Can you hear formants? Should you?

[TimR]

Thanks, that helps.

[MasterBlaster]

Sorry Owen but you are confused,

To answer your last question, formants are basically inaudible until they have a harmonic lined up near them. So if the harmonic is not tuned to the formant, the formant isn't heard.

False. In order to hear a vowel you must have at least 3 of the vowel format freqencies present in the sound otherwise it would not sound a like a vowel.

EG. "Bat" must have f1 660hz, f2 1700 hz and f3 2400 hz present in the sound at any fundament pitch to make it sound like "bat".

The voice works just like any other instrument except the vocal track can be adjusted (tongue,lips, larynx) to produce various formant regions. Ah has 3 specific formant region and OO (440, 1000, 2250 hz) has 3 specific formant regions. Ah sounds like ah because the when you position your tongue/lips it allows the ah formants to pass through to give you an ah sound. Same with oo, you have to put your mouth in a completely different position to allow the oo frequencies to resonant and pass through.

Most other instruments have fixed formant regions giving the instrument it's unique sound. There is no Ah or OO sound on the violin for a given pitch because the formants regions are fixed. The violin has 3 to 5 main fixed resonant formant frequencies that are present in every violin note otherwise it wouldn't sound like a violin. A violin sound likes a violin at any pitch in it's range.

It's important to know FORMANTS ARE NOT NECESSARILY MULTIPLES OF THE FUNDAMENTAL HARMONIC! If they were then every instrument would sound like a sign wave which sounds so plain because there are no special format regions just multiples of the harmonic. Sometimes the formants happens to fall close to a harmonic multiple but not always. If a formant happens to be a multiple of a harmonic then yes the formant will get a boost.

FROM STANDFORD.EDU

Formant Analysis

Most acoustic instruments produce prominent formant frequencies. Formants are resonances that are characteristic of a sound. Phonemes can be characterized by 3 prominent formants or frequency regions. A particular set of formant frequencies characterize each vowel and are relatively independent of a voices pitch. Female and male voices obviously have different formant frequency ranges, however the ratio between formant frequencies is consistent across males, females, adults, and children.

Some formant plots for various instruments here:

https://ccrma.stanford.edu/~jmccarty/formant.htm

So say you're singing a A3 at 220hz and the vowel is a kind of schwa "uh" with a F1 of around 500 hz and F2 of around 1500 hz. 500 hz will not be heard! Neither will 1500 hz! Because those frequencies do not coincide with the overtone series of a fundamental of 220hz.

False. They will be heard. That's why a "uh" sounds like an uh. How else do you think an uh hits your ears. It's because it has it's own unique formants that create the uh sound. And you hear these formants because the vocal track works like a specialized frequency filter allowing only the "uh" formant frequencies to resonant and pass through to your ears.

[Felipe Carvalho]

There is some confusion in this matter because some authors will define the resonance as the formant itself. While others will use the term as Martin mentioned.

But, if you hear a vowel, thats how the formants sound. And its always the interaction of resonance and source harmonics.

The idea of "formant tuning" is about using a vowel in such way that F1 or F2 amplify H2, H3 or H4.

Or, in the more correct way of saying it. F1 or F2 are the result of the interaction of H2, 3 or 4 with R1 or R2.

You can simplify it and read "going loud" :/.

[TimR]

Okay, I'm confused again.

Will think about it and post again later.

[VideoHere]

way, way, way, over my head.

i'd rather just practice and feel my way to the overtones.

lol!!!!

[Owen Korzec]

Sorry Owen but you are confused,

To answer your last question, formants are basically inaudible until they have a harmonic lined up near them. So if the harmonic is not tuned to the formant, the formant isn't heard.

False. In order to hear a vowel you must have at least 3 of the vowel format freqencies present in the sound otherwise it would not sound a like a vowel.

EG. "Bat" must have f1 660hz, f2 1700 hz and f3 2400 hz present in the sound at any fundament pitch to make it sound like "bat".

