3/3 time?

27 replies [Last post]
mcgruff
User offline. Last seen 1 year 3 weeks ago. Offline
Joined: 2008-08-30
Posts:

Is it possible to set an Ardour tempo in 3/3 time?

peder
User offline. Last seen 49 weeks 3 days ago. Offline
Joined: 2007-05-08
Posts:

There is no such meter in music.
It is all based on multiples of 4, so you can have 3/4, 7/8 or 21/16 but not 3/3, 8/7 or 16/11

John E
User offline. Last seen 15 weeks 4 days ago. Offline
Joined: 2006-10-27
Posts:

The second number in a time signature (e.g. the '4' in 3/4) indicates the length of the musical note - in this case a "quarter note" or crochet. In 3/4 time there are 3 crochets per bar. In 3/8 time there would be 3 "eighth notes" (i.e. three quavers) per bar. There's no such thing in music as a "third note". 3/3 time doesn't exist.

Vodage
User offline. Last seen 2 years 48 weeks ago. Offline
Joined: 2011-05-18
Posts:

I just created an account in order to post this. 3/3 time does indeed exist, but the processes involved in composing a piece of music in this time signature are very intensive.

To give an example:

Music is, to some extent, expressed exponentially. So if I start with 4 parts per beat, and 4 beats per measure, I'm looking at 16 parts per measure.

3 parts per beat and 3 beats per measure gives me 9 parts per measure. It's less, parts, so it should be simpler, right?

But the exponential base for beats in a "four" time signature isn't actually four. It's two.

So if i start squaring those numbers I get
2 4 8 16

If I start squaring 3's, however, I get
3 9 27 81

So in other words, when you're composing a piece of music in 3/3 time, your movements will become very long very quickly.

It becomes difficult to wrap one's mind around melodies like this.

It can be done but computers aren't really set up to jive with this sort of time signature.

My suggestion is to screw around with it as much as possible. Go and get "take 5" on paper, and try to add a fifth measure to the loop, so that you end up with a piece of music in 5/5 instead of 5/4.

The idea that a time signature simply "doesn't exist" is erroneous.
Time can be divided in any way you like. Just don't try to divide by zero or you may create some sort of musical black hole!

Vodage
User offline. Last seen 2 years 48 weeks ago. Offline
Joined: 2011-05-18
Posts:

Take the song "Happy Birthday" for example.

It is a waltz, Written in 3/4 meter.

1. Happy Birthday To you,

2. Happy Birthday To You,

3. Happy Birthday Dear Na-Ame,

4. Happy Birthday To You!

Sing the song to yourself out loud.

Now sing the song to yourself again, but this time, after you finish singing 3, go directy to 1. Skip 4 altogether.

That is 3/3 meter.

paul
paul's picture
User offline. Last seen 18 hours 41 min ago. Offline
Joined: 2006-03-16
Posts:

@mcgruff: same was as setting any other meter. Right click in the meter track -> New Meter -> set Beats Per Bar to 3.0 and Note Value to Third. Done.

paul
paul's picture
User offline. Last seen 18 hours 41 min ago. Offline
Joined: 2006-03-16
Posts:

Its not correct to say that computers aren't really set up to deal with kind of time signature. Ardour is setup to deal with any time signature, including time signatures with non-integer note values and beats per bar. Its a long time joke that in Ardour you can create a piece that is "in Pi!" To a piece of software, its all just numbers. To a musician, it might be more difficult.

chazza
User offline. Last seen 1 year 14 weeks ago. Offline
Joined: 2011-04-20
Posts:

I disagree with Vodage - the 4 in a 6 4 time signature doesn't represent a unique duration, if it did there would be no need for a bpm value. '4' is just a representation of part of a convention. The convention does not include '3' or '5' or 'pi' and is therefore meaningless. There's no reason to assign a value to it of one and a half quarter beats, it could just as easily represent a non-integer such as pi.

Similarly there is no note type associated to minus three.

Good thread though :)

peder
User offline. Last seen 49 weeks 3 days ago. Offline
Joined: 2007-05-08
Posts:

Like Paul said, "it's all just numbers", and since, say, triplets and quintuplets are sort of an abomination in standard notation I guess you could concoct some alternative meter that would make the notation of those easier. And I'm pretty sure you couldn't find a musician that would be able to read it.
No matter what elaborate meter you'd invent you could always boil it down to the standard notation so except from really slow drone music or totally chaotic avantgarde I can't really see the point.

But if there are some documentation on the net about it I'd really like to read it.

mcgruff
User offline. Last seen 1 year 3 weeks ago. Offline
Joined: 2008-08-30
Posts:

Thanks everyone for your help.

It was indeed a waltz time signature, 3/4, I was looking for - my mistake.

I couldn't find the UI command to change time signature because I always have the meter track switched off... doh! Got it now.

John E
User offline. Last seen 15 weeks 4 days ago. Offline
Joined: 2006-10-27
Posts:

Posted by chazza
I disagree with Vodage - the 4 in a 6 4 time signature doesn't represent a unique duration, if it did there would be no need for a bpm value. '4' is just a representation of part of a convention.

That depends on whether you're talking about musical durations or durations of time. The '4' indicates "quarter notes". In 6:4 time, each measure (or bar) would last for exactly six "quarter notes" - or 6 crochets. Of course, that doesn't mean that every bar has to be composed using 6 crochets. But what it does mean is that the duration of every bar is exactly the same duration as 6 crochets. In 6:8 time, each measure has the same duration as 6 quavers etc. The actual (time) duration of a crochet or quaver depends on the tempo. The faster the tempo, the shorter each measure becomes and correspondingly, the shorter is each crochet and correspondingly, the higher is the bpm

[Edit...] Vodage - "Happy Birthday" is actually in 6:4 meter. The shortest note is a quaver and the longest note is a minim but the accent falls on every sixth crochet. Therefore the time signature is 6:4.

seablade
User offline. Last seen 2 hours 10 min ago. Offline
Joined: 2007-01-22
Posts:

That depends on whether you're talking about musical durations or durations of time. The '4' indicates "quarter notes". In 6:4 time, each measure (or bar) would last for exactly six "quarter notes" - or 6 crochets. Of course, that doesn't mean that every bar has to be composed using 6 crochets. But what it does mean is that the duration of every bar is exactly the same duration as 6 crochets. In 6:8 time, each measure has the same duration as 6 quavers etc. The actual (time) duration of a crochet or quaver depends on the tempo. The faster the tempo, the shorter each measure becomes and correspondingly, the shorter is each crochet and correspondingly, the higher is the bpm

While from a technical standpoint when speaking strictly of duration, you are correct, 6/4 and 6/8 are both compound time signatures so the actual meaning is better represented as... 2 dotted half notes, for 6/4 and 2 dotted quarter notes for 6/8. The end result when speaking of duration is obviously identical, but when speaking of notation, beaming, etc. and also playing, is very different.

Seablade

chazza
User offline. Last seen 1 year 14 weeks ago. Offline
Joined: 2011-04-20
Posts:

I still think there's some confusion here: take Waltz time, 3:4, three is a number but 4 is not a number, it's a conventional symbol for crotchet. If we happened to use European note names - minim (M), crotchet(C), quaver(Q), semiquaver(S) and so on, 3:4 would become 3:C, 6:8 would become 6:Q and we wouldn't be discussing this.

Similarly the note names we use (A, B, C, D, E, F, G) have nothing to do with the alphabet, again they are conventional symbols. We could just as easily use cat, dog, hamster. A, B, C etc give a useful idea of the relationship between frequencies but so would P, Q, R, S, T, U, V.

Still, it's brilliant that all this came about because mcgruff wrote 3/3 instead of 3/4. Lead on, mcgruff

NB, I know a bit about music but I'm still an absolute beginner when it comes to Ardour.

paul
paul's picture
User offline. Last seen 18 hours 41 min ago. Offline
Joined: 2006-03-16
Posts:

@chazza: it is merely convention that says that the denominator is a standard note value (and typically just one of a small range). there are musical traditions around the world that have rhythmic constructions that are not well represented by a description of the form "there are 3 beats per bar, and a beat is a crotchet/quarter-note".

now, in deference to the facts that (1) these are somewhat obscure traditions among the users of a DAW and more importantly (2) that even within these musical traditions, most rhythmic constructions end up using patterns that can be reasonably represented with standard integer-fractions of a whole note, Ardour does limit the denominator to the normal western values. internally, however, it uses floating point values for both the numerator and denominator, and one could imagine a GUI that allowed you to specify any value for the denominator.

