6

We occasionally have a situation where one speaker is wearing a wireless microphone and another speaker is not. (Usually a translator would be using the wireless headset, standing next to a guest speaker).

When the speaker wearing the wireless headset moves away or nearer to from the fixed mic, a sweeping phase sound is audible.

No one has complained so far and I hope I'm the only one hearing it, but I'd like to fix it for the recording (at least).

So far I have only found static plugins like InPhase, Auto-Align, and so on, but these do not add or subtract delay dynamically with the moving mic. I'm sure it shouldn't be hard to do the following part (phase-locked-loop style of follow the leader), but I guess that smearing the difference in delay out smoothly could be a relatively heavy process.

I've been searching with various keywords like "Doppler compensation", "dynamic phase correction" and so, but maybe I'm using the wrong ones? The closest I could find was Pi but the description is a bit fuzzy about what it's actually doing.

A similar thing happens with a mic above a white-board and an overhead "shotgun" mic pointing to the whiteboard from the sealing. When the speaker walks to-and-fro between these two mics there would also be some phase shifting. Though I havn't been able to audibly identify these (yet) I'd still like to keep them in-phase.

It would be even greater if we could have such correction in the Live setting, but for now I'd settle with a plug-in when mixing down the final recording.

Michael Hansen Buur
  • 3,802
  • 15
  • 31
Louis Somers
  • 388
  • 1
  • 12
  • 2
    Why not simply use a more directional mic for the stationary speaker so that you don't get the cross talk issue at all? If you can isolate the channels so each mic only pics up the relevant speaker you will get better sound quality and the phase issue will go away. – AJ Henderson Dec 03 '13 at 21:41
  • True, we have considered a more directional mic there, but it would also constrain some speakers who like to move a little while passionately expressing themselves. Having each speaker wear a headset would be the "42" solution, but unfortunately that is not always feasible (with different speakers following each other up, it would be quite a hassle). We finally settled with this trade-off. Occasionally I also fade-in an overhead shotgun mounted on the sealing when a speaker walks away too far from the main mic. –  Dec 03 '13 at 22:00
  • You can still use the existing solution as a backup and avoid the problem 90-95% of the time. :) – AJ Henderson Dec 03 '13 at 22:38
  • Yep, but the consensus is generally "It's working great without headset, so why bother wearing one...". It also seems like hardly anyone is aware of the phasing effect anyway, so I just want to fix it for the recording-mixdown... No-one knows of any existing plugin so far, so maybe I should patent this idea and build one myself? :-) –  Dec 05 '13 at 15:37
  • My guess is that there are technical reasons that make it difficult to do this, but I don't know enough about the signal processing involved to say for sure one way or another. Moving signals with different frequencies seem like it would be very hard to manage to identify phasing issues on the fly without weird artifacts though, but that's a gut feeling rather than anything based in fact or knowledge. – AJ Henderson Dec 05 '13 at 15:40
  • Getting the initial lock would be easy, just shift the two tracks from a minimum to a maximum offset and pick the position where the volume spikes. Once you have a lock it's easy to follow, until the signals differ significantly (when the person without headset starts speaking). The tricky math comes in when compensating for detected movement, involving stretching or pitch-bending. Maybe it's easier to just eliminate one of the tracks and compensate for the volume drop on the other when detected... I don't seriously think I can build anything usable in my meager spare time though. –  Dec 05 '13 at 16:21
  • Yeah, that's what I meant about the artifacts. When you start trying to slide things around on the fly, there is probably some potential for some weird issues to crop up in terms of what it would sound like. The cure may be worse than the disease, or I could just be completely wrong. – AJ Henderson Dec 05 '13 at 16:36
  • Have you considered using side-chain comrpression? –  Dec 20 '13 at 20:22
  • @Don Nickel, I'm using compression (from the channel strip) which helps a little. Guess I'll just leave it as is, no one seems to notice it any way. –  Dec 21 '13 at 12:50
  • 6
    I mean use the compressors to gate the mics. So, speaker #1 talks, it gates speaker #2. –  Dec 23 '13 at 17:56
  • Didn't think of using a compressor that way (it'll probably not be possible with the built-in channel-strip compressors, so a side chain would be needed). On the other hand, the translator often starts translating before the sentence is finished (they overlap a bit), so I'm not sure if I'm going to try this, maybe if I can tweak it to get a gradual slope without fully muting. I recently removed some compressors from our 19" rack, thinking we no longer need them with the new mixer having them built in. Great idea! –  Dec 26 '13 at 01:36
  • 3
    @DonNickel If you add your tip as an answer, I'll mark it as an answer! Finally got the chance to test this yesterday using the "duck" function on the main mic triggered by the translators mic. I tweaked it so that the ducking effect is inaudible but enough to get the phase shifting effect inaudible as well. It worked out great! – Louis Somers Jan 25 '14 at 13:41
  • Vocal align is one plugin that can help you , if u think about how it manupalates the phase. – Jay v Apr 28 '15 at 23:55

1 Answers1

1

What you actually need here isn't a "phase aligner", it is an automixing solution like the Dugan automixer. This is built in to some of the higher-end recorders like the Sound Devices 788T or 688T but is also available as a plugin - I believe that Waves supply a Dugan automixer.

An automixer has the ability to automatically manage the gain of a suite of microphones so that phase artifacts are minimised. Only an active microphone is going to be mixed in to the output bus, however it is clever enough to always leave at least one mic live so that you don't get dropouts.

Mark
  • 8,870
  • 1
  • 15
  • 27
  • 1
    Sorry I'm a little late to respond, didn't get to digging into what a "Dugan Mixer" does until an hour ago, but Thanks for this tip! I first checked out what a Dugan Mixer was, (thinking it would be a clever expander or so), and starting be become convinced, bumped into a link to a youtube introduction of a new firmware update for the mixer we are using, which actually has a Dugan Mixer added as the main new feature, downloadable for free! The demo's really look convincing and I don't need to purchase any extra hardware or software! Wow, you got me exited now :-) Thanks again! – Louis Somers Dec 08 '16 at 22:20