ruaridh@marciablaine.com

I've just started having a potter about with max, worked through the tutorials etc and have been putting together some basic patches to get a feel for how to do the basics. One thing i seem to be having terrible trouble with is getting my objects to synch perfectly whether using metro or tempo.

In the basic patch i've got (which i can either copy in text to the forum or upload to my site), I'm trying to synch an uber-basic sequencer made up of triggers to a loop-player playing a 1-bar sample (using groove~). However every now and again the sample will start slightly delayed or the beat from the sequencer will slip, making it fall out of sunch for a bar, before recovering again afterwards.

Ive tried using both tempo and metro as i say, and tried it with overdrive on and adjusting the performance settings to favour the requisite processes. still it keeps happening. i'm starting to think either i'm being phenomenally stupid (possible as i really am a beginner here) or else it may be my pc.

my hardware: acer travelmate laptop runnign WinXp, 768mb ram, 2.6ghz processor, and an emu 1616 pcmcia soundcard.

let me know how best to show the patch and ill get that on display.

hello
I've just started having a potter about with max, worked through the tutorials etc and have been putting together some basic patches to get a feel for how to do the basics. One thing i seem to be having terrible trouble with is getting my objects to synch perfectly whether using metro or tempo.
In the basic patch i've got (which i can either copy in text to the forum or upload to my site), I'm trying to synch an uber-basic sequencer made up of triggers to a loop-player playing a 1-bar sample (using groove~). However every now and again the sample will start slightly delayed or the beat from the sequencer will slip, making it fall out of sunch for a bar, before recovering again afterwards.

let me know how best to show the patch and ill get that on display. 

basic-beginners-question-regarding-timingsynchmetro

There are a couple of things you can experiment with to 

determine if the timing of your events is solid.

Under DSP status, play with setting the "Max scheduler 

in Overdrive" and "Scheduler in Audio interrupt". 

If you still hear delays in your event timing after

setting both on, you may need to review your design and 

see if there are any places that something might get delayed.

In gerneral if you really need time critical events to 

take place, use an audio rate signal to sync things instead 

There are a couple of things you can experiment with to 
determine if the timing of your events is solid.
Under DSP status, play with setting the "Max scheduler 
in Overdrive" and "Scheduler in Audio interrupt". 
If you still hear delays in your event timing after
setting both on, you may need to review your design and 
see if there are any places that something might get delayed.
In gerneral if you really need time critical events to 
take place, use an audio rate signal to sync things instead 
of metro.

You won't get everlasting perfect sync by relying on messages, so I'd reccomend looking into using audio rate signals to do this instead, as Anthony suggested.

To post a patch, just select all and copy-paste into your post, et voila.

You won't get everlasting perfect sync by relying on messages, so I'd reccomend  looking into using audio rate signals to do this instead, as Anthony suggested. 

To post a patch, just select all and copy-paste into your post, et voila. 

Didn't I read somewhere that these timing issues were a result of OSX being multi this and preemtive that ? Is my Intel Mac not going to solve these problems (bad timing even on light work)?

On 9-Jul-2006, at 9:14, Stuart Grimshaw wrote:

> Didn't I read somewhere that these timing issues were a result of 

> OSX being multi this and preemtive that ? Is my Intel Mac not going 

> to solve these problems (bad timing even on light work)?

In what way does an Intel Mac change the OS? You're still running 

atop BSD Unix, it's still multi-tasking, it's still pre-emptive, it's 

If anyone thinks a change in processor is going to change timing 

issues, he's in for a let down. Even hoping "at least the 

inaccuracies will be faster" is pushing your luck.

Don't matter if it's a VW or a Daimler, when you drive on cobblestone 

Peter Castine +---> Litter Power & Litter Bundle for Jitter

Interface Building for |home | chez nous|

Max/MSP Extremely cool |bei uns | i nostri|

On 9-Jul-2006, at 9:14, Stuart Grimshaw wrote:
> Didn't I read somewhere that these timing issues were a result of  
> OSX being multi this and preemtive that ? Is my Intel Mac not going  
> to solve these problems (bad timing even on light work)?

In what way does an Intel Mac change the OS? You're still running  
atop BSD Unix, it's still multi-tasking, it's still pre-emptive, it's  
still all the things it's always been.

If anyone thinks a change in processor is going to change timing  
issues, he's in for a let down. Even hoping "at least the  
inaccuracies will be faster" is pushing your luck.

Don't matter if it's a VW or a Daimler, when you drive on cobblestone  
it's still a bumpy ride.

--------------    http://www.bek.no/~pcastine/Litter/    -------------
Peter Castine            +---> Litter Power & Litter Bundle for Jitter
                                Heavy-Duty Mathematics for Everyday Use
iCE:  Sequencing,  Recording &
       Interface  Building  for                   |home    | chez nous|
       Max/MSP   Extremely cool                   |bei uns |  i nostri|
       http://www.dspaudio.com/                   http://www.castine.de

>On 9-Jul-2006, at 9:14, Stuart Grimshaw wrote:

>>Didn't I read somewhere that these timing issues were a result of 

>>OSX being multi this and preemtive that ? Is my Intel Mac not going 

>>to solve these problems (bad timing even on light work)?

>In what way does an Intel Mac change the OS? You're still running 

>atop BSD Unix, it's still multi-tasking, it's still pre-emptive, 

>it's still all the things it's always been.

but suddenly (yes, took me time !!) this post/question made me wonder 

if people using maxmsp on windows do have the same timing problems 

than we have on macs??? - when a medium-level patch (i certainly did 

not build the MOST.... etc max patches) has great timing problems, 

when a single metro is NOT something you can truct in any way (except 

for banging at random, from time to time, which actually is often ok 

__I never paid attention if the (numerous) posts about timing 

problems were only from the mac users... if so, for patches which 

need a good (or normal) timing accurancy, would it be better, on new 

>On 9-Jul-2006, at 9:14, Stuart Grimshaw wrote:
>>Didn't I read somewhere that these timing issues were a result of 
>>OSX being multi this and preemtive that ? Is my Intel Mac not going 
>>to solve these problems (bad timing even on light work)?
>
>In what way does an Intel Mac change the OS? You're still running 
>atop BSD Unix, it's still multi-tasking, it's still pre-emptive, 
>it's still all the things it's always been.
>

but suddenly (yes, took me time !!) this post/question made me wonder 
if people using maxmsp on windows do have the same timing problems 
than we have on macs??? - when a medium-level patch (i certainly did 
not build the MOST.... etc max patches) has great timing problems, 
when a single metro is NOT something you can truct in any way (except 
for banging at random, from time to time, which actually is often ok 
with me)?

__I never paid attention if the (numerous) posts about timing 
problems were only from the mac users... if so, for patches which 
need a good (or normal) timing accurancy, would it be better, on new 
macs, to boot on Windows??

kasper
-- 
  Kasper T. Toeplitz
noise, composition, bass, computer
http://www.sleazeArt.com

On 9 Jul 2006, at 14:19, Peter Castine wrote:

> Don't matter if it's a VW or a Daimler, when you drive on 

nick rothwell -- composition, systems, performance -- 

> Don't matter if it's a VW or a Daimler, when you drive on  
> cobblestone it's still a bumpy ride.

   nick rothwell -- composition, systems, performance -- http:// 
www.cassiel.com


On 9-Jul-2006, at 16:22, Nick Rothwell wrote:

> On 9 Jul 2006, at 14:19, Peter Castine wrote:

>> Don't matter if it's a VW or a Daimler, when you drive on 

Exactly my point. It's the suspension, not horsepower nor fuel type, 

Peter Castine +--> Litter Power & Litter Bundle for Jitter

iCE: Sequencing, Recording & |home | chez nous|

Interface Building for |bei uns | i nostri|

> On 9 Jul 2006, at 14:19, Peter Castine wrote:
>
>> Don't matter if it's a VW or a Daimler, when you drive on  
>> cobblestone it's still a bumpy ride.
>
> Not if it's a Citroen.

Exactly my point. It's the suspension, not horsepower nor fuel type,  
that makes the difference.

--------------    http://www.bek.no/~pcastine/Litter/    -------------
Peter Castine             +--> Litter Power & Litter Bundle for Jitter

>On 9-Jul-2006, at 16:22, Nick Rothwell wrote:

>>On 9 Jul 2006, at 14:19, Peter Castine wrote:

>>>Don't matter if it's a VW or a Daimler, when you drive on 

>Exactly my point. It's the suspension, not horsepower nor fuel type, 

so what about the windows suspension?????

>On 9-Jul-2006, at 16:22, Nick Rothwell wrote:
>
>>On 9 Jul 2006, at 14:19, Peter Castine wrote:
>>
>>>Don't matter if it's a VW or a Daimler, when you drive on 
>>>cobblestone it's still a bumpy ride.
>>
>>Not if it's a Citroen.
>
>Exactly my point. It's the suspension, not horsepower nor fuel type, 
>that makes the difference.
>

On 9-Jul-2006, at 14:29, Kasper T Toeplitz wrote:

> this post/question made me wonder if people using maxmsp on windows 

> do have the same timing problems than we have on macs???

My (modest) experience with Max on Windows indicates that there are 

also timing issues. I'm not sure they're identical.

It seems likely that when the problem is audio processing competing 

with the scheduler for CPU time, you will have similar problems on 

either platform. And similar solutions (tweak Performance Options...) 

However, if the question is the OS preempting the CPU when the 

scheduler needs processor time, then both OS's can be difficult but 

there will be differences in the details.

I don't know if anyone involved with implementing the Max/MSP 

scheduler wants to go into the gory details of how the various 

architectures compare. I'm sure it is pretty gory.

On 9-Jul-2006, at 14:29, Kasper T Toeplitz wrote:
> this post/question made me wonder if people using maxmsp on windows  
> do have the same timing problems than we have on macs???

My (modest) experience with Max on Windows indicates that there are  
also timing issues. I'm not sure they're identical.

It seems likely that when the problem is audio processing competing  
with the scheduler for CPU time, you will have similar problems on  
either platform. And similar solutions (tweak Performance Options...)  
However, if the question is the OS preempting the CPU when the  
scheduler needs processor time, then both OS's can be difficult but  
there will be differences in the details.

I don't know if anyone involved with implementing the Max/MSP  
scheduler wants to go into the gory details of how the various  
architectures compare. I'm sure it is pretty gory.

don't know if I am wrong but the original question was by an user

> so what about the windows suspension?????

> I've just started having a potter about with max, worked through the tutorials etc and have been putting together some basic patches to get a feel for how to do the basics. One thing i seem to be having terrible trouble with is getting my objects to synch perfectly whether using metro or tempo.

> my hardware: acer travelmate laptop runnign WinXp, 768mb ram, 2.6ghz processor, and an emu 1616 pcmcia soundcard.

> let me know how best to show the patch and ill get that on display.

don't know if I am wrong but the original question was by an user
using windows xp...

> I've just started having a potter about with max, worked through the tutorials etc and have been putting together some basic patches to get a feel for how to do the basics. One thing i seem to be having terrible trouble with is getting my objects to synch perfectly whether using metro or tempo.
> my hardware: acer travelmate laptop runnign WinXp, 768mb ram, 2.6ghz processor, and an emu 1616 pcmcia soundcard.
>
> let me know how best to show the patch and ill get that on display.


Don't go through my hardships, metro/tempo are WAY inaccurate and get even more innaccurate the more things you do as max needs time to process all your other things. Don't even bother...

convert your BPM to milliseconds (hint: 60,000/x) = time for one beat

then depending on the resolution you want, use a signal (square wave?) to do the same thing that tempo/metro will do. except for the fact that everything will be in time because of signals using rates i believe.

correct me if im wrong before i dive into it headfirst :)

This audio rate business is all very well, but my little programme is a sociable chap, and needs to communicate with the outside world... VIA MIDI.

And on OSX, the midi sync I'm sending is, well, adventurous to say the least. I was hoping that my Intel Mac running Windows (sorry, I lost track of the car analogy) - which I haven't bought yet, hence the questions - would prove to be more reliable.

There IS no way to avoid the scheduler when dealing with midi.Or am I missing something ?

The scheduler handles all message/control data, the audio interrupt

handles all audio data. Why don't you have an audio 

rate signal trigger a midi messages. Something like a 

The scheduler handles all message/control data, the audio interrupt
handles all audio data. Why don't you have an audio 
rate signal trigger a midi messages. Something like a 
phasor~->threshold->noteout.

