What Happened When I Tested Alpha Labs De-Feedback Plugin

You know that moment during announcements when the pastor steps up to the podium and you're already riding the fader, waiting for the feedback to start? The mic is barely loud enough to hear, and every time you push it up a little more, there it is…that ringing 2 kHz frequency you've been fighting for months.

I've been there. A lot of church techs have. Ringing out microphones with cutting EQ in those sensitive spots helps, but it can only go so far.

Recently when I was scrolling social media, Alpha Labs De-Feedback Plugin started being all over my feed. Engineers I respect were talking about it and showing it in use at big shows and the claims of the plugin sounded amazing. I'd seen enough that I wanted to test it myself, so I bought the Option 1 with the recommended Focusrite Scarlett Solo 4th Gen and put it through its paces on a handful of corporate shows in large convention center ballrooms.

The result: 10 dB more gain before feedback. That's what I personally measured, both on stage and standing in front of the PA with a lav mic clipped to my shirt.

On one of the shows, I had 4 wireless lavs, 2 podium mics, 6 tabletop mics, and 4 wireless handheld audience Q&A mics. All of them routed to one speech bus with De-Feedback inserted. Even the Q&A mics were out in front of the PA. The whole group got the benefit of that single instance.

I verified the results with Smaart. De-Feedback doesn't touch your EQ curve. If you do a bad job ringing out your mic, it's still going to sound bad. What it gives back is the EQ bands you would have otherwise burned fighting feedback nodes. On a Yamaha CL5 where I normally use most of my PEQ bands and notch filters cutting problem frequencies, I was down to two notch filters and one parametric band.

Everyone wants to know about latency. Alpha Labs rates the Option 1 at 4.9 ms roundtrip. My measurements came in at 4.83 ms at 48 kHz with a 16-sample buffer. That's fast enough that you won't notice any delay.

I tested every major sample rate and buffer size combination so you don't have to. The full data is in the blog post.

I also put together a complete setup guide for using De-Feedback with the Behringer X32 using the Aux insert method along with the Option 1 setup that Alpha Labs recommends.

The Behringer X32's Aux path adds about 0.9 ms of latency, bringing the total roundtrip to 5.73 ms at 48 kHz with a 16-sample buffer. Still well under the threshold where you'd notice any delay on speech or singing.

The post covers physical connections, gain structure (including why you should use Input 1 on the Scarlett Solo instead of Input 2), and step-by-step routing for inserting on a single channel or a speech bus.

Until next time,

Drew

Here’s how I can help: