I am trying to show the dynamics, with these velocity values obtained by creating a midi with musescore
PPP 0-22
PP 22-41
P 41-56
MP 56-72
MF 72-88
F 88-104
FF 104-119
FFF 119-127
The problem I have is that I don't know how strict I should be when assigning the dynamics, if I have a note with velocity 20 and the other with velocity 25 it would show me two different dynamics, so I added a tolerance of +- 15, but still So I get more dynamic changes than musescore. Is there a correct way to do this? The file I'm testing with has many velocity variations almost for every note (attached file).
I've got MuseScore here, although not the latest version, so I don't know if there are any extra options in that.
MS will read in the note info from your midi file, and will act on that. The assignment of the pp/ff etc marking is literally as per the midi input. However, within MS you can edit this info, note by note, or by block, as you see fit.
You can handle this in any way you like, to achieve the effect that you want. That's what is important. But remember, anyone later performing your piece, or your arrangement, may well 'interpret' your instructions as per levels as they see fit, to get the effect that they want. That's what performers (conductors/musicians) do. Like it or not.
If you play a piece of music and record the midi, then the velocity levels will probably be all over the place. That is normal. The piece has been totally 'humanised'. If you then adjust the velocities to get then looking 'regular', then you may find that the playback is much less 'humanised', even becoming robotic, and the result may not be any use. You may need a compromise.
If you can get a midi recording made by a good musician, there may well be less variation in the velocities. You need to study the variations, and note how the patterns of the variations of the velocities relates to the structure of the music, the bars/phrases etc, and consider why that is. Why the performer has chosen to do it that way. I'm sure it's not accidental. If you wish to change the velocities, then OK, you can do so, just like a different performer might do, but you should have a musical reason for doing so.
Does this help you any?
Geoff
Hm, just listened to the midi file.
Fairly brilliant, I'm no expert, but it sounds exactly like it ought to sound like, and VERY well played by Ken - whoever he might be.
Totally natural, no evidence of any 'robotic' effect, so I'd assume that it was played/recorded to midi directly. Maybe some fluffed notes might have been corrected manually, but I didn't spot anything.
Well, I'm sorry that the MIDI velocities were not tidy enough for you, but you might take that up with Mozart. I'm sure that he'd be quite happy with it. All within the bounds of normal 'artistic licence' of course.
Geoff
I think Carlos' concern is that his software that he is writing is ending up with much more variation in velocity displays than the corresponding variation notations displayed by Musescore. So where MS might show a section as having, say, a part that only shows it starting "f" and transitioning to "ff", Carlos ends up with "f" to "ff" to "f" to "mf" to "f" to "ff". Or something like that.
Rather than using tolerances (or in addition to) it sounds like you need to average the values together over a certain length, like maybe a measure or two, and then base your notations on that value rather than on individual notes?
[quotePost id=15374]I think Carlos' concern is that his software that he is writing is ending up with much more variation in velocity displays than the corresponding variation notations displayed by Musescore. So where MS might show a section as having, say, a part that only shows it starting "f" and transitioning to "ff", Carlos ends up with "f" to "ff" to "f" to "mf" to "f" to "ff". Or something like that.
Rather than using tolerances (or in addition to) it sounds like you need to average the values together over a certain length, like maybe a measure or two, and then base your notations on that value rather than on individual notes?[/quotePost]
That's exactly the problem, I'm going to try something like that. The idea is not to fill the screen with lyrics and lose as little information as possible, as for the audio, it would not be altered by this.
Apparently I've gotten confused, because now when I open the midi with musescore it doesn't show me any dynamics, somehow I ended up believing that https://musescore.com/classicman/scores/49143 and the attached midi were the same.
In the attached image there is an example (on the same track because I don't see how to separate them correctly).
One program use, an old one by Yamaha, has the following dynamics embedded within it:
1-15 ppp N.B. not zero, a value of zero is a note off in MIDI
16-31 pp
32-47 p
48-63 mp
64-79 mf
80-95 f
96-111 ff
112-127 fff
So, a width of 16 in each velocity band.
See below in the right hand column.
It may be a reasonable guide.
JohnG.
[quotePost id=15379]One program use, an old one by Yamaha, has the following dynamics embedded within it:
1-15 ppp N.B. not zero, a value of zero is a note off in MIDI
16-31 pp
32-47 p
48-63 mp
64-79 mf
80-95 f
96-111 ff
112-127 fff
So, a width of 16 in each velocity band.
See below in the right hand column.
It may be a reasonable guide.
JohnG.[/quotePost]
Thanks, better to use these speeds for dynamics, with a fixed range. forget that velocity 0 is a noteoff but in midi 2.0 noteon with velocity 0 is allowed :p