Vienna Symphonic Library Forum
Forum Statistics

200,882 users have contributed to 43,216 threads and 259,147 posts.

In the past 24 hours, we have 2 new thread(s), 9 new post(s) and 60 new user(s).

  • Best way to setup a matrix?

    Since I´m quite new to VI, I´d be really greatfull for a hint about this, cause everytime I think, I have a good matrix setup, I figure out, that there are better setups for a better workflow.

    So how would U guys, having more experience with the VI interface, setup e.g. a single horn-4 matrix consisting of

    sus, leg slow, leg fast, short notes, upbeats, sffz, dim/cre of different length?

    Would U move stuff like glissandos or trills to different VI´s?

    Which controller do U assing to which target? I like to use the modwheel for velocity crossfades, but maybe there are bette things to use it for?

    I don´t use VI for live playing, but I found out, that just editing can be a mess or easy, depending which setup U choose.

    Thanks for your help!

  • Hi Felix. Using CC or PG switching, one could sub-divide strings, brass and woodwind articulations by type into blocks of ten - something like this:

    00-09 Single notes (straight) *
    10-19 Single notes (mute / sordino / stopped)
    20-29 Perf intervals, perf trills and perf repetitions (straight)
    30-39 Perf intervals, perf trills and perf repetitions (mute / sordino / stopped)
    40-49 Single notes (pizz / col legno / ponticello / harmonics)
    50-59 Dynamic change (dim / cresc / fp)
    60-69 Tremolo, flutter tongue & flautando
    70-79 Glissandi, portamento, slides & bends
    80-89 Grace notes & mordents
    90-99 Upbeats
    100-110 Fast repeated notes
    110-119 Runs
    120-127 Arpeggios

    (* Suggested allocations: 00 sus vib, 01 sus vib fA, 02 sus no vib, 03 marcato, 04 portato / detaché long, 05 portato medium, 06 portato / detaché short, 08 staccato long, 09 staccato short.)

    I've probably forgotten some styles but you get the idea! You'd have to tweak the system for percussion because there are so many more perc performance variations (and no perc perf interval patches), but there are many areas in common (mute = damped, tremolo = rolls, etc.) I haven't implemented this myself, but the idea is that one would eventually be able to memorise the numbers and the standarisation would hopefully help the workflow. However, the approach would fall down a bit if you wanted (say) fifteen different types of crescendo on one instrument.

    I would try to use only one instance of a VI per instrument / section to 'keep it real', help scoring and ease the burden on the computer.

    Using this system, the switching for your suggested horn matrix would look something like this (I tend to use CC#4 for switching):

    sus: CC#4 00
    sffz: CC#4 02
    short notes: CC#4 07
    leg slow: CC#4 20
    leg fast: CC#4 22
    dim/cre of different length: CC#4 50-59
    upbeats: CC#4 90

  • This is a very interesting topic. I've not come up with a standard model after several months of working with the Cube-- seems like each project requires such different setups that each project has been rather from scratch.

    I'd be curious to see how differently various users set up their matrices.

  • Hi Conquer,

    I did not quite get, in which cells I should put e.g. the single notes (straight).
    Are U using 10 cells per block? Did I get that right?

  • I usually use a line of horizontal cells - the first on the left (1A) would be for straight notes, then I'd add other variations as required in 1B, 1C etc., following the order I suggested above. (I don't use 'blank cells': I only create a new cell when I need a different style.)

    Your matrix would look like this:

    1A sus: CC#4 00
    1B sffz: CC#4 02
    1C short notes: CC#4 07
    1D leg slow: CC#4 20
    1E leg fast: CC#4 22
    1F dim/cre of different length: CC#4 50-59
    1G upbeats: CC#4 90

    That leaves the vertical cells free for switches between two related styles, say:

    1A sus
    2A sus fast attack

    A different controller is required for the vertical switching - in the last example, 'Speed' (= playing speed) would be a good choice, or you could use a different CC number, say CC#5. Keyswitches are another option, good for real-time switching but some people don't like them because they appear as notes in the score.

