In a DAW nowadays you’ve got several types of delay compensation, sometimes also called latency compensation. I would like to have a consensus on how to use these and their tiering.
The first one is the “automatic” delay compensation, which you can restrain via the button in the inspector, this disables all plugins in the signal path that REPORT (not cause) latency
The 2nd one is the midi delay in the inspector, where you can nudge midi alignment forward or backward.
The 3rd is the delay in ms adjustment in the Audio Connections window
The 4th one is where, on external fx, you can push the button to test the latency for the DA<>AD roundtrip
Then as a 5th one, there are 3rd PTY solutions like Innerclock, who not only give you a super tight clock pulse to external sequencers (this has no benefit when you sequence in Cubase) but also have global delay compensation to pull these sequencers on the grid with sample accuracy.
What are your best practices to set this up: Use the midi delay in the inspector? Use the delay in the Audio Connections?
I for example record all the synths individually with a rhythmic 4/4 and then measure the amount of samples late to the grid and put them in the audio connections delay.
In this case Cubase sequencing should be as tight on the grid as possible and external sequencing should be more of a clock jitter thing than a global midi delay thing.