It’s basically ALL i use, unless I’m recording audio.
There’s a bug in C15, on my end, where the buffer either puts out nothing and just a sliver of a MIDI event… or it actually puts the notes played but the MiDI event start point isn’t correct. It’s early.
The problem seems to go away after futzing with it. It’s almost like switching to a new MIDI track causes the problem.
Anyone else experiencing this? It sucks when you can’t rely on it.
Are you in cycle mode when you use it? Are you inserting linear, or cycle retrospective recordings?
Do you have any specific steps one could use to reproduce the issue?
The few times I’ve used it in Cubase Pro 15, it’s worked fine. Usually I cycle and record until I get the take I’m after.
Hi @shagazulu ,
I experienced something similar. The latest stuff played wasn’t saved at all. I didn’t have time to investigate any further.
It was a simple VSTi (Halion) which I played for a couple of minutes - not much going on in the project and the only instrument track at that point.
C15.06 Win10
C15 Help:
The buffer captures up to 10000 MIDI events. This can amount to a MIDI recording of around 2 minutes and 30 seconds. However, if you use a keyboard that produces a large amount of MIDI controller events, such as the ROLI Seaboard, this only corresponds to a recording of around 20 seconds.
Maybe buffer overflow? If it’s a FIFO, the beginning may be missing.
I had a similar problem in the past, caused by the fact that I have the RME sound card controller among my MIDI devices. All I had to do was disable it from the MIDI input devices in Cubase’s preferences, and everything was resolved.
I suggest you check the MIDI devices listed in your connected devices list and perhaps try disabling any that aren’t needed by Cubase. Apparently my controller was generating MIDI data that was needlessly consuming the MIDI buffer.
I hope this helps.
Just a FYI, this is not about my case:
Here, Retrospective Recording always worked with the exact same setup, same buffer size, same keyboard, no keyboard routed through RME via MIDI cable etc. Apart from that: the last bit (if not all of it) was missing!
Back to @shagazulu !
I’ve had this issue since Cubase 8.
I think there are two issues at work -
1.) Midi buffer is overflowing (Steinberg needs to increase the maximum size here)
2.) If “return to start” is activated, and perhaps when not, Cubase will lose track of the region start time and give you a small random region upon stop. Unrolling the region will reveal the midi data either ahead or behind. In this case, its all there, but in the completely wrong spot.
Thanks to all for answering.
Not really, no. But it definitely has something to do with first attempts once I change over to a new MIDI track. The problem does goes away.
Simplifying my setup and just having one MIDI controller connected for testing this is a great idea. Thx.
That’s an interesting thought. I’ve tried unrolling to the right, but not unrolling to the left. Hmm. I do do a lot of activating and de-activating of return to start - well, at least some. That could be messing with things.
Is your setup just one keyboard connected via USB and no other external MIDI controllers?
Let us know if you solve and how 
Yes, more keyboards and Controller via USB. It never happened with C14 with the same setup and that was the fist time. I will take a closer look next time it happens - might have been a mere coincidence. I don’t think it had something to do with another device, though. But we’ll see.
EDIT: RR works reliably on my second C15 system with Win 11. As long as it doesn’t happen again it might as well have been a one time thing or it’s related to Win10, a controller pedal - could be literally anything.