Chord Track recording oddness

(I split this off from another thread: https://www.steinberg.net/forums/viewtopic.php?f=250&t=118596 )

I was testing that, and found a bug, I think. Though it’s got little to with this specific issue, it does involve the chord track.

Stay tuned.

Owwwww, a master of suspense.

Okay smart guy. :wink: :stuck_out_tongue: Here’s a repro to play with.


New Instrument track, chord track, enable record on both.
Hit record and play some some chords one after the other.

Chord track gets multiple chords for each one I play in. An X chord and a couple others.

you?

If I play the same chord on quarter notes & then change the chord at each bar:

  1. I get multiple chords on the Instrument track, one for each played chord (duhh)
  2. On the Chord Track I get one single chord at the start of each bar where I’ve changed to a different chord. No X chords at all.

By the way it’s “smart a–” not “guy” :wink:

Hm. Playing chords on the piano, not the chord pads?

edit: (yeah I said “guy” 'cause I didn’t want phpBB to make me appear to say stuff like knuf flower rock quiz witch take your choice! :laughing: )

Oops. Yes keyboard. Will redo using pads when the DAW wakes up.

It is the keyboard I was having trouble with, not pads.

It’s intermittent. If you could be kind enough to try recording a bunch of four note chords, I’d be much obliged. I’ve tested with two keyboards, and two midi interfaces. If you can repro this I won’t have to tear apart my gear…

Well, just for completeness, the pad input worked as expected - a single new Chord Track Event when the chord changed.

Playing 4 note chords on the keyboard I got no X chords. Mostly I got one instance of each chord I played. I did on occasion get 2 similar chords one shortly after the other. But I think this was more related to my sloppy keyboard skills because it would be something like Em followed by Em7 where it seemed to detect a 3 note chord quickly followed by a 4 note chord. In one case I had a Dm/F & then an F6 where the Chord Events were at the exact same location. But I could tell that Cubase thought the F6 was after the Dm by the order I could arrow through the Events.

I did have one instance that seemed a bit odd. I’d played a single G7 chord (doubled checked by looking at the midi on the Instrument Track) but ended up with two G7 Chord Events on the Chord Track separated by about 70 ticks.

Not sure how Cubase decides where the Chord Events fall on the timeline. Comparing the Events to their corresponding notes on the Instrument Track all the notes start before the Chord Event. Maybe that’s just how long it takes to calculate the chord name.

That’s exactly it. If you play carefully it will keep doing it anyway, yes?

And it never happens with three note chords?

Yes 3 note chords seem to be always good. When I do play 4 note chords “carefully” (there is a high margin of error on what that might mean) sometimes I’d get 2 Events & other times just 1. In any case, as I think about it, it seems the new chord detection is much too sensitive. Stylistically there are lots of times when the 4th note gets played substantially after the 1st three. Playing chords and holding them it seems visually that the event creation is initiated by the note-ons. I think the event location should should be determined by the 1st note-on, but the event itself should be created when the note-offs occur.

If I create a new version of the Chord Track and in the Key Editor use Create Chord Symbols the new Chord Events are as expected. So this is only a real-time problem