Interesting question.... in my own work the distinction between orchestration and performance is somewhat fuzzy.  I'll explain:  Aside from the traditional tools of close listening and reading scores, I've been engaged in building my own tools as part of a project I call the Flicker Generative Orchestra.   It's a kind of "computer-aided" composition system that encodes a number of theories and procedures that reflect my own understanding of how music works... a kind of personal theory of music if you will.  It draws on personal intuition and experience, along with insights from cognitive science, music theory, musicology, linguistics, traditional and contemporary approaches to composition from a variety of eras and cultures, as well as a host of other sources.  Along with a number of processes that generate structure in time, pitch and dynamic spaces, there are a few passes at orchestration based on mapping those structures to decisions related to orchestration, e.g. deciding on grouping of players/instruments, leading, and following roles, dynamics, transitions, transformations, etc., all guided by reference to the structure of any number of entangled multi-level metric/rhythmic hierarchies.  The performance stuff is based on assessing a number of significant features of the generated melodic phrases: first, last, high points, low points, repetions, agogic moments, span of rise/fall, as well as significant metric features inherent in the metric/rhythmic hierarchies mentioned above. There's also a first stab at producing a kind of phrase arching for overal dynamics as well as sustained events using curves with variable peak positions that produce a surprising amount of expressive variety.  The most recent development in this long-term project took up most of the last year and a good chunk of this one in developing a hierarchical ordering of the articulation sets for all of the VSL instruments I have at my disposal, which is a fairly comprehensive set.  That involved producing a machine-readable document, more than 600 pages long, that represents the libraries as a hierarchy:

Orchestra

    Section  (Woodwinds, Brass, Strings, Percussion, Mallets, Keyboards, etc.)

        Category  (Playing methods specific to the instruments in a section, e.g. Strings here have Arco, Ponticello,
                         Tasto, Sordino, etc.; Brass have Standard, Mute, Stopped, etc.

            Type    This groups articulations into Shorts, Sustains, Performances, Repetitions, etc.

                Sub-Type    This is where things get interesting as this is where the grouping happens that allows
                                  "intelligent" searches and the influence of global characteristics of the music to be applied

                                   groups include obvious orderings like dynamics at each level from long to short,
                                  dynamics by intensity, repetitions by tempo, upbeats by tempo and number,

                                  other groupings are based on orderings based on globals characteristics such as
                                  connectedness, speed for instance which would order articulations in groups based
                                  their smoothness, or suitability for local tempo applications.

                    articulation1    MIDI specification: program change = matrix, 2 keyswitches for x, y position 
                    articulation2    ditto
                    articulation3    ditto
                    etc.                  etc.

The results are promising.  I've been using it to produce some short string quartet studies as these instruments represent the most rich articulation sets VSL offers, as well as the violin family having the greatest expressive powers, imho.  I reasoned that if I could make it work for those challenging instruments in one of the most sophisticated ensemble forms, I'd have done the hardest work up front.  And it's perhaps in the string quartet that the boundary between orchestration and performance is most fuzzy.  With such a rich articulation set at your disposal, deciding on articulations really is tantamount to orchestrating in this ensemble, and vice-versa. Or perhaps it's all just composition in the end, particularly in a form such as this, in which the composer is responsible for every detail of the music from structure, through instrumentation, orchestration and ultimately the  expressive performance of the music that others will hear.

The performance system works well enough that I'm in the process of pulling it into a standalone tool for providing quick performance sketches of a MIDI file you can load into the tool, and then try out different "playing styles" of your own devising based on the significant melodic features analysis I described above.

I'm attaching an example of one of the string quartet studies from a few months back, so you can hear what this all ends up sounding like.  

Regards,

Kenneth.

Many thanks for your long reply

Kenneth,

That is an impressive amount of work you've done on that.  While it is true that orchestration expresses a basic part of a composer's ideas and is therefore very creative - for examples, look at Ravel or Debussy, masters of orchestration who create musical ideas with the orchestration itself -  what you are doing can be viewed as a way of expressing your own concept of orchestration in digital form.  It also sounds like it could be developed into an actual marketable system of instant orchestration in various modes and styles.  

I just download "Band in a Box"

I will report later

Thanks William.  Yes it's definitely a labout of love, and you're absolute right about the Flicker Generative Orchestra (FGO) being a personal concept.  I think of this kind of work as a kind of cyborgian construction.  But in place of the typical body-machine hybrid that the word "cyborg" typically conjures, here it's more of a mind-machine hybrid.  Since starting this project in earnest, a little over four years ago, I've tried to keep it focused on what makes intuitive sense to my own musical sensibility.  That's why I mentioned in the first post that it's a kind of personal theory of music.  It answers the question: "How does music work?"  But the answer is focused on my own understanding of what music is.  It's not trying to be a comprehensive theory of music, which I doubt is even possible given the variety of forms and cultural differences that make one come back with "Which music?".  To which I'll answer: "My music."  I don't care, at this point, for instance that it doesn't know much about functional harmony, because I'm simply not that interested in that in my own work. I have other approaches to pitch relationships.  On the other hand it does, I think, do interesting things with the relationships between parts that reveal my interest in counterpoint.  The fact that it's doing a pretty good job of sounding like performed music is, I believe, because I've been playing music my whole life.  I've also spent a lot of time aligning the theoretical insights of others with my own embodied sense of what's going on when I play a line "expressively".  If the theory doesn't align with that I'm skeptical of it's usefulness.  If it then also fails to produce something my ears are pleased with I move on.  I'm a firm believer that our music software must be musical.  That is, it has to be grounded on an encoding of musical knowledge.  That's why I long ago lost interest in the use of randomness, cellular automata, strange attractors, and the host of other algorithmic approaches to generating music because then almost universally produce unmusial results.  Just google "generative music" on youtube for a list of, to my ears, unmusical results of these kinds of experiments.  Or perhaps I should put it a bit more diplomatically:  That's not the way (my) music works.  I want to work with a system that I can interact with in terms of the language of music, not mathematics.  I want to be able to think in the language of music and express as much as possible those thoughts with my computational assistant in that same language. 

And yes, I've had moments of foolish abandon, and under the well-meaning influence of family and friends, in which I considered starting a crowd-sourcing campaign to push the development of some these techniques to that marketable state you mentioned.  The automatic performance tool seemed like the most viable piece to start with.  A constant refrain we hear on this and other related forums is form people who are first starting out using these comprehensive libraries.  Where to start?  Why would I choose this articulation over that one?  In what context might it be useful?  How is this one better than that one? etc.  Add to that someone with a limited knowledge of orchestration or even experience playing an instrument that demands the detailed embodied knowledge that a violin requires to do with "style", and it's easy to see why there's also a demand for "out of the box" solutions.  The work I'm doing with FGO is perhaps a proof-of-concept for an more advanced form of Humanizing that could be developed for Vienna Instruments Pro.  The talented engineers at VSL are no doubt working on these kind of techniques behind the scenes. There's been good research into automatic performance systems around for years now.  See the work of Friberg and Sunders on this. They worked out a set of rules for rendering expresssive performances that influenced some of the startegies in FGO.  Even without encoding it into an automated system their work is worth a look for composers using comprehensive sampe-based libraries looking for articulation selection strategies.  It seems to me that by now there should have been some tools produced based on that and other research in this area.  I get impatient and start building my own.

I think more artists with programming skills should do the same to help guide the direction the industry takes, which is mainly why I responded to Cyril's open request for feedback on "tools".  An automatic performance tool is an obvious missing link in VSLs ecosytem of fabulous tools.  The best advice I've seen on this forum, aside from encouragement to learn as much about music theory, composition, instrumentation, orchestration, harmony, counterpoint, performance, etc. in order to make good choices on how to bring your music to life, can be summed up as "map it to the form" and "mix it up".  Mark the important structural points in the music, including melodic form, and try to use as much variety as you can.  There's even been suggestions that one use a different articulation on virtually every event.  That is, indeed, a very good way to get a richness and lively quality to the music, particularly if you know something about how the instruments you're working with are played expressively.  It's just depressingly labour intensive.  I've been thinking of this problem of how to make smart selections from the large pool of options the VSL libraries offer as a small Big Data problem.  It's crying out for a role to be provided for an intelligent assistant.  That's at least one of the reasons for the specialization of composers, orchestators, conductors, score preparators, section leaders,  performers, etc..  It's a huge and complex task with loads of expertise and time spent in mastering each part of the whole thing.  Enter the thought machine.  We're now on the cusp of an era in which our computional assistants are skilled enough, fast enough, and know enough to be viable assistants in this process, allowing someone to realize the potential of the wonderful resources provided by these libraries.  Of course, some will object, raising concerns about the machines taking over tasks that are distinctly human or precious to us in the sense that these are the things that make us human.  I take a much more optimistic view that we're allowed to think at a higher level with the computational system taking on a lot of the grunt work.  It's skilled work, yes, but grunt work nonetheless.  Particularly when you consider that our current crop of super-computers can do that work in the blink of an eye, work that would take a composer weeks if not months of labour.  I'd rather be dreaming up new musical ideas for my computational assistant to chew on. It's a kind of "intelligence amplification".

So, as soon as you start seriously considering what is involved in making such a tool truly responsive to the needs of a variety of users, the demands of keeping up with cross-platform issues, operating system changes, getting it to work with other libraries, dealing the bugs and failures that are constantly being introduced by the inherent instability of our computational landscape.  Witness the heroic efforts by our friends here at VSL to get Vienna Ensemble Pro 6 up and running smoothly.  That requires a team of talented people who have no time to do anything else during those times when you introduce something new and complex to the already overly complex computational landscape of hardware-software interdependencies.  I understand that world, having spent a few years working in the game industry coding audio behaviors.  In short, I don't have time.  I need to stay focused on the music.   So I live with the instabilities inherent in experimental software.  Try to keep a balance between the time spent on design, coding, integration, testing, refining and actually producing music.  I'm not willing to give up that last part to perfect an engineering problem.  Gotta keep the focus on the music now.

That's my story... and I'm sticking to it ;-)   Now back to work on things musical.

That is extremely interesting Kenneth and there is definitely a huge potential for development in what you're doing. One could say that such a system is the ultimate extension of humanizing.  I have noticed how in the past I always "humanized" by actually playing each line in a MIDI performance and not quantizing, but with the detail of VSL's humanizing functions - with individually selectable parameters and intensity levels - it has actually become unnecessary.  I love the idea of some hypothetical system in the future (maybe not too distant) in which one could select various orchestrational/compositional styles.  Not that one should be unaware of these styles personally, but as experimentation and inspiration that could be fantastic!  

Yes, the objective is not to work with a set of pre-defined "styles" but rather to have the option to quickly define some stylistic features to get a sense of what works in what context.  It takes long enough just to do one good take on styling a piece by hand with a rich set of articulations, much less being able to try different approaches. I really do believe that it's a kind of Big Data problem on a small scale.  Small because in terms of the numbers the amount of data is not that large.  I think my current VSL set of libraries totals close to two million samples.  I haven't tallied the number of articulations that's divided into.  but it's big data in the sense that one person has to deal with all that and it's just very very time consuming to do it well. The way I'm intending to move forward with this is to use the more promising outcomes as starting places for further development.  Import them into Logic and use the score editor to develop the ideas into more fully formed pieces.  I like the idea of a "composer's assistant", always ready with a new idea when the creative doldrums settle in as they all too often do.  That just reminded me of Leonardo's advice to young artists.... if you find yourself unsure of how to start, what your subject is, how it might be represented... simply throw some paint at the wall and use your imagination to see in it a suggestion as to how to proceed.  A musically knowledgable tool could provide a much higher level of such "splatter".  Splatter with formal, metrical, harmonic, performative coherence that points to things you might not have considered.  A friend of mine who researches design among other things, was encouraging the development of what he referred to as "design space exploration tools".  He noted that in anything worth exploring, there are usually a vast set of possibilities that make it very difficult to find the optimum areas with fruitful potential.  The current work at IRCAM with the Orchids tool for automated orchestration addresses this issue head on.  Its authors note that the number of combinations of instruments in the symphony orchestra is so vast that it's very difficult to know what will work and what won't in achieving a particular objective.  You simply can't try them all out and will likely miss the best combinations to achieve the sound you want.  Of course this is moving away from traditional standards of orchestration as handed down by the masters.  Orchids uses some advanced search techniques based on evolutionary algorithms to optimise the search of the timbral spaces provided by the various instruments of the orchestra to provide "solutions" to the question of "how could I best orchestrate this sound?"... It's an experimental approach.  They used it to support Jonathan Harvey in the composition of his orchestral composition "Speakings".  Analyzing the human voice and using their techniques to suggest orchestrations that would most closely approximate those voice sounds.  It's remarkable effective in moments.

There's short piece on how the piece was realized on the IRCAM Forum site:

http://forumnet.ircam.fr/tribune/making-an-orchestra-speak/

Cyril, Band in a Box is a completely different beast.  It's really designed to be a music-minus-one, anyone can be a musician, kind of thing that combines pre-composed phrases using time-stretching and pitch shifting to be able to provide an instant accompaniment for a chord sequence or melody you provide.  In my humble opinion you get more creative freedom with Garage Band on a Mac than BiaB.  

@kenneth.newby said:

Cyril, Band in a Box is a completely different beast.  It's really designed to be a music-minus-one, anyone can be a musician, kind of thing that combines pre-composed phrases using time-stretching and pitch shifting to be able to provide an instant accompaniment for a chord sequence or melody you provide.  In my humble opinion you get more creative freedom with Garage Band on a Mac than BiaB.  

Hello Kenneth

Did you try BIAB ? the 50 classical styles ?

On Atari there was Big Band, it was a little buggy but it was giving you material to help you to make your orchestration.

Hi Cyril,

A couple of things make it difficult for me to want to spend much time with BIAB. First there's the sound quality of the MIDI examples that are on their website.  Awful.  Then there's the styles, which, in the classical section, is pretty conservative.  I know you can make their MIDI playback to your own libraries but I can't imagine it having much nuance given the quality of their demos.  

I think you'd learn a lot more by listening to the great orchestrators of the late nineteenth and twentieth centuries, Rimsky Korsakov, Berlioz, Debussy, Mahler, Stravinsky, and Ravel wrote some of the most brilliantly orchestrated music I've ever heard.  In a more contemporary vein, the orchestrations of Toru Takemitsu, particularly his later works are gorgeous, he's definitely leaning on a love of Debussy.  Another master of orchestral coloring is Lutoslawski.  Just doing a close listening to the works of these composers is a learning experience.  Then if you can get a hold of some scores, then you can get an insight into the combinations they used, the way they bring out a line, pass it around, manage dynamic changes.  

Thinking more about your original question and shifting it from tools supporting the creation of orchestration to learning it I remembered another really very cool study resource that's quite fun to spend time with  These are the music apps that Touchpress are putting out.  They have one in particular called The Orchestra that gives you complete movements from the works of Haydn, Beethoven, Berlioz, Debussy, Mahler, Stravinsky, Lutoslawski and Salonen, played by the Philharmonia Orchestra conducted by Esa-Pekka Salonen..  So it covers a broad time span in the music written for symphony orchestras.  The app is very well designed and provides some really useful ways of getting inside the music.  A floor plan for the orchestra used in each of the pieces is provided that shows circles for each muscician in each section, with each section color-coded to make it easy to follow.  The circles pulse in time to the playing each player   so you get a really clear indication of the orchestration as it flows from section to section, thinning out within a section, splitting into divisis, etc.  Then there are two versions of the score that follow the music... a "curated" score that provides the main elements, melodies, accompaniments in a condensed form, then a complete score which, given the screen-size constraints of my iPad Mini is pretty hard to use if you really want to follow the foreground elements due to the need to be constantly scrolling up and down to reveal those parts as move from section to section, instrument to instrument. Further, each of these two versions of the score can be displayed using standard music notation or piano-roll style notation for those who prefer that style.  Finally there's four channels of video content, synced to the music of course, the conductor and whatever three elements the producers thought to provide to highlight the features of the music.  The camera changes its focus to, again, follow the musical surface as it moves within the orchestra.  Then, there are commentaries that can be played along with the music in subtitles or recordings of the speech of the conductor and selected musicians.  As if this was not enough, there is also a section that reminds me a bit of the VSL Academy section of this website, with information on the sections of the orchestra, as well as details about the individual instruments.  There are video segments with the players of each instrument providing some insight into the technique and role the instrument plays in the orchestra.  There's a 3D photo-realistic model of each instrument that you can rotate to see every detail.  Playing ranges are given on the staff, and a piano keyboard showing the range is provided that you can play to hear the way the instrument sounds in different registers. Fun facts about instrument history or construction are given, and a brief showcase video from one of the eight compositions shows the instrument in a foreground role doing what it does best.  Overall this is a great learning tool, and a useful resource to refer to when you're working with new instrumentation, or interested in checking out the many different ways the instruments and sections of the orchestra can be combined.  The quality is of the highest order, in the repertoire, the musicianship, and the production of the app itself.  I can't recommend it enough.

They've also got a similar app that features the Julliard String Quartet playing Schubert's Death and the Maiden quartet.  Pretty much the same set of features. One nice tweak in this one is that you can tap on the video of any  of the four players and the level of that part will be boosted in the mix so you can hear it in much more detail. Kind of a dimension strings approach.  This the great for the cello and viola parts which can be a bit harder to hear at their normal level.  And it's a fantastic performance.  

I think they made a similar app for the complete 9th Symphony of Beethoven with four different orchestras you can switch between, piano music of Liszt, Reich, Eliot's Wasteland, etc.  I suspect there are more of these in the works. 

Hello Kenneth

Thank you very much for your detailed answer.

I will not use the BIAB sounds, they are horrible :(

I went for 4 year at the Music University of St Denis in France

We had a course of "Orchestration", the teacher came only twice, and gave us as an exam a piece of "Musique contemporaine", it was a disaster because I have added a bar so Logic sets all the instrument and I forgot to remove it before printing. The teacher did not like it !!! i had a bad mark !!! stupid teacher !

I have play the piano from 6 to 11, learn flute but I lost all I have learn

When I compose I record a theme singing on my Iphone.

Put it in Logic

Add chords, bass, choir, drums, a few variation...

I manage to make pieces of 3/4 minutes

I have a lot of problems to write a piano accompagniement, I can use the arpegiator but it is too mechanic !

I am starting to use violin, brass and winds, but it is quite difficult to use all the instruments of the orchestra

I have to look at VSL to use there phrases, I am at the moment on holiday  and did not dare to take my VSL dongle with me so I use the EXS instrument ...........

The idea to use BIAB is to take some small phrases that I can export to my Logic song, using, of course, VSL instruments.

I will edit those phases to correct mistakes assign instruments, transpose, reverse .......

I also use synth of Arturia, K5, Omnisphere...

My music influences are : Vangelis, Yes, Genesis and of course Classical Music

Best

Cyril

Hello Kenneth

I gave a trial to BIAB.

I wrote a melody and it's chords and some orchestration and I gave the chords to BIAB to digest

After period of disapointments, I had quite good results*

I generate a few style and inserted into Logic

Then I begin selecting passage to improve my orchestration, deleting unwanted parts, notes... 

This on my MacBook Pro using EXS.

Next step I have to transfer all that on my Main Computer to use VSL

Best

Cyril

* you have to be carefull with your midi channels and to chords allignment 

Can you post an example?  i'd be interested to hear what you're getting.

Hello Kenneth

What do you think of what BIAB is generating ?

Best

Cyril