MIDI delay recording

[Amaury]

Amaury,

I will do your test tonight with my Axiom 49, sorry I didn't have time yesterday. However I tried it with a computer keyboard. Man, I cannot help but hit the keys to the metronome, not to the sound I heard (the Impulse All purpose kick drum). Without the metronome I noticed, I was compensating to have the kick sound come in time with the imaginary beat in my head.

As for your favorite pianist, was he playing solo or with an orchestra? It seems, it makes a big difference.

Hi,

Thanks for your input. I undersand what you're saying. It helps to know how people work. Have you tried recording music on top of pre-recorded beats, music? Do you listen to what you are playing, or concentrating on your fingers?

About my pianist? Well, it is for every pianist in the world the same. Solo or with the orchestra, theinstrument is the same. So when with an orchestra, the playing adapts. But when playing solo too, because firstly the latency varies, and the pianist need to listen to what he does, not look at his finger.

Regards,
Amaury

[michkhol]

Hi,

Thanks for your input. I undersand what you're saying. It helps to know how people work. Have you tried recording music on top of pre-recorded beats, music? Do you listen to what you are playing, or concentrating on your fingers?

About my pianist? Well, it is for every pianist in the world the same. Solo or with the orchestra, theinstrument is the same. So when with an orchestra, the playing adapts. But when playing solo too, because firstly the latency varies, and the pianist need to listen to what he does, not look at his finger.

Regards,
Amaury

Amaury,

I did your test yesterday. First I used a drum kit with latency (compressor II). With monitor ON the notes were all over the place (some late, some early, some on grid) and playback to the metronome was not comfortable. After I switched to the less latent kit, the notes were much closer to the grid. Still when I recorded a clip with monitor OFF and then compared it to the prerecorded drum loop, the beat matching was better. I don't know, may be it is just me (I'm not a drummer) and other people do it differently.
As for your question, I concentrate on my fingers and expect the sound to come out in sync. If it doesn't I'm confused.

[Amaury]

Hi,

Thanks for your input. I undersand what you're saying. It helps to know how people work. Have you tried recording music on top of pre-recorded beats, music? Do you listen to what you are playing, or concentrating on your fingers?

About my pianist? Well, it is for every pianist in the world the same. Solo or with the orchestra, theinstrument is the same. So when with an orchestra, the playing adapts. But when playing solo too, because firstly the latency varies, and the pianist need to listen to what he does, not look at his finger.

Regards,
Amaury

Hi,

Thanks for your detailed message. True, the main goal, appart from the discussion here, is to reduce the latency as much as possible. It IS unconfortable to play with latency, and it certainly takes some skills to be able to play anything sensefull while experiencing latency. It is far easier to record in 'deaf' mode, with monitor set to off.

Regards,
Amaury

Amaury,

I did your test yesterday. First I used a drum kit with latency (compressor II). With monitor ON the notes were all over the place (some late, some early, some on grid) and playback to the metronome was not comfortable. After I switched to the less latent kit, the notes were much closer to the grid. Still when I recorded a clip with monitor OFF and then compared it to the prerecorded drum loop, the beat matching was better. I don't know, may be it is just me (I'm not a drummer) and other people do it differently.
As for your question, I concentrate on my fingers and expect the sound to come out in sync. If it doesn't I'm confused.

[bry2k]

Amaury my friend...you talk too much. With all due respect, you have beaten this subject to a bloody pulp waxing philosophical about how you think people play an instrument ahead of time to deal with latencies. I had to take a break from it.

Bottom line: I agree with everything popslut wrote. His most eloquently stated point was the one about how all musicians are used to always hearing the sound later than the instrument, since sound necessarily takes time to travel to the ears only AFTER the note has been phsyically played/plucked/struck on the instrument. That is exactly the point I was trying to make.

Furthermore amaury, you seem to neglect to take into consideration the concept of groove and the predictive nature of a musician. If I hear 1 2 3 4 at regular intervals, you can turn the speakers off and I can play along for a few bars, turn the speakers back on, and I will still be in time (if I'm on top of my game that day).

All I care is that the note gets recorded on the timeline exactly where I played it relative to the other instruments. Playback delay compensation is a *separate* issue.

Having thought about all Amaury wrote for the last several pages, let me add this thought:

Live should NOT be trying to compensate for the latency of the musician, which is essentially what Amaury has been suggesting.

Live should only be compensating for the actual computational latency of the DAW. Host latency. Softsynth latency. FX latency. Etc. All added together. Nothing else. That is what all other DAWs do.

Pro Tools is the Bible for how to do it right. Don't reinvent the wheel. Just copy what Digidesign did.

