Breath Control--here's how to set up instruments

[Bruce A. Richardson]

This got buried in another thread, so I thought it might be helpful to give it a topic of its own.

Here\'s how to set up instruments to respond properly to breath control. Open the desired GIG file, and follow along...

First, I copy and paste a new instrument in the GIG file, and prepend the existing instrument name with BC-...that way when I do a search I can get all my breath control instruments on the screen quickly.

Select the \"response\" tab and set the library up for Linear/Low velocity response. That takes velocity out of the picture for volume changes and allows you to map it to the wind controller.

Next, select the Mix/Layer tab, and change the \"attn control\" from none to Breath Control. In the case of GOS, you change the same thing, but with the EXP instruments, this will actually be tying the crossfade, not attenuation, to the Breath Control.

Next, go to the EG/LFO1 tab, and look at the release time. Try something like .05, and see how it goes. Since your breath will now be \"sculpting\" a lifelike release when you stop blowing, you don\'t need the artificial ramp down that\'s generally programmed for keyboards.

If you want to set up some filters, go to the filter tab, enable the filter, and choose Breath Control in the Fc MIDI Ctrl drop-box. For instance, a little Lowpass filtering will give you some timbral shift, going darker when you blow softer and brighter when you play louder.

That\'s it...between attenuation and filtering, you can get an excellent breath response from almost any library. Pretty easy, huh?

OK, not that easy. Here are the complications. If the library is mapped from \"single volume perspective\" samples, that is, the pianissimos are actually pianissimo rather than attenuated full-amplitude recordings, you will get a \"double attenuation\" when you add breath. There are a couple of things you can do. If you\'re lucky and they did not use the disk compression, you can blow out all the samples and normalize them, then re-build the instrument (careful, this will destroy the previous mappings...but you DO have legit CDs of the library, right?).

You can also get a little more detailed in your Breath control design and start diddling with each layer\'s response. I won\'t go into it any more, because this could turn into a book.

The basics are here, and even the most difficult instruments can generally be set up well for breath control, you must just get your hands dirty and take them on.

And, here\'s a PLEA to sample developers. PLEASE support the addition of Export functionality for compressed sample libraries. When we users are \"locked out\" of editing the waveforms, it does a tremendous disservice to us. We need access, we need to be able to edit sample libraries as we wish, right down to the sample level. I have cursed Nemesys under my breath mightily for this oversight.

[Bruce A. Richardson]

Oops...rather major oversight:

Of course, you must SELECT the regions/dimensions in question and hit the \"apply\" after your edits. Usually, you can just select EVERYTHING, do the major functional edits, then when it comes to fine tuning of individual notes and velocity layers, etc., you\'ve already got the funtional stuff happening and you can concentrate on smoothing out the response. Breath control is pretty \"forgiving\" though, since you\'ll find yourself automatically responding with air to what you\'re hearing. I usually don\'t end up doing much tweaking at all.

[KingIdiot]

I thought I might add a little:

I\'ve been patiently waiting for my breath controller and Ztar for a few months but have been tweaking with programming ideas for a while for it.

If you are working with ensembles (or solo samples with mock ensemble patches). A combination of Layers and filters really works out well.

Consider GOS\'s EXP control that allows you to fade between Piano, MP, MF , and Forte. Set this control to Breath Controller instead of Mod wheel (or re map the breath control CC to Mod Wheel in your sequencer) and you can probably get some great results.

Now with Brass Add different filter settings to each layer and assign the cutoff frequency to be controlled by Breath Controller, on top of Crossfade control between layers. This creates a very playable instrument, with less \"quack\" from the filter use (since you\'re using a combination of filters settings).

With Dan Dean Brass ensembles I built one with one piano layer and no filters, and 3 \"forte\" layers with different filters assigned to each.

Another good thing to try and do is calibrate the filters and layers to a crecendo sample if there is one in the library.

Bruce I wholeheartedly agree about the \"lockout\" of samples. However I understand that developers want tosave space with Nemesys compresion. There are a couple of trasnlators that export the *.WA_ files, I believe. I also believe it will convert back from *Wav to *.WA_. Not sure if you can do a \"replace samples\" after the conversion....hope so...otherwise its pretty useless.

------------------
Really...I am an Idiot

[Duncan Brinsmead]

Bruce, I\'ve read your recent posts with great interest. I just got a Yamaha BC-3A last week and I\'ve been trying to set it up to work with gigasampler. My first problem is programming the midi solutions breath crontroller processor. I need to send it system exclusive messages, although I\'ve never done this before. I assume there is some way to do this using Cakewalk. Does anyone have any experience with this?

I want to try using a breath controller, as opposed to a wind controller. I like the polyphonic capability of the keyboard and also there is the possibility that velocity could be used in some way... perhaps switching between samples with and without attack portions for better legato.

Comments were made earlier about how wind controllers trigger attacks using breath, while breath controllers use the keyboard as the trigger, thus causing timing problems between tonguing and hitting a key. Might it be possible to use a realtime midi effect in Cakewalk to mimick what happens in a wind controller. If so, is it possible to create a realtime midi effect for Cakewalk using the cakewalk programming language..CAL?

Duncan

[This message has been edited by Duncan Brinsmead (edited 02-11-2002).]

