how to deal with ternary signatures and tempo ?

Hello,
I’ve been looking around about this issue for a bit and i can’t solve it alone…
I’m working on a piece with a 6/8 key signature. the tempo is 120, i want the reference beat to be the dotted quarter, not the quarter.

hydrogen takes in the 120 tempo sent by the clock as the dotted quarter seeing i have set a ternary beat.
but ardour considers 120 to be the quarter.

i want the metronome to beat the dotted quarter and the midi clock to be sent this way to.
Is this possible ?
How can i work with ternary subdivisions of the time in an other way ?

thanks for your help.

Mikaël

We support beats other than quarter notes/crotchets, but not dotted quarters, unfortunately. You can select the beat note type in the meter editing dialog, but you will only find “whole subdivisions” there. I would ask that you file this as a feature request so that it does not get forgotten. It is relatively simple to add new note types for this purpose (to the source code of Ardour), but someone with the time and expertise would need to verify the functionality.

If that’s going to happen, let’s allow any pattern of main beats to be defined manually. Many dance tunes from the Balkans are in irregular time signatures, though perfectly consistent from one bar(measure) to the next. 7/8 and 11/8 are very common, and I know one which is 25/8 grouped as 3+2+2+3+2+2+2+2+3+2+2. Those are the main beats, often described as long and short beats.

Ideally, it should be possible to have the time signature change anywhere, because there’s lots of music that does that too.

anahata: what you just described is very, very much more complex than what I mentioned. In fact, it is really a fundamental reaimagining of what a description of meter looks like. Although Ardour’s current version of that attempts to be very flexible, it is still rooted in the western canonical tradition in which there are bars and beats and the distribution of beats inside a bar is consistent, rather than just beat patterns (as is true in carnatic music and many other musical traditions from around the world). There’s really no reason to describe Balkan music as 11/8 (for example) - it is a collection (and repetition) of beat patterns as you describe. The 11/8 designation is just an attempt to impose the western canonical view onto the structure.

on the other hand, there is the argument that since the beat pattern does in fact repeat after 7, 11 or 25 beats (to cite your examples), and because Ardour doesn’t really do anything notable with the internal beat structure (other then the option for emphasis on the first beat), it is reasonable to argue that a 7/N, 11/N or 25/N representation is entirely adequate.

if Ardour was to offer more functionality with respect to the beat structure of a bar/measure, then certainly we’d have to offer the ability to define this beat structure as something other than “N beats uniformly spaced within the measure”.

If I have understood, couldn’t what anahata is suggesting be implemented as a normal western-style metre, as Paul describes, e.g. 25/N, but with an option to choose which beats the emphasis comes on separately? Maybe some checkboxes for yes/no emphasis on each beat (or something), which only appear when you enable the proposed emphasis mode for the current metre. I assume only the metronome playback would need to use this, since the underlying metre is still a western /N, and this is already accounted for as far as everything else is concerned. However, it would allow for arbitrary beat groupings within the metre from the perspective of the user when the metronome is enabled.

it might be desirable to draw the measure lines differently if you consider a piece to be made up of 4+3+2 rather than being in 9/8 …

Yes, measure subdivisions and different sounding metronome beats (and possibly dynamic emphasis) are areas where these subdivisions could have an effect.
Speaking of 9/8, a good and well known example of unequal main beats in that time signature is Dave Brubeck’s “Blue Rondo à la Turk” which is 2+2+2+3 whereas the usual division of 9/8 is 3+3+3.

Or anything else in Balkan music 3+2+2+2 or 2+3+2+2. For Balkan music it is very important to know the subdivision for playing a piece correctly. This is often written in the top of the score

I mess around with progressive rock and do a lot of stuff that resembles what anahata describes. Often the 15 or however many quaver beats are simply too fast to be much use – being able to set the metronome to tick on the pulses, the first of each group of two or three, would be really useful.

Currently I end up doing a drum loop in Hydrogen to get the feel I need, export, import into Ardour etc, and use that instead of the metronome. It does the job but is a bit laborious.

I don’t know if this still works, but it seems like the klick metronome is what you would need:
http://das.nasophon.de/klick/
Here is the manual page:
http://das.nasophon.de/klick/manual.html

I saw this post googling about that point. It didn’t help me for my purpose that was to give a 6/8 feeling in a 4/4 song. I finaly found the issue by myself simply changing the grid value as third of time instead of half or whole time (that are equal in terms of use) in places I needed.
I don’t know about Balkan music but in case of Dave Brubeck complex measures experiences in 5/4 for Take Five or 9/4 for Blue rondo a la Turk" we should not talk about x/8 because the time is already divided in quarter notes. These are binary signatures. It is just the number of times per mesure that is involved, not the inner time feeling.
When you talk about 6/8 it really means a different ternary inner time feeling where you divide the time in 3 or 6 instead of 4 or 8.
In Jazz, the 4/4 time signature is a simplification of time writing that should be noted 12/8, as 4 times divided by 3, precisely.
About metronome considerations, the metronome is not a rythme feature but a tempo feature. It should just mark the beat whatever how the beat is divided or the measure is composed.
Not to confuse with accentuation, in whitch Ardour should not be involved. It’s the composer role to put an accent if needed on beats I think. For example for “Blue Rondo a la Turk” whitch could be signed as 9/4 you could put a click on 1, 3 and 5 beats. Or 1, 3, 5 and 6 bits. But this can be done with any rythme box.
Hope that would help

I started reading the thread da capo and thought that it was something I ought to comment on… and lo and behold, I had done :slight_smile:

Just to say again, I think it would be very useful to be able to set the click not to play every subdivision of a meter and to be able to set the pattern of duplets and triplets it does click on.

Take for example something in 13/8 grouped as 2+2+3+3+3. Hearing all 13 beats each bar might not be desirable, depending on the tempo. The piece I have in mind I feel as two short pulses followed by three long ones: being able to select which of the 13 beats are actually sounded by the click (in this case 1, 3, 5, 8 and 11) would give the correct pulse.

I’ve thought of a way this might be able to work, from a user perspective (I’ve no clue how easy or not it would be to implement). As standard, each tempo marker works as it currently does, each bar is subdivided into x beats of N note value. On the context menu for a tempo marker have an option which pops up a dialogue box (called it “Feel” or “Pulse” or something like that). On the dialogue are x checkboxes, which the click “reads”, playing a click on the ones that are checked and remaining silent on the unchecked ones.

Alternatively, the click could play all subdivisions as it currently does, with the 1 emphasised as it currently is, but the checkboxes could be used to select additional subdivisions for emphasis within each bar, creating the uneven pulse.

Either method would allow every possible pattern of emphasis of beats in a bar in quite a simple to use manner. I think I’ll create a submission on Mantis…

Sorry to double post… all of which is to say that I disagree with kmchen. I think it would be very useful if Ardour’s click/metronome DID (or more properly, COULD) get involved in accentuation.

As mentioned in my post from 2016, for me, it would save me having to programme something in Hydrogen to import into Ardour.