"this one has Pi beats per bar, and 1 beat is 1/sqrt(2) of a whole note" :))

Ricardus
Ricardus's picture
User offline. Last seen 5 hours 4 min ago. Offline
Joined: 2009-05-31
Posts:

Steve Smith formerly of Journey who now plays drums in his jazz fusion band Vital Information spent some time in India learning all about eastern music, and when he got back, the first thing he did was write a song in 7 and a half. Sure, 7.5 and 3/3 don't really exist in western music, but they can be played. In the west we would take that 7.5 and turn it into 15/8 or something, but then the song loses its feel since the measures are twice as long. They came to my home town and played this song. Steve even taught us how to count it. In this video he taught the crowd how to count it, but they edited it out, but you can hear Steve speaking as he begins the song, and that was the tail end of his teaching how to count it:

http://www.youtube.com/watch?v=r_-dkeWOcnI

Ricardus
Ricardus's picture
User offline. Last seen 5 hours 4 min ago. Offline
Joined: 2009-05-31
Posts:

As my prog-rock drummer friend, Tim says, "Any time signature can exist, it just depends how weird you wanna get."

chazza
User offline. Last seen 1 year 14 weeks ago. Offline
Joined: 2011-04-20
Posts:

Paul - "this one has Pi beats per bar, and 1 beat is 1/sqrt(2) of a whole note" :))

Would love to hear that one :) I can imagine it existing, but not how it might sound, not the way I can hear a melody if I sight read it from a score. Anyway I think if we sail much farther we might fall off the edge of the world.

An afterthought, re Ricardus and '7 and a half : 4', there'e a whole tradition of dance music from the Balkans written in 7, 9, 11, 13, and 17 beats to the bar which sounds completely natural - as an example the 11 beat breaks down into 1-2, 1-2, 1-2-3, 1-2, 1-2 semiquavers, bpm dependent on amount of alcohol consumed but always pretty fast.

paul
paul's picture
User offline. Last seen 18 hours 41 min ago. Offline
Joined: 2006-03-16
Posts:

@chazza: the thing with half-beats-per-bar is that its often combined with a melody line that is in done in the 2x multiple. so you have a percussion line in 7.5 beats per bar and a melodic line with 15 beats per bar. this creates an interesting "phase shift" for the 2nd 7.5 beats as the percussion shifts by 1/2 beat relative to the melody. my guess is that there may be variants that reverse the percussive/melodic roles too. i've heard this done with indian music using 9.5 beats per bar, and its a very different effect than just using the 2x multiple.

thorgal
User offline. Last seen 1 year 21 weeks ago. Offline
Joined: 2007-08-03
Posts:

@Ricardus

I don't know Steve Smith at all, nor do I know his music and whereabouts, but I did compose a rock song in 7.5 (15/8) which I find really cool (never had time to finish it, but I will one day :). It really sounds at first like you can go along with the beat but once in a while, a beat is missing so you have to double up your "going along" with the beat not to lose it. A few friends got disturbed by that :D I just want to say though that the 7.5 time sig did not come to me out of intense thinking but because I made a mistake when improvising some guitar sequence. I ended up using the mistake on purpose as an intrinsic part of the song :)

chazza
User offline. Last seen 1 year 14 weeks ago. Offline
Joined: 2011-04-20
Posts:

I feel I'm wasting valuable time that could be dedicated to packaging an alpha release into an iso so a not-so-technical person like me could do some testing, but until someone tells me to stop I'll keep up my end of the discussion.

I listened to the Steve Smith piece (great playing) and I find it hard to hear anything that would indicate that the piece isn't in 15:8 with a beat of 4,4,4,3. There's a secton where two bars are run together to give 4,4,4, 3,3,3, 3,3,3, and a bit around 2m44s where the beat gets away and it all slips into 4:4 for one bar, also some sections that he divides into 30:16 then slices and dices that into all sorts of interesting divisons, but so do many other composers writing in standard time signatures (Beethoven's 5th? Da da da daaaaa .. it's in 4:4). It reminds me of Blue Rondo a la Turk from Dave Brubeck's album 'Time Out' (http://www.youtube.com/watch?v=kc34Uj8wlmE) which is in 9:8 - or Could It Be ... 3/3 ??? mcgruff, you're a star.

