Midi timing suggestion on recorded VST instruments.

Como, I think it doesn’t matter which VST instrument - When a note is on the grid it plays back on the grid. Cubase compensation sees to that all instruments and audio channels plays back where it is placed.
I think Chases explanation was well put.

From seeing the OPs video and answering the part above highlighted. No. What the video shows is midi timing completely out of whack. Nothing to do with what played when. Looks like something is just putting the midi miles out and that just is not happening to many users.
However, other users (the other posters?) who see similar but lesser timing discrepancies might actually be seeing a different problem that might indicate that there may actually be a bug with the timings that is not apparent with purely musical Projects (ie: where the timing does not have to be critically accurate).
So far it doesn’t look to me to be clear where the problem lies, we just know it involves erroneous placing of midi notes.

Cubase hears a note on and places it according to when it arrived (post latency). Now if that note on is erroneously reported by the keyboard generating it or midi filters :bulb: blocking the midi sysex etc. you might find that a confused Cubase could get a bit arbitrary about note positioning.

^

OK …

Though we do know that users say they play ahead of the click in order to get the sound to emerge together with the click.

How about this? …

Suppose someone did another test, where - using the same PC, etc - they recorded in time with the click, but while listening to a sound unaffected by latency - eg using a keyboard instrument that outputs its own sounds.

And, further, suppose it were found - consistently, over several recordings - that (a) when recording using a zero-latency instrumental sound, the MIDI was recorded properly on the beat, but (b) when recording using an instrument with latency, the MIDI was logged too early.

It seems to me that that would provide quite strong evidence that the reason for the MIDI being recorded early was that the performer was anticipating the beat in order to make the sound come in time with the beat.

(While doing the recording using a zero-latency sound, the VSTi ought to be still operational in Cubase, so as to eliminate its presence/absence as part of the cause of any observed effect.)

An all to likely possibility! :smiley:

And certainly not the first and inevitably not for the last time.

I’ll try to watch it a couple of more times, too.

However, nothing can convince me that recording midi while playing along with what you hear through Cubase, will not have PDC effect the placement of the recorded midi data … exactly as changes in overall latency does.

Otherwise, why would all who have looked discover that lower latency improves recorded midi timing?

And, early placement of the recorded midi notes is exactly what I am predicting, no?

Como

Yes, Experts, please come forth and tell us the formula.

And to make it even more complicated the distance varies a few ms from each note, but all notes and data has in common it’s always early. Even when recording with low latency. I get back to what we discussed before… Could it work like the midi data actually was “translated”… to be put at the actual sound start / start of reacting on cc?

Someone might say… Well keep the latency extremely low, and you’ll hardly notice. I don’t like that argument. because…
For the first, we often raise the latency a bit when having large projects, to let the system breath.
For the second, Yeah! we could do with 23 bit audio instead of 24 because no one would ever notice.
For the third, the workflow and the overall musical feeling would be improved.

Then it would work fine! You got that!

:smiley:
The ironic thing about this statement is that, you say the notes are completely out of whack! I agree about that.
But something didn’t just put them out of whack! I played them out of whack to make the sound come right!

Yes, asked and answered. It could work, but it would create other potential issues when the instrument used to record the original track was changed or the total PDC of the project was changed later.

  1. Midi timing problem one: the notes are always recorded early to what some people think they should be, but to others properly when the effect of audio latency and PDC are taken into account.

  2. Midi timing problem two: recorded midi notes show variability of relative placement on the grid, independent of all appearing early. A true problem, difficult to trouble shoot between midi keyboard behavior, midi connection behavior through USB or an Audio Interface with standard midi cable connections and the sequencer behavior. The only way to tell would be to use the identical first two components (Yes, Dorothy, there is variability even between identical units produced by the same manufacturer.), recorded into different DAWs and then compare the recorded data.

  3. If you confuse the two issues, you will never sort things out.

BTW, has anyone … which I have not … taken the time to see if this outcome of early recorded midi data is different in any other DAW?

