I said I was done with this thread ... but, I cannot help myself from jumping back in.
Conman, I might be mistaken, but it seems to me that part of the reason for the friction that's just arisen here is that your posts have seemed to ignore the OP's notion of what's going on. So ...
1) Do you actually agree that a possible - or, perhaps, even a likely - reason for the MIDI notes being recorded early is that they were played early? --- played early (probably, without conscious thought) to make their sound come in time with what's being heard (whether existing tracks or the Cubase click).
-- That explanation is what the OP and some others in this thread favour.
This must be the case, as it is the only way you can catch up with the beat of what you hear played through Cubase with latency. The playing early includes both midi latency adjustment and audio latency, although the former will be masked by the latter, since it will generally be a smaller value. That is, they are not additive, but parallel latency issues, and playing will simply adjust to the highest, or determinant, source of latency.
chase wrote:2) Suppose, for the sake of argument, that that explanation turned out to be correct. Would you then agree that it would be a useful option in Cubase to have the placement of the MIDI notes automatically delayed by however many milliseconds early the performer had to play the notes in order to get the VI's sound to align with what was being heard? (It seems that Cubase ought to "know" how many milliseconds that would be.)
This is where I continue to think there is misunderstanding. If you previously altered you playing to make the played instrument remain on the beat as heard ... not as seen, if you move the notes forward, i.e., towards the right, and leave everything else in a project as it was, don't you think the notes must now be heard late, even if they appear to be properly synced to the grid?
Think about this. We all understand tha Cubase uses Plugin Delay Compensation, based upon the highest latency plugin instantiated in a project. How often is that high latency plugin a VSTi? If it is, and you change that out for another instrument that is no longer the highest PCD plugin in the project, your timing of that playback will be different than in the first case.
Cubase, in my view, properly syncs the recording of a VSTi within a specific project. The midi file reflects playing that instrument in that context. Anything you subsequently do to change the projects PDC, increasing it or lowering it, will change the timing of the specific recorded midi track.
How could it possibly be any different?
If an adjustment were made for moving the instantly recorded midi track forward according to PDC of the project this would happen:
1. The specific VSTi would no longer play back in the project as you recorded it ... it would be heard late on the beat.
2. If you changed the PDC of the project by removing the high latency plugin or adding a higher one, the track would also no longer be in sync with the audio. (This is an issue in its own right, possibly leading to some thinking the software did something wrong when they change a VSTi and it doesn't seem to playback the same.)
3. This suggests that you probably should start a project and do all your VSTi at a stage and then render or record by internal summing all your VSTi and be done with it. Now you can be sure any tracks are in sync before you start changing timing with use of plugins in mixing.
4. Finally, as I tried to explain above, none of this has anything to do with the midi timing issues related to variability of when notes are recorded and/ or played back due to midi buffer issues.
Win7x64 C6-32 Intel MOBO Dual Dualcore Xeon 3ghz Processors - RME Multiface II PCIe - Echo Layla24 - Mackie 824 Monitors - Mackie Controller Universal Pro - Mackie Big Knob - 2 UAD2 Quads - Novation SL 61 MK II - StudioLogic 880 Pro - Roland XV5080 - Korg Radias R - Emu CS PK-7 - MOTU Midi Timepiece AV Parallel - A Designs Pacifica - Mackie Onyx 800R - Mics - Controllers - Software - Stuff