bry2k wrote:Wow. I found this thread very interesting. Glad I found it.
I don't know why Ableton finds this such a difficult issue to resolve.
In Pro Tools, it is very simple. There is a setting called "Auto-compensate for delay after record pass."
What is means is simply that, while you are recording, and monitoring with some amount of latency, you are responding to what you hear, which is later in time than the current playback position of the timeline.
No problem. When you hit stop after done recording, the track you just recorded, whether it is audio or MIDI, is automatically shifted backward by the amount of the total delay compensation.
So for example, using big numbers to make the point obvious:
1) You have a hi-hat penciled in to use as a metronome 1 & 2 & 3 & 4 &
2) You have latency of lets say 1 second (1000ms) as an extreme example.
3) You hit play. Pro Tools plays beat 1. One second later you hear it and play a kick drum on your MIDI keyboard. The track you are recording on has delay compensation turned off (automatically in Pro Tools) so you here what you are playing instantly. But the timeline is one second past what you are hearing, so your performed note is going to be recorded one second later in the timeline that it should be.
4) No problem! Because when you hit stop, Pro Tools automatically shifts the whole recorded region (clip) backwards in time 1 second. Or 11.5ms, or 1743 samples, or whatever the total latency of host and plugins is. So the end result is that you both hear what you are playing correctly (because delay comp is off for the track you are recording on), and your notes are placed where you heard them, because Pro Tools auto-compensates AFTER the record pass.
Therefore no laws of physics are violated as Amaury has suggested. And everything works perfectly.
Of course there are real world caveats to the above example, such as if you have a bunch of plugins on your master fader that induce delay to the outputs you are monitoring on (such as a mix bus compressor/EQ with high latency) then you are going to have a hard time monitoring your performance while recording it because there will be a delay between what you play and what you hear. The workaround is to minimize total latency during recording by deactivating plugins or shutting off delay compensation on tracks that add a lot of latency. In Pro Tools, delay compensation can be easily turned on/off globally or on a track by track basis.
Basically, Ableton has boxed themselves in with some very poor methods of dealing with delay compensation that they will have to completely re-work for it to ever work intuitively and as we expect. I suggest they review how Digidesign has implemented it in Pro Tools, which is as perfect a solution as can be implemented.
But to summarize, the simple fix for the primary problem discussed in this thread is for Live to auto-compensate for latency after the record pass is completed. MIDI notes should always be placed on the timeline where you heard them during the performance. I was kind of surprised after reading all 10 pages of this thread that nobody at Ableton has replied that they understand this.
-Bryan
Hi,
So, from your description, as I understand it, Pro tools would let you hear your performance as you play it, and then would shift the data to a place you never intended to play? I doubt it is the case, but I'm not the one in the house who studies all of this.
The goal is to provide something that:
-let's your hear something meaningfull, e.g without latency or with a minimum latency
-leave it as you heard it so you can trust what you play
-or if you do not monitor via the software, record without latency as it is the case right now. (read our lessons about driver error compensation, it shows how something is recorded with no latency at all, though it is audio)
Then another point I can't believe: if you have a latency of, say, 1 second as in your example, it is there both for the input and for the output, that is so. So, if you play and monitor a soft synthesizer for instance, with no other effects on the track or on the master track, you'll have the latency of one second for the soft synth audio out needs that time to reach the speakers.
As you noticed, mainly the problem arises when you put more and more effects on the track you are recording to or on the Master track. Different software use different techniques to shut down the effects when you monitor, or ignore the problem etc... and we are also aware of that.
The main problem we are focussing on in that thread is pure MIDI recording though, even before talking of soft synthesizer. The use of soft synth is a step forward of pure MIDI recording, so it is included in the problem. But we need to concentrate on the core of MIDI recording, where a lot of things are invlolved.
For instance, Digidesign only works with its own devices, and now M-audio devices, but I bet it developped its own drivers for the M-audio devices too. The biggest problem, and I don't want to go too deep into details, as I'm not the most skillfull person (with is good for all of us
), is a triangular relationship between the devices drivers, the computer operating system, and the sequencer.
Anyway thanks for your input,
Regards,
Amaury
I asked exactly that question.
