Details of musical vibrato

I expect there are some dopers who are highly educated in this sort of thing. Rare indeed is the subject with which no doper is familiar!

When a singer or musician uses vibrato, i.e. variation in the pitch of the note he/she is either singing or playing, does the pitch vary above or below the fundamental note? Or, as I suspect, could it be either way?

I ask this because I am trying to sort out the design of an electronic instrument that can sense the pitch of a note being played/sung in real time (i.e. this is not a “tuning device,” but a performance or ear training aid). When we listen to a note with vibrato, our brains have no trouble figuring out what the note is supposed to be, even though its instantaneous pitch is varying all over the place. Computers are notoriously bad at that sort of thing, however. Hence, this will be quite a challenge. I would like to know how our brains “know” what the note would be sans vibrato. How the heck do we know whether the singer is flat when her instantaneous pitch actually is flat due to vibrato?

Any information on the precise mechanics of vibrato will be most helpful. For example, stuff like, what is the range of frequency or variation (probably ranges from about just under 1Hz to “Stevie Nicks”), what is the range of frequency shift (probably about a semitone, give or take a few cents, I would guess, but with pretty wide variation between individuals), that sort of thing.

And anyone who understands a bit about the psychoacoustics of it all–how our brains process this stuff and sort out whether a performer is on pitch–would be the most valuable of all to me.

Thanks much!

Violinist here…

Vibrato is certainly around the pitch of the note. Hopefully this means you’ll be able to calculate an average frequency across a set time period. I’ve no idea how our ear does manage to do this itself, that’s not my speciality :wink:

As for the details of vibrato itself…I guess a very wide vibrato would be reaching perhaps a quarter-tone either side of the main pitch. Mostly it’ll be narrower than that, though. The frequency is probably not below around 5Hz, and I shouldn’t think above around 12Hz. Vibrato on higher-pitched notes is generally faster, both for physical reasons (the hand doesn’t need to move as far to create the same pitch modulation), and because it just sounds better that way.

Don’t the other notes in the tune act as audio cues? Your brain knows what ke you are in and knows what to expect.

Please excuse me a moment whilst I imagine a gorilla playing a violin-- :eek: --and thanks for the reply!

Are you saying that when you play, say, an A-440 (I don’t have the charts for instrument ranges in front of me, and the particular note is irrelevant), that you vary the pitch both above and below the A?

If so, I presume that your (left?) finger’s excursion is roughly equal in both directions. And I’m quite surprised to hear that. Somehow I thought that vibrato had the fundamental as one of its “endpoints,” as it were. Now, instead, it seems that our brain somehow averages out the pitch, which departs from the fundamental both above and below, and which makes it even harder for me to understand why vibrato is aesthetically pleasing. After all, you’re playing the “right” note only a tiny fraction of the time (if you get what I mean), and you’re never truly returning to it–only passing through it on your way to the other extreme!

When I sing, I depart from the fundamental only in the negative direction, returning upward to my “indended” pitch with each cycle of vibrato. (I shoulda put scare quotes around the word “sing,” too, but that’s a whole 'nuther subject…)

This is becoming truly fascinating. I hope someone can shed some more light on this for me.

I’m sure you’re right, j_sum1, but I don’t understand anything of how that occurs. And, if so, I would think that there must be a distinct note for it to be a cue, and they’re usually all modulated by vibrato to some extent.

Maybe no one understands this, but I surely hope that’s not the case.

What? I must be misunderstanding you. You seem to be saying that every note you play/sing has some vibrato. I know when I sing, I only use vibrato if it comes naturally, which is usually on the higher, louder, longer notes. And most instruments can’t do vibrato.

If you mean that all notes are slightly off-pitch, that’s quite possible. I remember reading somewhere that we can’t hear pitch variation of about 5 cents or less. (A cent is a hundreth of a semitone, I think.) So you could be off by 3 cents and probably nobody would notice. Many people are even less sensitive to pitch variation; just ask your non-musical friends to tune a guitar and you’ll see that.

To approximate the vibrato of a string player: place one of your fingertips on a stationary surface. Now move your hand from side to side **without moving the finger tip. **You can see that your finger tip sort of rocks from left to right, with the original position in the center.

Unless you’re talking the most miniscule amounts, the pitch of the note is not supposed to change. Vibrato is simply an interrupted (albeit extremely rapidly) note, not one that is changing its pitch. To demonstrate, hold out a note, any note. And now do what you’re talking about; rapidly change notes. If you’re doing it right, you’ll sound like Tarzan. Not good.

So, vibrato isn’t changing pitch, it’s simply inserting partial, extremely quick (almost like a dot matrix printer) interruptions in the same tone.

What? All the winds & strings I’ve played over the year (many, but none really well) are capable of it. Even the old reed organ I rebuilt had a Vox Humana to 0produce a vibrato.

What hammerbach said. There’s only a very few number of orchestral instruments that can’t do vibrato (and I’m not talking non-pitched instruments here). The piano is the big one, and the xylophone, timpani, a slew of other percussion instruments can’t do it, but all the others can.

And I’d like to elaborate a little more on my previous post. Obviously the MHz of a note that is being vibrato’ed is changing a little… I just got the impression you thought it was changing so much that it was becoming a different note. I thought you were saying, for example, if you’re singing an “A” with vibrato, that you were actually singing an “A-flat” and a “B-flat”, the adjacent notes, as well. And no, that’s not true. If it helps, the distance between the notes on a 12-tone scale (i.e. the one everyone knows and uses in most of the world) is 20 MHz. So between an “A” and a “B-Flat”, for example, it’s 20 MHz. So you can draw from that the vibrato has to be changing the original note less than that, because if you change it that much you get the Tarzan effect I described earlier.

