Mixer interface

• Mar 23, 2021 - 08:22

The interface for sounds in the Mixer really needs improvement, IMO. Also maybe consideration should be made as to whether it should be linked closely to the score editor, so that changes to instruments in one are linked automatically to changes in the other - though maybe that's a step too far.

Where things seem to go wrong for me is where sounds are changed for instruments in the Mixer. Then reverting them back to the previous sound doesn't appear to work. Further since many of the soundfonts are based roughly on GM - but some would appear to use different numbered representations, changing soundfonts can really mess things up. For example changing from the standard MS GM soundfont to Musica Barocca, and then back again to the standard causes mayhem.

Having to scan through a large dropdown menu to find sounds which get incorrectly "reassigned" is really painful, and there really ought to be a sensible grouping of sounds in that menu - with submenus. Even better would be a search feature - which at least exists for instrument changes in the notation editor when an instrument change is required.

If things are going to be radically different and improved in MuseScore 4 then maybe we can wait, but really the menu for the mixer and issues relating to the Mixer and soundfonts are in need of serious improvement IMO.

It would also be good to be able to select sounds from different soundfonts - something which I think is not possible in the current versions. As many computers now have a lot more memory than comparable ones a decade ago it ought to be possible to support multiple soundfonts - or similar "banks" of sounds.


Comments

You can most definitely have multiple soundsfonts loaded at once. Their entries will be appended into the mixer sound list in the same order as they are listed in the synth window.

You will have to wait for MS4 for such a big UI change, as that is exactly one of the two main points MS4 is all about (see also https://musescore.org/en/MuseScore4 ). There are currently no further 3.x releases intended.

In reply to by jeetee

I have sat on this for a while. How does this work? Does MS scan through the loaded soundfonts in order (from the top of the list down - in the Synthesiser panel) to find one which matches? So if the top soundfont has (say) a trombone does that override any other soundfont which might be further down the list? Equally - if the only copy of a Zither is in a soundfont further down the list, will MS find that? If not then it's not possible to really use multiple soundfonts at the same time.

Even if it is possible that MS works in the way I've suggested, there is still a possibility that some combinations won't work without constructing a new soundfont which selects individual "instruments" from existing soundfonts.

I would be interested to know how (and if) it is possible to use instruments from different soundfonts in the same score.

In a command line type notation it surely wouldn't be too difficult to do this - but obviously MS is mosty now menu driven.
For example: Sound 1=Soundfont/instrument 1, Sound 2 = [Default soundfont /Instrument 2 etc.
Explicit mention of the soundfonts to be used would avoid that masking/overriding of instruments in different soundfonts.

Maybe this really is a feature - or a potential feature - for MS version 4.

In reply to by dave2020X

No, a trombone near the top of the list does not over ride one further down. I use different instruments from different fonts all the time. And parts of sfz fonts. The whole synthesizer thing takes a bit of getting used to, but it works.

In reply to by dave2020X

It's purely a number, as defined by General MIDI. Trombone is number 58 according to that standard, so the first soundfont that offers a sound with that number wins. Zither is not defined by General MIDI at all, so you'd have to select the sound you want to use yourself - MuseScore would have no way to know which sound from which soundfont might happen to be sampled from a zither.

Anyhow, yes, it's easily possible to use different sounds from different soundfonts in the same score - just select the sound you want for each instrument in the Mixer. The sounds are listed in soundfont order, so to get to your second loaded soundfont in the list, just scroll down until you see its sounds.

In reply to by Marc Sabatella

Experimentation suggests that running a loop on a score which is playing allows one to change the soundfont dynamically in the Synthesiser, and then also cross-check with the Mixer to find out what is actually playing.
Attached are two examples of screen shots of the Mixer and Synthesiser panels - taken from dynamic playback of a piece ostensibly written for piano.
In both the "instrument" apparently playing is a Piano.
in the file Mixer-Synth1 the Sound is Violino, while in Mixer-Synth2 the Sound is English Horn SJO.

