Jump to content

Unraveling the Mysteries - The theory used in practice.

Rate this topic


stew503
 Share

Recommended Posts

Hi all,

Due to the Unraveling the Mysteries - Formants and Harmonics - Vowel Modification thread - Robert has asked for his voice to be laid bare and used to show some of the theory behind Harmonics (1*, 2*, 3* ... etc), Formants and how this lot relates to differing types of techniques.

I have used from Roberts available wav's, Boomy, Neutral (because I managed to catch an off guard piece - HOWEVER - I have asked Robert to re-record a neutral by asking a change of tongue position, once I have it - i'll re-work), Overlay and Husk. All these are available through his main page.

I have spend a few hours graphing, cutting / pasting two pictures that we can use to show the voice. The first uses the VVT software (used on trial).

As can be seen the harmonics of the Fundamental are pretty consistent (A4 - 447Hz for Robert)

So 447 - A4 - H1 1 times fundamental - is the fundamental. .. I have added Steven's theory so you see it applied.

891 - A5 (447*2 = 894 (so close enough)) - H2 2 times fundamental is the octave above the fundamental

1383 - F6 (447*3=1341 (again, close enough) - H3 3 times fundamental is the octave + perfect fifth above the fundamental (perfect 5th of A is E - so close enough)

1787 - A6 (447*4=1788 (pretty much bang on) - H4 4 times fundamental is the double octave above the fundamental

2295 -D7(ish) (447*5=2235 (close!! - Steven - can you explain this difference)) - H5 5 times fundamental is the double octave + major third above fundamental (Major 3rd of A is C# (in key of A))

2737 - F7 (ish) (447*6=2682 (Close!! - Again Steven explain) - H6 6 times fundamental is the double octave + perfect fifth above fundamental (and twice H3).

NOTE, the 'neutral' sound is "closer" at 5 times and 6 times.

Hopefully the math/theory is right - been doing this for a quite a while - and if it helps then great.

.

Steven - I'm not a noise expert - Can you please expand on Overlay and Husk.

....

As for the Harmonics (1st part of this post) - if you look at the other thread post 33 - it shows an excellent diagram of the above based on the theory. Nice to see that support the theory and application. (just step the A2 pointers to A4)

I've removed the formant side of things as I am researching where F1 should be, seems speech analyser may be telling porkies. I'll put back on once I have the right info.

Back in again ... As I had a niggle - due to one graphing I had vs. another, but seems like the figures match,

Coming to Formants - This was done using Speech Analyzer.

The F1's are consistent. Spot the differences in F2, 3 and 4's

So note Steven's point, "When the harmonic matches the peak perfectly, it is quite prominent. The resulting harmonic peak in the spectrum is a 'formant'."

So F1 - 892 - A5 (Very close to Harmonic at 891)

F2 - 1784 - A6 (Very close to Harmonic at 1787)

F3 - 3123 - G7 (although 2700 (Sharp of E7) at Neutral) - Throughout Neutral and Overlay close to harmonic at 2737 (slightly flat F7))

Steven - Can you expand on the differences F2, 3 and 4's throughout and what you feel the F2 difference between Boomy and neutral is (although, as have said I have asked Robert to re-record a neutral note)

More will no doubt follow, enjoy all.

Stewart

Link to comment
Share on other sites

The reason for the niggle was, I have a soprano A4 which I have graphed and F1 appeared to be in a totally different place than expected ... It was either that or the above F1's were wrong. i.e. is F1 sitting on H2.

Anyway - I did some more graphing and have this from Robert's vocal - (Robert Boomy).

Shows the peaks and where "possibly" the peaks should be (Steven can you please expand on that one, is that just the program plotting where they should be). And we can see we have F1 sitting on H2.

I hope Steven (or yourself Sep) with his/your experience can go into where F1's usually sit for male / female and say use an A4.

p.s. Thanks sep - no it wasn't harmonics, that I understand and the harmonic portion was to show the theory based H1, H2 (as shown on the other threasd) .. etc, vs. a voice. It was formant F1 and where it sits.

Link to comment
Share on other sites

What spooked me is below.

The program shows F1 at 879.3 (So sitting on / close to H2) ... HOWEVER, the little graph shown bottom right, shows 1st peak at H1 - 478Hz. (p.s. it's at a slightly different place in the vocal so hz's are slightly different).

It's that I don't understand ... Where in this instance is F1 sitting ? (and where it sits (in this example) for experienced female (Late) 20+ soprano) ?

Stewart.

I want to add to this post - rather than clog up below Sep, however this is in reponse to his comment. I found this on the VoceVista Pavarotti site though,

"In the female high voice the tracking of harmonics with formants is also of great importance, but there it is the first formant that follows the fundamental, H1. This may sound similar, but an important difference is that F1-H1 tracking is a natural phenomenon and relatively easy to implement. Tuning F2 to a higher harmonic is by contrast elusive"

Link to comment
Share on other sites

...Shows the peaks and where "possibly" the peaks should be (Steven can you please expand on that one, is that just the program plotting where they should be). And we can see we have F1 sitting on H2.

