Augmented instrumentation

[Dave2002]

Clearly since (maybe) the 1960s (or before) composers and performers have been experimenting with electronics, and perhaps mixing the sounds of synthesisers with other more standard instruments.

Synthesisers have in the past tended to have quite distinctive and artificial sounds - think Tomita and Carlos - Switched on Bach, but nowadays some electronic kit can make a fair pass at producing sounds which are quite close to some standard instruments. Indeed in some cases it may be possible to produce sounds which are close to an instrument, but which might nevertheless be very difficult to play on a physical instrument.

I'm guessing that theatre musicians may have experimented with this kind of thing already, due to cost constraints and other practical considerations.

While Mahler managed to bring employment to mandolin players, and also to many other players for his large scale works, works on a very large scale, or which use unusual/specialist instruments would generally be rather difficult to put on - if only for reasons of expense. Now that electronic equipment is becoming cheaper is there a scope for smaller orchestras, but with a few musicians armed with suitable computer kit to augment orchestras or ensembles, so that either new instrumental effects can be achieved, or a much larger sound?

The results would probably not be as good as if every musical line had a dedicated and skilled performer, but might be sufficiently good that enjoyment would be given to those willing to listen - and pay.

[Richard Barrett]

It's a shame you seem to take so little interest in what's actually going on in contemporary music, because everything you mention has been going on for many years.

[Dave2002]

It's a shame you seem to take so little interest in what's actually going on in contemporary music, because everything you mention has been going on for many years.

I am sure it has, but not in areas which I am either exposed to or have an interest in. There is relatively little broadcast on BBC channels which represents some of the things I have mentioned. I suspect that in the pop/commercial world there is a lot more use of technology, and I have already mentioned theatres - where possibly cost considerations would be an additional driver. Opera - well some companies have large amounts of funding, but probably don't do much that's too innovative regarding the sound, though may invest in complicated on stage machinery.

There are people who develop new instruments, or try to improve older designs. Probably many of us are aware of the technology which "improves" recordings - particularly in commercial genres, but also for some classical music performances. To some extent that may be what distinguishes classical music and musicians from others. Many players really do have tremendous skill, which works most of the time. I suspect that if a flute player (for example) turned up at an orchestra rehearsal, not with a flute, but with a box which could read the music, and produce exactly the sounds that he or she could, and more, that there would be problems. If the same player turned up with something which looked like a flute, but had a slightly "better" tone, and only small mechanical modifications, so that some notes became easier to play, or more in tune, then such a developement might be welcomed or tolerated.
What would happen if someone turned up with something which looked like a flute, but was capable of producing sounds which any normal player on a "normal" instrument would not have a hope of playing, or playing at speed, I can barely imagine.

The same would apply to most instruments and players in any orchestra. At the moment I think there is agreement, at least amongst some musicians, that skilled players and "real" instruments are better, but that may change. There could also be some considerable skill in playing extended instruments, or orchestras augmented in new ways.

I forgot also to mention artists producing sound installations, including some interactive ones, for venues like art galleries. They may represent a small part of the creative work force, but some are definitely working in that area, as I am sure you are well aware.

[Serial_Apologist]

What would be the purpose in subordinating music to technical reproductive facilities, other than to meet some nightmare scenario productivity goal that dispensed with the expense and trouiblesomeness of live human performers? There are areas of music in which computerised means of sound production have achieved degrees of complexity which apparently lie beyond human capability. But in thinking of Techno, Drum 'n' Bass, Jungle and associated genres, such computer-generated metres and rhythms are really speeded up variants on Funk - artificially-contrived variants on pre-existing types which some drummers have then sought to adapt within live human-scale performances, where genuine interaction can take place with audiences and between performers, rather than having the IT effectively dictating main prerogatives, which is akin to a dead end for real creativity.

[Richard Barrett]

What would be the purpose in subordinating music to technical reproductive facilities, other than to meet some nightmare scenario productivity goal that dispensed with the expense and troublesomeness of live human performers? There are areas of music in which computerised means of sound production have achieved degrees of complexity which apparently lie beyond human capability. But in thinking of Techno, Drum 'n' Bass, Jungle and associated genres, such computer-generated metres and rhythms are really speeded up variants on Funk - artificially-contrived variants on pre-existing types which some drummers have then sought to adapt within live human-scale performances, where genuine interaction can take place with audiences and between performers, rather than having the IT effectively dictating main prerogatives, which is akin to a dead end for real creativity.

For what it's worth the way I see the issue is this: the advent of electronic music has from its beginnings engendered new ways of thinking about instrumental music, just as, in the course of the seventeenth century, the possibilities of vocal composition expanded through the influence of the development of independent instrumental music. One example, as you say, is the way some drummers have absorbed the influence of digitally generated rhythms, although there are examples going right back to the 1950s where Stockhausen's experience in the studio had a decisive and well-documented influence on the way he conceived and realised instrumental music. There's no reason to think about digital techniques supplanting instruments any more than improvements in instrument-building put paid to vocal music three centuries ago. On the other hand, in areas such as film composition where most listeners are assumed neither to know nor to care about the difference in sound/articulation/expression between acoustic instruments and digital simulation of them, you might be surprised at how much simulation actually goes on. This is one reason to be involved in forms of music which emphasise the physicality of acoustic instruments and their relationship to the players, and indeed taking a comparably idiomatic approach to electronic music, in using its resources to do those things that instruments in principle can't do. Given the rarity with which new orchestral music is likely to be performed, creating such complex and multilayered bodies of sound using digital technology might seem to be an attractive solution, but if the result just sounds like an imitation of an orchestra what's the point?

[ahinton]

For what it's worth the way I see the issue is this: the advent of electronic music has from its beginnings engendered new ways of thinking about instrumental music, just as, in the course of the seventeenth century, the possibilities of vocal composition expanded through the influence of the development of independent instrumental music. One example, as you say, is the way some drummers have absorbed the influence of digitally generated rhythms, although there are examples going right back to the 1950s where Stockhausen's experience in the studio had a decisive and well-documented influence on the way he conceived and realised instrumental music. There's no reason to think about digital techniques supplanting instruments any more than improvements in instrument-building put paid to vocal music three centuries ago. On the other hand, in areas such as film composition where most listeners are assumed neither to know nor to care about the difference in sound/articulation/expression between acoustic instruments and digital simulation of them, you might be surprised at how much simulation actually goes on. This is one reason to be involved in forms of music which emphasise the physicality of acoustic instruments and their relationship to the players, and indeed taking a comparably idiomatic approach to electronic music, in using its resources to do those things that instruments in principle can't do. Given the rarity with which new orchestral music is likely to be performed, creating such complex and multilayered bodies of sound using digital technology might seem to be an attractive solution, but if the result just sounds like an imitation of an orchestra what's the point?

Very well said!

[Dave2002]

What would be the purpose in subordinating music to technical reproductive facilities, other than to meet some nightmare scenario productivity goal that dispensed with the expense and trouiblesomeness of live human performers? There are areas of music in which computerised means of sound production have achieved degrees of complexity which apparently lie beyond human capability. But in thinking of Techno, Drum 'n' Bass, Jungle and associated genres, such computer-generated metres and rhythms are really speeded up variants on Funk - artificially-contrived variants on pre-existing types which some drummers have then sought to adapt within live human-scale performances, where genuine interaction can take place with audiences and between performers, rather than having the IT effectively dictating main prerogatives, which is akin to a dead end for real creativity.

I agree with pretty much all of this, though I think it would be possible for there to be human input, which would shade the results.

I'm not thinking of robotic performances on synthesisers or computer kit, but if you look at the way even some practitioners on modern synthesisers work, you might realise that there can be significant human expression. There are also what might be considered fairly minor effects, but which weren't possible on other instruments. One such is so-called aftertouch on keyboard instruments. With a piano, once a key has been struck there isn't too much which can be done after that, other than to move on to the next note. With a synthesiser with aftertouch, other effects are possible. Whether those effects are useful, or desirable, I can't say - but some people seem to like them.

I'll give one more example of where some electronic assistance might help - though this would perhaps be unusual right now. Suppose a player has practised a section to death, and is now in a live performance. Most of the performance goes well - using the physical instrument, but then the player realises that he/she can't manage the next bit - so just hits a button and the augmentation kit takes over a for a few bars. In recordings this isn't a problem, as the player would just carry on, or skip, and another take could be inserted later on, but that doesn't work for live performances. Note that this could also impact on composers, depending on how they want their music to be presented. Some composers might be happy to write almost impossible sections, in the hope that sometimes they would come good - and if recordings were to be made that might not matter, while others might want to ensure that all the music was feasible for musicians playing in a live situation.

I am not proposing subordinating music to technical reproductive facilities, but augmenting the capabilities, and retaining the human elements of control.

[Richard Barrett]

I agree with pretty much all of this, though I think it would be possible for there to be human input, which would shade the results.

I'm not thinking of robotic performances on synthesisers or computer kit, but if you look at the way even some practitioners on modern synthesisers work, you might realise that there can be significant human expression. There are also what might be considered fairly minor effects, but which weren't possible on other instruments. One such is so-called aftertouch on keyboard instruments. With a piano, once a key has been struck there isn't too much which can be done after that, other than to move on to the next note. With a synthesiser with aftertouch, other effects are possible. Whether those effects are useful, or desirable, I can't say - but some people seem to like them.

I'll give one more example of where some electronic assistance might help - though this would perhaps be unusual right now. Suppose a player has practised a section to death, and is now in a live performance. Most of the performance goes well - using the physical instrument, but then the player realises that he/she can't manage the next bit - so just hits a button and the augmentation kit takes over a for a few bars. In recordings this isn't a problem, as the player would just carry on, or skip, and another take could be inserted later on, but that doesn't work for live performances. Note that this could also impact on composers, depending on how they want their music to be presented. Some composers might be happy to write almost impossible sections, in the hope that sometimes they would come good - and if recordings were to be made that might not matter, while others might want to ensure that all the music was feasible for musicians playing in a live situation.

I am not proposing subordinating music to technical reproductive facilities, but augmenting the capabilities, and retaining the human elements of control.

Again, if you made it your business to see and hear what is going on in the area, you would find that all of these issues have been given much attention by many people over a long period of time and have informed a great deal of musical production in all kinds of genres. I fail to see why you don't follow up your obvious interest in such things with some actual fact-finding.

[Sydney Grew]

nowadays some electronic kit can make a fair pass at producing sounds which are quite close to some standard instruments. Indeed in some cases it may be possible to produce sounds which are close to an instrument, but which might nevertheless be very difficult to play on a physical instrument.

There are two or three types of computer music, which may be categorised in accordance with their starting points.

1. Purely electronic music begins with the sounds of an oscillator or similar, and manipulates these with electronic means.

In the fifties and sixties of the last century a number of composers tried their hands at things like that, but almost all of them gave up the attempt, finding the result horrid and unmusical. Indeed I myself wrote to Herr Stuckhosen as early as the fifties, advising him not to persevere, but he did not heed my warnng.

2. Another early approach was to begin not with the oscillator but with "found sounds" such as the roar of a motor bicycle or the song of the cuckoo. Regrettably this produced results only similar to those of the first approach.

3. A third way, the most successful so far, is to start with musical samples. It is possible to obtain a set of thousands and thousands of recordings of every instrument playing every single possible note in every possible way, each with every possible attack, every possible after-touch, every possible way of joining to the neighbouring note, item and so on.

There is one tiny audio file or sample for each of these thousands of elements, played by skilled Austrian musicians.

To produce music from these tiny samples one needs a Sequencer, which specifies what is to be played, in what order and for how long, how long for each part of the note, the volume, the speed or tempo, and so on. 0ne can use a commercial sequencer or write one oneself. A performance may be done live or the sequence may be saved permanently and edited. I have tried this approach, and it works very well with all instruments, even the strings (solo or ensemble), though not so well with singers or speech.

A lot of this sort of thing has been done over the past ten or twenty years.

Now although these thousands of samples are derived from real-life playings, there is nothing to stop one from editing them so as to alter the sounds little by little in various ways (in addition to the alterations enabled by the sequencer). Thus one begins with what is purely musical, rather than with what, in the first two approaches, is not - that is the great advantage.

Despite all that, there is plenty of scope for good non-computerised instrmental music to be written; so it would not do to get too carried away with computer music.

Here are a couple of links that may be useful, the second with examples I believe.

https://en.wikipedia.org/wiki/Vienna_Symphonic_Library

http://www.vsl.co.at/

[Dave2002]

There are two or three types of computer music, which may be categorised in accordance with their starting points.
...

Here are a couple of links that may be useful, the second with examples I believe.

https://en.wikipedia.org/wiki/Vienna_Symphonic_Library

http://www.vsl.co.at/

You discuss what you have classed as three different types of electronic music, oscillator based, sample based, and instrument samples with articulations based on physical instruments.

Oddly the world at large has not dismissed the first two types as you have done. Some pop musicians have embraced deliberately introducing distortion into their playing, which seems odd to me, though sometimes it can be effective. What you have called oscillator based can produce most waveforms, though some of the simplest sound like the whistles one could hear with an old AM radio set. Some composers have employed sounds like that, and of course instruments like the theremin and ondes martenot do produce sounds of that nature.

What is surprising is that the oscillator approach can produce some very interesting and pleasing sounds, though out of the space of all possible sounds, those that some of us find pleasant or interesting are a very small proportion of the total range. Trying to find the parameters to achieve such sounds can be very difficult, but that doesn't mean it's impossible. Some musicians are clearly able to work with synthesisers and obtain reasonable results, though often these musicians may be willing to accept sounds which are either bland and boring, or horribly distorted.

Synthesisers, either hardware or software, are capable of producing many of the sounds which any harmonic process is able to produce. It is known that periodic waveforms can be approximated quite well by Fourier series, based on sine and cosine waves, though some cannot be approximated accurately. Square waves, for example, cannot be – there is always a ringing overshoot at the start and end of the square wave envelope – explained by the Gibbs phenomenon. Arbitrary waveforms can be described fairly accurately in a digital representation, sublect to quantisation/resolution accuracy, but if these are then used to drive physical equipment – electronic amplifiers and loudspeakers, inaccuracies will appear again. A near perfect square wave generated by a digital system will still generate waveforms with transient ringing – due to the interactions with external analogue systems. This is unavoidable.

Not all synthesisers are based directly on Fourier composition. Additive synthesis, based on summing a large number of sine and cosine series, would require a large number of oscillators, to give good results. Subtractive synthesis is an alternative approach, in which a complex waveform with many harmonic components is filtered. This can generate a large set of possible periodic waveforms with quite high accuracy – particuarly if the generation is all done digitally. Other approaches to synthesis include FM synthesis, which has been shown to be capable of generating a very wide range of possible waveforms, and physical modelling synthesisers, which can give extremely accurate representations of vibrating strings, or stretched membranes (drums). There are also synthesisers which mimic the functionality of a human voice – vocoder synthesisers. Many modern digital synthesisers now contain several of these different approaches to synthesis, which means that in practice it ought to be possible to mimic almost any sound with modest resources. The big difficulty is finding out how to do that. Finding a set of parameters for a given synthesiser which will produce a desired sound may at times be worse than finding a needle in a haystack.

For instrument like sounds there are now many virtual instrument libraries, including the VSL library you have mentioned, but also others such as those from Spitfire. There are also some other specialised libraries, for example of what to us in the UK would consider exotic, with instruments from Asian countries such as India, China and Japan, and no doubt African and Australian virtual instruments are also available.

Some of the orchestra libraries are routinely used by composers writing film music, and may even be included in low budget films. As has been noted, many film goers don't really care whether something that sounds to them like an orchestra has human players or not. However, it does still seem to be acknowledged that the best results are obtained by asking musicians to record the music which has been written, with high budget films typically using musicians based in Los Angeles or London, while some lower budget films might use musicians from eastern European backgrounds.

Whether such film music is then subsequently edited and tinkered with further I can't say – though I suspect that it will be modified (or mauled) depending on what the film producer decides. Sometimes the tinkering/editing will be fairly benign, while at other times one might wonder why expensive and highly skilled musicians were recorded at all.

Where musicians definitely score over computer generated instruments is in the interpretation from (typically) a printed score – but some scores may have extra and unusual instructions. Give a group of trained musicians a new score, and they won't play it exactly as written – they will spot phrases, put in very small pauses and very small inflections in tempi, and vary the dynamics and tone of their instruments. There may indeed be computer progams which can also do this, but they may be still in development in research labs, rather than in general everyday use.

An alternative, but tedious approach, is for one or more human editors to shape the digital sound generated automatically from a notation program to make it sound more expressive. In DAWs this is often called automation, which can either be controlled by using a graphical user interface, or a hardware controller. Note that Midi may be one representation of the music, to control outputs, but converting Midi back into standard music notation may lead to some very odd representations, as what may have been interpretative expressive features are turned back into a rigid notation.

[Richard Barrett]

in practice it ought to be possible to mimic almost any sound with modest resources. The big difficulty is finding out how to do that. Finding a set of parameters for a given synthesiser which will produce a desired sound may at times be worse than finding a needle in a haystack.

Not if one knows what one is doing. Finding the correct pitch on a violin is pretty difficult too, that's why people put so much time into learning technique and practising. Why should creating sounds on a synthesizer not require a comparable degree of knowledge and skill? But most people who are serious about using synthetic sounds in composition aren't interested in the idea of mimicking something that's already familiar.

Give a group of trained musicians a new score, and they won't play it exactly as written – they will spot phrases, put in very small pauses and very small inflections in tempi, and vary the dynamics and tone of their instruments.

Which of course is exactly what composers count on, and expect. The idea of playing music "exactly as written" is an illusion. No composer before Beethoven's time, for example, gave a precise indication of tempo, and not that many have done since his time either. As we were discussing on another thread, playing a Viennese waltz involves playing the second beat of each 3/4 bar slightly early, which isn't specifically indicated anywhere in the notated music. Similar considerations apply to most other notated music to varying degrees.

[Dave2002]

I "shouldn't" do this but I'm doing it anyway ....

Notwihstanding previous comments in this thread, I've tried a few more things.

Someone - has made available part of a score of Ravel's Gaspard de la nuit - Scarbo. If you're interested you'll find it.

This can be "played" on a simulated piano - and we can argue about whether a pianist can do better - I'm not going there. It can sound spectacular on a piano.

However, the "instrument" can be changed, and some of the synthesised instruments work quite well for parts, and not so well for others. I'm not too surprised that the harp does well on some parts, and the organ, harmonium, accordion, concertina provide some quite scary sounds at times. The marimba does very well on repeated notes. Then of course one can try voices ...

The effects of reverberation etc. are interesting - and blur things a lot, which is why "instruments" like the marimba seem to do well on repeated notes, and harps seem to manage quite well with arpeggios.

Maybe nobody wants to play these games ... but I've given up worrying about what others think.

The resulting sounds are interesting, and it would be possible to break sections up and assign to different combinations.
I very much doubt whether any of this would be playable on most instruments other than the one it was written for, but is is interesting to push the boundaries slightly - which available technology allows - even to those of us who are simply using a regular laptop to explore. It's not difficult!

[Dave2002]

There is already an orchestrated version of Scarbo (at least one) and here is a Youtube version - https://www.youtube.com/watch?v=UZ7qgB4yqZI

I suspect it could also work with a choir - but that would make it even more expensive and problematic to perform. It does start to sound a bit like Daphnis and Chloe - which is perhaps not too surprising.