It's not quite clear just looking at the images which sound fonts are being used, as in both cases the Symphonic Sounds.sf2 is highlighted. I think the soundfont being used is, however, the one at the top of the list, so the Violino sound is from the _Musica_Barroca.sf2 file.1

If everything is currently linked to GM midi numbers, it might be a helpful improvement to have that visible for each actual sound in the Mixer. That might at least help users to understand why in one system their Trombone might sound like something completely different, and in yet another not sound at all.

I agree that by an extended trial and error session it should be possible to assign sounds to instruments and to specific soundfonts, but this is somewhat haphazard. Clearly some "midi numbers" do not exist in some soundfonts, as there are instruments in some which disappear when the soundfont is changed.

Trying to link everything to Midi is perhaps not the best approach, though I agree that it would have been expedient at the time when MS was first being developed.

Trying to find instruments in the drop down list by typing the first few letters only works if you know that the instruments are organised in that way, and what names have been used. Examples: Kalimba might also be called Mbira in some soundfonts, and also some instruments may be much more effective soundwise than others. Yesterday I found one soundfont in which a Marimba (IIRC - which I may not) is a pretty decent representation of a Harpsichord - which is something which can only be found by experimentation. Experimentation is not necessarily bad, but if people are busy and want to do things quickly and effectively it really is not a good approach. If a more systematic and organised approach can be developed that would be preferable, and not just for developers, but also for users.

Over the holiday weekend I watched Martin Keary's video on Dorico (MK aka Tantacrul has inputs to MS). Some of his comments about Dorico and the general approach to music engraving also apply to MS - though if you search you'll also find an earlier critique by Tantacrul of Musescore. One observation is that it is difficult to come up with tools which solve so many problems - and also that there is a massive learning curve even for experienced users of notation software in trying to make things work in a different system. In the video he specifically addresses the issue of new users, though as an experienced user himself he also comments on problems that music professionals might have.

It's worth watching - if you have time - https://youtu.be/S-3wEC6Fj_8

I'll probaby refer to this and other videos by MK again.

Attachment Size
Mixer-Synth1.png 284.86 KB
Mixer-Synth2.png 313.97 KB

In reply to by dave2020X

As mentioned many times, there is really very little we can diagnose from pictures. We need a score and precise to reproduce a problem. But in this case,. we'd also need links to the specific copies of the soundfonts you are using in order to reproduce this issue you are perceiving.

And yes, we know that the dropdown list search mechanism isn't ideal, and that better organization of the Mixer would be desirable in the future. I'm just trying to help get the job done today.

Anyhow, this isn't really my area, so I wouldn't a good person to figure out what might be going wrong in your score or your system configuration, but maybe someone else can. All I can say is in general, it definitely works exactly as I've described.

By the way, it's not totally clear from your reference of his videos - you are aware the Martin is actually in charge of MueScore design and development? So it's safe to say we're all quite familiar with his videos, and that his suggestions will be taken into consideration for future development :-)

In reply to by dave2020X

I agree that the mixer is a problem. I think having two or more large fonts loaded is a mistake. It is indeed a chore to load a large score. Fortunately, my instrument needs are simple. However, there is no solo horn or trumpet sound in the General font. Or section sounds for flute, clarinet, etc. I have extracted those from other fonts. Your loop setup seems over complicated to me. I'm not sure how necessary it is.

"and also that there is a massive learning curve even for experienced users of notation software in trying to make things work in a different system."
There are days when I would consider this an irresponsible over simplification. I come from Sibelius. As a result I understand the basics of how notation software works.
On the other hand, of course there is a learning curve to figure out different software. It's even worse if you are trying to force one software to act like another one.

Note also that you don't have to visually scan through the dropdown list. Like all dropdown lists, it responds to keyboard searches by typing the initial letters of the element you are looking for.

It's not clear what you mean about reverting sounds not working. If you have a specific example of this, best to start a new thread with a sample score and precise steps to reproduce the problem, so we can investigate.

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