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.

I got some voice over files today from a new studio, and my co-worker was confused as to why the sound wave looks like this.

His concern is that the wave should be pretty even above and below the -∞ line, but the lower part of the wave appears to be half-peaking along a consistent line.

Anyone know what's happening here? Is this normal, or a result of faulty/poor equipment?

sound wave of human voice

EDIT: as a reference, here is how we would expect a wave to look

enter image description here

share|improve this question

migrated from avp.stackexchange.com Jan 24 at 12:01

This question came from our site for engineers, producers, editors, and enthusiasts spanning the fields of video, and media creation.

3 Answers 3

up vote 10 down vote accepted

There can be multiple reasons for such a waveform:

  • Transient-heavy signals from the source itself, the transients being asymmetric. As said by Friend Of George, this can easily happen when recording drums, but also on voices – however it's mainly pop noises there, which should be removed with a pop shield and/or high pass filter anyway. The signal does not look like that's the problem: not only the transients but also the sustained parts are asymmetric.
  • Phase-coherent signals. These occur mainly in simple synthesizers and sometimes in instruments like DI-recorded fretless electric bass, but should not happen in any signal recorded through a microphone, because acoustic transmission does not keep the exact phase relations. However, it may sometimes be accurate enough to retain some of the asymmetry of someone's vocal cords' oscillations.
  • Asymmetric distortion, as produced (in a quite pleasant manner) by tube amplifiers but sometimes also by poorly designed solid-state circuits. This is not usually visible in the waveform because such distortion will often be phase-scrambled and debiased by following filter circuitry and/or loudspeakers+microphones, but not so if the culprit is the AD converter or directly before it. This seems the most likely cause for your waveform.

As I said, it is disruption of the phase relation between the harmonics that usually prevents asymmetric waveforms. The phase relations are changed in any normal equalizer, so it's quite possible that the waveform will look perfectly normal after EQing (especially low-cutting). If that's not enough without making it sound bad and you're still worried about the asymmetric waveform, you can also mangle the phase relations without affecting the sound, by means of an all-pass filter. You may need more than one, set to different frequencies in the range 80-800 Hz. Other things that will probably mend the asymmetry:

  • Multiband compression
  • Symmetric clipping, soft or hard/brickwall limiting
  • Modulation effects (but I guess you don't want that on a voice-over)
  • Heavy short/gated reverb (also not so nice for a voice-over)
share|improve this answer
1  
Excellent answer - I learnt stuff :-D –  Christopher Woods Aug 16 '11 at 17:25

How very interesting! I do pro voiceover, and my raw trace looks asymmetrical when I compare it with the work of peers (I can't check their original recordings of course). First time I thought about it was when I happened to step behind a figure of eight ribbon mic and noticed a quality change along with inversion of the asymmetry.

Some BBC research of long ago found that male voices in particular can be asymmetric by up to 8dB, and in USA the mighty Robert Orban among others used phase rotation, or scrambling if you like, to ensure maximum modulation on AM. In Britain it probably happened anyway through the BBC’s then use of enormous numbers of transformers one after another in the national broadcast chain! I asked my tutor: he said 23 and counting.

Back to my ribbon: With the help of phase-swapping line transformers, I confirmed my equipment was indeed adding distortion and found it preferred the mic input to be phased such that some cancellation could be achieved. Not very nice!

Now, with a completely new set of equipment there’s no audible distortion, and quality-wise it doesn't care about polarity, but the loppiness persists. It looks much less obvious if the bigger peaks go upwards, so I have set the desk to do that, although you could just use polarity reverse in the software.

As leftaroundabout said, one remedy is peak limiting and yes I do that. Just enough to trim the infrequent tall spikes. Which are all on one side. Looks like a crew cut.

share|improve this answer

I have noticed this same phenomenon when recording drums using an Shure SM-58 microphone. I only tend to see it during the louder parts, so my best guess is that the sound pressure is so great it is pushing the diaphragm of the microphone in more than it is being pushed out.

share|improve this answer
    
That makes sense that that would happen, but I can't imagine that being the case with a simple voiceover :-/ –  Jeremy White Aug 13 '11 at 5:48

Your Answer

 
discard

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