I hope Steven (or yourself Sep) with his/your experience can go into where F1's usually sit for male / female and say use an A4.

p.s. Thanks sep - no it wasn't harmonics, that I understand and the harmonic portion was to show the theory based H1, H2 (as shown on the other threasd) .. etc, vs. a voice. It was formant F1 and where it sits.

Stewart: I think the program is just looking at the higher-amplitude harmonics, and calculating where the formants are based on those amplitudes.

On the question about ...where F1's usually sit for male or Female for A4... that depends on the vowel. Singing different vowels is how we move the resonances around. Or, phrased the other way, moving the resonances around is how we get the different vowels.

For a particular voice, R1 will range over certain frequencies as the different vowels are produced. /i/ and /u/ (ee and oo) have the lowest R1, /a/ (ah) has the highest. All the other vowels have R1s in between.

We will be getting to that point soon in the discussion in the other thread. For example, I am going to do a series of spectrograms of my own voice on a few vowels, and show exactly where my R1 is for them... and you will be able to see the F1 which results when the harmonics align well. Anyone with spectrographic software will be able to do the same experiments.

I hope this was what you were looking for by way of response.

Link to comment
Share on other sites

  • 4 weeks later...

I am not sure I fully understand formants - I'm a bit thick when it comes to science...

So, when a person is singing a note with a strong formant, is that when we, as the listener, hear that operetic "ringing" in someones voice?

Here is why I ask.

I never hear my own voice ring. I have recorded myself singing over 100 songs, and never heard that "ring" in them. When my band plays a show, however, I CAN hear a ringing in my voice, but only through the FoH PA system. Last night , I went to "The Haunt", a national stop for national signed bands, for Karaoke night. The DJ uses the FoH PA, and his own monitors. I could hear my voice ringing away on certain vowels, especially "ee", sung above c4 (any note above c4) Is this ringing I hear a formant? And if so, why do I not hear it when I am singing without a PA system? Does it take amplification to hear the harmonics better? Or, am I way off on formants and just do not have a clue?

~Keith

--------------------------------------------------------------------------------

www.drop-head.com

Link to comment
Share on other sites

I am not sure I fully understand formants - I'm a bit thick when it comes to science...

So, when a person is singing a note with a strong formant, is that when we, as the listener, hear that operetic "ringing" in someones voice?

Here is why I ask.

I never hear my own voice ring. I have recorded myself singing over 100 songs, and never heard that "ring" in them. When my band plays a show, however, I CAN hear a ringing in my voice, but only through the FoH PA system. Last night , I went to "The Haunt", a national stop for national signed bands, for Karaoke night. The DJ uses the FoH PA, and his own monitors. I could hear my voice ringing away on certain vowels, especially "ee", sung above c4 (any note above c4) Is this ringing I hear a formant? And if so, why do I not hear it when I am singing without a PA system? Does it take amplification to hear the harmonics better? Or, am I way off on formants and just do not have a clue?

Hi, Keith. Depending on what the singer is doing, there can be 3 or more strong formants in the voice at the same time. This is particularly true for male singers. A voice with strong formants gives the impression of 'ringing', because some of the harmonics are greatly louder than others, and the ear picks up these disproportionately loud harmonics.

Its likely the reason you do not hear the ringing is that high frequencies are directional, often absorbed by the room, and your mouth is out in front of your eaars... the ringing tones go straight away from you (to the listeners) and don't reflect back as well as lower frequencies. However, when PA's are involved, you have a better chance to hear them.

If you want to hear your own voice ringing, there are things you can do to increase the amount of sound that gets to your ear while you sing. One of the simplest is to cup your hand loosly in front of your ear, and with your wrist out to the side of your mouth about 3 inches... that catches much more sound from the vicinity of your mouth and carries it up.

Its possible to learn how to hear the ring without assistance, too. If you are interested to do that, speak up, and we can start a new thread about learning to hear the ring of the voice.

I hope this is helpful.

Link to comment
Share on other sites

Thanks Steve. I think a new thread would be awesome! It really is awesome to your own voice ring like crazy on certain notes. Just hearing my own ring at a show filled me with so much enthusiasm. I'll let you start a new thread :)

Link to comment
Share on other sites

Steven / Keith,

Maybe this is one for the other Unraveling the Mysteries thread as it contains loads of useful information and is a valuable resource, where Keith et al(ia) can add to knowledge on formants. (I was thinking about bumping it). And is certainy one for a sticky.

Possible to add to that thread too is Pure Vowel vs. Mods.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...