odd metro behaviour

Learn about building and using Max for Live devices.
Post Reply
fisk
Posts: 36
Joined: Wed Jan 28, 2009 1:15 pm

odd metro behaviour

Post by fisk » Thu Oct 01, 2009 5:20 pm

if i create a metro in MfL as such:
metro 1n @quantize 1n @autostart 1 @autostarttime 0

then the metro does not ever output a bang on the first hit, it picks up on the second and from then on its fine


if i create one like this:
metro 1n @quantize 4n @autostart 1 @autostarttime 0

then the metro usually misses the first hit, but sometimes can miss the first 4 hits, only picking up properly after 1 bar

is this a bug?

fisk
Posts: 36
Joined: Wed Jan 28, 2009 1:15 pm

Re: odd metro behaviour

Post by fisk » Thu Oct 01, 2009 5:22 pm

sorry that second metro is supposed to be:

metro 4n @quantize 4n @autostart 1 @autostarttime 0

benbracken
Posts: 68
Joined: Tue Sep 01, 2009 4:07 pm

Re: odd metro behaviour

Post by benbracken » Thu Oct 01, 2009 6:16 pm

Out of curiosity, what is Live's buffer size set to (Preferences->Audio)?

-Ben

fisk
Posts: 36
Joined: Wed Jan 28, 2009 1:15 pm

Re: odd metro behaviour

Post by fisk » Thu Oct 01, 2009 6:21 pm

506 samples

i'm also noticing some pretty appalling timing errors when comparing a metro with an Live midi sequence, especially when using sends and returns. this is even worse when the sends are on a different channel to the returns.

i thought the idea here was to have one transport system, i didn't realise there were going to be any kind of 'internal' sync issues, kind of surprising.

fisk
Posts: 36
Joined: Wed Jan 28, 2009 1:15 pm

Re: odd metro behaviour

Post by fisk » Thu Oct 01, 2009 6:28 pm

after experimenting with several different buffer settings i have managed to make a few nice flange patterns

i mean its nice but its not really what i was looking for

fisk
Posts: 36
Joined: Wed Jan 28, 2009 1:15 pm

Re: odd metro behaviour

Post by fisk » Thu Oct 01, 2009 7:21 pm

ok after more testing
its tempo-dependent as well as buffer size, change it 0.1 bpm and it makes a different flange-moire pattern.

the flange is between about 0 and 20 ms

its ridiculous, even just listening to a metro like this, with no sync reference
(metro 16n @quantize 16n @autostart 1 @autostarttime 0)

you can easily hear it drifting about, its worse than syncing max/live over over rewire and maybe on a par with midi beat clock.

benbracken
Posts: 68
Joined: Tue Sep 01, 2009 4:07 pm

Re: odd metro behaviour

Post by benbracken » Thu Oct 01, 2009 7:36 pm

Thanks for the report, we will look into it. There appears to be some problems with metros and buffer sizes that are not multiples of 64 samples.

Do you have problems when you set the buffer size to, say, 512?
-Ben

fisk
Posts: 36
Joined: Wed Jan 28, 2009 1:15 pm

Re: odd metro behaviour

Post by fisk » Thu Oct 01, 2009 7:41 pm

512 is pretty much the same
the only time i managed to get it to stabilise (or at least seem to) was when it was set to 252 samples and running at 150 bpm, but move it 0.1 bpm either way and its off again.

Post Reply