The voice works just like any other instrument except the vocal track can be adjusted (tongue,lips, larynx) to produce various formant regions. Ah has 3 specific formant region and OO (440, 1000, 2250 hz) has 3 specific formant regions. Ah sounds like ah because the when you position your tongue/lips it allows the ah formants to pass through to give you an ah sound. Same with oo, you have to put your mouth in a completely different position to allow the oo frequencies to resonant and pass through.

Most other instruments have fixed formant regions giving the instrument it's unique sound. There is no Ah or OO sound on the violin for a given pitch because the formants regions are fixed. The violin has 3 to 5 main fixed resonant formant frequencies that are present in every violin note otherwise it wouldn't sound like a violin. A violin sound likes a violin at any pitch in it's range.

It's important to know FORMANTS ARE NOT NECESSARILY MULTIPLES OF THE FUNDAMENTAL HARMONIC! If they were then every instrument would sound like a sign wave which sounds so plain because there are no special format regions just multiples of the harmonic. Sometimes the formants happens to fall close to a harmonic multiple but not always. If a formant happens to be a multiple of a harmonic then yes the formant will get a boost.

FROM STANDFORD.EDU

Formant Analysis

Most acoustic instruments produce prominent formant frequencies. Formants are resonances that are characteristic of a sound. Phonemes can be characterized by 3 prominent formants or frequency regions. A particular set of formant frequencies characterize each vowel and are relatively independent of a voices pitch. Female and male voices obviously have different formant frequency ranges, however the ratio between formant frequencies is consistent across males, females, adults, and children.

Some formant plots for various instruments here:

https://ccrma.stanford.edu/~jmccarty/formant.htm

So say you're singing a A3 at 220hz and the vowel is a kind of schwa "uh" with a F1 of around 500 hz and F2 of around 1500 hz. 500 hz will not be heard! Neither will 1500 hz! Because those frequencies do not coincide with the overtone series of a fundamental of 220hz.

False. They will be heard. That's why a "uh" sounds like an uh. How else do you think an uh hits your ears. It's because it has it's own unique formants that create the uh sound. And you hear these formants because the vocal track works like a specialized frequency filter allowing only the "uh" formant frequencies to resonant and pass through to your ears.

Have you ever looked at a voice on a spectrograph before?

Unless the singer is applying some kind of distortion or airiness to emphasize the noise of the voice (for instance Steven Tyler did this on one of his screams that Steven Fraser from this forum analyzed a couple years ago), you will never see the formant resonance itself show up anywhere near as prominent as the multiples of the fundamental.

Try it, sing a clean note (no airiness no distortion etc.) into a spectograph and try to find any spikes within the vowel region that aren't even order harmonics. It will be very difficult if not impossible.

Don't get me wrong, they WILL be SLIGHTLY heard and SLIGHTLY show up - they just don't come out as strongly as the harmonics of the fundamental, so it's pretty much negligible and what really matters is how these formants affect the intensity of the multiples of the fundamental. On a violin or guitar or something sure maybe these formants will come out very strong and almost as prominent as the harmonic multiples. But it's just not so in the voice.

What formants do MORE is affect the intensity of the harmonics around it. And that is how an "uh" sounds like an "uh" - the placement of formants affects the relative intensity of the harmonics and that's how we hear it - and as a result, in our head we can kind of hear the formant position (even if we can't hear the formant resonance itself) as a specific tone - it always emphasizes the same range of frequencies regardless of pitch, that's how we recognize it, NOT because it's producing a complete second set of harmonics!

This also explains why the higher you sing the harder it is to phonate pure vowels, a well known fact in vocal pedagogy. At a low pitch, the vowel formants affect the upper harmonics that are close together, so no matter how change the vowel, the formant is probably going to affect several harmonics, strongly highlighting a specific tonal quality that we can identify as a vowel. But at very high pitches, the formants start to line up with the lower harmonics which are further apart and then there are less harmonics to emphasize the formant position, so we don't hear the vowel color as clearly.

One more thing. If we could hear the formants themselves, that would enable us to create chord-like qualities in the voice which is not possible except throat singing which RELIES ON THE MULTIPLE HARMONICS to get the extra tones.

For instance in that uh example if I actually heard the 500 and 1500 hz it would sound like there are two resonances in my voice that are slightly out of tune with the note I'm singing. That just doesn't happen.

