First principles for MIDI/VST recording

I am very inexperienced at Cubase (but an experienced musician), and i have some very basic questions of approach.

I have a song which I’ve recorded with drums (Groove Agent), bass (Halion Sonic), piano (The Grand) and synth (Halion Sonic). These four tracks sit in my project. I’ve tidied up the errors by tinkering with the MIDI and I’d like to add some more instruments.


  1. Am I better off leaving EVERYTHING as it is (i.e. as a MIDI driving a VST) until it comes time for the final mix? Or would I be better off exporting tracks to audio files when those parts are complete and reimporting those instead?

  2. If YES to exporting to audio, should I apply FX (reverb, EQ etc) before export, or should I export dry and apply the sound treatments to the reimported audio files?

  3. If anyone knows of an online site/resource or book that might help me, I’d be very grateful. Most manuals seem to concentrate on audio recording, but I don’t have any need or desire to do any audio recording at this stage.

Many thanks, and stay safe everyone.



  1. Better to keep it as it is. You have much more freedom to edit MIDI data than Audio signal.

  2. Apply the FXs to the Instrument tracks.

Thanks Martin. I guess my issue is to do with the processing power needed to work the VST tracks compared with the audio tracks …

You can create a stereo mix to work with. Temp disable the vsti tracks but keep them in your project, or export them and reimport them later. Switch back to the original tracks and higher buffersetting for final mixdown.


What about to Freeze the tracks?

I don’t really understand freezing … I’ll have to do some research. Thanks for your help Martin and Jb

Agreed about leaving the tracks as MIDI until the last minute, Edward. My experience has been that the more tracks you add, the greater a chance there is that things will become cluttered.

I’m a piano player and did a song recently with piano as the main rhythm instrument and some guitar doing complimentary licks. Of course, I recorded the MIDI tracks separately but afterward discovered that I had some severe overlaps of the piano and guitar notes. They were fighting for the same musical space. With MIDI it is so easy to correct the problem. It can be done with audio but you would need to do a punch-in and punch-out recording.

With MIDI you have one more crucial control that you don’t have in audio, and that is Velocity. Velocity and Volume are very much related but Velocity can also change the sound of the instrument not just how loud it is. For example. if you play a low piano note at a high Velocity, you may also trigger extra sounds of the piano cabinet itself, depending on how the original voice was sampled. You won’t achieve that just by increasing the Volume of the piano audio track.

After I have my MIDI tracks 100 percent right, I Render them to audio and apply my effects and EQ. All my MIDI tracks are recorded dry.

  • Lee

Thanks very much Lee - the kind of things that you are doing are similar to me (I just got Garritan Personal Orchestra today so my space is going to get even more cluttered). But being able to tweak the MIDI so precisely is why I don’t want to lose it just yet… however, I have previously had some strange drop-outs etc when trying to mix from multiple MIDI tracks driving different VSTs with effects.

Glad to help, Edward. The only time I experienced dropouts was a long ago when I recorded with a Korg M1 and Roland Sound Canvas. If I had too many notes triggering at the same time, I exceeded the total polyphony of both instruments. Back then (early 90s) each device only had a 16 note polyphony. I solved it by delaying one or two less critical notes from triggering by a 1/32 note. The human ear couldn’t tell the difference and the polyphony issue was solved. I suppose you could say that I caused the dropouts by quantizing 100 percent all the time.

I’m not sure what would cause your dropout issue. Your computer specs seem very good. Perhaps someone else could chime in on this :slight_smile:.