• Register
  • Help
Results 1 to 8 of 8

Topic: How do I make a Gig that doesn't use voices?

Share/Bookmark
  1. #1

    How do I make a Gig that doesn\'t use voices?

    I send different program numbers to multiple channels in order to change instrumentation on the fly while improvising. Sometimes I\'m sustaining notes on 8 channels while playing a melody on 8 other channels. I use DSP to fade channels in and out as I desire in order to get the mix I want, but of course, all instruments are using up voice resources, no matter what their level is in the mix. Of course keyswitching would lighten the load, but doesn\'t give me the subtlety of mixing instruments at different levels.

    Depending on which giga instruments I have loaded, I am more or less successful in playing without voice-robbing or mild breakup (pops and crackles.... not really enough to be objectionable).

    Once I\'ve loaded voices into all 16 channels, I no longer have the option of sending a program change that will detatch an instrument from a particular channel. Or maybe I do? I can\'t figure out how to do that with a midi program change command.

    If I can\'t, what would be the structure of an instrument that puts the least load on GigaStudio (in terms of actively playing voices)?

    I\'ve thought that maybe assigning a tiny sample to only one region of an instrument, and putting that region at the extreme end of the midi note range (either top or bottom I suppose), might do the trick. Is there a better way? Can I somehow detach all instruments from a particular channel with some midi command instead?
    Gabriel

  2. #2

    Re: How do I make a Gig that doesn\'t use voices?

    Originally posted by gabriels:
    I send different program numbers to multiple channels in order to change instrumentation on the fly while improvising. Sometimes I\'m sustaining notes on 8 channels while playing a melody on 8 other channels. I use DSP to fade channels in and out as I desire in order to get the mix I want, but of course, all instruments are using up voice resources, no matter what their level is in the mix. Of course keyswitching would lighten the load, but doesn\'t give me the subtlety of mixing instruments at different levels.

    Depending on which giga instruments I have loaded, I am more or less successful in playing without voice-robbing or mild breakup (pops and crackles.... not really enough to be objectionable).

    Once I\'ve loaded voices into all 16 channels, I no longer have the option of sending a program change that will detatch an instrument from a particular channel. Or maybe I do? I can\'t figure out how to do that with a midi program change command.

    If I can\'t, what would be the structure of an instrument that puts the least load on GigaStudio (in terms of actively playing voices)?

    I\'ve thought that maybe assigning a tiny sample to only one region of an instrument, and putting that region at the extreme end of the midi note range (either top or bottom I suppose), might do the trick. Is there a better way? Can I somehow detach all instruments from a particular channel with some midi command instead?
    Gabriel
    <font size=\"2\" face=\"Verdana, Arial\">Remove at least four of your arms (two each of your left and right), and at least three fingers from each remaining arm until only two arms, each equipped with five fingers, remain.

    You can use any leftover two-fingered arms for MIDI patch and control switching as you improvise with your two five-fingered limbs.

    The way I see it, the issue you\'re having is not with available voices; rather, it is with the scheduling of the various finger movements required to improvise on eight channels using your first eighty fingers whilst sustaining on eight additional channels with your other eighty--all the while manipulating various conrol and DSP parameters using whatever remaining digits you may have...

    Sorry gabriels, I just couldn\'t help it!! :-)

    Question: are you using GS96 or GS160? You\'ll get multiple sets of 16-channel ports (GS160 provides four 16 channel ports).

    \"level\" settings and \"number of voices\" are unrelated. One describes the volume of a single voice and the other describes polyphony (how many voices are available). How \"loud\" a single voice is has nothing to do with how many other voices are playing.

    You\'ve made 100+ posts: am I insulting you by misunderstanding your question?

  3. #3

    Re: How do I make a Gig that doesn\'t use voices?

    I think what you\'re trying to do is best accomplioshed either with, as you more-or-less said, a sample of perhaps one byte of zero level easily produced in Cool Edit or the like. \'m not awaqre of a midi command that will detach an instrument, only one to replace it. Have you tried sending a patch change to a non-existant patch? That might effectively detach it, assuming you\'re not using all 127 patches.

