Florent Ghys

. The argument 5n for a metro doesn't exist according to Max Documentation, so how can I have a metro playing quintuplets using this new Tempo Relative feature in Max 5 ?

. Is this feature really a big change in time processing in Max or is it just an easier and more "human-friendly" way to deal with several time objects avoiding sends and receives ?

. Thanks everybody for this great forum, I learnt a lot just reading the posts here.

1. The argument 5n for a metro doesn't exist according to Max Documentation, so how can I have a metro playing quintuplets using this new Tempo Relative feature in Max 5 ?

2. Is this feature really a big change in time processing in Max or is it just an easier and more "human-friendly" way to deal with several time objects avoiding sends and receives ?

3. Thanks everybody for this great forum, I learnt a lot just reading the posts here.

5n-argument-for-a-metro

1. The note abbreviations 1n, 4n etc are just conveniences for people, they are by no means meant to be exhaustive. Use ticks. 1920 ticks (per bar) / 5 = 384.

Since triplets are possible by providing eg the argument 2nt, quintuplets according to this logic should be notated as 2nq and a septuplet as 2ns or something like that. Apparently this is not supported. But then again, who is going to be able to dance on quintuplets?

Quote:1. The argument 5n for a metro doesn't exist according to Max Documentation, so how can I have a metro playing quintuplets using this new Tempo Relative feature in Max 5 ?

For my second question, I really feel now a metro based on the object transport is more steady than a regular one based on a fixed ms value.

The idea of ns, nq... is nice. It would avoid to do simple maths by ourselves.

But sending the message n ticks to metro works perfectly. 

The idea of ns, nq... is nice. It would avoid to do simple maths by ourselves.
But sending the message n ticks to metro works perfectly. 
thanks

jvkr wrote on Tue, 08 September 2009 09:13

But then again, who is going to be able to dance on quintuplets?

I bet everybody can dance to "Take Five" easily...

The notation of xlets is not that important, they are translated to ticks anyway...

jvkr wrote on Tue, 08 September 2009 09:13But then again, who is going to be able to dance on quintuplets?

The notation of xlets is not that important, they are translated to ticks anyway...


Hi guys, I realise this topic is a couple of years old now, but I'm now dealing with the following, related issue. As Andrew pointed out, a quintuplet can be achieved as follows: 1920 ticks (per bar) / 5 = 384. Nice integer.

However, a septuplet would deliver 1920 ticks (per bar) / 7 = 274,28571428571429… That outcome is a truncation anyway, multiplied by 7 again, it does not make 1920.

Anybody who has been dealing with this? What are the consequences of Max' limited temporal resolution?

Hi guys, I realise this topic is a couple of years old now, but I'm now dealing with the following, related issue. As Andrew pointed out, a quintuplet can be achieved as follows: 1920 ticks (per bar) / 5 = 384. Nice integer.
However, a septuplet  would deliver 1920 ticks (per bar) / 7 = 274,28571428571429… That outcome is a truncation anyway, multiplied by 7 again, it does not make 1920.
Anybody who has been dealing with this? What are the consequences of Max' limited temporal resolution?
Thanks,
JB


I can see your point. That is not much fun but I guess it will have to be that way... ;-)

4nq - Quarter note quintuplet - 384 ticks

16nq - Sixteenth note quintuplet - 96 ticks

32nq - Thirty-second-note quintuplet - 48 ticks

64nq - Sixty-fourth note quintuplet - 24 ticks

That leaves us with 7 and 9-tuplets, and above ten: 11, 13, 17, 19-tuplets, etc. - and multiples thereof.

While in composition, the above tuplets >10 are hardly ever used (Yes, ok, Ferneyhough), the 7 and 9 are very, very common.

Even then, the Midi resolution of a system would have to be ridiculously high to deal with these.

Other systems have this restriction as well. Logic Pro X for example, runs at 960 ticks per beat, that's 3840 per whole note - also not sufficient for 7-tuplets. Indeed, I did run into theoretical problems with DAW's in the past. I'lld like to stress "theoretical", since the workarounds are inaudible, which often has been, in the end, the aim of the developers.

The subject brings highly interesting filosophical issues regarding binary environments to the surface...

Thanks Andrew!
I can see your point. That is not much fun but I guess it will have to be that way... ;-)
The only tuplet 
1nq - Whole note quintuplet - 1536 ticks
2nq - Half note quintuplet - 768 ticks
4nq - Quarter note quintuplet - 384 ticks
8nq - Eighth note quintuplet - 192 ticks
16nq - Sixteenth note quintuplet - 96 ticks
32nq - Thirty-second-note quintuplet - 48 ticks
64nq - Sixty-fourth note quintuplet - 24 ticks

That leaves us with 7 and 9-tuplets, and above ten: 11, 13, 17, 19-tuplets, etc. - and multiples thereof.
While in composition, the above tuplets >10 are hardly ever used (Yes, ok, Ferneyhough), the 7 and 9 are very, very common.
Even then, the Midi resolution of a system would have to be ridiculously high to deal with these.
Other systems have this restriction as well. Logic Pro X for example, runs at 960 ticks per beat, that's  3840 per whole note - also not sufficient for 7-tuplets. Indeed, I did run into theoretical problems with DAW's in the past. I'lld like to stress "theoretical", since the workarounds are inaudible, which often has been, in the end, the aim of the developers.
The subject brings highly interesting filosophical issues regarding binary environments to the surface...


Mmm, Andrew. I'm looking at your approach again and I'm running into timing problems. The attached patch will show what I mean.

Sorry, Andrew, once again, a newer version for ease of analysis... What happens on my system is that the very first pulses are always together, but after that, the tuplet starts late. If we restart (through your original toggle), it does the same.

Yeah I think you need to specify the interval in ticks or it defaults to ms

Aaaah! Of course, thanks a lot. We'll soldier on.

Sorry to have to say this, but that doesn't solve the problem, as became apparent after pasting your code into my tester:

I think the problem is that attrui sends ms instead of ticks.

Yep, that solves it. And your assumption about the reset is correct, I think: The first beats stay in sync.

Why all this? I'm patching a m4l algorithmic composition assistant, so a correct binding with Transport is essential.

It will need all my cunningness (in desperate solitude!) to get your solution implemented, but it works so thank you very much, Andrew.

Yep, that solves it. And your assumption about the reset is correct, I think: The first beats stay in sync.
Why all this? I'm patching a m4l algorithmic composition assistant, so a correct binding with Transport is essential.
It will need all my cunningness (in desperate solitude!) to get your solution implemented, but it works so thank you very much, Andrew.


Going to let it run for an hour to see if it falls out of sync

this seems to do the trick
Going to let it run for an hour to see if it falls out of sync


5n argument for a metro