Can I sync Live to Augustus Looper?

[Axl]

Hi,

I've been trying to setup Ableton Live 7 as a live looper. My main obstacle is Live's inability to set its tempo according to the first recorded loop making me dependent on some kind of a click track.

My question is: Can I slave Live to Augustus looper's tempo clock via a midi loop so that I can determine its tempo by recording into the looper? The manual doesn't say anything about it although I think this would be a killer feature worth highlighting.

Thanks in advance,
Axl

[Axl]

To answer my own question: yes, it is possible. I've had some issues though.

First, Live didn't like Augustus's evaluation time running out while slaved to it. This, I think, won't happen anymore once I have a licence.

The other thing was more serious. It has happened several times now, that clicking the tap rec button resulted in a 0.0 length loop. I really don't think it was caused by accidentally double clicking.

Has anyone had success with a setup like this?

Axl

PS: OS X 10.4. on a 2nd gen. MBP

[os]

The first answer is yes, as you've found. NB you might want to use the "rounded tempos" option, as described in the manual.

I can't think how you'd end up with a zero length loop. Where you actually clicking the button in the Augustus Loop GUI, or triggering the recording via MIDI (or OSC)? Or using Live's default plug-in parameter interface? Was it the first or second press of the button that behaved in a way you didn't expect? The more precisely you can describe what you did and what happened the better.

cheers,
os.

[zerocrossing]

There's a fellow named Per Boysen who hangs out in the Looper's Delight list that has had success with this I think. Might want to do a search on their archives. The list can be useful sometimes too, especially his posts. He's a super great guy that's helpful and I love his music as well.

Me, I go the other way as I seem to remember the technique wasn't worth the reward as ultimately I'm fairly sure I want something that's "X" bars long before I start anyway.

[james0tucson]

There's a fellow named Per Boysen who hangs out in the Looper's Delight list that has had success with this I think. Might want to do a search on their archives. The list can be useful sometimes too, especially his posts. He's a super great guy that's helpful and I love his music as well.

Me, I go the other way as I seem to remember the technique wasn't worth the reward as ultimately I'm fairly sure I want something that's "X" bars long before I start anyway.

One of the reasons I want a JamMan is because I often don't know what I want until I've already played it. I've personally been hoping for a looper pedal emulation that can work with, say, my FCB1010, and provide the same "drooling drunk" kind of simplicity. Then I read threads like this and I get scared to death. If a looper pedal is more complex to use than "Stomp on it, groove on your guitar, stomp on it again", it's a failure.

[os]

I'd like to think Augustus Loop can be that easy to use - it's basically pretending to be a loop of tape, after all. It's just that it has an awful lot of other stuff to it that you can use *if* you want to.

Try it, if you've not already - I'd love to hear your feedback on how I can make it easier to pick up.

[Axl]

I get the above mentioned behaviour after I've recorded a loop and Live is synced to it. All is fine until I reset the state var and start a new loop. The first tapRec tap always results in a very short loop (a couple of samples) and the Midi Clock tempo shoots up to some absurd value. Can you reproduce that? I'm basically using your own example script. I would strongly suggest you build in some kind of reasonable maximum value for the Midi Clock. I had to restart Live a couple of times now.

[os]

Can you send me the exact script you're using now, just so I'm completely sure I know what you're doing?

[os]

Thanks for sending the script. It works fine for me.

Can you download e.g. MIDIMonitor from http://www.snoize.com/MIDIMonitor/ and watch the CC messages being sent from your controller - specifically, check it's not sending out a double message (e.g. one when you press the button and another when you release it).

[Axl]

Thanks for sending the script. It works fine for me.

Can you download e.g. MIDIMonitor from http://www.snoize.com/MIDIMonitor/ and watch the CC messages being sent from your controller - specifically, check it's not sending out a double message (e.g. one when you press the button and another when you release it).

I'm monitoring my midi messages with MidiPipe and everything is as it should be. I've also experienced the problem when I mouseclicked the buttons, so it doesn't have anything to do with the midi messaging. I'll try to give you more detailed information tonight.

[Axl]

I've made some screenshots of what's happening.

Here is my Live set before I start looping:



Live happily syncronized to Augustus:



After I've clicked the state reset and tapRec:



Can you see something I'm doing wrong?

[/img]

[os]

The MIDI appears to be coming from the IAC driver. What's on the other end of that connection? Is there a reason you're not sending MIDI directly from your controller to Live?

After I've clicked the state reset and tapRec:

Can we dig into that a bit? By "clicked the state reset" you mean you sent MIDI CC 126 to trigger the second event handler in the script? And then are you pressing tapRec in the GUI or via MIDI somehow?

Are you waiting for the Clear Loop to finish before you try to do anything else?

If you like we could do a Remote Desktop session so I can see your desktop directly - maybe that would get to the bottom of this faster.

[Axl]

The MIDI appears to be coming from the IAC driver. What's on the other end of that connection? Is there a reason you're not sending MIDI directly from your controller to Live?

I use an FCB1010 foot controller. I need MidiPipe because I also want to access some Mackie Control features from the controller. So I have to route the corresponding Midi to a dedicated iac bus. Anyway, for testing purposes I've taken MidiPipe out of the chain, but the results are the same.

Can we dig into that a bit? By "clicked the state reset" you mean you sent MIDI CC 126 to trigger the second event handler in the script? And then are you pressing tapRec in the GUI or via MIDI somehow?

What I do is: send CC 126, wait for the Clear Loop button to turn off and then send CC 125. After that, nothing works anymore until I click Pause Clock.

If you like we could do a Remote Desktop session so I can see your desktop directly - maybe that would get to the bottom of this faster.

I've never done anything like that. I've googled it and it sounds a little scary. How would we do this? Would I have to give you admin rights for my computer?

[os]

What I do is: send CC 126, wait for the Clear Loop button to turn off and then send CC 125. After that, nothing works anymore until I click Pause Clock.

Could you put a print in the handleEvent1 function so we can see exactly what's being called and when? e.g.

Code: Select all

local function handleEvent1( channel, cc, value )
	print( channel, cc, value, state )
	if state == 0 then

Something must be setting TapRecord back to zero after handleEvent1 sets it to one. Either handleEvent1 is being called again, or something else entirely is happening e.g. some other parameter automation or remote control in Live.

Is it possible that you've recorded some parameter automation in Live and this is controlling TapRecord?

I've never done anything like that. I've googled it and it sounds a little scary. How would we do this? Would I have to give you admin rights for my computer?

If you're running Leopard, we just use iChat, and that's it.

[Axl]

I did some more testing and disabled all Midi ports in Live except for the one the foot controller is sending to. It seemed to work when clearing and restarting the loop for the first time. But I remember on rare occasions it has worked before. On the second attempt I again had bad results.

I also put the print() in the handleEvent1 function as you suggested but found no sending of double events whatsoever.

Then I fiddled around a bit with the dropdown menus of the clock section and while setting the 'start' bit to 'none' didn't help, doing this with the 'stop' thing does the trick. At least, I haven't had a problem since I tried it some 10 minutes ago.

Now what do you make of it? As far as I can see the problem must lie somewhere in Augustus. I've excluded all possible external sources of error I could think of. By the way, I don't understand what the dropdown menus do. The manual doesn't explain that.

I'm still running tiger.

Greets,
Axl