    I must admit that like JWL, I've not found it possible so far to devise a standard matrix, but I do always adhere to the principle of putting the simplest single note style (like sus) in the first cell and introducing more complex ones as I move from left to right.

    >Are U using 10 cells per block?

    No - the suggested blocks of ten are just a way of grouping stylistic variations into broad categories and have no relation to the number of cells I might end up using.

  • Seems like a lot of similar thinking going on here.

    Don't know if if this is so much a standard for VI, but more of an extension of my brain: I seem to work more fluidly when samples are organized with the longest samples on the left and gradually getting shorter as they spread to the right.

    My first matrix row tends to be Perf samples. The second row will mirror the first, except they are not Perf samples.

    One problem I find with standardizing matrices is the music itself. It's one thing to organize a matrix as my brain works, but a particular violin part, for example, may demand very different patch switching from one project to the other. I've set up matrices with all needed articulations only to find that the needs of the music work somewhat against the layout of the matrix.

    I don't know if anyone else experiences this, but the more vertical rows I have, the easier it is for me to lose my place. Seems that once I find my workflow in one project, another project demands a whole new way of thinking.

    But, it as a nice problem to have, if that can be said.

  • Hi JWL.

    >I seem to work more fluidly when samples are organized with the longest samples on the left and gradually getting shorter as they spread to the right.

    Yes - the opposite of how they appear in VSL's lists, though their way is equally logical.

    >One problem I find with standardizing matrices is the music itself.

    If only we could find a way of getting rid of that troublesome musical element we could all have perfectly standardized set-ups! I guess there will always be exceptions to a rule and one must be flexible . . . but it's good to have a method, it helps me get started.

    I know what you mean about vertical rows. It's hard enough to keep track of a single horizontal line of cells without adding that second dimension. However, as you say it's great to have so many options.

  • Seems logical.

    One thing I could not find a solution for is the different volume of sus- and dim-samples:

    Example: The VI_perf-legato in cell 4 has two velocity layers. I´m using the mod-wheel for velocity-crossfade. The problem is, that the VI_dyn-me is too loud to be played after a silent leg-note for a smooth ending, even with the mod-wheel down.

    Should I use NoteOnVelocity to have an extra control for the volume of the dim-samples?
    For me this doesn´t seem to work, cause the overlapping note gets volume-trimmed as soon as the dim-note with the lower velocity attack is triggered, resulting in a kind of click-sound.
    So how are U combining a legato-note with a dim-note except using the cell XF?

  • last edited
    last edited
    Hi Felix:

    Not sure if this will help or not-- but check this link and look for:

    >6.1: Legato After Single Articulations<
    http://www.beat-kaufmann.com/vitutorials/vitipstricks/index.php#53248297b5030e502

    @Conquer said:

    ...If only we could find a way of getting rid of that troublesome musical element we could all have perfectly standardized set-ups!....


    That darn music!! AAAGGGH!! [:D]

    yes-- another thing that throws me is where Chamber Strings, Appassionata Strings, and Orchestral Strings differ with some of its articulation choices. You think you can set up a parallel instance with the exact same cell organization so that all of your previous controller data would *nearly* work with a little less fussing about with two or three different volumes (App, Chamber, or Orch) of string sounds.

    Suddenly, you'll find an A/B cresc/decresc is not quite as effective on one instrument at a particular moment in the score than perhaps some other technique to the same effect in another instrument.

    Then you get to the point where you believe you've gotten your ideal string template sorted out. Suddenly you find yourself doing something akin to Barber's Adagio for Strings just as you've finished off something akin to Grieg's Holberg Suite.

    It just doesn't end! [[:|]]

  • Excellent input from all posters. Videos of this and like topics would be welcomed by all, I'm sure.