On 11-Jul-2006, at 23:08, Stuart Grimshaw wrote:

> Is my Intel Mac not going to solve these problems

If you say "Intel Mac" without anything else, it is reasonable to 

assume you mean "running OS X" and that's the context of the follow 

- AFAICT Max timing is no better on Windows than on Mac OS.

Caveat: I have not made extensive comparisons, this is just

- Windows XP (and several generations of its predecessors) are

every bit as preemptive, etc. as OS X, but without the

- Buying an Intel Mac to run Windows is a relatively

expensive approach to XP unless you have reasons for

- Changing OS is pretty major culture shock. Sort of like moving

from Hannover to Magdeburg 20 years ago. (For that matter, even

On 11-Jul-2006, at 23:08, Stuart Grimshaw wrote:
> Intel Mac running Windows

You said:
> Is my Intel Mac not going to solve these problems

If you say "Intel Mac" without anything else, it is reasonable to  
assume you mean "running OS X" and that's the context of the follow  
up so far.

- AFAICT Max timing is no better on Windows than on Mac OS.
   Caveat: I have not made extensive comparisons, this is just
   gut feeling.
- Windows XP (and several generations of its predecessors) are
   every bit as preemptive, etc. as OS X, but without the
   geeky tweaking things Unix allows.
- Buying an Intel Mac to run Windows is a relatively
   expensive approach to XP unless you have reasons for
   wanting both Mac OS and XP in one box.
- Changing OS is pretty major culture shock. Sort of like moving
   from Hannover to Magdeburg 20 years ago. (For that matter, even
   today).

Quote: Peter Castine wrote on Wed, 12 July 2006 10:42

> None of those machines were doing DSP on the motherboard. So it's not 

> as if everything was better in the old days.

I knew there was somrthing I'd forgotten.

> If what you want is a tight MIDI beat, the best platform would be an 

> absolutely minimal OS that lets one application take over the 

> machine. None of today's desktop OS's fills that bill.

I still have to use Finale on OS9, and pressing the play button hogs da whole show. That's just what I need to do - have my programme behave impolitely. The Macs that can run OS9 are too slow, but maybe an older Windows ? On a Mac Mini, meine ich.

> Maybe it's because if I do something complex, the complexity goes 

I have a barage of ctlin's (which don't get used that heavily), but apart from that it's just the sync~ to midiout. Hardly complex, even by my standards. Even if I have audio rate CPU load at, say, 30%,I'm still getting major problems.

Is this maybe a problem unrelated to Max/OSX ?

And yet. Logic, Reason et al all run with no problems. Is it just the fact that Max is not built to be tight but flexible ?

> Dell or Sony will probably have something vaguely similar with a 

> lower price tag. Comparable? Comparing computers depends on so many 

Comparable = small, will fit into the wooden box that The Other Stuff is nailed into :-) Not a tower.

Quote: Peter Castine wrote on Wed, 12 July 2006 10:42
ra).
>
> None of those machines were doing DSP on the motherboard. So it's not  
> as if everything was better in the old days.

I knew there was somrthing I'd forgotten.
>
> If what you want is a tight MIDI beat, the best platform would be an  
> absolutely minimal OS that lets one application take over the  
> machine. None of today's desktop OS's fills that bill.

> Maybe it's because if I do something complex, the complexity goes  
> inside an external written in C.

> Dell or Sony will probably have something vaguely similar with a  
> lower price tag. Comparable? Comparing computers depends on so many  
> factors.

thanks for the responses; i've since realised that some more diligent searching of the list archives would have found me some great strating points for audio-rate metronomes etc. My problames are now sorted and i'm able to make some headway. I think the confusion lies in the fact that the manual suggeests that using 'tempo' or 'metro' is the best way to go about timing...had i not posted to teh list i'd have probably been ripping my laptop apart trying to find where it had (yet again) let me down :)

hello
thanks for the responses; i've since realised that some more diligent searching of the list archives would have found me some great strating points for audio-rate metronomes etc. My problames are now sorted and i'm able to make some headway. I think the confusion lies in the fact that the manual suggeests that using 'tempo' or 'metro' is the best way to go about timing...had i not posted to teh list i'd have probably been ripping my laptop apart trying to find where it had (yet again) let me down :)

Basic Beginners question regarding timing/synch/metro