I recorded my dog growling and shifted the pitch down -12semitones to get a really deep growling rumble for a monster sound. I put this sound in a video and I got a comment asking if it was an MP3. I recorded at 24bit 96k so Is drastic pitch shifting bad for quality?

Here is the link to both sounds. http://soundcloud.com/shaunkelly

4 Answers 4


It can be based on the quality of algorithm used to drop the pitch. Digi's factory Pitch Shift plug in is faaar below the quality of number crunching compared to some of the higher quality (and more expensive) plug-ins like Serato Pitch 'N Time and Ableton Live's Pitch Shifter.

Another factor is sample rate - drastic changes to pitch or time can degrade sound quality to the extreme based on the above stated factor. To combat this, try using 192k on an HD system if you have access. As you know, the more samples = the higher definition for processing on input and allowing a higher definition output. Bit depth doesnt matter I wouldnt guess as its main purpose is in the dynamic range. A sampled frequency is a sampled frequency regardless of where it is on a scale from amplitude range. Quality is about # of samples.

Another option is to pitch the dog down only say half as much, and use a low end harmonic exciter/generator such as LoAir, LowEnder (I think thats what its called), or iZotope Ozone harmonic exciter to boost the low end material. If the low end is there originally, boost it - if it doesnt have much low end spectrum to start out with, use a subharmonic generator for supporting the sound.

Let me know how things work out!


  • So in general: it does affect the quality, because the technical procedure is about some level of numerical approximation or exchanging speed to quality or quality to speed. That's why offline rendering (which also has different choices for quality and speed) is generally noticeably better, because real-time has to be computationally more light weight. Commented Jun 9, 2013 at 10:34

LIstened to both samples, doesn`t sound bad to me. But remember, that when you pitch down a sound your highs will probably get your mids. Your may get your low rumble, but you will loose your high end, this is maybe the reason, why that certain person thoght, that it is an mp3. Try to layer the unprocessed version above it, this way you will get both: your low rumble and your high end.

  • That's a very good idea. Though I'd be curious to hear the result, as you might end up with something sonically dissonant. Commented Nov 22, 2010 at 22:02
  • I did what you suggested. It sounds better. I layered the original with the processed then EQed them both and compressed them together. Then added a De-esser to get rid of some highs that sounded weird. soundcloud.com/you/tracks Its the top track
    – ShaunKelly
    Commented Nov 23, 2010 at 5:06
  • WadioSound it can sound dissonant, it even can sound musically. thats why I do in most cases -12 semitones or -24. this way its an octave and it doesn`t sound musically at all. Commented Nov 23, 2010 at 17:07

Many microphones are specced up to 20kHz frequency response. They do capture frequencies above that, but the levels would require EQ-ing. We can't hear frequencies above 20kHz so EQ makes sense post pitch-shifting. I'd expect rebalancing to be necessary with most sounds re-pitched. It's not news but just a reminder to refer to the Nyquist rate as a general rule whether/how-much re-pitching will affect high frequencies..


Every processing of audio, even if you just move a fader reduces the quality of the audio. But of cause not massively. You should only be concerned if you do too many processing steps to the same audio.

But of cause there are plug ins that have a bad algorithms. Just hear how it sounds with a different pitch shifter.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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