[MU4 Issue] Muse Sounds transitions smoothly between dynamics, even without crescendos/decrescendos

[erinic04]

Describe the bug Muse Sounds transitions smoothly into new dynamics, even when it is not specified, or even if specifically specified not to (subito)

To Reproduce Steps to reproduce the behavior:

1. Make new score
2. Make one whole note at pp, one note at ff
3. Hear crescendo, even though there isn't one written

Expected behavior There should not be a crescendo

Screenshots In context:

https://user-images.githubusercontent.com/21023353/197378736-98744d5b-5a9b-4ba8-8594-68119826c089.mp4

Isolated:

https://user-images.githubusercontent.com/21023353/197378827-2690c94e-ac04-43c9-9795-f08f7fcf3a27.mp4

Platform information

• OS: [e.g. macOS, Windows, Linux]

Additional context Add any other context about the problem here.

[Tantacrul]

Here is a (slightly amusing) example of how a user suffered from this problem: https://github.com/musescore/MuseScore/issues/14378

[LuisLiszt]

I wonder why this is not a critical problem, I think it is one of the biggest problems MS4 has right now.

[mccoydtromb]

Responding to the request to test - still a problem for me. Win 10.

[Tantacrul]

I wonder why this is not a critical problem, I think it is one of the biggest problems MS4 has right now.

It's in our release project, so I don't see much reason for complaint. With the latest changes to the Muse Sampler, (which requires the Muse Hub to be opened), it's working for me.

[Tantacrul]

Responding to the request to test - still a problem for me. Win 10.

Can you go to your Diagnostic menu in MS4 and tell me what version number of MuseSampler you have?

It should be version 0.3.0.

If you are still having problems, then can you share the score you are having issues with, along with your OS type?

[erinic04]

I am having problems with this score - also yes, I have the MuseSampler 0.3.0 Super Mario Galaxy.mscz.zip (specifically the second measure in the parts that go to mp)

[MarcSabatella]

I can confirm this on that particular score. After some experimentation, I think I found the trigger - it's the rest on beat 1 of the second measure of the phrase.

Check out the following simplified excerpt:

https://user-images.githubusercontent.com/1799936/204857903-ed86b8cb-d9d5-4386-b71b-f2e5984fe199.mp4

The first phrase plays as expected (except the final ppp note gets swallowed by the tail of the previous note, perhaps the reverb?). The second phrase show a sudden unexpected change in dynamic & sound between the first & second bar.

[LuisLiszt]

@MarcSabatella with the latest nightly and MuseSample 0.3.0: when I rebuild your first staff I can confirm that there is no automatic crescendo/diminuendo like before. However when using Muse Keys and Muse Percussion instruments the issue still persists. It seems fixed for almost every MuseSounds instrument except the ones mentioned. Can you confirm this? @Tantacrul

[Tantacrul]

Yes. It still does it for piano. Thanks @LuisLiszt !

@davidwhearn ^

[MarcSabatella]

For the record, I was using tenor sax. So yes, with piano, both lines play incorrectly. With tenor sax, the first line is good, the second bad.

The issue with the piano might be related to https://github.com/musescore/MuseScore/issues/13877 (even though that issue was reported for MS Basic I believe).

[LuisLiszt]

MuseSampler 0.3.1 fixed the issue for MuseKeys!

[Tantacrul]

Yes, I can't reproduce anymore.