dynamics over single note - a request

• Feb 26, 2018 - 23:07


I was reading from the pull request that Nicholas finally recently said that this feature will not be in 2.2, due to code stability, but hopefully it would be in 3.0. I too waited for months for this feature, as it was announced to be reviewed once 2.1 were released.

Please, Johannes (@hpfmn), would it be possible to indicate what was the musescore branch that was used to compile the MusesScore executable used for the demonstrations in the Crescendo and Soundbanks screencasts (https://musescore.org/en/user/527826/blog/2016/06/12/crescendo-screenca… and https://www.youtube.com/watch?v=Bx4AtSmv95I) ? and https://www.youtube.com/watch?v=TlhB6RBWIJA (as the final result I guess about this GSoC 2016 work)

Do your fork at https://github.com/hpfmn/MuseScore matches the demonstrations above? is it available? after seeing and hearing the above videos, I don't mind at all if the build used was unstable. It's a highly-sought enhacement to the playblack.


In reply to by mdi1972


Just tested OK a build of Johannes work - sound nicely!

For the record of anyone interested, the branch was really this: https://github.com/hpfmn/MuseScore/tree/CC11 . It first implemented the CC#11 expression controller via hairpins. I had problems with the MidiSpanners branch (MuseScore crashing every time I checked add MIDI controls in Edit->Preferences->Score)

This branch compiles to a release that matches AFAIK the video Johannes posted at https://www.youtube.com/watch?v=TlhB6RBWIJA&feature=youtu.be . In this video, he explains very nicely the new capabilities, with a practical example (it involves saving the score as a .mscx, and adding a and tags to the Instrument section.

I'm attaching a screenshot in my PC, and the playback from this build (it's the fragment in the screenshot)

Attachment Size
CC11-hpfmn.png 152.38 KB
kio3.zip 3.92 MB

In reply to by mdi1972

Just to leave documented, for anyone interested to build this branch:

-My build was Windows platform, using the steps in the Developers manual for windows builds.
-Used MinGW 4.9.2, QT 5.6.2, CMake 3.10.2; and downloaded the LAME/Ogg/portaudio/JACK/etc. packages.
-Adjusted the PATH variables
-Adjusted the main Makefile.mingw32, including a MINGW_DIR variable, pointing to MinGW's install/"win32" folder.
Adjusted the build/mingw.mingw32.cmake, just to include the MinGW paths ("
set (CROSS C:/MinGW492/mingw32)
set (CROSSQT C:/Qt/Qt5.6.2/5.6/mingw49_32)"

Before building, I had to include a "cmake_policy(SET CMP0071 NEW)" in the top-level "CMakeLists.txt" file. Without it, many code will be discarded by the compiler.

Other than the above, the branch compiles cleanly.


In reply to by mdi1972

For master you should be using Qt 5.9.x (or later?, even if min version is set to 5.8), for 2.2 Qt 5.4(.x).
In the top-lever CMakeLists.txt we explicitly set that policy to old.

# Don't process generated source files with AUTOMOC
if(POLICY CMP0071)
      cmake_policy(SET CMP0071 OLD)
endif(POLICY CMP0071)

and do so since a39935e4, to keep CMake behave the same as before version 3.10

In reply to by Jojo-Schmitz

Yes, Jojo, for the current MuseScore master branch, it's obvious the above indications are not correct. For the current MuseScore, the reference are of course your comments and/or the official Documentation regarding builds.

I hope no one gets confused: I'm not refering to the current master - the build notes I just dropped here are valid only for that hpfmn's "CC11" branch, developed by around 2016 (branch which at the end didn't get committed into the master)

I didn't tested later QT releases with this CC11 branch, those may work, or not. I'm not interested testing newer QT releases against this precise branch - I have now a working - thought of course a bit dated - MuseScore environment that, in my opinion, wonderfully implements the single-note crescendo-decrescendo, and that's all.

And I wish to give thanks not only to Johanness, but to you all in the developer team, since you all contributed to this branch. Even it was not pulled in the master (a more advanced solutions seems to be "in the works"), it's a great work of you all.

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