-Bryan

PS - Last thing re musicians playing ahead. Lee Sklar, world renowned bass player, whom I've recorded many times...he will consistently pluck his bass notes about 40ms after the attack of a kick drum, which looks wrong when you zoom in on it in a DAW, but it sounds perfect, because the attack of the kick cuts through the mix before the bass guitar frequencies come along to step all over it. Analyzing how Lee plays with your ears and not your eyes is essential. Also, Lee just barely wants to hear himself while he records (as is the case with most other good bassists/guitar players). He just wants to feel the groove of the drums he is listening to and rip apart his bass strings in sync with them. His timing is executed: AT. HIS. FINGERS. He is not responding to the sound of his instrument, or waiting to hear the last bass note he played strike his ears in order to decide when to pluck the next note. He is responding to the other instruments he is playing along with. This is typical of all the really good session players I know. They do not play ahead of the groove to compensate for latency.

PPS - Ok, I'm done with this thread.

[bry2k]

About my pianist? Well, it is for every pianist in the world the same. Solo or with the orchestra, theinstrument is the same. So when with an orchestra, the playing adapts. But when playing solo too, because firstly the latency varies, and the pianist need to listen to what he does, not look at his finger.

Regards,
Amaury

Ok, I lied. One other comment: Amaury, I would disagree with your point about pianists as well to some degree. There is instantaneous tactile feedback when playing a piano (or any instrument), and great piano players respond as much to the feel of the keys & hammers and the vibrations of the instrument at their fingertips as they do to the sounds it produces at their ears.

Cheers,
Bryan

[Amaury]

Amaury my friend...you talk too much. With all due respect, you have beaten this subject to a bloody pulp waxing philosophical about how you think people play an instrument ahead of time to deal with latencies. I had to take a break from it.

Bottom line: I agree with everything popslut wrote. His most eloquently stated point was the one about how all musicians are used to always hearing the sound later than the instrument, since sound necessarily takes time to travel to the ears only AFTER the note has been phsyically played/plucked/struck on the instrument. That is exactly the point I was trying to make.

Furthermore amaury, you seem to neglect to take into consideration the concept of groove and the predictive nature of a musician. If I hear 1 2 3 4 at regular intervals, you can turn the speakers off and I can play along for a few bars, turn the speakers back on, and I will still be in time (if I'm on top of my game that day).

All I care is that the note gets recorded on the timeline exactly where I played it relative to the other instruments. Playback delay compensation is a *separate* issue.

Having thought about all Amaury wrote for the last several pages, let me add this thought:

Live should NOT be trying to compensate for the latency of the musician, which is essentially what Amaury has been suggesting.

Live should only be compensating for the actual computational latency of the DAW. Host latency. Softsynth latency. FX latency. Etc. All added together. Nothing else. That is what all other DAWs do.

Pro Tools is the Bible for how to do it right. Don't reinvent the wheel. Just copy what Digidesign did.

-Bryan

PS - Last thing re musicians playing ahead. Lee Sklar, world renowned bass player, whom I've recorded many times...he will consistently pluck his bass notes about 40ms after the attack of a kick drum, which looks wrong when you zoom in on it in a DAW, but it sounds perfect, because the attack of the kick cuts through the mix before the bass guitar frequencies come along to step all over it. Analyzing how Lee plays with your ears and not your eyes is essential. Also, Lee just barely wants to hear himself while he records (as is the case with most other good bassists/guitar players). He just wants to feel the groove of the drums he is listening to and rip apart his bass strings in sync with them. His timing is executed: AT. HIS. FINGERS. He is not responding to the sound of his instrument, or waiting to hear the last bass note he played strike his ears in order to decide when to pluck the next note. He is responding to the other instruments he is playing along with. This is typical of all the really good session players I know. They do not play ahead of the groove to compensate for latency.

PPS - Ok, I'm done with this thread.

Hi,

You're certainly right, I may speak too much (write, in that case).
I was only wanting to make sure of this:

-As it is, in Live, what you hear during recording with monitor IN, is what you get while playing back, after recording.

Now, I understood that it is NOT what you want. Thanks for the input.

Regards,
Amaury

[iain.morland]

Hi again,

I was looking at this again yesterday and have a numbskull question:

If I use track delay to add a negative delay to a MIDI track that was recorded with monitoring on, and if that negative delay is the same as the output latency of my soundcard - as reported on the audio options panel - will the track play as if it had been recorded with monitoring off?

To be clear: I'm not talking about adding the negative track delay while recording, merely while playing back.

Thanks
iain

[Amaury]

Hi again,

I was looking at this again yesterday and have a numbskull question:

If I use track delay to add a negative delay to a MIDI track that was recorded with monitoring on, and if that negative delay is the same as the output latency of my soundcard - as reported on the audio options panel - will the track play as if it had been recorded with monitoring off?

To be clear: I'm not talking about adding the negative track delay while recording, merely while playing back.

Thanks
iain

Hi,

Sure. If the track contains no devices, your case will work. Otherwise, you'll have to add the devices delay (the devices on the track you are monitoring, and the master track).
It all depends on your use case: what are you trying to achieve?

Any comment about 'what you hear is what you get?'

Regards,
Amaury

[bry2k]

I was only wanting to make sure of this:

-As it is, in Live, what you hear during recording with monitor IN, is what you get while playing back, after recording.

Now, I understood that it is NOT what you want. Thanks for the input.

Regards,
Amaury

I have absolutely no idea how you came to that conclusion.

I don't know how else to explain what I've already said. Hopefully Ableton will examine the issues we've discussed here and adopt the methods other DAWs such as Pro Tools use to handle latency and delay compensation.

Thanks,
Bryan

[Amaury]

I was only wanting to make sure of this:

-As it is, in Live, what you hear during recording with monitor IN, is what you get while playing back, after recording.

Now, I understood that it is NOT what you want. Thanks for the input.

Regards,
Amaury

I have absolutely no idea how you came to that conclusion.

I don't know how else to explain what I've already said. Hopefully Ableton will examine the issues we've discussed here and adopt the methods other DAWs such as Pro Tools use to handle latency and delay compensation.

Thanks,
Bryan

Hi,

I came to that conclusion, because Live's way of recording MIDI (we are talking here of a monitored track containing a soft synth and eventually some effects), ensures that what you hear while you monitor the track is what you'll hear when playing back what you've recorded. That is with Delay compensation ON of course.

Pro Tools does not feature Delay compensation but in Pro Tools HD, so the problem is different.

Cubase does not let you hear the same thing between the recording and the play back of what you've recorded. Their only strategy is to reduce the latency as much as possible so the problem is not noticed. But if you, on purpose or because you have no choice, use some devices that add latency to the track and record something while monitoring, you'll hear something different on playing back.

Thus my conclusion. It's only a way to phrase it, I want to make sure we know what we are talking about.

Regards,
Amaury

[iain.morland]

Hi,

Sure. If the track contains no devices, your case will work. Otherwise, you'll have to add the devices delay (the devices on the track you are monitoring, and the master track).
It all depends on your use case: what are you trying to achieve?

Any comment about 'what you hear is what you get?'

Regards,
Amaury

Ok, I think I'm *nearly* understanding this - but the business about the device delay is a bit confusing.

Isn't one problem of the 'what you hear is what you get' method that when devices on a track are changed, you no longer get what you heard?

On the other hand, isn't the point of PDC (as opposed to the 'what you hear is what you get' feature) that changing the devices wouldn't change the timing of any MIDI playing back on that track?

With regards to where to go from here, can I suggest a very straightforward change to how Live operates (drawing on other people's suggestions): simply add an option for the user to choose between: 'always record what you hear' or 'always record what you play' or 'ask me every time'. Please pass this suggestion on to your programmers!

[Amaury]

Ok, I think I'm *nearly* understanding this - but the business about the device delay is a bit confusing.

Isn't one problem of the 'what you hear is what you get' method that when devices on a track are changed, you no longer get what you heard?

On the other hand, isn't the point of PDC (as opposed to the 'what you hear is what you get' feature) that changing the devices wouldn't change the timing of any MIDI playing back on that track?

With regards to where to go from here, can I suggest a very straightforward change to how Live operates (drawing on other people's suggestions): simply add an option for the user to choose between: 'always record what you hear' or 'always record what you play' or 'ask me every time'. Please pass this suggestion on to your programmers!

Hi,

I know it's a bit hard to grasp, but no, changing the devices after the fact does not change a thing: the sound plays where the note is, regardless of the devices and their latencies (with Delay compensation ON).

So, in our case, while recording, the sound comes later that the strike of the note. The note is recorded where the sound is heard. On playing back, the sound plays where the note has been recorded, where the sound was heard.

We're now well aware off what some users seem to want, or to be used to.

Regards,
Amaury

[popslut]

With regards to where to go from here, can I suggest a very straightforward change to how Live operates (drawing on other people's suggestions): simply add an option for the user to choose between: 'always record what you hear' or 'always record what you play'...

Precisely.

[nowtime]

Hi - I am a drummer and a keyboardist. I have experimented as thouroughly as possible with Battery and Impulse. Live is the only sequencer I know. I have given up trying to play drum grooves on my keyboard into Live.

Although it not too diffuclut to get accurate parts recorded by playing early, i have never been happy with the results. I am talking about the GROOVE FACTOR. It just never sounds great. If I record drumming from a no-latency monitored keyboard it's phat. But if I try to use all these fantastic drum kits from Battery or impulse, forget it. It's useless.

So I'm learning MIDI drum programming! It's very exciting!

If you have any specific questions for me feel free to ask.

[iain.morland]

I know it's a bit hard to grasp, but no, changing the devices after the fact does not change a thing: the sound plays where the note is, regardless of the devices and their latencies (with Delay compensation ON).

Ok, thanks for your response, but isn't it the case that the sound of the note (while I'm performing) is delayed by the devices, because PDC operates only on playback?

Therefore isn't the recording of the note (since it's recorded where it's heard) affected by the devices? (you said earlier that "you'll have to add the devices delay (the devices on the track you are monitoring, and the master track)").

My brain hurts again now. Seriously, if an option could be added simply to turn this on or off, so I could experiment more easily, then I'd upgrade from Live 5 to 6 in an instant.