Forums > MaxMSP

transport bpm illogical?

October 12, 2008 | 11:57 am

In Max’s new transport object, you specify tempo in beats per minute (bpm). However, in reality what you’re setting is quarter notes per minute. So, if you’re in a time signature with a denominator other than 4, then your tempo is still set in quarter notes per minute.

For instance, if I’m in a time signature of 4/4 at a tempo of 120 bpm, then I progress one beat every half of a second, for a total of 120 beats per minute. However, if I change that time signature to 4/8, all of a sudden my tempo doubles, and I’m actually progressing 240 (eighth-note) beats per minute, since my tempo is still 120 quarter notes per minute. And if I change the time sig to 4/16, now I’m going at 480 beats per minute.

In my musical training (and in the labeling of the transport), the denominator of the time signature defines what the word "beat" means. If it’s a 4, then the quarter note is the beat, if it’s an 8 then the 8th note is the beat. For instance, you’d rarely divide a measure of 6/8 into three quarter notes (since 3/4 would be a much more obvious choice if that’s what you wanted to do).

Does anyone else agree with me or am I being stupid? I believe that one of the following should be changed (in order of preference):

1. The transport should specify its tempo in true beats per minute, where the "beat" is defined by the denominator of the time signature.

OR

2. The labeling of the tempo input on the transport (and on the transport window) should be changed from beats per minute to quarter notes per minute.


October 12, 2008 | 12:04 pm


October 12, 2008 | 4:19 pm

Perhaps the labelling may appear inconsistent, but the behaviour is the same one you have grown to know and love from all your other audio apps.

-A


October 12, 2008 | 4:21 pm

We tried to model the relationship between time signature and tempo on other software. If there are inconsistencies between the way the transport works and the way Live, Logic, or other software works, we will consider changing the transport to match. Music software, not "musical thinking" is the most relevant model.

David Z.


October 12, 2008 | 6:14 pm

Quote: David Zicarelli wrote on Sun, 12 October 2008 11:21
—————————————————-
>If there are inconsistencies between the way the transport works and the way Live, Logic, or other software works, we will consider changing the transport to match.
>
—————————————————-

It seems consistent to me, but somehow it never occurred to me that sequencers behave this way, and it’s true that it’s not musically correct. In Live, if I record a sequence of quarter notes in 4/4 at 120bpm and then change the meter to 4/8 the sequence plays back the same but the metronome will be going twice as fast. One would think it should be that the metronome stays the same and the sequence should play at half the speed. But I would say don’t change Max because then it would not be consistent with sequencers.


October 12, 2008 | 6:39 pm

For me, "musically correct" is a nightmare. 6/8 and 9/8 are counted in 2 and 3, for example. The "beat" becomes a dotted quarter. If sequencers were to address this whole scenario, then with every time sig change, you would need to specify the relationship between the beat in the previous section to the next. No thanks.

-A


October 12, 2008 | 7:11 pm

>
> For me, "musically correct" is a nightmare.

Perhaps that is a bit exaggerated, since classical musicians have managed these notational conventions quite handily for several centuries now :)

> 6/8 and 9/8 are counted in 2 and 3, for example.

That is indeed the convention. Accordingly there are three metric levels to treat: the underlying pulse (an 8th note in both cases above), the beat (either 2 or 3 pulses above) and finally the bar itself.

>>
The "beat" becomes a dotted quarter. If sequencers were to address this whole scenario, then with every time sig change, you would need to specify the relationship between the beat in the previous section to the next. No thanks.
>>

Actually, this is pretty easy to do; that’s how the sequencer works in my dmach~. In any case, most of the metric models I’ve seen for sequencers are basically "toy" systems. Perfectly fine for waltzes or four on the floor techno, but not much use for the complicated metric/tempo relations found in the music of Elliott Carter and Brian Ferneyhough among others. So I reckon that the Live/Max/Logic model is perfectly fine for the music it is designed to support.

Eric


October 12, 2008 | 7:15 pm

Correction:

>> 6/8 and 9/8 are counted in 2 and 3, for example.

> the beat (either 2 or 3 pulses above)

The beat is 3 pulses in both cases above. For 3/4 the beat would be 2 pulses.

Eric



kjg
October 12, 2008 | 7:22 pm

Quote: stringtapper wrote on Sun, 12 October 2008 20:14
—————————————————-
> Quote: David Zicarelli wrote on Sun, 12 October 2008 11:21
> —————————————————-
> >If there are inconsistencies between the way the transport works and the way Live, Logic, or other software works, we will consider changing the transport to match.
> >
> —————————————————-
>
In Live, if I record a sequence of quarter notes in 4/4 at 120bpm and then change the meter to 4/8 the sequence plays back the same but the metronome will be going twice as fast. One would think it should be that the metronome stays the same and the sequence should play at half the speed. But I would say don’t change Max because then it would not be consistent with sequencers.
—————————————————-

"One would think it should be that the metronome stays the same and the sequence should play at half the speed."

Not really.
In Logic, when I create a one bar part with 4 quarter notes at 4/4 meter and then change the meter to 4/8, the part plays back the same BUT is now 2 bars long. It is now a _2 bar part with 4 quarter notes in it. This makes sense. Because if I make a piece and then change the meter (because that is more convenient for scoring or practizing), the note values will not change but less or more of them will fit in to one bar.
Tempo/MM/bpm has nothing to do with this, really.

In scores, tempo is indicated as a note value followed by a MM (metronome measurement) value. The note value/MM combo should be defined in a way that is most practical for scoring/counting/practicing/rehearsing. It says "we are counting in this type of notes, and we will have this many of them per minute". This is NOT always quarter notes.

The MM indicates how many _pulses_ per minute the metronome should be set to. The metronome is not set to any specific note type per minute, but just a number of pulses. How these pulses are interpreted by the musicians depends on the meter.

The tempo indication in Logic (and other sequencers) is not a MM value (pulses per minute) but a BPM value (quarter notes per minute). If I change to from 4/4 to 4/8, existing material will now be twice the bar length it was before AND the metronome will now be indicating eight note pulses instead of quarter notes, like it should. The material still plays back at the same tempo since the BPM (the amount of quarter notes per minute) is still the same, as it should.

You could have a fast piece that is notated in 12/8 but with _quarter note = 160 as MM. This way you can notate a fast piece without bar numbers running insanely high AND without a metronome uzi-ing your head of at 320 pulses per minute while practizing…

Ok… Logic does what it should do meter and metronome wise and indicates tempo as quarter notes per minute (not pulses). Luckily, The metronome follows the meter, like it should. From my memory (which could be flawed) other sequencers behave the same. I imagine (but haven’t checked it) that score programs work with a true MM instead of BPM implemented as quarter notes per minute.

I would personally prefer Max to use a true MM system and not follow the four to the floor and/or loop based school of thought that sequencers follow. If it would stay the way it is now, it should be stressed in the documentation that the BPM is actually QNPM (which it has been for ages in practice, but not by definition.

Regards,
Klaas-Jan Govaart


October 12, 2008 | 7:54 pm

Quote: kjg wrote on Sun, 12 October 2008 14:22
—————————————————-
>
>
> The tempo indication in Logic (and other sequencers) is not a MM value (pulses per minute) but a BPM value (quarter notes per minute). If I change to from 4/4 to 4/8, existing material will now be twice the bar length it was before AND the metronome will now be indicating eight note pulses instead of quarter notes, like it should. The material still plays back at the same tempo since the BPM (the amount of quarter notes per minute) is still the same, as it should.
>
> Regards,
> Klaas-Jan Govaart
>
>
>
—————————————————-

I guess I was unaware that BPM meant "quarter notes per minute" as you seem to indicate here, I simply took the word "beat" to mean what it does in scored music. I have always learned and taught that the denominator of the meter is the duration that takes the beat, therefore 120BPM in 4/4 means 120 quarter notes occur in one minute, and accordingly in 4/8 at 120BPM, 120 eighth notes occur in one minute.



kjg
October 12, 2008 | 8:45 pm

Quote: swieser1 wrote on Sun, 12 October 2008 13:57
—————————————————-
> In my musical training (and in the labeling of the transport), the denominator of the time signature defines what the word "beat" means.

I see where you are coming from, but this is not completely correct. The number of pulses and the note value one pulse indicates is determined by the Metronome Mark (not Metronome Measurement, like I erroneously typed earlier) and in usually this follows the denominator of the meter, but not always.

>
> Does anyone else agree with me or am I being stupid? I believe that one of the following should be changed (in order of preference):
>
> 1. The transport should specify its tempo in true beats per minute, where the "beat" is defined by the denominator of the time signature.
>
> OR
>
> 2. The labeling of the tempo input on the transport (and on the transport window) should be changed from beats per minute to quarter notes per minute.

I agree with you, and you are not being stupid :)

I would personally prefer a true MM system where one sets the number of pulses per minutes AND the meter AND what note value the metronome indicates. This should include whole, half, quarter, eight, sixteen and dotted and triplet values.

Second best option would be your first, where the metronome just follows the denominator of the meter. If a "different metronome than pulse situation" would occur (rehearsals of complex meter and/or fast music), this could still be done pretty easily by dividing or multiplying the pulse by two or three.

If both these option are for some reason not possible or preferable, the documentation should indicate that transport does not use a true BPM, but the dumbed down BPM (which is really QNPM) we all "know and love" from the mainstream sequencers. These are like Eric Lyon stated somewhat toy like when it comes to meter because they are mostly used for 4/4 pop stuff.

I don’t agree with Eric that Max should follow this "either 3/4 or 4/4" path. In my opinion Max should be easily and practically usable in even the most complex/strange/wacked out of avant-garde/experimental music.
Like I would even use Max for making techno… That would be a typical Live job, maybe with a few pluggos for the "freakier" stuff.

Max != sequencer.

Max should be flexible and usable in any musical context, and we should therefor go with the traditional, music composition convention, not the oversimplified "beat producing" convention.

Please go MM Cycling 74, not dumbed down BPM.

With kind regards,
Klaas-Jan Govaart

PS: Thank you for bringing this up swieser1, I hadn’t noticed it yet since I’m still on Max 4 mostly, but I feel very strongly about this.


October 12, 2008 | 8:49 pm

Quote: kjg wrote on Sun, 12 October 2008 15:45
—————————————————-
> Quote: swieser1 wrote on Sun, 12 October 2008 13:57
> —————————————————-
> > In my musical training (and in the labeling of the transport), the denominator of the time signature defines what the word "beat" means.
>
> I see where you are coming from, but this is not completely correct. The number of pulses and the note value one pulse indicates is determined by the Metronome Mark (not Metronome Measurement, like I erroneously typed earlier) and in usually this follows the denominator of the meter, but not always.

> With kind regards,
> Klaas-Jan Govaart
>

Well of course, in compound meters the denominator is the subdivision of the beat, forgot to say that part in my own post.



kjg
October 12, 2008 | 8:54 pm

Quote: stringtapper wrote on Sun, 12 October 2008 21:54
—————————————————-

> I guess I was unaware that BPM meant "quarter notes per minute" as you seem to indicate here, I simply took the word "beat" to mean what it does in scored music. I have always learned and taught that the denominator of the meter is the duration that takes the beat, therefore 120BPM in 4/4 means 120 quarter notes occur in one minute, and accordingly in 4/8 at 120BPM, 120 eighth notes occur in one minute.
—————————————————-

BPM does mean Beats Per Minute. But it seems that from 90% of all western music being 4/4, 9.9% being 3/4 and only 0.1% being something else the BPM convention has become QNPM in practice. No problem when you do anything poppy or traditional in Logic or Live, but since Max is often used in more avant-garde and experimental or at least serious, scored "art-music" settings I think it should not follow the sequencer herd here.



kjg
October 12, 2008 | 9:02 pm

Quote: stringtapper wrote on Sun, 12 October 2008 22:49
—————————————————-

> Well of course, in compound meters the denominator is the subdivision of the beat, forgot to say that part in my own post.