    Dasher

  4. #4

    Re: How do I make a Gig that doesn\'t use voices?

    Sri-bubba and Dasher,
    Thanks for the laughs (Sri) and the insight (Dasher). To paraphrase someone with about 20 times as many posts to this forum as I have \"Really, I am a moron!\"

    or...... am I perhaps .... A SUPERGENIUS!!!!!!! Bwahaaa...haaaaa...haaa...

    Dasher... the unassigned sample slot thing doesn\'t do the trick. The current sample stays there. Good thought though. The short sample method might be working. I\'ll have to devise a hearing test to determine if it does.

    No offense taken, Sri. That was the funniest thing I\'ve read in a long long time. Tears rolling down my cheeks... the works!!!

    But what\'s a person who writes \"whilst\" doing in New Hampshire anyway?

    But seriously speaking, for a second here, it\'s all true.

    - I use GStudio160 and load up 16 channels (4 on each port - hoping to spread the midi load though I think this is probably nonsense because once the midi commands are in the computer as opposed to being dumped out a midi interface, there\'s probably no problem with midi clogging).

    - Each channel gets loaded with a different instrument.

    - I have sort of a midi saxophone of my own design which communicates with my midi rig (including Gstudio) via a PC.

    - The PC is running a program which was written to my spec.

    - The sax can output on several channels... melody notes on one channel, chords on another, arpeggios on another, etc etc.

    - But actually I am only sending 1 channel of melody to the other PC which is running GigaStudio. Well, actually, two channels.

    - When I play, I can sustain a note, and then as I continue playing, subsequent melody notes played on top of the sustained note are transmitted on a second channel

    - The secret, of course, is the use of Midi-Ox, which intercepts midi messages and duplicates them on other midi channels. Channel 1 melody notes go out 8 channels, to be played by the different instruments assigned to them. If I press a certain button, the channel 1 notes are all sustained and subsequent melody notes are transmitted on Channel 2 (on the other 8 channels.

    - If all channels were to be mixed into the final output (using the DSP section of GStudio) then, of course they\'d all be present in the final mix. However, I control the individual levels of the different channels to get an expressive solo voice.

    - I _could_ do that with a multi-amputated arm with only a stub on the end because I use a spring-loaded rectilinear pot to generate 16 separate streams of continuous controller messages.

    - The way this happens is that the digitized output of the pot is simultaneously \"reading\" 16 independently programmable curves (drawn into 16 graphs). So, in response to movement of the one pot, CC1 could output a constantly increasing value, CC2 could output a wiggly curving value, CC3 could remain at a steady value, etc etc.

    At worst it sounds like a combination of Lawrence Welk and bad elevator music... you know, the kind in which each successive phrase is played by a different instrument.... xylophone then clarinet, then harp, then cello ad naseum. At best, I get an instrument with really flexible timbre. It just depends how successful my choice of instruments is, and how subtly I am able to mix them.

    Sri... In regard to your statement that

    >\"level\" settings and \"number of voices\" are
    >unrelated. One describes the volume of a
    >single voice and the other describes polyphony
    >(how many voices are available).

    You\'re absolutely right. What is also true, though, is that if you use 16 channels for crossfading of instruments (as opposed to keyswitching them), the layering takes up lots of voices, and that is the problem I was trying to address.

    As it turns out, (or so I\'ve heard) a sample releases the midi voice it is using, as soon as it has finished playing. If the sample is really short, say one wavelength of a reasonably pitched note (played at 0 volume so that there\'s no \"click\") then the note is over very quickly, and the voice is released, which allows higher polyphony since these the released voices are again available for assignment to a new midi note.

    > \" You\'ve made 100+ posts: am I insulting you by >misunderstanding your question?\"

    No way. I still can\'t stop myself from smiling when thinking of the multiple amputated appendages. Say, wasn\'t there a movie called the
    5,000 Fingers of Dr. Terwilliger... about a mad music teacher?

