Assigning a staff to more than one sound

• Feb 1, 2023 - 10:04

I have been testing MS4 - most recently with vocal sounds. Recently I also bought the Spitfire Epic Strings library, and that also works with one of my test files - substituting string sounds for voices.

It occurs to me that it would be quite useful to have some of the [original] vocal lines in the scored doubled by string [or even other ...] sounds.

I can clearly do that by generating new staves in the score, copying them, and assigning each stave to a different sound.

Is there a way to get this desired result - "doubling" of a staff with different sounds, without having to create duplicated staves? The new mixer seems more sophisticated than the one in MS3, so I wondered if this is possible.


Comments

The problem that I see with this is that you would lose control over the the individual sounds. What if one sound is too loud or soft. Not to mention panning.
It's an interesting idea. I'm not sure if there is a way to make it work.

In reply to by bobjp

I don't think that's insuperable by any means. As far as I can see most DAWs provide something very similar, and they do work. I can check this again with several of the DAWs I have working - but I'm pretty sure it's not a major issue. It may complicate the "behind the scenes" working and increase the processor load, but otherwise it'd probably be feasible.

I can see Marc's point about engraving vs. playback, but maybe it is possible to have both. Also, sure - if the score is finalised then it will have to be engraved, and parts produced for the whole score, and for the parts, but while it is being developed a feature such as suggested could be helpful for some creators. Unless harmony is very complex, many pieces do tend to have some duplicated note sequences across several staves, so why not exploit the technology to make things easier?

In reply to by dave2020X

Anything is possible indeed, but again, it's about prioritization. MuseScore focuses on optimizing things actually needed in production scores, and there is already a very long list of features to add or refine and bugs to fix just to reach that goal. So for every feature not relevant to that task that the developers take time out to design and implement, something else that is vital to real world usage has to wait that much longer. And that's why it is crucial to understand and assess what the actual use case is.

In reply to by bobjp

That is not really a problem at all. The section samples are balanced by the engineers when they are recorded. They are already panned and have extensive adjustments for microphone placement and room acoustics. Most even have attack and release settings as well as various articulation settings. Besides, if you need to bring out a single "solo" instrument, that is very simple to do simply by layering the solo instrument. So actually you have much, much more control than you would working with the stock MU-Sounds.

This has come up for discussion in the planning for 4.1, but so far no one has succeeded in finding a real world use case for this. Normally, in published music, vocals would never read off the stroing part or vice versa - they'd be separate parts. But it's certainly possible some sort of use case might exist for this - cases where there really should be only a single part but it should be connected to two different sounds. If someone can think of an actual case in published music where this would make sense, this would be a good place to show examples.

In reply to by Marc Sabatella

The use case would be for composing music. Typically composition and orchestration are two different steps. First create a sketch and then afterwards come back and orchestrate the sketch.

I recently posted a simple example of how this works using section sounds, which were available in MU3 (but missing in MU4). You can do the same thing in MU4 if you ditch Muse-Sounds and use something like Spitfires "Epic" series of strings, winds, and percussion.

The way you would go about it in MU4 is to set up a separate staff for "Woodwinds & Brass", Percussion, and Strings. You then assign Woodwinds and brass to the first staff, percussion to the second staff, and strings to the third staff. In MU4, each staff can have sub-staffs that share the same sound sample so it is easy to have several treble and bass staves. Then you can easily sketch your composition ideas.

Once you have the ideas for your composition outlined in the sketch, then you can orchestrate the ideas on a normal score setup using which ever sound samples work best to convey your musical ideas.

This would be very simple to use if Mu4 would allow choosing a specific sample or would add section samples to their selection. As it is, it will work very well using the Spitfire Audio "Epic" libraries which are very inexpensive. Another issue is that MU4.0 does not allow changing the names of the sub-staves, but for a sketch, that can be handled by using staff text (or by simply remembering since this sort of sketch is not generally intended for public distribution).

See the attached template.

Attachment Size
Composition Sketch Template.mscz 20.71 KB

In reply to by bobjp

I sketch for the same reason I outline when I am writing prose. It keeps me on track with the main idea. If I don't have a pretty clear idea, I can get lost in all the nice licks and forget where the piece wants to go. I think of "Music" as what is left after I delete all the stuff that is not music. So I like to work through that before I waste a lot of time orchestrating.

