Breath mark playback, Release memory on exit, ...
Hi, I just downloaded r1667 for my first trial. Many thanks for great program and support! Feels like Sibelius yet better organized...
I've used it for a few minutes, and thought I'll do my part to provide feedback, so you all can decide if any feature is useful:
- I played back the demo, it sounded very mechanical. May I suggest, short-term: add playback support for breath mark. Longer term: add humanized playback?
- When I play back, I do not see an obvious way to change the timeline as to where to restart playback.
- After I quit the program, mscore.exe still remained in memory, eating up precious memory that I never have enough. (Even though it has relatively light footprint.) Can we put in an option to not let it remain?
Thanks for listening and a great tool.
Comments
1. Currently slurs, staccato, tenuto, and arppegios are interpreted, plus markings for tremolo strings, pizzicato, and muted trumpet. Other markings (such as breaths) have not been implemented yet.
2, Click on the note you want to start playback from. Then press play. (If you are in note entry mode then you must leave note entry mode first). If you prefer a timeline then open up Display > Play Panel. (At the time of this writing there is a least one bug related to the timeline in the play panel: http://musescore.org/en/node/1256 )
3. When I quit the program mscore.exe does not (and should not) remain in memory. Was it a one off or does mscore.exe remain in memory every time you quit ? Make sure you wait a few seconds if your computer is busy at the time.
In reply to 1. Currently slurs, staccato, by David Bolton
3. It's not one off. Happens every time. I have to use Windows Task Manager to kill it.
Thanks for the quick responses. I was aware of 1. previously by reading other people's posts. Just thought breaths are easy to implement, so suggest it separately. Thanks for 2., I'll try it out...
In reply to Reply about quit by newsong4life
Could you describe the method you use to quit MuseScore? Also maybe you could share the specs for your computer (version of Windows, amount of RAM, etc.). We'll see if anyone else has the same problem you are experiencing.
In reply to Could you describe the method by David Bolton
To quit: I clicked the X on the window, and answer the dialog to save if needed.
My system: Windows XP Home SP1, it reported Pentium 4 CPU 3.00GHz, 1.00 GB RAM
Thanks for looking into the issues.
When you say that playback sounds mechanical, what exactly do you mean? I've had some success in making things sound better by putting an invisible tenuto mark on the last note of a "musical paragraph", and adding rubato using slight tempo changes. Unfortunately any type of dynamics are not supported at the moment, and these tempo changes are not interpreted correctly if you have repeats.
If you want to interpret the breath mark, I would suggest making the note before the mark staccato. Then make the dot invisible and add a tempo change for that note to make it a bit longer and give the break you're looking for. Using these sort of tricks, I've been able to get a good deal of realism out of my scores, even though it does take a fair amount of work. In a cadenza, for example, I've put a tempo change on almost every other note, and in some places on several notes in a row.
I would welcome the ability to tinker with the note duration, velocity, and aftertouch in the note properties box.. This would go a long way to making things more realistic; letting the user provide an answer to "Just how staccato?", and "Exactly how much accent?"
In reply to How would you play these? by MDMilford
@MDMilford: The prerelease that newsong4life is using does not have issues with tempo changes and repeats. Werner fixed this today (see 21 March ChangeLog )
In reply to @MDMilford: The prerelease by David Bolton
Wonderful! I'm downloading it now!
In reply to How would you play these? by MDMilford
@MDMilford Bravo. I appreciate your attitude and ideas!
Unfortunately, that kind of control gets old after a while. What I'm looking for is that the first time I write the score, without any user control, it sounds "natural". I want to minimize time spent. Since the app is 0.9 only, I'll post some simple suggestions on a separate thread later.
Of course, I want all the capabilities MDMilford mentioned above to continue to grow. If I really want a natural sounding result, I'll have to do a natural modeling (simulation), which means: for one violin part, I'll clone it into 4 violin 1 and 4 violin 2, I'll split the notes to violin 1 and 2, then for each violin section, I'll modify their note entry, duration, velocity, etc. I'll even change the timbre of each violin if that is available. That's the right way to do it if I really want a controlled natural sound... And anything that helps in this is also appreciated. (Later I'll post separately as feature request.)