I figure you would understand this if you have actually seen a voice in a spectrograph before. The spikes are almost always multiples of the fundamental and this is very obvious visually. But if you haven't then that is probably the source of your confusion on this point. Other than that you do have a good understanding of formants and I agree with pretty much all of the rest of what you wrote.

[Owen Korzec]

We are starting to get real nitpicky by the way, none of this knowledge will exactly make you a better singer...it's just as effective to just consult with a vocal coach with a great ear to learn the right vowel shades to use at different moments of singing and get familiar with the tonal quality by ear, without ever having to go into the science, numbers, visual, etc.

[ronws]

I will play around with resonance until I find a bright spot or sweet spot where my head rings like a bell.

[Felipe Carvalho]

Don't get me wrong, they WILL be SLIGHTLY heard and SLIGHTLY show up - they just don't come out as strongly as the harmonics of the fundamental, so it's pretty much negligible and what really matters is how these formants affect the intensity of the multiples of the fundamental.

You have the source, H1, H2, H3 ... Hn

You have the resonators, R1, R2, R3

From the interaction of Rx with the source, one or more of the harmonics will be affected, when they are, you will have peaks of intensity on the spectra. These peaks are the formants.

Since you never take your head of your neck, they are always present during phonation.

When the fundamental is above the frequency that a resonator is affecting, then this formant is "lost", and as a result you start to lose vowel information (check sopranos high range).

Some authors call the resonator itself a formant, thats the source of the confusion present in this discussion. If you do so, then its correct to say that you will not see/hear the resonator frequency, but the harmonic that is being affected. However its more precise to say that the affected harmonic itself is the formant and to keep in mind that the resonance center is most likely NOT that peak of intensity.

All in all, this can be resumed to: vowels.

[TimR]

Ah, so there are two uses of the word formant.

One is a system resonant frequency, which is adustable by changing the vowel which changes the shape of the space. This is not an actual sound, it is a frequency that will amplify a sound if one happens to be input.

The other use is the actual sound that is amplified and heard. This sound is an overtone of the fundamental. It may not be at the precise frequency of the resonance, but if it is close enough it can be amplifed and heard. If an overtone aligns closely with a resonance this sound may be prominent.

The second use is what I thought the term meant. But I didn't understand that the source is an overtone from the fundamental.

Am I getting it now?

[Owen Korzec]

Don't get me wrong, they WILL be SLIGHTLY heard and SLIGHTLY show up - they just don't come out as strongly as the harmonics of the fundamental, so it's pretty much negligible and what really matters is how these formants affect the intensity of the multiples of the fundamental.

You have the source, H1, H2, H3 ... Hn

You have the resonators, R1, R2, R3

From the interaction of Rx with the source, one or more of the harmonics will be affected, when they are, you will have peaks of intensity on the spectra. These peaks are the formants.

Since you never take your head of your neck, they are always present during phonation.

When the fundamental is above the frequency that a resonator is affecting, then this formant is "lost", and as a result you start to lose vowel information (check sopranos high range).

Some authors call the resonator itself a formant, thats the source of the confusion present in this discussion. If you do so, then its correct to say that you will not see/hear the resonator frequency, but the harmonic that is being affected. However its more precise to say that the affected harmonic itself is the formant and to keep in mind that the resonance center is most likely NOT that peak of intensity.

All in all, this can be resumed to: vowels.

Awesome explanation, rep point

[Owen Korzec]

Ah, so there are two uses of the word formant.

One is a system resonant frequency, which is adustable by changing the vowel which changes the shape of the space. This is not an actual sound, it is a frequency that will amplify a sound if one happens to be input.

The other use is the actual sound that is amplified and heard. This sound is an overtone of the fundamental. It may not be at the precise frequency of the resonance, but if it is close enough it can be amplifed and heard. If an overtone aligns closely with a resonance this sound may be prominent.

The second use is what I thought the term meant. But I didn't understand that the source is an overtone from the fundamental.

Am I getting it now?

I prefer the first definition personally, at least if we're talking scientifically. If we're talking vocal pedagogy, sometimes teachers will refer to the auditory resonance itself as a formant, especially with singer's formant. But since that kind of disobeys the original scientific definition it can get confusing IMO

In reality that auditory resonance we hear is a harmonic whose intensity is increased by being lined up around a formant resonance. Formants only operate as a sympathetic resonance.

And what sympathetic resonance is in this case is, when the formant approaches a harmonic (or vice versa) that has a similar resonance to it, it becomes active because of the commonality, the synergy - they can now combine and multiply their intensities to create a louder sound. Whereas if the formant and harmonic are not similar frequencies, they can't double up and combine to create a louder resonance.

[Martin H]

@Owen

You still seem to use the term formant in different ways. For instance you don't align a formant with a harmonic. You align a resonance with a harmonic and THAT will create a formant. :)

[TimR]

Still some disagreement, I see

A formant is a resonance, OR a formant is a harmonic that is amplified by being aligned with a resonance.

[Martin H]

@TimR

Yes there are a few different definitions. However, the last one is the most precise in my opinion. Because you can have a resonance without a formant, but you can't have a formant without a resonance. So a resonance is not the same as a formant. Also, it's a bit nonsense to say, that "a formant is a harmonic amplified by a formant". ;)

[Danielformica]

what came first the chicken or the formant?

[ronws]

"Living easy,

Loving free.

Season ticket on a one-way ride ..."

[Owen Korzec]

@Owen

You still seem to use the term formant in different ways. For instance you don't align a formant with a harmonic. You align a resonance with a harmonic and THAT will create a formant.

aaaah okay. Got it. I understand that definition now.

[TimR]

While this may seem confusing, I've learned a great deal from the discussion and want to thank everyone who contributed

When I joined this forum I asked what were the fundamentals I needed to work on first

The suggestion was breath support, then vowels, then head voice, then mixing

This thread seems to confirm the importance of vowels in the early stages

[gno]

Seems like the topic is closed - but here is a nice graphic about formants from this web page:

http://hyperphysics.phy-astr.gsu.edu/hbase/music/vowel.html

Notice how it says that formants remain the same regardless of pitch. However, the formants do change a bit throughout our range, like in the passagio where we modify the vowels formants.

[Steven Fraser]

A nice, detailed article by Dr. Brad Story on this topic, soon to be published.

http://sal.shs.arizona.edu/~bstory/bstory_handbook_singing_R1_preprint.pdf

[ronws]

Thanks for sharing that, Steve. We should start a petition to have your name changed to Steven Awesome Fraser.

For it speaks to the real act of filtering, which happens regardles of what one does. As certain freqs are enhanced, others are filtered out or attenuated by lack of the proper resonating shape, or seem to have been filtered out because they are not in prominence like the F0 and maybe H1 and/or H2 are.

I think it is interesting to note and please, please, correct me if I am wrong. The "general" formant, as it were, centering around 2.5 kHz that seems to be present in all male voices. That doesn't mean each voice is creating a fundamental there but some of the higher sounds or at least brightness of voice is due to this. A basso can be doing a bass note but some of the clarity comes from a partial that is in the 2.5 kHz area, right? Yes, I am way out of my league and may have misread.

And a tenor may be closer to that, making him especially bright and full in some notes that would have an F0 near there but it is not a barrier to "lower-centered voices," as I like to call them, from making notes there, which has everyday proof, with any number of "baritones" here and elsewhere, singing notes in the upper 4th and 5th octave.

And that the total sound one is making is not a result of merely adduction at the folds which, by itself, is a buzz, but is the product, both figuratively, and maybe literally, as hinted by the formula for pressure, glottal freq times the resonant frequency in the "right" shape of vocal tract.

I also notice that the farther back the articulation or narrowing, the closer is f1 and f2. That when the vowel is articulated at the lips or end of vocal tract, the f1 and f2 move farther apart. So, rather than open throat, let's call it open mouth. And it's not the cavernous shape of the mouth that brings the volume but the close alignment of f1 and f2 by articulating or vowel shaping at the back. Which is primarily shape and height of the hump of the tongue.

I could be wrong and I may come up with more misconceptions on this, never fear. :lol: