Yesterday on a Zoom depo with five attorneys and a witness on speakerphone, I had constant overlap and shifting levels; I ended up slapping temporary IDs and timestamping the rough every 2 minutes. I bounce between arraignments and depos, running Case CATalyst on a Luminex II with a USB boundary mic — any tricks to keep readbacks clean in that chaos without killing realtime?
Kill the speakerphone: turn on Zoom’s ‘Original Sound’ and set background noise to Low, then force push-to-talk so your USB boundary mic hears one at a time. If you’re host, enable ‘Separate audio tracks’ for verification later without touching realtime: Official Zoom Support | Help Center. If you can, move the witness to wired earbuds instead.
Level shifts are usually Zoom’s AGC; uncheck ‘Automatically adjust microphone volume’ and set your USB boundary mic level manually in the OS so it stops playing whack‑a‑mole… If the witness has to use phone audio, ask for wired earbuds with inline mic instead of speakerphone — does that keep your readbacks cleaner?
Quick example: in Case CATalyst, I pre-map one-stroke speaker stamps (W:, AT1:, AT2:) to F-keys and fire them the instant two start, so my readbacks stay clean even if the audio goes mushy… Since you were “timestamping the rough every 2 minutes,” bind one key to Insert Timestamp and another to a bold ‘Overtalk’ marker, then smooth it once the question lands. It’s a few minutes of setup per panel, but it saves realtime — want my hotkey map, @OP?
What finally helped me was routing the call through VoiceMeeter Banana (VB-Audio VoiceMeeter) with a gentle comp + brickwall limiter so the ‘overlap’ stops spiking and my readbacks stay sane. On Luminex II/Case CATalyst, I also mapped a single stroke to drop a [TS] tag so I’m not hand‑timestamping every 2 minutes — do you have a quick macro for that?
‘Original Sound’ on + Low suppression; ask witness off speakerphone. Helped my five-counsel circus — tried that?