I took a “Science of Sound” course a couple years ago at college, and saved the textbook, so I’ll see what I can do…

What we’re talking about is termed “pitch discrimination.” The “just noticeable difference” is, as you might guess, the smallest difference between two sounds that the ear can perceive. For pitch, this is largely determined by frequency, sound level, duration, and suddenness of the frequency change. The jnd from 1,000 Hz to 4,000 Hz has been found be be about .5 percent of the pure tone’s frequency, or 1/12th of a semitone.

What that tells us is that vibrato HAS to vary by at least 1/12th of a semitone, or else the ear wouldn’t be able to pick up the difference.

No it isn’t. Vibrato is varying the frequency or amplitude of a held note. That’s all. How do you insert a partial interruption? It’s either interrupted or it isn’t.

This is false as well. The 12-tone scale is logarithmic. In a tempered scale, between A 440 and the nearest B flat there’s a difference of 26.16 Hz, but if you go up just one octave, the difference is 53.22 Hz.

What you seem to be describing here is tremolo, not vibrato.

Tremolo can be both the rapid repetition of a single pitch, or the rapid alternation of two pitches. For some reason they call the whammy bar on a guitar a “tremolo” bar, when in fact you can do all kinds of things with it, since it can bend the pitch both below and above the fingered note as rapidly or as slowly as you like.

As a choral singer with a strong vibrato, it seems to me that vocal vibrato is variations in intensity, not pitch.

Good point. The human voice produces both, and the amount of each varies from person to person. A child’s voice rarely has either. That’s what makes childrens’ (usually it’s boys’, for historical reasons rooted in the Catholic church) choirs so enjoyable to so many people; the tones are purer than adults can make. When an adult sings, producing neither tremolo nor vibrato, s/he is singing “from the throat”, and not “from the diaphragm.” I dunno if the change is a byproduct of puberty, and doubt it has ever been investigated. :slight_smile:

As one who has had considerable to do with vocal music, usually choirs, I can tell you that the amount of tremolo produced by a “correctly produced” singing voice goes up decade by decade. It’s why most female opera singers hafta retire sometime in their 50s, and the men sometime in their 60s; the voices of the elderly include increasing tremolo - more in singing than in speaking, but it’s no longer as “enjoyable” (certainly as defined by the standards of western musicology). Someone who is willing/able to … uh … add in a “throat” component can damp out both tremolo and vibrato, but opera singers take lessons throughout their singing careers. I suspect that the knack of “throat singing” gets lost somewhere along the way, 'cos their coaches certainly wouldn’t permit them to do it.

The amount of vibrato in an adult human voice varies from individual to individual. I have heard singers whose vibrato variation was more than a quarter tone :smack: (ya don’t want to stand next to them unless you can mentally “tune them out”). :rolleyes: Of course, such singers are rarely asked to do solo or ensemble work; it’s too noticeable. However, anyone who can carry a tune, and whose voice isn’t actively unpleasant can sing in a choir. When you hear a choir, you’re hearing the composite sound, and, when listening from the proper distance only a person with a very good ear can hear individual voices (or to a recorded performance, and then only if it hasn’t been electronically balanced).

This is wrong for a few reasons, but most obviously: 20 MHz is way too high. The range of human hearing is approximately 20 Hz to 20 KHz (20,000 Hz).

I hate it when physicists teach people music. :wink:

With a series of solitary exact pitchs, perhaps a sine wave generator, then yes, there’s some such limit to how close they can be before we fail to discriminate.

In a musical context, there’s, well, context. There’s the ear’s expectation from one moment to the next. There’s other things happening simultaneously. If I could play out of tune by a 12th of a semitone and nobody noticed, I’d be sorted :smiley:

There’s also a fluctuation of timbre - any alteration of the vocal cords or mouth during a single note will change the sound quality.

Sorry about the mis-type. I of course meant Hz, not MHz. And I wasn’t talking octaves… I was just talking about the confines of one octave to give him some idea of how far apart each note was.

Re vibrato: according to the Manual of Musical Terms from Schirmer, vibrato is "a tremulous effect caused by very rapid partial interruptions of the tone", so I dunno.

My grand point in the whole scheme of the post was to illustrate that vibrato is not changing a note to the point that it is becoming a new note, as the OP suggested. Rather, it is just dancing around a particular note. The OP asked “I would like to know how our brains “know” what the note would be sans vibrato”, and answer is because it is still that note, just tiny variations thereof.

Yes, but my point is that you can’t give an idea of how far apart the notes are since they’re on a logarithmic scale. You can say “Between A 440 and the nearest B flat is a difference of 26.6 Hz,” but you can’t say “the distance between two notes in a 12-tone scale is 20 Hz,” because it’s not. The difference between the C two octaves below middle C and its corresponding C[sup]#[/sup] is less than 4 Hz, but the difference between middle C and its C[sup]#[/sup] is almost 16 Hz (close to your guess of 20). The difference between C two octaves above middle and its C[sup]#[/sup] is over 50 Hz.

Even within one octave the range of values between two notes varies tremendously. For example, in the octave starting on middle C and going up, the average distance between the notes is 21.11 Hz, ranging from 15.55 Hz to 27.72 Hz. If you go up just two octaves, the average distance between the notes is 84.45 Hz, ranging from 62.2 Hz to 110.8 Hz.

You simply cannot say “between any two notes, the distance is about X” because it’s just not true. You can’t give anyone some idea of how far apart each note is unless you know which two notes they’re talking about. Or at least which octave.