A lot of well known pieces that do this kind of thing without using non-integer beat lengths, e.g. The Rite Of Spring, 1913 (anyone interested in seeing a bit of the score look here: http://en.wikipedia.org/wiki/The_Rite_of_Spring ), Carmina Burana part 6, 1940, (various 11 and 14 beat sections over a thumping 2:4 beat) and how about Kashmir by Led Zep? No half beats here but the rhythm's clearly in a different time signature to the guitars and vocals. Also check out the solos in Rock and Roll.

I would really love to hear some music that could only be in a time signature such as X:7.5 but I'm not convinced yet.

Thorgal, can we hear a few bars of your 7.5?

paul
paul's picture
User offline. Last seen 18 hours 41 min ago. Offline
Joined: 2006-03-16
Posts:

@chazza: (1) an ISO isn't going to help anyone test a3-alpha in a useful way, i think (2) i don't the points above were about X:7.5, but about X.5:Y. if you want to hear something with 9.5 beats per cycle, check out L. Shankar's Raga Aberi. Not my favorite piece of his, and it doesn't demonstrate the shifting between the percussive and melodic lines as clearly as some pieces I've heard him perform live, but it does have the basic structure where the extra half-beat is key, and its not just "half of 19 beats per cycle".

chazza
User offline. Last seen 1 year 14 weeks ago. Offline
Joined: 2011-04-20
Posts:

OK, will listen to that, thanks for the direction.
Re ISO, you can easily tell I don't know what I'm talking about

thorgal
User offline. Last seen 1 year 21 weeks ago. Offline
Joined: 2007-08-03
Posts:

@chazza:

It is unfinished work. I would prefer releasing my work (or part of it) when finished but if I were to count along a few bars, I would go like:

one-two-three - one-two-three - one-two-three-four - one-two-three-four-five
one-two-three - one-two-three - one-two-three-four - one-two-three-four-five
...

or, if B = bass drum and S = snare drum, the basic sequence is:
| BBS - BBS - BBBS - BBBSS |

Parts of my song are based on this basic pattern, with variations of course. I speeded up the tempo at some point, for a fast ending sequence (144 BPM, 15/8 time sig)and it was a bit of a challenge to play that rhythm on the guitar and bass without thinking. I can now play it naturally but it took me a while :)

chazza
User offline. Last seen 1 year 14 weeks ago. Offline
Joined: 2011-04-20
Posts:

@thorgal
I know what you mean, I don't like doing that either. When you get it finished remember me and let me know where to go to listen. I'd love to hear it. regards, chazza

DavePhillips
User offline. Last seen 1 week 6 days ago. Offline
Joined: 2006-04-19
Posts:

The denominator in a conventional time signature is an indication of an orthographic representation, i.e. a 16th, an 8th, a quarter-note, etc. It's there for the sheer convenience of the notator.

I like to pose this trick question to my students: Tap a steady beat on the table top, with no particular accent pattern, just ta-ta-ta-ta-ta-ta-ta without discontinuity. Now tell me the time signature it's in. Of course, the rhythm may be divided into any meter I like, merely with the placement of an accent, i.e. TA-ta-ta-TA-ta-ta-TA-ta-ta. Now we have a clearly articulated meter of 3 beats - now tell me what the beat representation is supposed to be in the written notation. 3/16 ? 3/8 ? 3/4 ? All correct. The decision is a notational decision. 3/3 will be incorrect because there is no accepted representation for the duration of a "3rd" note.

