MIDI Input Lag - Issue has still not been fixed after 6 years

• Mar 15, 2022 - 05:05

Hi all, this stupid issue with MuseScore has still not been fixed.
With a high-quality USB MIDI Keyboard (Roland A-49) in my case, with MuseScore Stable 3.6.2,
you get around 300ms of input lag which is especially bad when trying to note input.
MuseScore simply erases all the rests and puts all the notes together, and even the audio lag is just unbearable.
Using ASIO4ALL isn't a solution, it's a work around that isn't part of MuseScore.
Why is it so slow with WASAPI and MME?

Why hasn't this issue been fully addressed since day 1?
It's been 6 years since this issue has been first reported, from what I can tell.
No clear fix has been made, only certain people claiming it isn't an issue, or staff claiming "duplicate issues" or "lack of information" to continue. MIDI Input lag is real, and you can't mark an issue as duplicate if the previous issue was marked as closed.

Please get this addressed.


Comments

In reply to by Jojo-Schmitz

https://musescore.org/en/node/103691
https://musescore.org/en/node/281924
https://musescore.org/en/node/177581
etc.
Before you dock the second one, the person who replied and the OP are different.
There's plenty of issues talking about this.

I have tested multiple different softwares and ASIO/without ASIO, and seems that Finale Notepad 2012 has the lowest latency without ASIO, and Ableton/FL Studio has insanely good latency with ASIO. Musescore is at the bottom of both lists.

In reply to by Jojo-Schmitz

There has been attempts, especially with other people finding out about how the MuseScore ALL Notes Off issue was adding latency.
https://musescore.org/en/node/322301
It's possible to fix this in the source but I'm not familiar with the codebase yet, so I'm not qualified.
It would take weeks to get myself familiarized with the codebase.
How irresponsible is it to tell the end-user to fix the issue by diving into the source code, rather than elevating it to a proper issue in the dev team?

In reply to by Jojo-Schmitz

Open Source doesn't mean the end user has to fix it.
If you want me to take the weeks to navigate the source code, sure.
But how incompetent is it really to drag in a new person to fix an existing issue,
rather than taking one look at the issue yourself?

Yeah, I'm incompetent since I haven't taken one look at the source code yet and thus cannot make changes until I'm familiar with the codebase and QT compilation.
Heck, I don't work full time as a developer, I'm a music producer that knows how to code.
You who is apparently a developer can't take one look at the source code and tell the open source community to fix it instead of taking a look.

The attitude here makes me laugh. You had a similar attitude with everything else earlier, I distinctly remember you docking me and other people for publishing pieces with invisible scores when the website was rampant with score stealing and plagiarism.

In reply to by Apprisco

I'm kind of surprised that JoJo is still responding to you. Think about it. "Developers" are very busy working on V4. And have been for more than a year. Was your issue "fixed" two years ago? No. Sorry.
Open source does not mean that the user HAS to fix an issue. It means that the user CAN fix the issue. If, as you claim, you can code, then maybe it would be worth your time to figure it out and stop insulting folks. If, as you claim, there are others with this problem, what you come up with might be helpful for them.

In reply to by bobjp

The point is that bug fixing is always more important than new features. Your logic doesn't hold up- you are telling me to fix the issue because I can, when I have to sacrifice so much of my own time to read through & get acquainted with the codebase for a program I'm actively paying to support. Again, the issue and exactly what causes it has already been discovered, it seems the developers have been alerted and lliterally has the solution implemented, but no one else other than the specific dev can access it because the people who were working on it moved to emails. Being able to code & being able to dive through a 263MB codebase to figure out an unfamiliar library & entire system are very different things. I'm being insulting because honestly the effort used to push terrible website updates could've easily gone into fixing these bugs. It's been found for 2 years. What has been done about it, when there is a solution? Nothing.

In reply to by Apprisco

> "The point is that bug fixing is always more important than new features"
If that were a truth, most software package would have to stop innovating.

There is also triaging of priorities, and while this seemingly bug has been (wait for it...) bugging you for a long time, it seems to be affecting only an extremely small number of users and is part of a not too often used feature compared to the entire user base.
Yes, that might be infuriating, but that's also how triaging works (even more so in an open source project where much of the work is being done by volunteers).

In reply to by Apprisco

The way open source software works is, anyone with the skills to fix a problem is welcome to do so. And lots of people have those skills. So the problems that get fixes are the ones that bother those people the most - either because they personally are affected by the issue, or because large numbers of users are politely and respectfully reporting that the issue is important to them. Apparently none of the people with the skills to fix this happen to be affected by it personally, so it comes down to the second part - having large numbers of users politely and respectfully reporting that the issue is important to them.

Developers certainly want to see MuseScore be successful, so they definitely do prioritize features that are reported politely and respectfully by large numbers of users. So far, that hasn't happened with this particular issue. That could certainly change, and you can help make that happen.

In reply to by Marc Sabatella

Hi, my name is Jeremy - Are all you nice folks referring to MIDI input delay when MuseScore 4 is trying to record or process MIDI signal in "timely" fashion from keyboard (or any external MIDI input source)... Such as playing a piano note (MIDI) and then very long delay below 1 second but definitely not realtime (I just need few milliseconds) delay that doesn't affect the score. I get the MIDI note but I can't play with the delay coming after my very fast hands, LOL (piano player here).

Compared to all my other apps, Cubase/FL Studio etc....I can't get any setting to work. I'm right here at studio desk with 3 foot USB midi out to it. Even old-fashioned midi cable routed to other things. I play a note, get about an 8th to 16th note delay. I know it's not my equipment as all fine and realtime in every other app....

If I'm in wrong forum, may someone post to correct. I can't fix this and need help. For now...I just record what I want in Cubase or FL Studio as it actually works, export them and import to Musescore 4 to finish things up. That is not supposed to be how a scoring and amazing program should be. Oh, and....Musescore 3 doesn't have this issue. Thank you all!

I am very fond of MuseScore vs Cubase score arranging. Now on to another forum about merging all to a grand staff vs all these instruments on separate tracks..

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