I use two USB microphones as an aggregate device to podcast into Audacity. Person A records into the left channel of the aggregate device. Person B records into the right channel of the aggregate device. They form into one stereo track and then I compress it down to mono so when you listen to it, one person isn’t just on one side of your headphones.
Here is my problem:
When Person A is talking (and recording into the left channel), you can hear their voice ever so slightly in the right channel because my partner and I are sitting in a small office and we aren’t that far away from each other’s microphones. The result is that you hear a slight echo in the broadcast because the signal is coming in twice, once on the proper channel, and then again, delayed, on the other person’s channel.
Do you guys know of any way to correct this in post-production? Noise Removal? Or anything else?
And if there is no way to correct this in post, how can I prevent this in the future? Do I need to turn down the audio level for the microphones? Sit further apart? It’s not like we’re sitting RIGHT next to each other - there’s a desk in between us.
It may actually be worse than that. Your microphone may picking up your partner’s voice from the walls. Room echoes are deadly.
Echoes are also one of the permanent fatal problems. No software cure. The grownups do shows like this in a room with no echoes and then they use directional microphones in front of each actor.
I have pictures of people doing this, but they’re not electronic or convenient.
You can go to headset microphones and put the microphone up close and personal. The room and everybody in it goes away when you do that.
This is where you figure out how serious you are about this.
You might figure out the polar patterns of your microphones and put yourselves in each other’s nulls. Cardioid microphones tend to have a good “hole” right behind them.
There is a free plugin called Kn0ck0ut (Windows only) which will spectrally subtract one track from another.
You could align the two tracks so the original voice on one track coincides with its echo on the other track, then use Kn0ck0ut to subtract the echo.
Don’t get your hopes up though , Kn0ck0ut creates digital artifacts which may be even more obtrusive than the echo.
I actually use headset microphones. It’s odd. The beginnings of the podcast, everything sounds fine. The echoing starts in the middle and gets pretty noticeable at the end of the podcast. I have a reasonably fast computer (Intel Core 2 Duo, 2.4ghz) so I don’t think it’s the CPU being shitty.
“skipping” can cause tracks to go progressively out-of-sync … Multiple Layers not syncing right
The acoustic crossover may not be noticeable when the tracks are in-sync, but as they become increasingly out-of-sync it becomes more and more obvious: the delayed acoustic crossover becomes an echo with a progressively increasing delay-time.
If the echo problem takes several minutes (“middle of podcast”) to appear then I don’t think it’s due to “enhancements”, whose effects become apparent in a very few seconds.