Como



Asked and answered… Fair enough, but the potential issues that you’re naming, aren’t they out of the blue, really? Cubase handles so that all midi notes play back at the note start, no matter how many plugins or what kind of instrument. It’s the delay compensation, which works really well.

Believe me. that should not happen. If you play them out of whack they should stay out of whack if unquantised as you say.
What’s going on here? What’s the point of this post?

Everyone’s assumptions here can be out of the blue as we look for solutions.

PDC is not a ‘fixed’ entity or value. It is determined in real time by Cubase measuring the longest delay among the audio chains presented to the master bus. Which ever chain is longest will determine the PDC at that moment in time.

What makes audio chains longer or shorter is the accumulated additional latency introduced by any plugin.

Which ever chain is longest will be the PDC value and all other measured chains will be adjusted accordingly based upon their individual total latency.

So, if you add to any chain another plugin which makes that chain now longer and the longest, the PDC will change … and how your midi data will playback in the project due to the change in PDC will also change.

You may be right, that this could be ‘set and forget’ on the first recording adjustment, but I’m not certain how subsequent changes in PDC will affect playback.

I’d still love to hear from someone knowledgeable about this issue in other DAWS: whether this ‘notes recorded early’ behavior occurs, and if it does not, whether the DAW has some built in compensation to adjust the position of the recorded midi data in a track. As far as I can see, the note must always arrive before the instrument sounds. Based upon my understanding of how Cubase works, the note goes through the track midi output to trigger the instrument. I must imagine the passage results in the midi data being recorded as it passes from the track midi input to the output … before it reaches the responding instrument.

Como
Como

Hi,

I’m new to Cubase, I come from Logic.

I never had to edit the midi data when recording with latency in Logic. Neither in Pro Tools. We tend to adjust our playing to what we ear so I’m not sure if I was playing earlier or not. But I know that I never had to nudge the data to the right to ear it in sync with the other tracks.

Is this a new behavior in Cubase 6.5? Should this be considered as a bug?

I usually work with dense orchestral arrangement with multiples VSTi so I need to be able to record with latency.

Did you ever have occasion to visually inspect the recorded midi in those other DAWs to see how the data lined up on the grid?

Or, did it just sound right and you never bothered?

Como

Did you ever have occasion to visually inspect the recorded midi in those other DAWs to see how the data lined up on the grid?

Or, did it just sound right and you never bothered?

Like you said, it just sound right and I never bothered.
I never had to think about it, playback was the same as what I heard when recording. So I really think I was adjusting my playing to be on time.
That’s the problem I have now: What I ear on playback is different of what I ear while performing recording.

Then we’re on to something and I may have a great banquet of crow to eat. :astonished:

Como

I just made a quick test.

I record a quick piano track with a click and 1024 buffer size using Kontakt in Vienna Ensemble. Trying to sound in time with the click.

I found that Logic and Cubase is not responding the same…

In Logic, midi notes are aligned on the grid and playing in time with the click.
In Cubase, notes are to much at left on the grid and playing too early.

Two DAW, different behavior. There’s definitely something wrong here.
(I will try it without using Vienna Ensemble soon.)

Do I have to eat the feathers, too, or can I pluck first?

Como

It might be an idea to download demo’s of other DAW’s and see if they have the same problem or not. Than it would be easy to tell if its a midi or software problem.

guys, the human factor trying to compensate etc has already been dismissed. There was a user some posts ago saying he recorded form a synnc external arpeggiator and the notes were recorded early too. I can see why you say that, and this in fact happens to some degree but with higher latencies the offset is just too much to be us compensating “naturally” (like in the video).

In my computer, Ableton Live records MIDI correctly. Its a Cubase problem. We just shouldnt be afraid to admit that.

And “it might be an idea to” read the thread before posting. :wink:

For instance, look at the post just two above yours.