Yes, but even in a simpler meter the composer might still use a different MM, just for practical reasons (scoring/counting/rehearsing).

Quoting myself:

"You could have a fast piece that is notated in 12/8 but with _quarter note = 160 as MM. This way you can notate a fast piece without bar numbers running insanely high AND without a metronome uzi-ing your head of at 320 pulses per minute while practicing…"

This is completely at the composer’s discretion.
It would be great if Max would be this flexible in its transport system too.


October 12, 2008 | 9:28 pm

>
I don’t agree with Eric that Max should follow this "either 3/4 or 4/4" path.
>

The transport system is somewhat atypical for Max, as c74 generally tries to avoid building music-conceptual models into Max, and when they do, it’s generally in minimalist fashion such as with kslider or nslider. And for good reason, I think. As this conversation suggests, had c74 implemented a more sophisticated rhythmic system, they would have been further inundated with complaints from educated users about how the system was just short of perfection, for all kinds of arcane reasons.

>
In my opinion Max should be easily and practically usable in even the most complex/strange/wacked out of avant-garde/experimental music.
>

I agree. But may I suggest that for whatever complex/strange/whacked-out idea you might have, it would probably take just as long for you to implement it with the primitives that Max already provides, as to learn how to use whatever higher-level tool c74 could write that might or might not be capable of doing just what you wanted.

But then again, you’re always welcome to propose feature requests :)

Eric



kjg
October 12, 2008 | 9:49 pm

Quote: Eric Lyon wrote on Sun, 12 October 2008 23:28
—————————————————-
> >
> I don’t agree with Eric that Max should follow this "either 3/4 or 4/4" path.
> >
>
> The transport system is somewhat atypical for Max, as c74 generally tries to avoid building music-conceptual models into Max, and when they do, it’s generally in minimalist fashion such as with kslider or nslider. And for good reason, I think. As this conversation suggests, had c74 implemented a more sophisticated rhythmic system, they would have been further inundated with complaints from educated users about how the system was just short of perfection, for all kinds of arcane reasons.
>
> >
> In my opinion Max should be easily and practically usable in even the most complex/strange/wacked out of avant-garde/experimental music.
> >
>
> I agree. But may I suggest that for whatever complex/strange/whacked-out idea you might have, it would probably take just as long for you to implement it with the primitives that Max already provides, as to learn how to use whatever higher-level tool c74 could write that might or might not be capable of doing just what you wanted.
>
> But then again, you’re always welcome to propose feature requests :)
>
> Eric
>
—————————————————-

Thank you for your comments Eric.

I am still very happily chugging along on Max4, and I never even felt the need for the transport system. Like you said c74 stays away from music-conceptual ideas most of the time, so it makes perfect sense to me that one builds a system like that himself when there is a need for it.

But now it IS there in Max5 anyway… In its current incarnation it is already usable in a lot of situations, but maybe it makes sense to do it "right". Right in the sense of more flexible and sticking to the convention of the composed music tradition and not to the convention of the "sequenzed beatz" tradition.

Just replying to the thread, Eric, not a feature request per se.

I’m currently looking into score following systems and it would be great if the transport could actually follow all the different tempi and meters (including metronome needed for practicing) without too much hassle.
But hey, I’m sure I’ll manage to translate or pre-chew some data from the score to get transport (and my own metronome) to output what I need.

I just don’t agree that Max should by definition follow the Logics, Lives and Cubases. _If you are going to implement a music-conceptual system, you might as well use a flexible, mature
and "musical" system.


October 12, 2008 | 10:12 pm

>I just don’t agree that Max should by definition follow the Logics, Lives and Cubases. _If you are going to implement a music-conceptual system, you might as well use a flexible, mature
and "musical" system.

