• Register
  • Help
Results 1 to 9 of 9

Topic: AO winds and DDSW: modwheel expression?

Share/Bookmark
  1. #1

    AO winds and DDSW: modwheel expression?

    I really am hooked using the modwheel to control the dynamics of an instrument. Do both AO winds and DDSW allow you to do this?

    Thanks

  2. #2

    Re: AO winds and DDSW: modwheel expression?

    anyone? Buehler?

  3. #3

    Re: AO winds and DDSW: modwheel expression?

    Currently only AO has Xfading woodwinds, there are some \"mock ensemble\" patches you can sue in Dan dean, but they still require some tweaking IMO since I dont liek how those were programmed.

    The problem with AO is that you\'ll get some weird phasing, and neither AO/Dan Dean will allow you to Xfade solo instruments very effectively. Its a fundamental problem with solo instruments, Xfading litterally gives you two waveforms and starts to sound like two instruments.

    There is some research going on to fix this. I\'ve developed some interesting methods on my own, as well Xsample seems to be doing somethign about realtime vibrato control (from a hint at their neww flute they are releasing). Not to mention Gary Garritan\'s ongoing research with others in this field. There are alot of options, and only time will tell which one users will prefer. Rest assured that I know personally that there are some brilliant minds on this issue, so its going to be intersting in the near future with regards to solo instrument samples.

    BTW you should look into physical modeling. I think that a combination of sampling and physical modeling will be the best end result for solo applications

  4. #4

    Re: AO winds and DDSW: modwheel expression?

    I\'ve developed some interesting methods on my own,
    <font size=\"2\" face=\"Verdana, Arial\">Well, you can never be sure who\'s already been there first. I had the bright idea of layering a lp-filtered forte sample against an up-faded unfiltered copy of the same forte sample, the idea being to keep *some* upper partials in the picture throughout most of the fade. About a week later I read in Bob Moog\'s column in an old Keyboard magazine (ca. 1986) a description of exactly that technique having been \"invented\" by a professor at Berklee, or somewhere.

    A couple of *really* zany ideas I\'ve tried: 1) band-pass filtering a sample into four separate samples, one containg partial 1, the next partials 2-3, the next partials 4-7, the last partials 8-and-above, then layering them, with the idea of being able to modulate them separately in a manner similar to \"group additive synthesis\"; 2) using a tool like Delay-Dots\'s Spectral Shaper to process a forte sample with the spectral curve of a piano sample, then crossfading the processed with the unprocessed version.

    OK -- so I\'m nuts. I got the inspiration for the \"group additive\" experiment after reading Andrew Horner\'s chapter on emulating a french horn using \"contiguous-group wavetable synthesis\" in the CSound book -- but you can\'t blame him for my nutty idea. I\'m pretty excited that he has finally published his long-promised \"Cooking with CSound\" book. If it\'s like his french horn CSound chapter, it may contain some tasty recipes for using CSound to synthesize lively, plausible acoustic-emulation sample sets. My wife\'s giving it to me for my birthday in a couple of days, and if it\'s as good as I hope, maybe I\'ll report on it here.

    Oksi

  5. #5
    Senior Member
    Join Date
    Jan 2002
    Location
    Dorset, UK
    Posts
    470

    Re: AO winds and DDSW: modwheel expression?

    I \'ve done that, in a different way. The trouble with the above method is that you miss out on the bow noise which is very evident on genuine p samples, which my method overcomes. I am now happily using a realistic MW XF dynamics version of DDSV. (I\'m finally convincing people). Except, there is too much vibrato on the p. (There is a reason why this is not too much of a problem with DDSV, but I won\'t go into that now).

    Vibrato on strings is complex. There are 3 components: amplitude, pitch and timbre (partials) modulation. The first is easy to remove. Pitch modulation can also be ironed out. I\'m stuck on the third. It\'s way beyond the scope of 4-band compression.

  6. #6

    Re: AO winds and DDSW: modwheel expression?

    Yah I tried both of those, and the stuff I\'ve been toying with is more of a spin of that.

    closer to what beckers is doing but with some of the \"additive\" synthesis ideas Oksi had.

    I get some good results, but its not in the vibrato control area, more in the dynamics control area which is better than nothing.

    However, there are some crazy people that are going beyond what we are talking here and REALLY twisting up the \"rules\" we are used to.

    Like I said, interesting times.

  7. #7

    Re: AO winds and DDSW: modwheel expression?

    Originally posted by Beckers:
    I \'ve done that, in a different way. The trouble with the above method is that you miss out on the bow noise which is very evident on genuine p samples, which my method overcomes. I am now happily using a realistic MW XF dynamics version of DDSV. (I\'m finally convincing people). Except, there is too much vibrato on the p. (There is a reason why this is not too much of a problem with DDSV, but I won\'t go into that now).

    Vibrato on strings is complex. There are 3 components: amplitude, pitch and timbre (partials) modulation. The first is easy to remove. Pitch modulation can also be ironed out. I\'m stuck on the third. It\'s way beyond the scope of 4-band compression.
    <font size=\"2\" face=\"Verdana, Arial\">Sure, you can layer in a \"noise\" sample, and I\'ve messed with that, too. There are tools out there (SMS is the one I\'m thinking of) for extracting the \"noise\" portion of a sample so that you could conceivably add it back in in a controlled way.

    I haven\'t yet figured out how to remove vibrato from samples, either. Supposedly some people know how to do it with analysis/resynthesis tools, such as SMS, but I\'m not that adept with SMS yet.

    I use a wind controller, and although I appreciate having vibrato-less samples to play with, such as XSample\'s, I confess I haven\'t yet learned how to \"lip\" the wind controller fast enough to do a convincing violin vibrato. The example of Sal Gallina doing violin vibrato with an EWI controller -- using FM synthesis, if you can believe it! -- is a real inspiration:

    http://www.patchmanmusic.com/SalGallina.html

    As Bruce R. says, it\'s the witch not the wand.

    Funny thing about vibrato: not long ago I ran several dry violin samples through Digital Ear, a piece of software I\'d had lying around but hadn\'t been able to figure out what to do with. Digital Ear\'s attempts to translate the pitch portion of the vibrato into pitch-bend events showed a definite \"DC bias\". In other words, the pitch warble appeared more sharp than flat. (I can hear the violin players snickering :-) I\'ve never seen a synth or sampler that enables you to set a \"DC offset\" to the LFO, and maybe that\'s part of the reason why LFO vibrato often sounds \"synthy\" to people (the other reasons being that there needs to be a special blend of amplitude and maybe even filter LFO as well for each instrument type, plus realtime control over LFO speed as well as depth).

    There\'s an interesting research article (in PDF format) on modeling violin vibrato, which may for all I know represent the state of the art in academia-land, at

    http://www.iua.upf.es/mtg/mosart/papers/p32.pdf

    but it doesn\'t mention this \"more sharp than flat/more flat than sharp\" question.

    Oksi

  8. #8

    Re: AO winds and DDSW: modwheel expression?

    Actually I think you can do this in Giga, with its +/- offset in the lfo tab.It will sharpen/flatten the tone then add the lfo....

    I think. Its been a while since I toyed with it.

    Its still problematic since the lfo\'s are too consistant, and the lfo pattern is a perfect sine waves.

    add to the problem with the fact that we need better filter setups. If I could get a couple of inverted notch filters as well as a slight volume/lowpass lfo, with pitch lfo all able to run on their own lfo pattern and a randomizer for max/min of the pitch within limits I set....... Maybe I could get something that would sound good. .... maybe...

    It would depend on the instrument of course. Winds are a different animal than strings for example. Physical modeling and samples could be the answer for both tho. Atleast a combination of samples with clever sound design, and physical modeled instruments to create added resonance and noise artifacts. Even a simple metal resonance PM instrument from Tassman could help a flute out a ton. It jsut has to be matched to the sample and the samples room ambience. (a difficult task in itself)

  9. #9

    Re: AO winds and DDSW: modwheel expression?

    Originally posted by KingIdiot:
    Actually I think you can do this in Giga, with its +/- offset in the lfo tab.
    <font size=\"2\" face=\"Verdana, Arial\">Doh! Will check it out asap. Thanks.

    Oksi

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
  •