    Thanks for the fun.
    Gabriel

  5. #5

    Re: How do I make a Gig that doesn\'t use voices?

    Wow, Gabriel, after reading your description of what you do, the image of a color wheel in front of a light comes to mind. It does sound like a lot of polyphony would be used up as sustained notes accumulate from the press of your \"certain button\" which, as I understand it, extracts notes from the melody and sustains them...is this correct?

    And yep, crossfading between layers would certainly do it to ya in short order.

    So, I now get that you\'re looking for a way to turn off some of the layered notes such that the texture of the piece stays at a manageable \"thickness\" and chording stays real without becoming (unintendedly) dissonant as sustained notes add up. And, that what you\'re looking for is a way to silence specific notes by detaching the instrument on the channel of interest.

    I risk going \'way off on a tangent, \'cause if I\'m misinterpreting you we\'re really going for a ride from this point on. Nevertheless, assuming we\'re congruent so far, a couple of alternatives present themselves. Each requires a second button!! :

    --a MIDI reset (all notes off) button aimed at the channel you\'re looking to silence. You\'d have to figure out a way to tell the button what channel to do the reset on. Maybe you could switch between channels with the button and add a standard sustain pedal which, when lifted, sends the actual reset message. Ah mean, with all them arms and digits a flappin\', yer gonna need ta use yer feet too!

    --as thesoundsmith suggested (...\"have you tried sending a patch change to a non-existant patch?\"), perhaps you could try adding a \"null patch\" button which simply sends patch #129, or an unloaded patch #, to the channel in question. I\'ve never tried to send a \"patch 129\" or and unloaded number; who knows what would happen--maybe you can tell us...

    Sounds awfully complex. How do you focus on the music with the technical stuff draining all the energy from your left brain?

  6. #6

    Re: How do I make a Gig that doesn\'t use voices?

    Originally posted by sri_bubba:
    Wow, Gabriel, after reading your description of what you do, the image of a color wheel in front of a light comes to mind. It does sound like a lot of polyphony would be used up as sustained notes accumulate from the press of your \"certain button\" which, as I understand it, extracts notes from the melody and sustains them...is this correct?
    <font size=\"2\" face=\"Verdana, Arial\">Sri.... Yes, that\'s exactly what happens.


    And yep, crossfading between layers would certainly do it to ya in short order.

    So, I now get that you\'re looking for a way to turn off some of the layered notes such that the texture of the piece stays at a manageable \"thickness\" and chording stays real without becoming (unintendedly) dissonant as sustained notes add up. And, that what you\'re looking for is a way to silence specific notes by detaching the instrument on the channel of interest.
    <font size=\"2\" face=\"Verdana, Arial\">Not so fast, there, Bubba. Actually, though it is true that I can accumulate notes to sustain by pressing, not the \"certain\" button which I call the Sustain button, but another button which I call the Pyramid Button. I only accumulate notes as long as I am pressing the Pyramid button. That way I get consonant (or dissonant) chords built up as I like.

    If I press the Sustain button, only the last melody note is sustained. So, my problem is not controlling the number of notes sustained.

    I can see why you thought I was always building up these dense clusters though, since I mentioned that, when I use the sustain function, I\'m holding 8 notes.

    Here\'s what\'s happening in that case. Even if I am playing only 1 melody note, I am having MidiOx output that note on 8 channels, each loaded with a different instrument. The same goes for the sustained note. It is actually 8 notes played simultaneously. But all 8 notes are the same pitch (in the case of a melody note - sent to 8 channels - and in the case of a sustained note -sustained on 8 channels)

    While playing,I can use a continuous controller to fade different channels in and out to create different timbres or to use different articulations. Since this is done with layering rather than keyswitching (so that I can mix any instrument into the final mix at any volume, instead of merely switching from one instrument to another), 8 instruments are playing each note, whether or not the instruments are audible in the mix. Each of these instruments may have layering in their internal structure also, so this can add up to lots of voices very rapidly. Now add in the sustained single notes, or the sustained chords, and I\'ve really got a bundle of voices being used. I\'m getting away with it, somehow... not to much trouble with popping or dropouts.

