Question of Pitch Perception

[guyman]

Post a paper, talk out of your a**, grab an oscillator- do a sweep and give an honest personal answer.

At what lowest and highest frequencies would you say your personal sense of definitive pitch begins and ends? When is it just a "feeling" or a timbre? We can hear frequencies from about 20hz-20khz, and can consciously sense/feel them lower and higher (given great enough magnitude)- but what is the window you would consider it a "tone" or "note" and not just noise, or vaguely sensed overtones, or rumble. You hear the tone and can nail the note, feel the harmony/melody.

What is the range?

[martinvicanek]

Interesting question. I think the result will depend on how you conduct the test. I would expect that the timbre of the tone has an influence. You’ll probably get a smaller range with a pure sine. Then, how do you actually test whether or not you perceive a pitch? I would suggest to play two or more different notes in sequence, perhaps a little melody, and see if you recognize it as such.
Another thought comes to my mind. The low B on a bass guitar is around 30 Hz, and I would not claim that I perceive a pitch in that region. However, in a band context with other instruments, you can hear immediately if the bass is a semitone off.

[tulamide]

Many things influence the answer.

For example, before I bought my monitor speakers I would have said that anything below 60 Hertz is not pitch anymore. Now with these higher quality monitors, it's anything below around 33 Hertz. The same is true for the high end. Before I had a hard time to recognize frequencies above 8 kHz at all - now they are clear pitches.

Age is another factor. You lose the ability to sense high frequencies with every year. At one point you are lucky to recognize anything at 12 kHz or above. Let alone recognizing clear pitches.

I'm afraid this is a topic that won't have any definitive answer. 20 Hz to 20 kHz was always only the best possible outcome, not a generic range for everyone.

[guyman]

I think we can get a decent answer given the right experiment and a decent data sample. Perhaps a wide range of different sounds (melodies, rich overtone wave forms, sines, even pitch shift materials) and perhaps conduct it like an eye test, take the results, put a best fit line to it.

I know that someone's personal subjective bias, and hearing loss would affect this, but I am looking for a generalized result so that's ok. That said I agree, getting solid data without control over the monitor system it's heard through would be futile.

Unless of course we accept the differences due to speaker systems.

After all, I wish to take this into account for my mixes, songwriting, and tool development- and when I give my mixed and mastered music out - I have absolutely no control over what speakers they listen to my music on. So we can just group the ears the mind and the speakers together as "subjective bias"

Sounds like a fun experiment, I will research more.

[trogluddite]

At the low end, there is a complementary question that might have a bearing on things. If you play a series of pulses at a slowly increasing tempo, at what point do you stop perceiving them as individual pulses (a "rhythm") and begin hearing them as a tone (a "waveform")?

What I find interesting about that is that the pulses can consist of a wide spread of harmonically unrelated frequencies (a complete lack of pitch information) yet at some point they will merge into a tone which does have a perceivable pitch (complete with many overtones).

IIRC, the upper limit for perceiving individual pulses is usually around 15-20Hz - suggesting that there's a crossover point between "tempo perception" and "pitch perception".

As for the top end: I used to have extremely acute high frequency perception, which lasted well into my early forties (a lot of other autistic people I know report the same thing). I used to get driven nuts by supposedly "inaudible" ultrasonic cat-scarers and bug-repellants, or by sympathetically vibrating parts in old CRT TV scan circuits and switched-mode power supplies.

However, at those high frequences, I definitely had no perception of a distinct pitch. It didn't even really seem like hearing a sound - more like a sixth sense that made me uneasy whenever the stimulus was nearby. On several occasions, I proved this to people by getting them to run a blind test, because they refused to believe that I was being disturbed by something which they couldn't sense at all.

[guyman]

Yeah trog that's what I'm imagining here, cross over points with gentle slopes, not abrupt lines, where tempo stops, and again where high "noise" starts, and I'd expect them to vary, but i'd wager with a thorough enough test, with best fit lines implemented, would illustrate where these " ideal perceptual filter" points could be. I'm thinking this is what a lot of auto tune plugins implement in their pitch detection circuits. Though I'm certain those are more limited to the "typical" ranges of instruments / the voice.

Martin, how did you select your cutoff points for your various pitch detectors over the years? Trial and error?

[RJHollins]

There was an interesting U-tube vid that demonstrated the 'transition' from slow 'pulses' rising CPS into the audible Frequency range.

Saw it quite awhile back ...

[martinvicanek]

Martin, how did you select your cutoff points for your various pitch detectors over the years? Trial and error?

Depends on the goal or application. A guitar has a four octaves range from E2 to E6. My voice goes from E2 to G4. My bass is B0 to G5. The lowest note determines the latency of your pitch tracker. You need a bit more than one cycle to detect pitch. That’s no issue for a tuner, but for a guitar synth it is.

[Spogg]

I’ve not done any research into this, but I do know that my hearing system can tell me that very low notes can actually sound very out of tune, so much so that an ascending chromatic scale can actually sound in the wrong order, to me anyway.

In fact, I once thought I had a weird frequency fault and I had to actually check the pitches produced! I should mention this was using pure sine waves in an additive system. I only realised it was me when I added higher harmonics and then the notes sounded as expected.

So maybe we can all perceive low frequencies as musical pitches, but not necessarily accurately. That’s led me to think about pitch “resolution” at low frequencies. The difference between C1(note 24) and C#1 (note 25) is only about 1.9Hz. At higher pitches the Hz difference becomes much greater. For example, the difference between middle C and C# is over 15 Hz.

[R&R]

Wonder if we perceive <20Hz and subsonics through resonance in the skull/bones and such? As higher resonance... but the mind/perception filters it somehow...

And in a band context as Martin mentioned... Maybe also percieve very low "off notes" by instead noticing some sort of secondary standing waves that feels all wrong or are themselves off note "for the context" or interacts all wrong with both any playing notes and all of the instruments bodies present in the same room?

I've had a few shakers/tactile tranducers (or Exciters/Resonators as Monacor describes them) installed in both home cinema and hifi setup. They can really trick the mind in all sorts of ways...