Large latency selecting drum pattern

[staplesyrup]

When i select a different drum pattern from the bottom right section of push in drum rack mode, Live takes up to half a second to actually change to the new pattern. I cant find any quantization setting that could be causing this, and its the same delay at any tempo. Why the big delay? Its pretty hard to change patterns live in a rhythmic way, unless im misunderstanding something?

[Bin ray]

I should add the patterns change near instantly if the transport is not running. When transport running the pattern takes 250-500ms to change. Any ideas?

[Because789]

1. make sure Repeat is off
2. make sure Auto Select on the drum rack is off (see here for more infos: viewtopic.php?f=55&t=197026, viewtopic.php?f=55&t=197042)

[Bin ray]

1. make sure Repeat is off
2. make sure Auto Select on the drum rack is off (see here for more infos: viewtopic.php?f=55&t=197026, viewtopic.php?f=55&t=197042)

No, that makes no difference at all to the latency in selecting a new pattern.

Its the "loop length controls" section that I'm having trouble with. Bottom right of the drum rack layout. It seems like it would be a good idea to have quantize settings for changing the loop braces. Then you can jam with it in a musical way rather than this weird half second latency? Looking at the screen, the loop brace moves near instantly, but the pattern doesnt change until a bit later, like the graphics are prioritised over the sound..... Please tell me if I'm missing something basic.

[Bin ray]

OK, according to this post viewtopic.php?f=35&t=146837&hilit=+loop+length+quantize
from 2010, there was a problem with controlling the play position from M4L, which appeared to be a problem on Live's end.
It was first described here http://cycling74.com/forums/topic/move_ ... gers-clip/

In clip view I can see the playhead doesnt appear in the newly moved brace until about 3/16ths of a bar (at 120bpm) after the brace moves

This looks alot to me like the same problem with Live retriggering a clip (?) whenever the playhead is moved, unresolved 4 years later, but now built into Push as a feature that doesnt really work properly. Am I wrong?

[Because789]

Ah, obviously I completely misread your first post...

It seems to me that it's not a lag but the desired behaviour, I guess selecting a pattern is quantized to beats in the Push remote scripts.

Check out PXT-Live for the clip chopping you have in mind: http://www.youtube.com/watch?v=dUWE5qenqyM

[staplesyrup]

I fail to see how this can be desired behaviour. The loop brace changes after a fixed interval that doea not relate to BPM , or any quantization that I can understand, nor am i aware of any quantize settings for this behaviour. It seems to be simply a large latency of 250 - 500ms

I want to use the functionality of the drum rack note sequencer, not the PXT audio clip looper. It would be interesting if PXTs looper can do midi clips (does it?) as a proof the concept of live midi looping can work, but I am interested in having the loop brace controllable, musically, as part of the drum rack set up, having the sequencer and pads available too.

[Because789]

Checked it again and if you set the tempo to 20 bpm it looks like it's quantized to 1/16th notes. Since global quantization hasn't an effect on the pattern selection, I guess that quantization is hard coded in the Push midi remote scripts.

The PXT clip chopper works for midi clips, too. Since the script works in User mode, you'll always have the Push default mode available with one button press (User).

[Bin ray]

Checked it again and if you set the tempo to 20 bpm it looks like it's quantized to 1/16th notes. Since global quantization hasn't an effect on the pattern selection, I guess that quantization is hard coded in the Push midi remote scripts.

Do it at a higher BPM and you'll see that interval is not quantized. For instance at 120bpm I see (on the push itself) and hear the playhead shifting about 3/16ths late, at 240bpm its 6/16ths late, and at 20bpm its less than 1/32nd late. The playhead shifts with a fixed latency which doesn't relate to any quantization or BPM.

[Because789]

Try the following:

- set tempo to 200 and rhythm to 4/4
- make a two bar loop, each bar with different 1/16 hi-hats
- switch between the two patterns in a quarter beat rhythm (1/4 pattern 1, 2/4 pattern 2, 3/4 pattern 1...)

If I do that acurately I get smooth transitions between the patterns without any hearable shifting. Whitch leads me back to my initial guess that there's kind of a quarter note quantization going on which is not bound to the global quantization. An that means if you want to start a pattern with the first beat you have to hit the button between 4 and 1 (which works).

Since with a low tempo 1/16th play a role, it looks like the whole thing is dependent on the tempo.

[Bin ray]

if you want to start a pattern with the first beat you have to hit the button between 4 and 1 (which works).

Since with a low tempo 1/16th play a role, it looks like the whole thing is dependent on the tempo.

If I want to start a new loop on the first beat I have to trigger it at least 250ms before the playhead reaches 1, which isnt very useful for live manipulation. I can use midi clips with quantize off, set to legato to do nice fast changes between patterns but the loop brace selector is way off the money on my setup.

There is no quantization as far as I can tell. The audio might sound quantized but thats because the loop is quantized midi notes, not audio, you can't start halfway through a midi note. The latency seems constant to me.

Does anyone else have any ideas about this?