    I send program changes on all the channels I\'m playing when I want to change over to another type of sound. So, if I happen to be playing a small chamber orchestral sound, and now want to move over to a more synthy pad ethereal sound and ultimately bring in a fuzzed out guitar, for example, I might do that by sending out a stream of midi program change numbers which would reload the channels with new instruments. In cases where the reprogramming uses some very \"big\" instruments, it would be nice to be able to unload some of the channels which might not be necessary, so that they are not \"playing along\" with their giga instruments. I can\'t seem to unload the channels with a midi command, even by sending a program number for which nothing is loaded. I\'m afraid I can\'t send anything beyond 0-127 as a program number with my setup, so that\'s out. So what\'s left seems to be just loading a very simple voice, and maybe putting the regions where they will never be played.


    Ah mean, with all them arms and digits a flappin\', yer gonna need ta use yer feet too!
    <font size=\"2\" face=\"Verdana, Arial\">Yuh, well, I do.... 8 footpedals, each digitized and mapped to any desired continuous controller through a table (much as the 8 pots on the woodwind controller are configured)


    --as thesoundsmith suggested (...\"have you tried sending a patch change to a non-existant patch?\"), perhaps you could try adding a \"null patch\" button which simply sends patch #129, or an unloaded patch #, to the channel in question. I\'ve never tried to send a \"patch 129\" or and unloaded number; who knows what would happen--maybe you can tell us...
    <font size=\"2\" face=\"Verdana, Arial\">Well, as I said, the unloaded patch thing doesn\'t work, and I can\'t send #129


    Sounds awfully complex. How do you focus on the music with the technical stuff draining all the energy from your left brain?
    <font size=\"2\" face=\"Verdana, Arial\">My man.... You have hit the nail on the head exactly. It\'s maddeningly complex, and the left side of my brain is probably twice as big as the right for all the years of developing this instrument. But, there is hope. As I refine the beast, I\'m always consolidating the controls....linking more and more continuous controllers to fewer and fewer pots. When I started, I didn\'t even know which pot would end up controlling which parameter of which device. Now, I have a much better feel for how to program the curves for the different cc outputs created by one pot, and how to co-ordinate the effect of one continuous controller with the effect of another one, linked to the same pot.

    The various controller messages still have independent curves since the pot is \"processed\" by curves that can be programmed independently for each output. There are 2 ... maybe 3 pots that are positioned so they are accessible without interfering with my finger movements while I play, plus a breath and bite controller. Each is mapped to control several parameters of giga and the other synths and outboard devices, but, as I said, independently.

    If all goes according to plan, I might be able to just go into a dream, play the music, and not think one bit about the complexity underneath, since the complexity will be taken over by the preprogrammed cc control. [img]graemlins/tounge_images/icons/smile.gif[/img]

    I think.....

    Thanks for the feedback.
    Gabriel

  7. #7

    Re: How do I make a Gig that doesn\'t use voices?

    Duh! No, you can\'t send a patch #129... [img]images/icons/rolleyes.gif[/img]
    Try 119, or other unused patch number...

    Dasher

  8. #8

    Re: How do I make a Gig that doesn\'t use voices?

    Double Duh. You\'re right of course! But sending an \"empty\" patch number doesn\'t do anything. The currently loaded instrument remains there.

    The \"tiny instrument\" approach seems to be helpful, though. If many notes with \"tails\" are played sequentially, they naturally start incrementing the number of voices being played simultaneously. Since a single cycle sample releases it\'s voice as soon as it is done playing through it\'s brief timespan, this type of instrument has a minimal impact on polyphony.
    Gabriel

Go Back to forum

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •