I’m trying to record a podcast using Vocaster 2, MacBook Air (Sequoia 15.6.1), Audacity and Google Meet.
I have BlackHole 2ch installed on my MacBook.
I have tried various settings.
If I have Vocaster 2 as the input device on Audacity, it only records my voice but not my guests.
If I have BlackHole set as the input device on Audacity it only records my guests voice but not mine.
On my MacBook I have tried both Vocaster Two and BlackHole as the input and output devices, but it still doesn’t work.
I also tried the option of creating an aggregate device in the Audio Midi Set Up (see pic attached) and have tried selecting that for the input and output devices on Audacity but it doesn’t record mine or my guest voice when selected.
Google Meet audio settings are Vocaster Two for input and output.
Hi, did you come up with a solution? If not, I’m leaning toward a browser limitation. Safari does not fully support setSinkId, so it can’t direct WebRTC audio to a specific output device. Chrome gets bagged on Mac for plenty of valid reasons, but if you are open minded it will do this.
Another way to think about it is your interface is working, it just can’t read the signal Core Audio is outputting. We think about audio in an analogue way, but in the digital world it’s data streams, not sound. Platforms design their services around how that data is managed within their ecosystem. Sometimes you can work around those limits so you are not dependent on the platform and its features. For example, you can change browsers, or use QuickTime to start a screen recording and split the audio afterwards.
I believe Zoom will still record all sides of a connection for you. At one time, they would supply you with individual sound files.
I have been able to do well in a quiet, echo-free room by placing a sound recorder on the desk and record the computer speaker and my live voice. You may have to juggle a bit.
You are fighting a bloody battle with Google Meet whose job it is to keep all the conversations perfect, clean, and separate, and suppress everyone’s room noise and feedback damage.
There is a slippery cheat technique. Bring up a second computer in the conversation. It has no microphone or has one that’s turned off. You can record its speaker feed to hear everybody. I did that where I used to work. I had to record several timezone’s worth of conversations, production development, and client instructions.
The one on the right thinks its sending the show to me and processing my voice. Really, it’s going to that tiny sound mixer and then on to the computer on the left which is recording everything (In Audacity).
The machine on the left has a sound channel left over, so I used that to play theme music, stingers, bumpers, etc.
That is a technical experiment, so it’s not at all polished. Denise is three times zones away.
Omg Koz, big ups for effort! Is that a legacy 802? We used to roll a similar rig - 1204FX into an 2496 sound card. However, audio is now 100% digital & managed by os. The interface acts as both an external sound card & input/ output for mics, instruments & ‘devices’. It makes the desk redundant for this stuff so I’m stuck on the why?
I hadn’t used Audacity for a while but I’ve got this going with a 20 year old Fast Track & Safari on an M1 iMac running Sequoia 15.7.2. The interface is using the os core audio. I’ve also tested it on a 2012, 2019 pro intel & 2024 m4. Thumbs up.
We did this sort of thing because of technical limits eg we’d live read over carts to avoid having to put it into a single wave editor like Cool Edit & then sequence it in sawclassic. We would have been on something like a p4 w/1.8 clock, 256 ram, onboard video etc. It was way easier to work in analog like this!!
HELP PLEASE!
