Monitoring the velocity of notes

• May 6, 2019 - 03:12

Is there some direct way to keep track of the velocity of a given note after several hairpins without having to compute it manually? I mean direct reading.

Besides, there is some ambiguity in the very definition of offset velocity. In the manual it says:

"Offset: Make the value shown in "Velocity" relative to the previous dynamic marking."

1) In general, "offset" is considered an increment or decrement from the current value of something, so if starting from mp at 64, an offset of 10 would imply a velocity 74, while starting from f at 96 it would yield a velocity of 106. However, the word "relative" suggests a percentage. Opening the piano roll seems to confirm this, since unlike the inspector, it shows a % sign. If this is the case, starting from mp 64 would mean an absolute infcrement of 6 (after rounding) so the result would be 70 instead of 74 and in the case of f 96 we would have an increment of 10 (after rounding), so we would have a velocity of106.
2) What does it refer to when speaking of the previous dynamic marking? Suppose I start from mp 64 a crescendo hairpin with "velocity change" 10 (I'm assuming here that the velocity change is an absolute increment, but it could also be a percentage, please confirm). So at the end we will have 74. If the next note has an offset correction of 10, will we have 64 + 6, 64 + 10, 74 + 6 or 74 + 10?
In the first two cases, I'm reading literally, the previous dynamic marking is mp, while in the latter two cases I'm assuming tht the "previous dynamic mrking" includes the outome of applying the hairpin. Adding 6 or10 dependes on how the hairpin interprets "velocity change"

As a final comment, since velocity seems to be proportional to psychoacoustic loudness (see discussion https://musescore.org/en/node/2880129), it would make more sense that "offset" meant actually an offset instead of relative or percentage increment, since at very low levels (pp) 10 % is bearly audible while at high levels (f) it is quite a loudness increase , while an increment of 10 units is roughly the same step across the full dynamic range.

Comments

• May 6, 2019 - 03:30

To be clear. All dynamics are indicated in velocity. All changes are calculated using velocities. Relative does not mean ratio it just indicates some sort of offset. In the case of a crescendo, the velocity is increased, for diminuendo the velocity is decreased.

The decibel levels can be calculated as in the discussion you referred to. These decibel levels are what was used when the unofficial standard velocity for a each dynamic mark was created. It's also why a velocity change of 15 from pp is far less noticeable than the same change from f.

Reply

• May 6, 2019 - 07:35

I've done an experiment (see attachment), since I discovered that piano roll allows in part to do what I asked for: It shows the absolute velocity in a graphical fashion. I could confirm the offset is not a true offset (in the sense of a specified fixed increment) but a relative increment in percent.

mf = 80
mf + velocity offset 20 = 96 (equivalent to f) instead of 100

p = 49
mp + velocity offset 30 = 64 (equivalent to mf) instead of 79

So the true target velocity is given by

V2 = V1 * (1 + v.o./100)

instead of

V2 = V1 + v.o.

where V1 is the previous dynamic marking velocity, v.o. is velocity offset and V2 the resulting velocity.

I also tested crescendo setting velocity change 20. Starting from mf = 80, the last note has a velocity close to 100 and it reaches 100 on the next note after the end of the hairpin. I added a last note with a velocity offset 10 and it reaches 110, confirming that the offset is applied to the outcome of the hairpin and not relative to the last written dynamic marking (mp, mf, etc.).

So I find these issues:
1) Calling offset to what is really a relative increment is misleading.
2) If the name is to be kept anyway, it would be nice that the percent sign % were attached to the figure (as sp is attached to the figure n the case of sizes or positions) in the inspector, the same as happens in the piano roll.
3) There is some inconsistency between the effect of the velocity offset and the effect of the velocity change in the case of hairpins. If I want the same final effect, in one case as a suden change and in other as a crescendo, I must think differently.

Attachment Size
Test velocity offset and crescendo.mscz 8.87 KB

Reply

• May 6, 2019 - 22:46

Indeed, this is the very reason why I started to dig a bit deeper: the effect seemed not to agree with what I expected... because what I expected was a true offset, not a relative increment. Relative increments have practically no effect at low dynamic level, and excessive effect at louder dynamics.

Now I can mention another inconsistency: "offset" is used differently in the Element tab for most items in the Inspector, in which we have a section where we can edit the x and y offset. These are values expressed in spaces to be added to the current coordinates of the element, for instance, the position of a dynamic mark. This is the expected effect of any adjustment called "offset" and should be also the effect of the velocity offset.

As a suggestion, the velocity setting could have three possibilities: User, Offset (relative) and Offset (additive).
The Offset (relative) would be the same as is currently (with back-compatibility purpose) with the addition of the % sign to avoid confusion, and Offset (aditive) would be a new option. The manual should reflect these, and so the piano roll window.

Reply

Do you still have an unanswered question? Please log in first to post your question.