Ableton Forum

Is it not there to compensate for effect plugin latency ?

Lets say Vintage Warmer says it's internal latency is 49ms on your machine. You then advance that track by 49 ms ahead of the rest. With a lot of other plugs you have to work this out yourself.

That's my understanding anyway.

Yep, why not.

Back to the original question:
I'm still waiting for one of the developers to really explain how this functionality is implemented

UKRuss wrote:

Negative values bring the whole track infront of the beat and positive push the track output behind the beat

I could see how that might be the case with one track. But the behavior changes depending on how many tracks are affected by individual delay-times, and in what relation this tracks are to each other (route them in chain, or bus them and the result is sort of unpredictable.)

Max.

I believe what happens is that if you bring track forward in time you are essentially just delaying all the other tracks by that amount.

I believe what happens is that if you bring track forward in time you are essentially just delaying all the other tracks by that amount.

Right. But that's still just one of the most basic scenarios. I'd like to know several more things:
What happens when you set these delays before recording into the track, compared to setting them after?
I'd also like to understand what governs the implementation of delay. Say for example you have several MIDI tracks routed to one "bus" track, and delay that track by 50ms. Additionally, you have three audio tracks, one of which you pre-delay by 15ms. Now:
One result could be that all MIDI tracks get delayed by 65ms, all other tracks get delayed by 15ms, and the pre-delayed track stays "in place". While we're at it, what happens to the click track?

Max.

For the 3-year anniversary of this post I'd like to bump it one more time to see if anyone with real inside knowledge of how Ableton Live handles the track delay function can explain how it is implemented.

I'd especially like to understand how this function adds to latency when processing the signal in real time.

Here's one example of a situation:
I'm looping and processing a drum set in a live show.
Let's say I have 40 milliseconds of compound (round-trip) latency on my setup.
I'm trying to lay down a drum groove (with the click in my ear, live on stage) to sync up with 4 already playing MIDI tracks, which are bused.
Should I:
(a) "pre-delay" the drum-loop track by 40ms ("-40")? or
(b) delay the midi bus track by 40ms? or
(c) delay each individual midi track by 40ms? or
(d) do something entirely different to get things in sync..?

I've tried all these things and have gotten varying results. Most noticeably, several of these methods have added to the actual overall latency.

As noted before, the manual is thin on the technical side of this matter.

thanks,
Max.

for real....why is it so hard to get an answer on this.

Someone has to know.

I'm interested in how the delay setting effects incoming recorded audio.

Example....I have a track set to receive audio from my Microkorg. After recording I realize the audio is 10 ms late from where I would like it to be. Should I set the track to -10, or +10.......

Say now i adjust the track delay to get the audio recording where I like it (not 10ms late anymore).....so now, do I need to take the track delay off to get it in sync with all other midi tracks?

Yeah yeah, use your ears...blah. A technical answer would be most relevant. Thanks!

Max Oepen wrote:the 3-year anniversary of this post

I don't know the answer but I'm sure that an email to support would de-mystify the issue for you.

post up their response if you get one