Volume levels are low on playback
Launch Musescore 4.01 (OS: Linux Mint 20.3, Arch.: x86_64, MuseScore version (64-bit): 4.0.1-230121751, revision: 9b70a8c)
Open score with choral voices using MS-basic (or MuseSound) voice instruments
Set playback mixer levels to 0 dB
Play score - sound is considerably less vs. the same score played in MS3 (188.8.131.528021370 rev OS: Linux Mint 20.3, Arch.: x86_64, MuseScore version (64-bit): 184.108.40.2068021370, revision: 3224f34)
If dynamics are included in the score (e.g. mp, p, f), the volume does vary accordingly but softer dynamics cause the sound to become effectively inaudible.
Workarounds: Don't include dynamics in score for playback purposes (save an an alternate version of the score with dynamics removed), increase mixer settings to the top of their ranges.
It is indeed a known issue that choral voices are currently very quiet.
This is a bit more general matter. When we open a MS3 score, which does not contain dynamics directives, and play it with MS Basic the sound levels seem to be OK. When we replace MS Basic with MuseSound, the MuseSound instruments play with a much lover volume. We have to write the forte or other dynamics in score for these instruments.
I have not made any all over test to hear if this behavior happens with all MuseSound sounds.
I would expect that the default sound levels are equal no matter what source (MS Basic, MuseSound or VST3) when we play a basic score (one without dynamics directives).
Elliot wrote that the softer dynamics are almost inaudible. So a thorough dynamics and sound levels checkthrough might be necessary.
You shouldn't be changing the dynamics just to hear louder playback. It's true Muse Sounds has greater range, so soft passages will be softer. That's a normal part of audio. If you'd like to hear things louder, simply turn up the levels in the mixer.
If your score currently lacks dynamics at all, though, then of course, you should add them if you want to control how loud human musicians - and MuseScore - actually play.
I can confirm that in 4.0.1 and 4.0.2 under MacOS, using the new voices renders all songs inaudible compared to the grand piano/chords. Using the old MS3 boxed voices solves the problem. It happens with old MS3-scores that are imported as well as with new scores created in MS4.
Instruments such as flute, oboe and clarinet also appear unbalanced compared to the piano.
This is an important issue to me, since I use MS mainly for choral and vocal scores with or without a piano part.