I made a BUNCH of mistakes recording this podcast. This was all so easily avoidable and I’m beating myself up about it, but I need to see if I can salvage what I did. I was using Streamlabs to record a video podcast over Discord. My guest’s audio sounds fine, but I accidentally had a second microphone active on my computer. So the program recorded myself coming from two microphones, my guest, and some system audio all into one track.
Is there any way for me to make this sound even halfway decent? I figure if I can get my audio sounding alright I can go back and edit the final audio track inserting my fixed audio as best I can. But I don’t even know where to begin with something like this.
That was not good. Don’t do that any more. Now you know why the audio engineer is always wearing headphones.
I never thought I’d say these words: The far side voice is fine, but yours…is not fine.
The Party Line is we can’t take out echoes. That’s basically removing your voice from itself.
However there is a secret incantation you might try.
Step 1. Strong coffee.
Step 2. Zoom into your voice and in particular find the ends of words where most of the sound is echo. It’s not easy. Drag-select that sound, copy, create a fresh, new timeline and paste it there. Do that to a bunch of different words and echoes one after the other. When you play that track by itself (SOLO), it should sound like you down a long tunnel. All tails, echo and distortion.
Select that track (on the left).
Effect > Noise Reduction > Profile > OK.
Select the voice track. Effect > Noise Reduction > 12, 6, 6 > OK.
SOLO the voice track and see what happens to the sound. It may only tighten up the sound a little, but it’s all I got.
Were you able to salvage your audio? I’ve got a similar problem on my hands, where my guest audio and the computer audio are both on a single track, and my guest’s audio (which sounded great live) is barely audible, and mixed in with the background footage’s audio. I’m wondering if it’s even possible to normalize its volume.
I’m wondering if it’s even possible to normalize its volume.
There’s a fuzzy rule that you can’t record a Skype, Zoom, Meetings or other chat on one computer. The chat program aggressively takes over the computer sound services and its goal is for everybody to hear each other, not make a recording.
I did it well with two computers and a small sound mixer similar to the way the broadcasters do it.
Both Skype and Zoom offer a service where they can make clear recordings on their servers and then you can get them to send you sound files after you’re finished. Zoom can even provide individual sound files for each guest. This was such a common problem that even older Skype can do it.
Another fuzzy rule: Audacity can’t split a mixed sound file up into individual voices, instruments, or sounds. Once everything is on one track, you’re kind of stuck.
There are exceptions, but not ones you’re likely to run into. I was at a dinner recently where someone did transcribing of Zoom calls from a single computer. I asked her how she did it. She said her company bought insanely expensive software that competes with Zoom for sound pathways and services … and wins. It wasn’t push-a-button in free Audacity.
One other production note. Most of the noise and distortion in chat calls is Zoom (for example) cleaning up hands-free conversations. If both sides are wearing headphones, cleaning up isn’t needed and the production is much cleaner and clearer. All that and it’s still true that if you record in a noisy room with echoes (wood floor, bare walls) your voice is always going to sound like you recorded in a bathroom. There isn’t good software for removing echoes in post production.
There is a beautifully produced explainer show I can’t watch because they recorded it in a bare room and it sounds awful.