In reply to by Jim Ivy

Don't get me wrong. I'm not against people sketching out tings first. It's just not the way I work. When I sit down to write, I seldom have an idea of what I want to do. I don't write melodies and then accompaniments. I believe they are integral. I might write a measure or two of what might be a melody. Then I might add a base line. But where those to things lead determines the next part of the melody. That way neither is an afterthought of the other. They depend on each other. And to make it work, I have to hear it on the intended instruments. I come to measure 36 and I think that a flute melody with string backing would be nice at that point. So I make it happen. But right after that I need a double horn part over the full orchestra. I tend to think in terms of the instruments that will make my ideas work, first. Of course, I make changes along the way.
I know it is standard to orchestrate from a sketch first. Oh well. The composition course I took in college didn't work that way. We could if we wanted to, but it was the end result that counted. We were graded on the results, not the process. This was a class that could only be taken by students that had already taken 3.5 years of college music courses.

In reply to by bobjp

Well, you have to do what works best for you. For me, it works best if I have an idea what I need to accomplish. Sometimes my clients ask for a sketch - many times they have specific things that they want the music to do. So, in those cases, it helps to have something that I can share with them.

If I am just tossing off a piano or combo arrangement, it is not as necessary as if I am doing a more involved multi-movement piece. But even with a simple piece, I like to know that the beginning, middle, and end are in proportion. I find that having a plan helps me to know when a figure should transition to the ground, that sort of thing. Like a novel or movie, music is a dramatic structure. I can't imagine sitting down to write even a long essay without have a plan, but that is just me. Everyone has their own style.

In reply to by Jim Ivy

I do understand. I get to write what I want. Not what I have to.
An essay? I'm not sure I can put two words together. And there are certainly those who are sure I can't put two notes together. Oh well. I don't (or can't ) write long pieces. Five minutes is about tops. Past that I feel like I'm writing filler. I make my point and get out. Having too much of a plan takes the adventure out of it.

In reply to by Marc Sabatella

Not sure if this is now the right place to post this. Maybe what is really needed is something on how to allocate Musescore instruments to more than one sound - possibly switchable.

This really is needed - as users might want to assign instruments in a full score to different libraries. OK - that's possible now, but it may also be very helpful to be able to switch instrument sounds dynamically from one library to another. That could be useful for several reasons:

  1. To check articulations. Some libraries have poor articulation, even if the basic sounds are good. Currently that's where some of the Muse Sound instruments are. Some libraries also use different ways to switch articulation, so being able to check compatibilities would be very helpful.

  2. To check levels. Ideally it would be great if all nominal instruments in each library generated very similar output levels - i.e pp in one has the same effect as in all the others - similarly for ff. This is not currently the case, so mixing libraries to get maybe a "better" sound [which is of course subjective] may give the desired sound quality, but unbalance the mix.

  3. To add "colour" to some instruments. For example, the Spitfire Discover library has good trumpet sounds, but they are very classical. OTOH the Orchestral Tools library - accessible via SINE Player has an interesting jazz style trumpet. It is inappropriate for much classical music, but it could be used together with the Spitfire trumpet to give a good effect.

At the moment instrument/sound linkage is done in a static sort of way, which I think requires stopping playback, resetting each instrument to a sound, or collection of sounds. However, if this could be done dynamically I am sure that some musicians would find that very useful. For example, 6 libraries could be preloaded - or at least set up for loading - and one 6 bit data value could choose out of the 64 possible library combinations - and if that could be done "on the fly" that would be really great.

At the moment trying to get this kind of analytic way to use sounds would require having scores with more than one stave - generating multiple copies of "an instrument", then assigning that instrument to different libraries. It would then be possible to hear the effects of the different libraries by using the Solo [S] or Mute [buttons] in the mixer, but this procedure could be simpler if libraries or combinations of libraries could be swapped rapidly while playing back the audio.

In reply to by dave2020X

I do this all the time using MU3 with Reaper as the sound engine. Others do it using Logic and other DAWs as the sound engine. All you do is route the midi out of MU3 to the midi in of your DAW, the load the sounds you want to use in the DAW. You can layer as many different sounds and effects as your processor and memory will support. You control the balance between the various layers in the DAW. Once you get a setup that you like, you save it as a template and you can use it over and over again. There are a number of how-to articles and videos on the web that explain how to do this with different OS and DAW combinations.

In reply to by Jim Ivy

I'm not denying that this can be done now, using additional tools. However what I'm suggesting is that the means for doing this kind of activity can be incorporated into MuseScore, and that a better - more sophisticated - interface could allow comparisons to be made very quickly and dynamically.

At a meta level, when considering changes and new designs to MuseSccore, decisions could be made not just on - "can we do it now?", or "could a knowledgeable person with expertise in other software?" do this with some modest effort, but rather "could we come up with systems and interfaces which would make some operations very much more effective for the human user?"

I can do this kind of thing now, without a DAW, by using multiple staves , copying and pasting, and careful use of the mixer - assuming that doesn't break the software at run time, but it's a lot more work than assigning several instruments to one staff and being able to switch the sounds dynamically at run time.

