Buzzing sound in voice

[juliansader]

Hi everyone,

I often hear a kind of 'buzzing' sound in singers' voices, whether the singer is male or female, amateur or professional.

To illustrate what I am talking about, I have uploaded a short example of the buzzing sound:

http://www.speedyshare.com/files/23323206/BuzzingExample-Male.wav

In this example, the voice starts out without buzzing, and then after a few second the buzzing enters.

I am very curious to know, what is this buzzing sound called, how is it produced, and (perhaps most importantly) how does a singer avoid this sound?

Any advice would be much appreciated!

[ronws]

I approach it from the opposite direction that you did. I have a nasty habit of reading most anything I can get my hands on. I have no advanced degree but I learn a little bit of each thing. So, here I go. Zoologically, man is a primate, specifically a great ape, because of our overall height. Relatively hairless and bipedal. We don't grow a thick pelt and we walk on our hind legs and call our front legs arms. Primates essentially make the same kind of sounds. Buzzes, grunts, and whistles.

I think the buzz is always there. What makes it go away is how a note is resonated and the overtones that receive dominance and cancel out or "cover" the sound of the buzz. I realize that I have oversimplified things greatly but I am good at that. I am a simple guy.

[Mr Bounce]

I don't know too much about it, but it sounds like what Brett Manning calls the "10k 'sheen.'" RnB and Soul singers sing with this timbre--which is compressed and also breathy--all the time.

I think it sounds fantastic. In the recording it sounds a little exaggerated, or maybe the recording apparatus is set up in such a way that accentuates these frequencies.

I could only suggest focusing on more compression/'cry' and less breathiness in your tone, but I'm sure expert opinion will soon set us straight :P

[VideoHere]

I don't know too much about it, but it sounds like what Brett Manning calls the "10k 'sheen.'" RnB and Soul singers sing with this timbre--which is compressed and also breathy--all the time.

I think it sounds fantastic. In the recording it sounds a little exaggerated, or maybe the recording apparatus is set up in such a way that accentuates these frequencies.

I could only suggest focusing on more compression/'cry' and less breathiness in your tone, but I'm sure expert opinion will soon set us straight

it sounds like a type of resonance not nescessarily undesirable.

[Steven Fraser]

Hi,

I'll download this and do a spectragraphic analysis at the end of the week. On vacation now :-)

[Steven Fraser]

Hi everyone,

I often hear a kind of 'buzzing' sound in singers' voices, whether the singer is male or female, amateur or professional.

To illustrate what I am talking about, I have uploaded a short example of the buzzing sound:

http://www.speedyshare.com/files/23323206/BuzzingExample-Male.wav

In this example, the voice starts out without buzzing, and then after a few second the buzzing enters.

I am very curious to know, what is this buzzing sound called, how is it produced, and (perhaps most importantly) how does a singer avoid this sound?

Any advice would be much appreciated!

Juliansader: I got to this earlier than I thought I would.

I put the recording through spectragraphic analysis, and here is the frequency distribution toward the end of the recording, showing the cause of the 'buzz'. I will explain how to interpret this image after it.

Explanation: The 'buzz' you hear is from the pronounced energy peak which centers on the 2nd red vertical line at 3200 Hz. It is commonly called "singer's formant' in classical singers, 'ping' or 'focus', It is produced by narrowing the epilaryngeal space, right above the vocal bands, forming a high-frequency resonator there. This resonance aligns very closely with the sensitivity curve of the ear, so its presence in the sound makes the voice more audible, more easily found on a 3-dimensional soundspace, like a stage, boosts the overall volume of the voice making it 'project' more, and (as a convenient byproduct) more easily sustained.

FYI, the buzz is there for the entire length of the note, but is not as prominent at the beginning as it is at the end.

The reason you hear this sound in so many good amateur and professional recordings is that it is the result of good, efficient, sustainable voice production. Many singers, including hard rock, want this sound in their vocal quality, though they would also likely want twangier vowels which would obscure the presence of this formant.

What kind of a sound are you looking for, that this sound does not jibe with?

[juliansader]

Hi everyone, thanks for the responses!

Steven, your spectrographic analysis inspired me to do a few analyses myself, and I got the following interesting results:

Below is a spectrograph of the voice after the buzzing sound started:

I noticed that the peaks between 6KHz and 9KHz only appear when the buzzing starts, and their height correlates with the intensity of the buzzing. (There were also a few soft peaks at about 15KHz that are not visible in this screenshot.) On the other hand, as you mentioned, the peaks at the singer's formant are present for the entire length of the note.

I therefore used an equalizer to suppress much of the frequencies between 6KHz and 20Khz, and voilá... the buzzing disappeared! (Except for a very slight remaining buzz that could probably be removed by even heavier equalization.) I have uploaded the new, equalized version of the example to:

http://www.speedyshare.com/files/23372618/BuzzingExample-Male-Without-6KHz-20KHz.wav

It therefore seems to me that the buzzing sound is not related to the singer's formant, but I have no idea what else might be the cause.

