phone and onto your Mac. That’s the step I had trouble with.


Ding! Got it.

Aggregate Device insists on you telling it which sound source to use as a timing reference. Given the quality of most home electronics, one of the two sounds is going to slowly drift off sync with the other over the course of a show—and I don’t think you can separate them.

This is where we wait eagerly for The Goal.


Thank you! I appreciate your rapid and comprehensive reply.

Goal: record an interview via a web-only remote chat interface, with a consenting remote participant who will not join a zoom meeting or allow a change in venue to a different medium.

Excellent. Nothing like flying blind in the forum from multiple time zones away.

My first reaction is the guest refuses to be interviewed. Full Stop.

However, I do have a question. How were you planning on talking to the guest? The purpose of Skype, Zoom, Meetings, etc, is so two (or more) people can talk to each other.

If your guest won’t join a Zoom call, then you two will be talking to each other on the phone. If the guest won’t change the medium, then that’s you visiting his house and doing the interview over coffee.


This is me doing a coffee interview with a phone on the table of a music-free restaurant.


Missed one. Such as what? How would I do that?


How new is the Yeti? You know it’s a side-fire microphone, right? This is a common error. You speak into the side grill just up from the company name. As a fuzzy rule, set the pattern on the rear for heart-shaped (cardioid). That’s record from the front (you) only.

You are asking a single-sound-channel computer to record multiple sound channels using single channel software. This is not easy which is why you couldn’t find any handy solutions.

Does the web-only remote chat interface have provision for recording?

If you do find a way to do this other than two computers, etc, post back how you did it.


There’s a blast from the past. I forgot about this one.

If your chat application will support three people, set one computer up in the corner either with a shut-down microphone, or no microphone, and tell it to record the far-side. That will be one channel with both you and the guest.

I did this for conferences where there was no way to straighten out multiple company offices and time zones. I set up one of the participants to not speak, but record everything. Share the conference sound with everyone who missed the meeting, or even if they didn’t.

The only down side was someone had to remember you were doing that, since there was no human associated with that computer.


Thank you again for all these ideas!
Currently, none of them beats using OBS to record a black screen and all the audio streams I want. I can then open the saved output from OBS in audacity and work with the audio.

The drawback there is that OBS is trying to use a GPU this machine doesn’t have to record a screen I don’t care about, it’s constantly complaining about lack of resources, and it’s sometimes slowing down the system.

Re the usage of the (side-firing) mic, cardioid mode: yes, I know how to use my microphone. Thank you. Audio quality on my side of the mix has not been the issue.

Re switching from Bluetooth to the wired output via the mic’s headphone jack: This does seem like a good idea. And I tried it. But it came out worse. I think I might try jacking into the onboard headphone port instead; the USB device seems to do funny things with system resources when it’s playing both sides of the sound flow (in & out).

This “ghost in the chat” approach is my favorite I’ve seen here so far. I do have another machine adjacent, with many attributes to recommend it for this purpose (and a few that have so far disqualified it as the one for conducting this interview). Worth looking into.

On the computer? That typically doesn’t work. You’re not listening to your microphone. You’re listening to your microphone round trip through Audacity Playthrough which will arrive late. This gives an an annoying reverb or echo that you can’t stop. That’s why the connection on the microphone is so handy. No echo.

Having said that, there is one software technique that allows you to listen to the computer for your own microphone. There are software drivers that grab your microphone sound as it arrives and immediately smash it back out the headphone port. No echo, or 85% less echo. You have to turn off the Audacity Playthrough for these systems.

There’s a lot to be said for the third computer recorder method.


Can I use that?


Can you describe this in more detail? It seems to me like if I can do this (and yes, clearly, turn off Audacity playthrough) it should be fairly simple to have Audacity point at that output stream (which should also have the output from the browser) as the target for recording.

No. I never needed it and I always had other processes and hardware available to do what I wanted. From Very fuzzy memory, it was part of a program or app. It wasn’t a stand-alone driver.


Fair warning anyone trying to do this job in local software. The chat or communications software assumes priority in setting computer sound services, and it always wins.

This is why the large commercial packages took off. They work no matter how badly you messed up your machine. They also recognized this condition early on and offered to record your work on their servers for you. Even the old, legacy Skype could do that.

The only reason you wouldn’t want that is if you weren’t supposed to be recording the work. And this snaps us back to Audacity can’t be used for surveillance, law enforcement, or conflict resolution.


This topic was automatically closed after 30 days. New replies are no longer allowed.