Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Is it possible to change the pitch of a voice?

  1. Jul 30, 2017 #1
    Yes, this is kind of a weird question to make, but I haven't found anyone discussing it so I decided to ask it myself.
    The question: Is it possible to change the pitch of a voice up two tones (for example), in order to create a harmonizer that has no electronics involved?
    The idea and where it comes from: I am a high school student from Argentina, and went on an exchange for one year to Houston, TX, USA, and now I'm taking all the required exams to get back into my school for this senior year. My music teacher has asked me to come up with an instrument, plan it, build it and classify it.
    It was going to be a guitar, but I like original creations and came up by accident with a tremolo for voice (easy to make) and thought it would be cool to present some sort of pedalboard for the voice, but with no electronics involved because it needs to be cheap (part of the assignment) and I have no knowledge in electronics (I suck).

    I just need somebody to tell me that it is by all means impossible (like, don't say I'm crazy), OR somebody to tell me hey, there's that already, somebody came up with it yesterday, and show me a link to it.

    Thank you beforehand, and if you want to help me with more ideas for hand made voice FX, that'd be awesome!

  2. jcsd
  3. Jul 30, 2017 #2
    The electronic way of doing it involves digitally sampling a sound, then messing with the digital numbers to get the desired effect,
    and then finally converting the massaged numbers back to an analog signal.
    I can't think of any realistic mechanical way of doing the equivalent process.
  4. Jul 31, 2017 #3


    User Avatar

    Staff: Mentor

    Welcome to the PF. :smile:

    What is a "tone" in this context? (Sorry if it is something I should know) Do you mean by two octaves?

    One thing you could try is to have a mechanical arrangement that generates harmonics from the original sound. The harmonics would generally be odd harmonics, so if the original frequency is fo, the harmonics would generally be 3fo, 5fo, 7fo, etc. The amplitudes of the harmonics get smaller as you go up in frequency.

    To generate even harmonics, you need to make the sound generation uneven for the two directions of the oscillation of the string or speaker or whatever you are using to make the sound. If you can make the motion asymmetrical somehow, you could generate even harmonics like 2fo, 4fo, etc.

    Here is a link that may help you start to understand harmonics of a base sine wave waveform:


    EDIT -- Or as shown in the Wikipedia link's diagram, you may be able to use a tuned cavity to generate even harmonics somehow, if the original sine wave frequency is fairly constrained...
  5. Jul 31, 2017 #4


    User Avatar
    Gold Member

    Inhaling helium does raise the pitch of the voice quite substantially.
    Maintaining the increase is difficult, although deep divers breathing helium/oxygen mixtures are examples.
  6. Aug 1, 2017 #5
    Dang....beat me to it.

    Any gas that has a different speed of sound than "air" will change the pitch.

    As far as modifying the pitch of a voice in the acoustic realm, you might want to explore doppler effect.............but that requires a bit of movement.

    I would have used the guitar and a slide.
  7. Aug 2, 2017 #6
    I would imagine he means DO-RE-ME, those two musical tones. A semi tone is from one key to the next closest key on the piano.

    Doppler effect is what came to mind for me as well. It is part of the sound of the Leslie rotating speaker. But that is a modulation, it can't produce a steady state shift. I can't imagine any practical way to use mechanical motion for a steady-state shift.
  8. Aug 3, 2017 #7


    User Avatar
    Science Advisor

    Doppler Ultrasound imaging of fluid (blood) flow speed is commonly done. The drawback I see is that the fluid has to move a substantial fraction of a wavelength per cycle of the sound.
  9. Aug 3, 2017 #8
    Several cultures have polyphonic singing traditions. This MRI research project video shows one of the techniques for singing two tones simultaneously.
  10. Aug 3, 2017 #9
    Fascinating, Here's another of hers, for my ears, the ascending/descending run at 0:40 ~ 0:50 is the clearest example where I can hear each of the harmonics being emphasized, along with the fundamental. In some other cases, they blend enough, it sounds more like one 'tone' to me.

    Maybe semantics, but I can't quite accept this as "singing two notes at the same time". What these singers do, is generate a low fundamental rich in harmonics. They use their vocal tract to emphasize harmonics that are already there - we all produce these harmonics when we speak/sing, so we could then say we are singing two notes at the same time. The trick is emphasizing select higher harmonics by forming the proper resonant cavity - quite a feat!

    This is pretty easy to simulate in an audio program like Audacity. Generate 30 seconds of a low frequency, harmonic rich tone like a sawtooth (odd & even harmonics). Use low amplitude so the harmonics can be boosted w/o clipping. Probably best to then use EQ to boost those harmonics, get it really 'buzzy' like her voice. Then select a few seconds, and apply a parametric EQ at 2,3,4,5,6,7,8,9,10,11,12x the fundamental. You will hear something much like what she demonstrates. But only one tone was generated, t's not like two separate 'generators' of pitch.

  11. Aug 3, 2017 #10
    There is a mean to do it. You can hear what i mean if you have a Ham SSB short wave radio. Anyway, to do exactly what you need using this method is not easy.
  12. Aug 3, 2017 #11
    There are many ways to do it electronically. The OP was looking for non-electronic methods.

    I think the SSB effect is much like what musicians call a "ring modulator", but I think the non-linear mixing to obtain sum + difference frequencies is done within (and/or maybe somewhat above) the audio band. That would allow for filtering out some of the products. Not sure where that name "ring" came from, I may look it up.
  13. Aug 3, 2017 #12


    User Avatar
    Gold Member

    There is a distinction between moving the frequency bodily upwards by a number of Hertz, as with SSB, or multiplying the frequency up by a factor, as by taking harmonics. The latter case will create a sound which is melodic because its harmonics are still correctly placed, whereas the former will not.
    By the way, the ring modulator circuit looks like a ring of diodes, as used for carrier telephony.
    If you multiply the frequency of audio, it tends to still sound the same, because the ear fills in missing sub harmonics.
  14. Aug 3, 2017 #13
    I think it is because the way four diodes are connected in this kind of modulator.
  15. Aug 3, 2017 #14
    The only ways to do it mechanically is to have a chamber tuned to use phasing to bring about a different note or have an object touch vocal cords much like a guitar slide.

    I would tell the teacher your findings and request to go the guitar route.

    Your declaring it impossible or near impossible with reasons should be enough.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted