Jump to content

Seminar: Principles of instrumentation (1st exercise)

Recommended Posts


Welcome everyone to this little experiment, to test the waters so to speak. This is a little "seminar" I'm going to be doing on what it means to arrange/Orchestrate/reduce music to a composer, how to tackle it and what does it look like.

With that said, the first thing to get out of the way is the notion that instrumentation is something separate from composing, which is entirely false. Instrumentation is, in reality, everything that concerns the actual production of sound, be it electronically realized or with a full orchestra. Since "instrument" can be literally anything these days, it's prudent to take a more abstract approach than to think strictly in terms of violins and clarinets.

But what does it mean that instrumentation is part of composition, when we really get down to it? It means, quite simply, that assigning sounds is intrinsic to actually writing music. When you work on a piece, you automatically worry about the end product, the sounds you'll be working with when it's all said and done. We'll be concerning ourselves with this, and how in part it's unavoidable to understand instrumentation as a fundamental part of the composition process (whichever it may be.)

We'll also be looking at examples where instrumentation may be free, uncertain or random, and styles of music (and eras) where the conception of adding or subtracting instruments from a piece was viewed through a different perspective. However, I hope to integrate all this in the next sections as not to make this too long or boring to read.

All in all, practical purpose of any course or exercise that concerns instrumentation is to be able to translate abstract musical ideas between different sounds and settings. It involves, therefore, manipulation of those ideas on an intellectual level so it's expected that people attending this are at least somewhat familiar with music history and styles.

Part 1: The beginning exercise.

So after that intro, you're probably wondering what exactly all this means. It's very simple, we'll be dedicating this entire little thread to just one exercise. But, this exercise is meant to demonstrate the actual work involved in getting the skill to translate ideas across instruments, to think with the instruments you're working with as well as being able to think without them if necessary.

The exercise consists of taking one of the 15 2 part inventions Bach wrote and setting it to string quartet. Here are the rules, however:

1) You may only have two voices simultaneously, just like the original piece (doubling is not allowed save for ONE exception.)

2) Related to the above, you may not add other notes or extra notes. You have to work with the same exact notes as the original, nothing more nothing less.

3) Dynamic markings are optional, as the original doesn't have them.

4) Don't worry about bowing technique or whatever extended techniques that can be done. The only exception may be Pizz if you think it fits.

Sounds simple, right?

Part 2: But it's never simple, now is it?

Having the freedom to write the melodies in any of the four string instruments of a quartet seems very limited at first until you realize that you can't just go about throwing things arbitrary around. No, just like Bach's work has proportions and a thought-out form, you need to make sure that your quartet arrangement can translate some of this despite having a different sound altogether.

So here's a few things to consider:

1) Which instrument should start? Which should follow?

2) Should certain melodies be played always on the same instrument they originally appeared? OR should they instead jump from an instrument to another?

3) When a melody changes instrument, since this is the same voice jumping instrument, how can this "jump" effect be minimized? Could an option be that both instruments share a "pivot" note, making the transition smoother (this being the exception to doubling mentioned earlier?) If so, where? If not, why not?

4) How do you handle the endings? Should they end like they started, or differently? Do starting instruments here play a role in the decision?

5) How long do the melodies stay on a single instrument? Is it dependent on the motives? Or on a larger section of sequences?

And so on.

The important part here is not to just copy paste the notes across instruments, but to think in terms of what it would actually mean for the composition itself to be played with the quartet setting. Would it be something you find actually good to listen to?

The line between exercise and an actual composition is never clearly defined, jump on both sides of the proverbial fence to get a perspective on the work you're doing. Remember, just because something is an exercise it doesn't mean it has to be without artistic merit what so ever.

Part 3: You're done?

Now you post it in this thread. You can make use of the network upload and make a PDF for posting here. You must post it in this thread for others to see, this is an open course and everyone should be able to see others' work. It's not important that you make a sound rendition of it, since we are concerned with something which is very easy to see visually.

Plus it's also good that we detach ourselves a little from all the instant-sound rendition that goes on with today's software and practices. Hear less, use our heads more, so to speak.

I'm going to let this thread sit until enough people have worked on this exercise and posted their work here before I begin giving critiques and answering questions.

Have fun.

Link to post
Share on other sites
  • 2 weeks later...

Here is my arrangement. Tear it apart!

Invention no. 4 (String Quartet arrangement).pdf

Invention no. 4 (SQA).mp3

Link to post
Share on other sites

Tis the season to be wickedly busy. So SSC just asking when do you think things will slow down for you to check out our exercises. And oh for an idea of a piece from you - do you have an opera scena? I could forward it to a board member of an organization who will be doing workshops here in NYC.