Only if you consider Max5 as an app which will never be more integrated with any of these other apps in any way.

You’re more than welcome to treat Max’s transport as simply a beat counter, and roll whatever complex layer of time structuring on top of that.

-A



kjg
October 13, 2008 | 12:17 am

Quote: Andrew Pask wrote on Mon, 13 October 2008 00:12
—————————————————-
>
> >I just don’t agree that Max should by definition follow the Logics, Lives and Cubases. _If you are going to implement a music-conceptual system, you might as well use a flexible, mature
> and "musical" system.
>
> Only if you consider Max5 as an app which will never be more integrated with any of these other apps in any way.
>
> You’re more than welcome to treat Max’s transport as simply a beat counter, and roll whatever complex layer of time structuring on top of that.
>
> -A
>
>
—————————————————-

If further integration with the mainstream sequencers is what’s ahead for the future, then great! I’m all for it. If a more "musically correct" transport would harm that integration then it is obviously a no go.
So in that case transport will use the same somewhat dumbed down definition of BPM, and whoever needs more will manage just fine just like in Max 4, with or without rewire, midi clock, MTC, score followers etc.
For classically trained users new to Max a little extra info in the help file to point out that BPM actually means quarter notes per minute as opposed to beats per minute might avoid confusion and posts to the list though.


October 13, 2008 | 12:58 am

One thing which might be helpful to understand is what tempo really means. It means the rate at which increments of 480 ticks occur per minute. The time signature is a way to map ticks to a counter. It does not change how fast the ticks occur. Since the time signature affects the counter, it also affects the mapping of bar/beat/unit values you enter, because the model is that bar/beat/unit values mean "the time when the counter is at a particular bar/beat/unit combination." This means that (as of 5.0.5) if you change the time signature as the transport is running, you are changing the absolute tick values at which bar/beat/unit values will occur. But if you had specified times in ticks, nothing will change. The meaning of the "note value" symbols such as 4n never changes when the time signature changes. 4n is always 480 ticks. 4n is a way to make data entry easier, and should not be confused with a "musical quarter note." 4n has no musical meaning — it only has a numerical meaning.

So, in a sequencer such as Logic, when you record some notes, they are assigned tick values. If you go and change the time signature later, the tick values of the notes you recorded are unchanged. The way counting happens, however, changes.

David Z.


October 13, 2008 | 1:42 am

Quote: Andrew Pask wrote on Sun, 12 October 2008 12:39
—————————————————-
For me, "musically correct" is a nightmare. 6/8 and 9/8 are counted in 2 and 3, for example. The "beat" becomes a dotted quarter. If sequencers were to address this whole scenario, then with every time sig change, you would need to specify the relationship between the beat in the previous section to the next. No thanks.
—————————————————-

6/8 and 9/8 are only counted in 2 and 3 when the tempo is fast enough, and only predominantly in western music. 9/8 could be counted as 5+4 or 3+4+2 or any other combination you want. These are merely groupings of beats for convenience, not the beats themselves. The denominator of the time signature defines what the actual beat is.

Quote: David Zicarelli wrote on Sun, 12 October 2008 18:58
—————————————————-
One thing which might be helpful to understand is what tempo really means. It means the rate at which increments of 480 ticks occur per minute.
—————————————————-

This is exactly my point. If I had created Max 5, I might have done this a slightly different way, but that doesn’t matter because I didn’t create Max. My point is that either Max’s way of marking time should be changed to my way (which is the selfish view that I fully realize has virtually no chance of happening) or it should be clearly documented that the "beat" in "beats per minute" will ALWAYS be a quarter note, will ALWAYS be 480 ticks. To me, this is counterintuitive at first and I would have benefitted from more clear documentation.

Thanks for the clear explanation and for not being discouraged by our constant and relentless criticism of your amazing program… ;)


Viewing 20 posts - 1 through 20 (of 20 total)