To the best of my knowledge, the sound of the vocal cords on their own is also a somewhat 'buzzy' sound, which is then smoothened by the resonance in the rest of the vocal tract. (Almost like the harsh sound of guitar pre-amp that is smoothened by the cabinet.) So, perhaps the buzzing sound that I hear is vocal cord buzz that somehow escapes the smoothening effect of the rest of the vocal tract?

Or perhaps it is just a bit of mucus vibrating in the throat?

What kind of a sound are you looking for, that this sound does not jibe with?

I am not actually looking for only one specific kind of sound, since I prefer the singers that I work with (I am not a singer myself) to adapt their sound to the requirements of the piece that they are singing, whether it is a soft, dreamy voice for an intimate love song or a loud, bombastic voice for the drunken abbot in Carmina Burana. However, this buzzing sound is one sound that I personally find extremely jarring and that I have never found any use for. I would therefore love to know how to avoid it (besides heavily equalizing the recording)...

[Steven Fraser]

I am not actually looking for only one specific kind of sound, since I prefer the singers that I work with (I am not a singer myself) to adapt their sound to the requirements of the piece that they are singing, whether it is a soft, dreamy voice for an intimate love song or a loud, bombastic voice for the drunken abbot in Carmina Burana. However, this buzzing sound is one sound that I personally find extremely jarring and that I have never found any use for. I would therefore love to know how to avoid it (besides heavily equalizing the recording)...

Juliansader: The buzzing you are identifying is very soft, by your chart, more than 20dB softer than the singer's formant.

It would be interesting to watch how its amplitude changes during the course of the note. Some writers on the topic of resonance have wondered if there is a '2nd' singer's formant. I'll have to go read up on that.

Also, I think it would be useful to correllate those frequencies with the other acoustic content of the voice, to see how the frequencies align. For that purpose, I will extend my own spectrum out to include up to 10K, using a linear scale (you used a logarithmic one) to see if the amplitudes up there rise simultaneously.

Good stuff.

[Steven Fraser]

juliansader: Here is an article from the Journal of Singing, 2003, in which Professor Ingo Titze analyzes this resonance, and provides an explanation for its presence.

http://www.ncvs.org/e-learning/nats-pdfs/2003_03_04.pdf

The net of this: its the secondary resonance of the epilarygeal space. The singer's formant is the primary. He also mentions that it will be 20dB softer than the singer's formant. So, in the particular recording you provided, the reason it is not prominent at the beginning is the same as the lack of prominence of the singer's formant. Toward the end of the note, both of them get quite a bit louder.

This attribution means that the buzzing cannot be avoided when singing with singer's formant. However, it will be highly directional, so will not be picked up well on a microphone that is off-axis. If the sound-board attenuates this area, (as you say) it will disappear almost entirely.

[juliansader]

Wow, this is extremely interesting and helpful - I have never before heard of the second singer's formant!

This probably also explains why I don't hear the buzzing sound so much in Pavarotti's voice. According to one of your articles, Pavarotti used (at times, I am not sure how often) H3 formant tuning without combining it with a strong first singer's formant.

I wonder why the second singer's formant sounds more harsh and buzzy (to my ears, at least) than the first. A quick Google search didn't show much literature on the second singer's formant, but I think I should study some recordings of tenors to see if it is possible to produce a second singer's formant that is strong as well as smooth.

[Steven Fraser]

Wow, this is extremely interesting and helpful - I have never before heard of the second singer's formant!

This probably also explains why I don't hear the buzzing sound so much in Pavarotti's voice. According to one of your articles, Pavarotti used (at times, I am not sure how often) H3 formant tuning without combining it with a strong first singer's formant.

I wonder why the second singer's formant sounds more harsh and buzzy (to my ears, at least) than the first. A quick Google search didn't show much literature on the second singer's formant, but I think I should study some recordings of tenors to see if it is possible to produce a second singer's formant that is strong as well as smooth.

juliansader: The 2nd one is much softer than the first, and the ear is much less sensitive in that frequency area. Unless a room was absolutely silent, I doubt that most folks would hear it at all.

The first singer's formant, on the other hand, is not only quite intense, but in the really sensitive part of hearing. That makes it not only loud sound, but sound loud :-) As Titze mentions in passing about the 1st, the presence of the epilaryngeal resonance causes formants 3, 4 and 5, which are ordinarily spread a bit from teach other, to cluster or bunch together, so that they strengthen each other.

As to the strength of 2nd singer's formant, remember that the amplitude of harmonics decreases the farther you go up the series. Also, note that some sound engineers 'squash' high frequencies when recording, mixing or mastering, especially for TV broadcasts.

On Pavarotti... sometimes his singer's formant is present in recordings and sometimes it is not. I am thinking these days that this is a good evidence that engineers were struggling on how to best record his voice, which I allude to as 'artifacts' in some of my other articles.

For trying out most recordings, I recommend avoiding TV (or youtube) videos... stay with .WAV files directly taken from CDs to get the best sound. Even then, the engineers can have already tweaked things.

Glad this all helped!