Super confused by note learn

I don’t know why I’m finding this so much harder to use than on the Pyramid. I’m sure I’m doing something wrong. But let’s say I want to enter a C4 via note learn. In step view, I click encoder 2. This fills the first to “squares” of the screen with the note learn interface. Cool.

Then I hold Live to bring up the “keyboard”. I play a C5. I release the Live button and everything snaps back to the step view. “C5” is displayed in the note learn UI. It all looks good.

Then I press a pad and instead of the learned C5 it enters whatever note that corresponds to the pad I pressed. What am I supposed to be doing to get a C5 into my piano roll?

I bet your in project scale mode. I find this confusing too. It makes learn in project scale mode pretty worthless. In track scale mode it works because step mode doesn’t restrict the notes to the scale for some reason.

Also we could be doing something wrong.

Edit: It does work for chord mode whether your in project or track scale.

1 Like

Okay, I figured this out. I didn’t understand that Note Learn works differently depending on whether it’s learned a single note or a chord. I didn’t see this explained in the manual so I’ll summarize here. Sorry in advance if this has been explained elsewhere.

When Note Learn is on and a chord is played in Live mode, everything works how I’d expect:

  • In step mode, the root of the chord “+X” is shown in the learn UI.
  • When a pad is tapped, the chord (as entered in live mode) is inserted in that column irrespective of which pad in the column was actually pressed.
Some nice details I'm glossing over:

When a cord is learned, the step mode’s viewport is adjusted so the root of the chord is the bottom row of the pads. But if your chord is an inversion with a note lower than the root, when you actually insert it the viewport jumps down to fit that lowest note instead… Pretty cool!

When Note Learn is on and a single note is played, however, tapping a pad in step mode does not always insert the learned note. It inserts the note for the pad you tapped, just as if Note Learn was off.

What the heck? So what’s the purpose of Note Learn when using single notes, then?

Well, what’s actually happening is the step mode is automatically setting the viewport so that the bottom row of pads is equal to the single Learned Note. So if we want to insert a “D” all we have to do is jump to Live, find our “D”, then jump back to step and tap the bottom row.

This is nice as it provides a way not just to enter the note we just played (it’s the the bottom row) but also to orient ourselves such that we can play other intervals without jumping around.

But (as @Luke mentioned) this “Adjust the bottom row of the viewport to be the learned note” behavior doesn’t work when a pScale is set.

Why? Because…

When pScale is set, the step mode’s viewport becomes locked so that the bottom row of pads is always the root of the pScale. Forever and ever, amen. You can move the viewport up and down, but only by octaves because the pScale root has to be the bottom row.

But remember, shifting the viewport such that the bottom row equals the learned note is the only thing Note Learn does when a single note is learned. Since, in pScale mode, the bottom row of the viewport has to be the root of the pScale, it can’t be the learned note, and thus Note Learn doesn’t do anything when pScale is set and a single note is learned.

Chords can still be learned and inserted in pScale mode, just like normal. The only difference is the viewport won’t move to accommodate them.

yeah, its a bit odd at first… but you kind of get used to it.
(excluding pScale, but thats a known issue)

I think the discrepancy really is simply…

  • single note allows for transposition
    as it moved viewport, but allows you to place not at the root)
  • chord disallows transposition
    it does move the viewport, to the root of the chord BUT pressing any pad in the column, always inserts at the root.

should the chord allow for transposition, so you have to get used to tapping the bottom row?
I think this would be more consistent, likely better?
but honestly, im not 100% others don’t prefer as is, hence ive not suggested it as a change. (*)

pScale, not sure how thats going to work… I guess it will have to be like chords are currently, so what ever row you press - it’ll insert the learnt note. (as viewport is fixed)

as an aside, another thought, that Im ‘undecided on’ is using the bottom row…
I kind of wish the viewport was focus in the centre… (say row3), so that you can see above and below the selected note…
though, if this was the case, I think the learnt/active note would have to be ‘coloured’ so its easy to find… and thats probably why its not been done this way :wink:


(*) I can see its quicker to enter chords just pressing any row… but Id assume you’d get used to just press the bottom row.

The note learn is indeed quite confusing, especially when it was so straightforward in Pyramid. The fact that it shows the learned value on screen but then refuses to place it anywhere makes it even more so. I had to read the manual section a few times to make any sense of it the first time, and after a few weeks of involuntary break that was all forgotten and had to reach for the manual again.

I’d assume there’s a use-case for it being this way, but I personally can’t imagine ever wanting any other behavior than “just place the chord I entered, damnit”, be it bottom row or whatever :sweat_smile:

1 Like

it works more efficiently for chords for me, at least in the current iteration (i believe the team intends to update Step mode to better reflect single note values as entered). but with chords in Learn, it is as simple as playing the chord notes on a keyboard and then just touching the point on the grid pads where you want to place it

This is a pretty big workflow problem for me. I come from a world where I’m sequencing midi with an octatrack, it’s extremely limited but somehow easier than this to enter the notes I want. I am able to hold down a step, hit a note on my keyboard, and that note will be in the octatrack. I love how powerful the Hapax is but I’m frustrated at the same time by a few little workflow things like this. There is no way (in pscale mode) to find a note by ear and enter it without doing math in my head. I just want to be able to enter the exact notes I want exactly on the grid where I want them, I can think of 3 things that would make that possible, but none of them work: Destructive quantize, note learn that actually works (obviously what I would prefer), or note preview. I don’t think I’ll always be using project scale mode and obviously this isn’t a problem otherwise, but it would be nice to be able to maintain a consistent sequencing workflow across my projects.