Loading...

So which of those is it or neither? Well, the former, but it could technically do both. If there's enough room to run AGC and compression and stuff after the NR, which is basically the way this thing works is it's a machine learning noise reduction. It's not traditional noise reduction. It's more like vocal isolation where it says, OK, here's noise, here is vocals. Let's just separate these out entirely. And the noise just basically disappears. So I do at some point, if I ever get to the point where I can make this work live, I'd love to try it on a repeater where people have kind of marginal signals. And after a lot of training, of course, to make sure that this can actually be done. It would sort of be interesting to see what that would do in a repeater situation. So that's it. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that. I'm going to go ahead and do that.