In the meantime, even with current software - several DAWs etc., I can't even find an easy way to reverse stereo channels, so there are even some very simple operations which are tricky to do for anyone who isn't an expert.
OK - I can just put my headphones on the other way round!

In reply to by dave2020X

In response to my own last comment I found this which works with MU4.

From Raising Jake studios - https://www.kvraudio.com/news/raising-jake-studios-releases-stereo-swap… though why such a simple plug-in has to be explicitly provided by third parties I can't say. Surely this kind of functionality should be built into integrated sound systems, such as DAWs and MuseScore 4. Having written that though, this plugin works surprisingly well, and does give some new insights into the sound even of previously heard works.

RJ StereoSwap copy.png

The image shows what it looks like in the MU4 mixer.

In reply to by Marc Sabatella

Example follows as requested by @MarcSabatella, but firstly a big Thank You to Marc, Jojo and the myriad others who do such a fantastic job of giving the world such an excellent piece of software - I salute you all! And here's the requested example... many Bach (Cantata) Chorales double instruments with voices on one staff - attached is final 2 pages of the NBA full score of BWV187, but other Cantatas have similar... Top line is Soprano + Oboes + Violins I, so that's three voices needed (!), then it's Altos + Violins II, then Tenors with Violas (and Basses get their help from the continuo part, which is separate, presumably to stop the figured bass from interfering). Of course, the printed violin, viola & oboe parts probably wouldn't have the words (although they should really, to help them get the phrasing right, but that's a different question...).

Attachment Size
BWV187 final Chorale.pdf 860.92 KB

In reply to by Marc Sabatella

There are two specific use cases that I can think of. One is for a four-part choir condensed to two staves, in which case the soprano and alto voices are written as separate voices on one staff, while tenors and basses are written as separate voices on another. It would be useful to assign each voice the appropriate sound.

The second use case is on string parts that include a soloist who isn't the main feature of the piece. In these cases, the solo will usually appear as a separate voice on the same staff that the ensemble uses, with each voice assigned by text reading something along the lines of "solo" and "others." The point where the soloist rejoins the group is then marked with the word "tutti." An example of this is on page 1 of a piece called Incantations, by Richard Meyer (it's sold by J.W. Pepper, and a preview of the score showing this notation is available at https://www.jwpepper.com/Incantations/2478662.item). Being able to split the staff into different sounds based on voices would be useful in this context because then the solo sound can be applied to one voice while the ensemble sound is given to another. This would be especially useful with the voice-specific dynamics you are planning to implement in Musescore 4.2.

Considering that text objects are linked to notes, perhaps the most flexible way to implement this would be through the addition of dedicated "solo" and "tutti" texts that are programmed to function like an instrument change, but limited to only the voice it is assigned to. That way, voices 1 and 2 can also be used in other areas for divisi.

Currently, there is a workaround for this, but it is somewhat awkward and error prone. It requires writing in the intended notes as a separate voice on the desired staff and manually disabling their playback one note at a time, then duplicating these notes onto a separate instrument and making them playable. This dedicated staff is made invisible for printing but kept in for audio rendering. Since you can't selectively copy/paste specific voices and therefore must make all changes twice, it is possible to make a mistake that will cause the audio playback to not match what the printed score shows.

An additional complication - though sometimes useful - is the use of some VST libraries for particular lines in a score. These may sometimes be generic - e.g. Woodwind, Brass - which can be useful for getting a quick mock up of what a piece may sound like - but those generic patch libraries take away quite a lot of the fine control which a composer or arranger has if he or she works with multiple staves. For example, they may eliminate the particular timbre of a [hopefully] well chosen instrument - [a flute sounds different from an oboe or clarinet, or even a cor anglais for example], and some of these ensemble patches can also change [or at least appear to change] the voicing of chords, so effectively change the voice leading and harmonies - perhaps even slightly.

AFAIK Musescore 4 does not include any such generic ensemble libraries, though it probably could if there were felt to be a need for that. Also - and it does seem to be topical right now - if in the future any form of real time AI instruments were to be included in the mix - then such could be both a blessing and a curse. It would be a blessing for anyone who wanted to get something up and running quickly, though there would be issues about repeatability [would the same passage get treated the same on each repeat?] and it would be a curse for anyone who really wanted to have more control over what is written. I have not seen so much about this in serious music production - though I bet some groups are working on it. In the photographic image domain I have seen some amazing images - which on close inspection can be identified as not real, but they do look very genuine at first glance, and may indeed have some aesthetic merits. This has of course also been a recent source of discontent within the film industry, with actors not wishing to have CGI versions of their images used because it is convenient for the production teams, and there are also very serious concerns about revenue generated from the use [misuse?] of AI.

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