I am using a basic piano with a default map (since it really only needs velocity / CC1).
I have noticed a frequent thing happening for me lately where when I copy music, the velocity will drastically change. This might be due to humanization settings but it doesn’t really make much sense, since the drop is significant and very noticeable, becoming a lot quieter. I drew a green line across the velocity to make it more clear exactly what I mean:
I have also observed it ocassionally go in the opposite direction, getting much louder.
For the record this is why I recently asked about fixed velocity for playback for a sketch piano, but in this case I am using a piano I want in the score and it’s tedious to have to manually change velocities every time I copy/paste music. Not sure what the root cause of this or if other people are able to reproduce it on their end. I’m assuming it’s a bug?
are you sure that the score dynamic is the same in both places? You’re not pasting from left hand to right hand or vice versa where the dynamics might be completely different? If not and you haven’t overridden with manual changes in the key editor, then certainly the pasted section should take on the prevailing dynamic of the section copied from.
Still, this doesn’t seem to be what you’re getting at and in a quick piano test using the standard EM, everything works as I’d expect. I suspect we’ll need to see a sample score after which with luck everything will become clear.
Well, it is my ears that brought me to viewing the key editor in the first place; it may not be obvious by the visual but from sound, with the velocity response of my piano (which is a linear curve), I would describe the difference from the original bar to the duplicated bar to be something like mf to p. So it’s a bit jarring to lose about two dynamic levels suddenly without changing anything else. As you can see in the score, I actually had no text dynamics written here, and at this stage of capturing the video I hadn’t done any manual overrides either.
I should say as well that this has occurred to me off and on for a few months, particularly on a piano grand staff. I have observed it to happen when duplicating from the bass/left to the treble/right, too. I think it has happened to me as well on a marimba grand staff a few times.
So I can at least rule out single-staff instruments; I have never heard this issue happen on a cello or flute for example. Therefore I think there must be something going on with grand staves for me.
I wonder, if it has anything to do with with the duplication of staves which overlap (bass/treble). For example if you see in my example, the final note of the treble clef is on the downbeat of m.3, exactly where I duplicate the downbeat of the bass clef.
I believe previously when I had a issue with this I was duplicating a piece which had a bit of alternating bass/clef arpeggio runs where they overlapped on the downbeat. Maybe there’s something to that where it’s forcing the pasted music to take on whatever the other clef is doing? Hope that makes sense.
It’s tricky because it is the sort of thing which seems to happen randomly and thus difficult to isolate. However, I do have saved a rough sketch from the piece I just mentioned above where I ran into this a handful of times. I can simplify the project and share a section of bars I know were having an issue – if you think that would be helpful?
In Library > Playback Options > Dynamics > Polyphonic Instruments, try unchecking Use automatic polyphonic instrument voice balancing. When I do this with your example, the dynamics in the first three beats of the first and third measures are essentially the same.
I assumed that had already been switched off as it’s part of the humanisation settings (the same applies to things like “use rules for contrapuntal music”) but does no harm to check! Certainly with everything in the Dynamics section disabled – which is essential for proper testing – I have no issues.
yes- I think it could be useful to see a project which of course includes both material copied from and to if in the meantime switching off all humanisation still hasn’t fixed it.
Ok, I’ll do that if for testing. One of the reasons I have not wanted to completely turn off humanization is that is not my goal - to lose humanization globally. Everything else is sounding great and natural. I just want duplicated music to not vary so wildly (of course a little bit would be human but not dropping two dynamic levels suddenly!). I’ll post something shortly and also look at John’s suggestion above too.
Unfortunately as I’ve dug through previous projects I cannot find an older “raw” version prior to me doing a lot of manual velocity overrides.
I also tried manually entering a few bars and recreating the issue, which it wasn’t happening. I even re-entered last night’s issue (from the post at the top) and even that is not giving me the same result - with no other setting changes, all else being equal! Which is why I sorta wondered if this might be a bug, due to its seemingly random and sporadic nature of occurence. I think the only way to be sure will next time I experience it, stop to take detailed notes of all my settings and observe the humanization settings too, and just keep a log for any time I witness it in the case I can see a pattern emerge. Like I said, the only constant variable I am confident about is that it’s always been on a grand staff.
This, however, does seem to have a drastic effect and I would say, with it turned off, a longer piano score (with a single dynamic marking and no hairpins) sounds a lot more balanced through and through. With it turned on that same score does sound a lot more varied in dynamics from bar to bar and phrase to phrase.
Interestingly, I like aspects of both, the more controlled and predictable version (unchecked) sounds a lot more clear and intentional; whereas checked, it does sound a bit more human, but almost too much so where ideas get lost. I think I will play around with the settings - I noticed there is a slider to adjust how much the effect is emphasized by the dynamic curve and other percentages. It’s possible I might be able to find a much more suitable balance for my ears and the default is simply too extreme with variations.
FWIW, while I often use NP for the full orchestra, I do not use it for most keyboard-type instruments, preferring my own pianos, harps, and marimbas (Kontakt etc). So these instruments might be responding to that above polyphonic setting with more drastic variation than the NP equivalents do (with NP’s custom map overrides too). Something I’ll have to investigate further.
Yeah A/Bing the playback of the same score to compare, there’s aspects of the polyphonic balancing where I thought “oh that’s nice, I would phrase it that way in reality” and then many times I thought, “what the heck are you doing? who on earth would play it like that?” – haha! If I had to guess this probably boils down to the issue I’ve described having on pianos, perhaps it is just too wily—and I am a too much of a control freak—so I might be happier with it turned off, manually adjusting dynamics to taste.
I’ve done a bit of testing with the polyphonic voice balancing and found initially it tended to go to extremes but later Dorico updates are rather more balanced in the default settings. There’s no doubt that on the whole (with some exceptions I won’t go into here) it does now give a more intelligent balance between melody and accompaniment than can be achieved switched off and can be a great time saver for piano music (unless of course you play in live). As with anything, the main thing is to be aware of which humanisation settings are actually activated and what they do so you can make things work for you.
