2

I'm trying to find a way to measure the speed of a song from the perspective of singing/rapping. So far all I can find is BPM (beats per minute), however there are at least 2 issues with that:

  • The BPM number depends on which beats are counted, quite often a website will report twice the BPM that I measure by myself.
  • The BPM is not always related to the vocal speed. For example, "Thoughts Of A Dying Atheist" by Muse has 179 BPM but the singing part is quite slow, while "Break Ya Neck" by Busta Rhymes has only 83 BPM, but a... neck-breaking rapping speed.

Is there a more accurate measurement for the vocal speed? I'm thinking something along the lines of (continuous) syllables per minute, excluding the instrumental parts.

  • 1
    part 1) perhaps you are miscounting beats. part 2) vocal speed does not have bpm – Doktor Mayhem Mar 26 at 16:29
  • 2
    @DoktorMayhem 1) perhaps I am, perhaps the person who published the number on the website is; 2) that's my point, how to measure it? – aditsu Mar 26 at 16:31
  • By the way, sometimes different websites show different numbers for the same song (usually one is double the other, but not always) – aditsu Mar 26 at 16:32
  • Syllables/second? Words/min? Notes/second? – user45266 Mar 26 at 17:13
  • There's probably a similar term for generic music note rates. Notes/second, perhaps? Liszt's Transcendental Etude No. 6 in G Minor ("Vision") has a slow tempo (bpm) but loads of ornamentation and therefore a deceptively high note rate.... – Dekkadeci Mar 27 at 5:25
7

There is the concept (not specific to music) of Speech Tempo. As you will see from the article, there is some degree of discussion about how this should be measured - for example, words per minute, syllables per second, or sounds per second. One could imagine even deeper levels of granularity - such as the inclusion of changes in pitch or timbre as sound 'features'.

Within rap music, it's common to measure rap performances in terms of syllables per second - such performances have been noted in the Guinness book of Records.

  • 1
    Interesting, it's hard to decide how to split things up, and the wide variety of singing styles makes that even trickier. I'm thinking about the ending of "Don't Cry" by Guns N' Roses for example, first the "Don't you ever cry-y-yy-i-i" then the famous endless "toniiiiiii...." – aditsu Mar 26 at 23:07
  • 1
    I think syllables per second (or minute) can be a good reference though, for most songs. I wish it was more common to measure that too, not just the BPM. – aditsu Mar 26 at 23:14
  • 1
    @DavidBowling I think we have very different scenarios in mind. You're probably thinking of a situation where you have the full sheet music for a song, and a metronome, and you're just figuring out how fast you need to sing. I'm thinking about a case where you are asked to pick any song that you'd like to sing, but it has to be reasonably fast. First of all, how fast is "fast"? There needs to be some kind of measurement. You pick a song that is reported to have over 140 bpm, but it gets rejected because the vocal part is too slow. Now what? – aditsu Mar 26 at 23:56
  • 1
    @DavidBowling BPM is useful both as a direction ("how fast do I play this?") and a metric ("I want something at such-and-such a tempo"). Something like Speech Tempo isn't so useful as a direction, but, in a world where it was measured, could nevertheless be useful as a metric to find pieces that fit a certain profile (without having to listen to them!) – topo Reinstate Monica Mar 27 at 0:27
  • 1
    @DavidBowling yes, BPM + "predominant subdivision" would be along the right lines. My initial thought is that I don't hear "predominant subdivision" mentioned as an analytical metric anywhere near as much as time signature or BPM / tempo indication, which may be why the OP is on a quest... – topo Reinstate Monica Mar 27 at 0:51
3

I humbly propose "maximum notes/second": Take the fastest part of the song and calculate the number of notes in that amount of time. If you have a passage that is super-fast in one part but really slow in another, it's way more important to know that you need to be able to rap/sing/play at that fastest part's speed to play the whole song. Compare the chorus of Eminem's "Rap God" to the fast part. Obviously, to tackle that song, one needs to be able to rap as fast as the fastest part, not just the average of the whole song.

Clearly, one would have to experiment with certain parameters, such as minimum measured time, and even definition of a note (slides? bends?) in order to get useful results.

Disclaimer: No one uses this, because I made it up. I think it's way more useful than average notes/second.

  • That's a good point, and could make it easier to measure. I think the average is also relevant, and the fastest part should probably have a minimum length (e.g. at least 5 seconds). – aditsu Mar 26 at 22:39
  • @aditsu Right, otherwise one-note passages approach infinite speed... – user45266 Mar 27 at 3:56
  • I wasn't thinking about single notes, but rather quick "bursts" that may last 1-2 seconds and are not representative of the whole song. – aditsu Mar 29 at 7:49
  • @aditsu but that's my point. To play the song, you need to be as fast as the bursts! – user45266 Mar 29 at 14:51
  • Well, if the fast part in "Rap God" was only 1 second long, I think almost anybody could do it. What makes it really hard is that it keeps going for about 15 sec. – aditsu Mar 29 at 16:35
1

I don't think there's a useful single value that you can use to summarise a song like this. Instead, it's probably most useful to consider a combination of the BPM, the length of rhythmic subdivisions, and distribution of rhythms. For example, Break Ya Neck has a BPM of 166*, and semiquaver/quaver (16th note/8th note) rhythms sustained over the period of several bars. In comparison Thoughts Of A Dying Atheist is 179 BPM but the rhythms are mainly a mixture of crotchets (quarter notes) and longer notes, with no consecutive passages of 8th notes.

*I know you said 83 in your question but I think 166 is more accurate. When/why BPM values can vary by a factor of two would be a good topic for another question!

  • Funny, I actually looked it up, and multiple sources said 83, and my own count also confirmed that. But I also see how it can be counted as 166. Anyway, I hope you get my main point: some songs have lower BPM but faster lyrics/vocals. – aditsu Mar 26 at 23:25
  • I'm not sure if "a combination of the BPM, the length of rhythmic subdivisions, and distribution of rhythms" is really practical. It may be accurate, but difficult to interpret as a meaningful measure of speed (for many people). E.g. a 5/16 note at 80bpm is longer than a 3/8 note at 120bpm, but it's hard to tell without a calculator. – aditsu Mar 26 at 23:40
  • @aditsu Maybe, but if you try something simpler I don't think it will capture what you want. However, also note that 5/16th notes aren't really a thing. Usually you would talk about 1/4, 1/8, 1/16 notes and their triplet variations in which case it is quite easy to compare songs since you can double BPM and note length to keep the actual rhythm the same. – rlms Mar 27 at 17:58
  • I think something simpler could work better and would be easier to use, but your suggestion is also valid. 5/16th is a quarter note tied to a 16th note, as shown for example at en.wikipedia.org/wiki/Tie_(music) , but I agree that it would seldom come up. Some songs are a bit harder to analyze though (without the sheet music), e.g. "Hoy" by Gloria Estefan. – aditsu Mar 29 at 7:22
  • I think that song is about 195 BPM with a mixture of 1/4 and 1/8 notes in the vocal line. – rlms Mar 29 at 11:34
0

if you look at it from a DJ perspective it will always be BPM value. Think of it as an acapella in sync with a rhythm track or beat.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.