Link to post
Share on other sites

Alright guys, time for some anachronistic Bach.

I do encourage you guys to check eachother's work, as you'll obviously have to if you want to read through what I'm going to be writing here anyway. But I mean it's interesting to see how the three arrangements are very different (as it often is the case with this kind of exercise, the result is always different depending on who's doing it.)

I'll do Jaw's first, since he seems to have gotten very creative with the arrangement.

===Invention 3 in D major==== (Why did you name it 4? 4 is in D minor!)

It's pretty clear the intention here is to make the most use out of the quartet while keeping with the only 2 simultaneous voice restriction. As such the voices jump a lot between the instruments and on some cases this is good and on others this makes actually following the melodic and motive bows difficult.

Examples of it being good is directly in measure 5 to 9, where he has the violins alternate on the same repeating figure which was most likely played on different harpsichord manuals (hence different sound.) In this case the figures themselves lend to this jumping around. The bass line being split between the viola and the cello like that is also interesting, if a little weird. The figure is also split like the right hand melody, which means that you could instead have it jump between both cello and viola, like the violins are doing (much like you would in the harpsichord too.) Instead he opts for a static arrangement where the instruments are doing a pedal. The effect is less dynamic, but it would sound "bigger."

I don't think it's a bad handling of the bass problem there, and the static balances out the jumping between the violins.

And for an example of how this separation could hurt the arrangement's musicality, there's measure 10-11. Clearly the idea to jump here to finish the first violin's line with the second is something that comes out of the previous sequence, but here it's important to remember that such a division can only work if the motives being worked with aren't part of a bigger phrase. In those two measures the ending phrase is cut between two instruments when the idea of that phrase, that is the interval characteristics of it, are meant to be handled in one single voice. Imagine those two measures with the same abrupt jump, but now played on a harpsichord's two manuals.

Why this is problematic musically is because when you have such small and tightly composed motives, specially like ending segments, the player (and by extension the audience) already expects the single idea to be handled as a phrase. This is similar to two people finishing off eachothers' sentences. You can do it for effect, but then you will have problems the moment you start demanding greater control over timing and melodic line in spite of the jumping around. Choirs are used to have jumping around in this fashion, but remember this is also a function of text to help cement a musical context here. Even if in a choir you can have one section finish the others' motive, it's usually bound together by more factors that just the melody.

With the previous example this isn't a problem since each instrument is playing the entirety of a small motive that repeats. With the bass figure the motive is split, but the repetitions ensure the musical stability of the arrangement. Here in those two measures (10-11) the violins seem to have just continued for too long with the exchange. It'd be better if only one of them handled the ending, and perhaps the next time it comes again (measure 22-23) have the other violin handle it.

Curiously however measures 22-23 you have only one violin handle the ending phrase, but again while it feels a little better this time around it's also worth pointing out that it generates an automatic inconsistency with the beginning (10-11) where the ending was split due to the construction of both endings. Musically their ideas are not different, structurally they are practically identical, but the arrangement treats them different.

Again with the same topic, I want to bring attention to measures 13-14-15 (right hand). Here you could interpret three different phrases, but looking at the bigger motive arc here there are only two separable motives. The starting jump is good to separate as it could very well serve as a pedal tone were this tone to be held, while the middle phrase split is a little awkward particularly because the motive itself is a bow. The phrase markings in these inventions aren't so much indications of where the actual musical phrases are, since obviously some of the motives are phrased longer than a single measure (such as this particular motive.)

With that said, however, the entire piece hangs on the motive introduced in the first measure, but this doesn't say that the musical intention is always handled in single measures. On the contrary, from that small motive there are larger structural formations that need to be treated differently than single independent motives.

As with the first example when the same figure is repeated on the bass, it's expected that the arrangement will reflect the first form used and here that's the case. Measure 47 to 50 reflect measures 5 to 9 at the beginning, so does the arrangement here which is good. What's not so good is the finishing two notes of the 52-53 ending motive. You ate measure 49 since yours is actually measure 50 in the original. Oops.

In the ending measures you actually do what I've been saying and have for the first and only time the entire two measure phrase on a single instrument (cello, measure 55-57.) The question here is, did you do this because you realized that if you kept up the jumping model it'd be a little too much for a rit. ending? But it then begs the question, why didn't you think this was the case with every other ending phrase cadence and the two-measure motive I already spoke about?

And finally to the overall use of instruments I like that you set out to start with the violins and end with the viola and cello, in a sense ending differently than where you started. However, it really depends if you see the piece in a bow-form where it would've been more fitting for the ending to have a violin. Though the last cello motive has a backwards dynamic bow which is a little awkward (having the p in the middle where normally you'd have the loudest point instead.) But in general I didn't look too much into the dynamics otherwise I'd be here all day.

All in all a good exercise, shows you actually put a lot of work into it.

===Invention 8 in F major====

This invention is also based on a two measure motive that is separated and used constantly, and as such the same observations have to be made as for the previous invention. The arrangement is very straight forward and I don't have much to say.

At the end however, in the closing measures the violins make jumps to the viola (probably because of the low register,) but this would then only mean you should seek to arrange this segment different to avoid having to jump to the viola in the middle of the motive. Also, in the ending you used all instruments simultaneously which I didn't really want since it made the choice of which instruments close the piece moot (it's always easier to have them all go for a forte ending than having to pick two only.)

In the sections where you used more than two voices (like measure 22) I assume it's because you wanted to keep consistency with the voice jumping to a different instrument and hence the previous just keeping a trio-like pedal. It's not a bad idea, since it's quite in style of what you'd have done at the time. In a sense however it turns the arrangement into more of a trio, which isn't what I was going for. But it's pretty easy to see you could've just erased those extra pedal tones and the result would be much the same so I don't mind.

Another thing that bugs me is that despite me not asking for dynamics, you wrote them anyway but you lazied out and half the piece doesn't have dynamic indications. Either do the whole thing or don't, since it's distracting.

In general it's a pretty by the numbers arrangement, which is good. It's got a couple of details that can be fixed, and you could've done a little more with the quartet (for example alternating instruments on the sequence in measure 21-22, etc.)

===Invention 2 in C minor====

This is a hard one to arrange, specially because it's hard to reduce the motives like you can with the other previous two, and since the theme is so long. Due to this, the arrangement ends up being split into large segments with instruments playing long phrases. It's not a bad thing either, considering too much jumping around in this invention would've killed any perception of motive unity and structure. But what other alternatives can there be to handling such a long subject and motives that are based on it?

This arrangement stands in direct opposite to Jaw's, who tried to give the quartet full use through splitting every single motive in cells, CO here actually handles it by larger sections. This has the benefit of it being easier to hear the form and musicality of the lines as they remain on instruments longer, but it also takes away some of the dynamic that changing instruments brings. In the bigger picture, they are both valid approaches, but what we're concerned here is to see if both can communicate the original idea. The F major invention arrangement is safer and is a little in between both, having a smaller motive but being divided also in medium sized portions for each instrument.

In all instances where the instruments change in this arrangement it's always due to a new phrase starting or ending, which is fine. It's a good way to keep instrumentation close to the musical material, and it makes it easy to hear the changes in the motives and phrases as they're also changing in the instrumentation. This kind of "group" instrumentation is what ends up being the staple of instrumentation by the Vienna school, where changes in orchestration are always tied to form.

But then there's a couple of instances where I'm not sure why the instrument change. Measure 26 right before the end (right hand) why the small motive on the second violin? It doesn't anywhere else in the arrangement, so why there? Which brings me to my next point, the sequences are handled unevenly. The motive in measure 3 is handled differently in measure 23-24, why? It's a little too close to the end to make that kind of change without some kind of logic behind it. This fragmentation of the sequence can work OK as seen in the first arrangement, but you need some consequence to it or you're actually making the instrumentation point out something different than what is there in the motives.

It shouldn't be the case that instruments feel fragmented without a structural meaning.

But otherwise, it's also an OK arrangement. The ending is also fitting with the violin ending how it started. It'd be only better if some corrections were made to the ending or a different approach was used to handle sequences which was consistent to the current ending.


Yeah, that took a while to get through.


Link to post
Share on other sites


Thanks. Looking at it now, it was a mistake to give the line to the cello in mm 23 so near the end - I had wanted to do an alteration in color with the motive at the beginning but I thought it best to keep the color uniform. Near the end, I was looking for a change of color but somehow take into account the viola's range - forgetting the structure. Also, the switch off near the end between 1st and 2 nd violin was to find a way to emphasize the dominant to tonic. As you noted, it was hard to decide what approach to take - assign long lines to one instrument or divide it up. Sticking with an approach to the end of a work is sometimes hard for me to do (call me ADD addled).

For some reason I am disallowed to view jason's pdf. I would have liked to see what he did.

As for questions - in my arrangement - would it maintaining the long line if the descending ornaments in the countersubject were split between two instruments? For example pick-up to measure 4 taken by the second violin, and then have the first violin take the pick-up to beat 4 in the same measure. Later, mm 21 - 23 viola, and then cello takes the latter part of mm 23. Then at the measure before the end, the violin 2 taking over earlier (pick up to beat 1 of mm 26) and ending in the same place would then make sense. I aks as in an earlier sketch I thought of doing this.

Link to post
Share on other sites
This topic is now closed to further replies.
  • Create New...