If camera audio and a lavalier each sound acceptable alone but become hollow together, check for delayed copies of the same voice before equalizing the mix. Near-time duplicate signals can produce comb filtering. Keeping one good dialogue source may be cleaner than adding every available microphone.
Listen to each source separately
Solo the camera track, then the lavalier, using the same phrase and a comparable listening level. Note background, clipping, handling noise and intelligibility. If the hollow sound appears only when both are active, the interaction deserves attention. If one source is hollow alone, investigate that recording separately.
DPA's microphone guidance explains comb filtering from closely delayed copies and discusses multiple microphones and reflections. The cause is not the same as simply reversing polarity, and a polarity button is not a universal correction for an arbitrary delay.
Compare timing without losing context
Choose a clear consonant or other shared transient to inspect waveform alignment. Different recording devices may have a fixed offset or changing synchronization, so check more than one location. A visual match on one peak is not enough when the sources drift later.
Keep the raw tracks and make a reversible short test. If alignment is appropriate, listen again rather than assuming aligned waveforms guarantee a good mix. The microphones also contain different room reflections, and a person moving changes relative distance.
A hypothetical interview repair
Suppose the lavalier carries clear dialogue and the camera captures useful room ambience. With both at similar dialogue level, the voice sounds thin. The editor compares the lavalier alone with a mix that uses the camera only in suitable ambience intervals or at a deliberately restrained level.
The review checks speech clarity, transitions and room continuity rather than maximizing the number of active tracks. This is a hypothetical mixing decision, not a measured demonstration that a particular alignment or level solves every recording.
Decide each track's role
Use a track as primary speech, backup, ambience or a consciously designed blend. Do not keep two full-strength speech copies merely because they were recorded. A useful backup can remain muted and available for a damaged word instead of coloring the whole narration.
Accept the final mix when required speech is clear and stable through the conversation, including movement and edits. Reject a hollow combination or an alignment that repairs the opening but fails later. Check the actual delivery's channel behavior as an additional review; mono compatibility is a separate test, not the only definition of comb filtering. Save source roles and edits with the project. These procedures belong to external dialogue editing and do not imply automatic multi-microphone correction in Imagild.
Sources
Prepared by Imagild Editorial with AI assistance and reviewed against the linked sources on October 13, 2026. Examples are hypothetical; external procedures do not establish Imagild features or guarantee results.