Now of course we can call the denominator anything we like if we're not abiding by the conventions of standard Western notation as represented by Gardner Read (or UE or Schirmer). And of course then you can spend extra time explaining your decisions to musicians who will probably wonder why you went to the trouble. :)

The top number in the time sig can of course be any number. Compound meters are often represented as sums, i.e. 7+8 over a 4 indicates a 15/4 meter with a subdivision after the 7th beat. But you all know this. Fractional beats, e.g. 7.5 over 4 might be better construed as 15/8 (as thorgal indicated) and are likely to be so construed by reading musicians. Fluent sight readers will correct your notational inventions in order to sensibly perform your piece. :)

Las refers to Indian rhythmic practice. I have a question regarding that practice: Is there a sense of a mensural beat in a tala ? The concept of the rhythmic cycle is straightforward enough, but when they refer to a 17-beat cycle or a 33-beat cycle, is there a unit of measure implied, i.e. do they think in terms of notational units similar to 8th notes, crochets, and so forth ?

paul
paul's picture
User offline. Last seen 18 hours 41 min ago. Offline
Joined: 2006-03-16
Posts:

@davephillips: i don't want to try to answer too authoritatively on this because I don't think I know enough from the studies I did, but I'm pretty sure that the rhythmic cycles ("talas") used by carnatic and other classical indian music have no notion of a mensural beat. the cycle is defined by a fixed number of beats and some subdivision, but the subdivision is (i think) always expressed in terms of the konakol phonemes ('bols'), not note lengths. so for example, the most popular tala, tinntal, consists of 16 beats in 4 equal groups of 4, with equal timing between each beat onset:

dha dhin dhin dha
dha dhin dhin dha
dha tin tin na
na dhin dhin dha

notice that the "tin" and "na" sounds are slightly shorter and pitched differently than the dha/dhin sounds (all 4 are different from each other), and so the real structure of this tala is the shifting time feel that comes from the "tin tin na / na" section.

the thing about a given tala is that its just a structure on which the real performance is based. if you ever hear anyone play tintal, it will be quite hard to pick out the structure above, because they will be subdividing almost every beat into phrases. So for example, one of the "dhin" beats might be replaced by "Te Te Tete".

the melodic instruments will be doing the same thing, and I am relatively confident that the performers will speak of note durations in terms of beats (or fractions thereof) rather than defined note lengths. of course, the actual terminology in different (its not english), but i think this that "beat" is an accepted translation of the terms used in India. thus, when playing this music, you don't measure time in note lengths, but via the length of the tala and the tempo.

allank
allank's picture
User offline. Last seen 6 weeks 2 days ago. Offline
Joined: 2006-12-07
Posts:

@Paul (and Dave Phillips).. this is also similar to music of sub-saharan africa where time is often determined by a cowbell pattern...
This has been reduced by many western musicologists as usually being in 12/8 - however speak to a traditional musician in Ghana and they may tell you that 12/8 is meaningless. (learned from my major in ethnomusicology many years ago.. I knew it would come of value somewhere :-) )...

For this reason the ability to have any denominator within Ardour maybe very useful ... Musical rules are and always have been a construct of theory (specific to time period and culture), that over time are often broken.. without breaking the rules we never would have had got as far as Beethoven, so it is important that the software we use doesn't limit us to the accepted theory and allows for anything..

This is equally true for rhythm as it is for Harmony.. (and thanks to certain sYnth manufacturer of the 80's for allowing custom programmed intervals)..

Well there is my 2 cents on this very interesting Friday afternoon read.

Rimbama
User offline. Last seen 2 years 43 weeks ago. Offline
Joined: 2011-05-18
Posts:

I do not feel there are any limitations rhythmically with Ardour. Our concept of musical time signatures is based on convention and emphasis. When we use a time signature we are relating to the listener what the rhythmic emphasis is. In Ardour, if we want a 5/8 bar we have the tools to relate this with one emphasis at the beginning of the bar, or we can set it up as 2 +3 or 3+ 2 and we could do 2+2+3 for 7/8.

Our western system of notation is very flexible and works for almost anything I can think of. I guess what I am trying to say is the numbers in a time signature are meaningless. We find the emphasis of the music and then relate that anyway we can. So far I have not run into something that cannot be done in Ardour.

Note: Being able to change the denominator in Ardour would be useful to change the tempo of the click as a whole. As a practice tool, if you wanted to be able to have complex rhythmic relationships without several tempo changes set up, then having more denominators would be useful. In other words, being able to change the initial speed and have it apply the correct math across the piece instead of doing the math yourself.

I am not sure if I am on subject, just wanted to give my opinion on the rhythmic aspects of Ardour. When I get home this evening I will post something that I have been working on in Ardour that is in x/4 time but has subdivisions at the level of the quintuplet and is difficult to tap your foot to.

R