Take the 2-minute tour ×
Sound Design Stack Exchange is a question and answer site for sound engineers, producers, editors, and enthusiasts. It's 100% free, no registration required.

Hey everyone,

So i shot a short film where one scene happens in the boot of a car. The actor was directed through a iphone earpiece but unfortunately the level was set too high, it bleeds into the recording quite obviously at times. Ive tried pretty much everything i can think of to get it out, eqs, izotope, x-noise etc but i never get it right. ive got the original directors take and tried to match it excactly to the bleed, even re-recorded it through the same earpiece and tried to use phase cancellation, but it just doesnt go away, not even dent a little bit (maybe im not doing it right) Any ideas on what i could do? and since i dont use phase cancellation a whole lot, i might be doing something wrong..

-g

share|improve this question
add comment

6 Answers

Phase Cancellation only work if the signal is EXACTLY the same. So the director speaking it in again won't help, because he can't repeat it exactly like it was. But here you need to have the phase at the position where the mic was. The phase of the signal arriving at the mic is obviously also shaped by the boot and the actors body. So if you had an original recording of the director speaking, you could try to get the actor in the same boot again and the mic into the same position and then just re-record the directors voices through the iphone earpiece. Then you have a chance of it working.

Realistically this is a case for ADR. I may be wrong, but I don't see any other way out of it. Noise reduction plug ins are always designed to keep the voices in and filter out noise. If the noise itself is a voice you have a problem.

share|improve this answer
add comment

Sound 1844 hit the nail on the head. Phase cancellation isn't really an option because the phase-inverted signal has to match up exactly, almost to the sample. The world is chaotic; nothing ever happens the same way twice, at least not to the degree that phase cancellation requires.

What kind of vocalisations/dialogue does your actor have? If he's talking, you might be able to get away with cutting tight around his dialogue - his voice might psychoacoustically mask the director, but that depends on how loud your bleed is. If your actor is just breathing or groaning, you might be able to use izotope RX spectral repair to minimise the director through the earbud (you'll be able to better isolate the high-end earbud bleed against smooth sounds like breaths and groans).

ADR will be your highest quality fix, though.

share|improve this answer
add comment

Yep, ADR, Foley, and Trunk atmo. ORR, use this f-up and cover it up with music...then you have a VERY sufficating scene! The audience would be dying with the character because they can't hear what they expect to hear.

share|improve this answer
    
I've worked on a couple of projects where ADR wasn't an option and the only tactic left was to slather music all over everything and hope for the best...it's not vastly rewarding. –  Joe Griffin May 5 '12 at 7:42
add comment

Thanks for your good answers, Ive got the original track of the director speaking and tried re-recording it through the same earpiece, matching it up perfectly but nothing changes. I guess it is because of what you guys been saying, that everything has to be EXACTLY the same, almost to the sample for it to work. I think i might look into other Ive got the actor to redo his whole performance in ADR which matches up pretty good, but it looks like i need to record the car again because the shot is a one shot performance while driving... this should be an interesting experience! cheers though, i think ill stop trying to use phase cancellation and focus on sorting the problem somehow differently

share|improve this answer
add comment

Covering it up with music is THE sound designer's joke.

share|improve this answer
add comment

Sorry folks, can I just be picky here and point out that you're really talking about polarity inversion here, not phase cancellation. They are really two separate audio phenomena. When I hear "phase cancellation" I think comb filtering, timing/distance issues, and multiple signal interference. Polarity inversion is simply multiplying an audio signal by -1, which I assume is what you're attempting to do to cancel out the unwanted audio. While phase cancellation occurs naturally in everyday sound propagation, polarity inversion does not -- unless that is you are naturally terrible at wiring cables ;)

share|improve this answer
1  
Phase cancellation is everything where an inverted frequency attenuates or completely obliterates an identical frequency, be it direct polar reversal with an identical source, reverbation, or comb-filtering. Polarity reversal as such is not audible at all, the human ear can't hear phase direction, it's only application is together with another signal, either as full-on cancellation, like in the balancing in mic-cables, or removal/lessening of cancellation, like in multi-mic'd drumkits. –  Christian van Caine May 5 '12 at 15:25
1  
When you are talking about the relative orientation of two different signal waveforms, (or the same signal waveform, from multiple sources,) then phase is the appropriate term. Polarity inversion is what they are doing, yes, to one of the input signals. That is what creates the phase differential that the OP was trying to generate. You are correct in that phase is often used when polarity is the more accurate term. However, in this case it is appropriate as we are discussing the relative orientation of two waveforms. –  TheFaderJockey May 6 '12 at 15:45
    
I realize I may not have answered the OP's question, but I still think it important to point out proper use of audio terminology. As others have used in responses and comments, "phase-inverted signal" and "inverted frequency" are really not accurate -- you can't invert phase or frequency -- they are talking about inverting polarity. Point agreed, phase is all about relativity between two signals, but as it is freq specific it would always ultimately result in cancelling certain freqs while boosting others, resulting in comb filtering and unable to achieve what the OP was requesting. –  schwartzsound May 9 '12 at 17:10
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.