Go back to Musescore 2

• Mar 1, 2019 - 16:54

Hi,

I installed Musescore 3 on my computer. To open my files they had to be put into Musescore 3 format. That done starting to work on some compositions I found that this version of Musescore is very slow/ sluggish. Note input with the mouse is very slow, selecting multiple measures is tardy, and I get the impression that there is a sort of time loop; what you change doesn't show immediately but a second or 2 later. Is something wrong with my computer or is it the kindergarten of Musiescore 3?
Anyway, I'd prefer to revert to Musescore 2, BUT, the files that I saved in Musecore 3 format can't be opened with Musescore 2. Is there a solution for this?

Thanks and regards,
Danker

Comments

• Mar 1, 2019 - 22:13

Are you using 3.0.3 or later? It is objectively speaking faster than MsueScore 2 on average, so if you have some specific score where it seems slower, we'd like to know so we can investigate why a problem exists for that particular score. Could you attach it?

Reply

• Mar 4, 2019 - 16:54

Hi. Couple of weeks ago I installed version 3.0.3. I converted files from Musescore 2 to the new 3.0.3 version. The problems I encountered were as I said slow note input with the mouse, it seems there's a time loop, what you do doesn't show immediately as in version 2 but a second or 2 later. Also selecting multiple measures was very slow.

Then I installed the newer version, 3.0.4.5763. What has improved now is selecting multiple measures, for example with Ctrl-shift arrow key or by selecting measures with Shift and mouse. Note input with mouse has improved slightly. But play back is fucked up although not in all my scores. Starting playback somewhere in the beginning of a piece is ok (but still with a small time loop), if I do that toward the end of the piece, it jumps after a couple of seconds, or it sounds a standing tone, then stops altogether. I attach 2 files, 1 is called Open Rev, the other Hearts of liberation. Playback anywhere in the first piece ("Open REV") is ok, but not at all in the second.

Thanks for your opinion/ advice.

Best,
Danker

Attachment Size
.Hearts of Liberation.mscz, 66.9 KB
.Open REV.mscz, 45.42 KB

Reply

• Mar 8, 2019 - 00:55

It sometimes is a bit less reliable in updating (a change or something affected by it is only (re)drawn once you do something else), which may make people perceive it as slower.

It’s also massively slower in continuous mode (which is known, I am aware), which many people used to prefer for inputting without having layout concerns or jumping hairpins. In theory continuous ought to be faster, as it doesn’t have to do page layout (a measure can thus never affect anything in a different measure).

MuseScore 3 is not really faster. It’s actually slower because it does more computations. It just tries to be faster by doing barely as much stuff as needed at a time, which is usually a good thing (modulo the overeager not-doing bugs I described above, and that it doesn’t work in continuous mode).

Reply

• Mar 8, 2019 - 03:35

Are you on 3.0.4? I just did a quick check using one of the larger scores I have handy. In continuous view it was maybe 20% slower (about 3 seconds per update versus 2.5). In page view 2.3.2 is about as slow as continuous view, but 3.0.4 is so fast I couldn't get an accurate timing - essentially instantaneous.

Would be lovely to apply some of those same optimizations to continuous mode. But, it's not true that no page layout is needed - it's just different page layout. One big long system, and changes in one measure most definitely can affect others. Worse, they likely affect from that point to the end of the score, whereas in page view it often only affects to the end of the system. of course, in most of those cases, the changes to the next measure are just of the form "take everything as is and just update the X-offset without changing anything else", except maybe spanner that need to update their lengths. This has been suggested before, and it's a good idea. Maybe that optimization would be feasible without a whole lot of effort.

Reply

• Mar 8, 2019 - 06:30

Hi Marc,
I know it is easier to say than to implement, but from a logical point of view, by storing positions relative only, a well programmed updating algorithm in continuous mode should really only need to update at maximum one or two measure to the right of the current measure.

Reply

• Mar 1, 2019 - 23:27

You also didn't get specific about which version you have. Version 3.0.3 is significantly faster than previous versions (2.x or 3.x). As Marc said, if you have a score that is slow, attach it so someone can find out why.

Reply

• Mar 5, 2019 - 19:57

Hello,

I have the same problem as you (musescore 3 is very slow)
did you find a solution to make it faster ?

Thank you !
Rayen.

Reply

• Mar 6, 2019 - 07:52

Hi, this reply is intended for all those who commented in this thread.

First: I uploaded 2 scores couple of days ago (Marc requested that) - these were the wrong ones as Shoichi pointed out in his comment; I didn't look well when choosing the files. Then I uploaded the correct files in a previous post.
Second: I use Musescore 3.0.4.5763 (that seems to be the latest version). Compared to my previous version of Musescore 3, speed is a bit better, but note input with the mouse and selected multiple measures is still slower than version 2. There is a delay of 1-3 seconds. Also playback to the end of a score does not work or is "hazardous" (but not in all scores).
Third: It seems that all of these problems can be solved when displaying the score in page view, not in continuous view (and this was not necessary at all in version 2). Why that is so, I don't know. And I dislike the page view. The continuous view provides better oversight of a composition. As far as I am concerned, the developers should work on this "continuous view" issue.

Another thing I don't like about version 3 is the Play Panel, the time indicator font is way too small.

Best,
Danker

Reply

• Mar 6, 2019 - 13:09

There are various behaviors. One is that it takes up to 5 or 6 seconds before the playback starts (if at all), the vertical moving bar shows up a couple of seconds thereafter. Another is that (same as above) playback starts a couple of measures further than where I wanted playback to start (up to 10 measures further down the road); and yet another is that (same as above) there is a long standing distorted tone/ sound that doesn't stop by itself (sort of a frozen sound). None of this happens in page view.

Reply

• Mar 7, 2019 - 07:19

Well, you're right in a way, I've just now run a comparison: there is a longer delay in MuseScore 3 than is in MuseScore 2, but there is a delay in either, 3-5 seconds in MuseScore 2.3.2, some 10 seconds in MuseScore 3.0.4. Not sure why though, it is the exact same soundfont as far as I know.
But as mentioned, using an SF2 version of that soundfont and the delay is no longer there, so it is definiltly the SF3 soundfont and its decompression in the background being the culprit

Reply

• Mar 7, 2019 - 14:31

I experienced that when I start up the PC and open Musescore for the first time, the first file that I open has a delay of 2-4 seconds in version 2, 10 or so in version 3. Opening a file in version 3 or 2 after that there is no delay. But when I start playback towards the end version 3 behaves wildly weird, it jumps forward by 8-15 measures, then starts playing, sometimes not at all, or there's this loud standing tone that doesn't stop or nothing at all happens. It's all random misbehavior! That's not so in version 2. But none of these problems occur in page view.

Reply

• Mar 7, 2019 - 22:15

I was VERY unhappy with Musescore 3.2. I compose hymns and needed shapenotes and found out (too late ) that the program did not support the plugin. Went back to 2.whatever and am composing away.

Reply

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