How to fix a boomy podcast voice that sounds like an echo
Tell Vizard Agent what you hear in your own words, even if you call it echo. It checks whether there is a real echo before choosing a fix. When the problem is low-end boom from a close mic and the room, it cuts those frequencies gently, compares before and after at the same volume, and leaves the picture untouched.
What do you type into Vizard Agent?
Describe the sound the way you hear it and say how much of the episode it affects. You don't need the technical word. Listeners often call boom echo, and Vizard Agent treats the word as a symptom to test, not as the name of the fix to apply.
There's an echo in the low tones of my podcast. Make the voices cleaner and crisper across the whole episode.
What does Vizard Agent actually do?
It separates two problems that sound alike. A real echo is a delayed copy of the voice and needs echo removal. Boom is too much energy in the lows and low mids, often from a mic held close or a resonant room, and needs an equaliser instead. In the session behind this article it:
- Pulled samples from four parts of a 75-minute interview, indoors and outdoors, and found the two channels almost identical (correlation 0.9987), so there was no delayed copy in one side.
- Measured the energy band by band and had the samples listened to with an instruction not to mistake dull sound for echo. The verdict was no audible reverb, just proximity boom and room resonance.
- Dropped the echo removal it had planned and used gentle cuts instead: a high-pass at 85 Hz, 3.5 dB less at 250 Hz, 1.5 dB less at 470 Hz, a small lift at 3.2 kHz and light noise reduction.
- Compared the original and processed versions at matched loudness, so cleaner could not simply mean louder, and listened for a thin or metallic sound.
- Processed the whole episode, lowered the limiter after finding one peak over 0 dB at 500.6 seconds, and confirmed by hash that the video track was unchanged.
When does this not work well?
When there really is an echo. If the voice comes back as a delayed copy, as it does in a bare room with hard walls, cutting the low end only thins the voice and the echo stays. That needs echo removal, which is why the test comes first.
Common questions
Is boom the same as echo? No. Echo repeats the voice. Vizard Agent measures which one you have.
Will the voices sound thin? Vizard Agent compares both versions at the same volume to check.
Does it change the video? No. Vizard Agent checks the picture is identical afterwards.