[Jeff Hurchalla]

Hi Duncan-
I\'ve been toying with the idea of doing exactly what you suggest as a feature in a stand alone (expanded) Maestro utility, but I wasn\'t sure there would be enough demand for it. Would it interest you, and for others reading does it interest you? Anyway, I think you should be able to program this with a CAL script. I think it should work well during playback, but CAL won\'t be as great during realtime work due to the latency.

[KingIdiot]

Interested http://www.northernsounds.com/ubb/NonCGI/images/icons/smile.gif


Tho I dont know if it wwould be really worth it if its not good in realtime. Part of the reason to have it would probably be for the \"feeling\" it makes you have when you play

------------------
Really...I am an Idiot

[Neal Acree]

I love the idea of a wind controller but I\'m apprehensive about spending $600 on an instrument I can\'t return if it doesn\'t work for me. Is there anything out there for less? (I do mean a wind controller not a breath controller). Also, which midi parameters do the wind controller basically effect by blowing a single note?

[thesoundsmith]

Neal, I\'d say the quick answer is eBay. A little patience could snag a WX5 or some such for less than $350, perhaps.

Bruce, thank you for the detail. I\'ll try this with my keyboard and breath controller tosee if it works.

in the other post you mentioned something about wind controllers not needing (or using) the initial velocity. I don\'t understand, the Midi note on command includes an initial velocity as part of the transmitted code. Are you implying there is another way of turning on the note, or a low initial velocity, or something other?

I really would like to try this without a wind controller, if it can be done, so I\'m willing to experiment once I can figure out the editor. This thread has helped that a lot.

Dasher

[astrahhan]

I am interested in purchasing a BC but are there any BC that you can connect directly on a PC\'s midi in? I think that MIDI Solutions has a product for this but this is an extra cost.

Thanks

[IGOR]

I am very intersted in this topic since I just did purchase a Synthophone. For those who do not know what this is, it is a wind controler build into a real saxophone. For saxophone players it is ideal because you just play it as a real sax. The down side of it, is that it does cost about 4000 $ (US). In order to try to play with more control, I did buy a velocity converter from midi solutions, the programing is pretty easy, but you have to have their utility in order to do it with any sequencer. For slow, melodic passages it is fantastic, but for fast playing I do not feel comfortable with it, because of some delay.

IGOR

[This message has been edited by IGOR (edited 02-12-2002).]

[This message has been edited by IGOR (edited 02-12-2002).]

[fmfgs]

To Neal:
Your question what comes out of a wind controller is answered with example on my old thread: http://www.northernsounds.com/ubb/NonCGI/Forum3/HTML/001249.html (\"http://www.northernsounds.com/ubb/NonCGI/Forum3/HTML/001249.html\")
fmfgs

[Jeff Hurchalla]

Hi King -
A maestro type software utility would do everything in real time with no latency issues. It was CAL scripts I was referring to that realtime might not be ideal.

Going by the problems Bruce mentioned for breath control, this utility would probably address them to make breath controllers functionally as useful as wind controllers. I\'m not sure yet if there\'s demand for it basically, so I can\'t say it will happen. What do the prices seem to be for an average breath controller vs. wind controller? For those reading, would it help you?

Here is the concept, based on what I understand as the problem Bruce had with breath controllers for keyboard:
Ordinarily a sample will begin playing when the user depresses a key on the keyboard, and after this initial attack the breath control will change the volume of the sample. This generally results in an attack that is uncharacteristic of a wind instrument, but does allow good control of expression while the note is playing. Keep in mind this attack problem doesn\'t exist for wind controllers because the sample begins to play when a breath is detected by the controller, and the keyboard is not part of the picture. Now what a software midi utility that I could program would do to fix the problem for breath controllers is to detect the note-ons from the keyboard without sending the note ons. This way it can \"know\" what note will be playing, and send the note on when the breath controller begins the breath - with the initial velociy determined by the breath (most likely). I haven\'t investigated it at all aside from this, but I suspect the end result is that it would give functionality exactly the same as a wind controller (aside from the fact that the notes to play are determined by keyboard).

[This message has been edited by Jeff Hurchalla (edited 02-12-2002).]

[Duncan Brinsmead]

Jeff, this sounds like what I\'m looking for.
However do you think one should set velocity based on the initial output of the breath? Given that the attack is shaped by an onset from zero of the breath, one would need to wait a split second from the onset of the attack to determine the initial strength of the attack( the first non-zero value from the controller might be rather random ). This wait would introduce a latency, as you can\'t turn the note on until the velocity is determined. It seems to me that one could leave the polyphonic velocity associated with the original keystrokes and use another channel for the breath intensity. Thus one has two dimensions of expression input( polyphonic velocity and breath ) not just one. By the way..I\'ve only heard of Maestro tools in conjunction with GOS. What is it exactly?
Duncan

[Bruce A. Richardson]

<BLOCKQUOTE><font size=\"1\" face=\"Verdana, Arial\">quote:</font><HR>Originally posted by Jeff Hurchalla:
Hi King -
A maestro type software utility would do everything in real time with no latency issues. It was CAL scripts I was referring to that realtime might not be ideal.

Going by the problems Bruce mentioned for breath control, this utility would probably address them to make breath controllers functionally as useful as wind controllers. I\'m not sure yet if there\'s demand for it basically, so I can\'t say it will happen. What do the prices seem to be for an average breath controller vs. wind controller? For those reading, would it help you?

Here is the concept, based on what I understand as the problem Bruce had with breath controllers for keyboard:
Ordinarily a sample will begin playing when the user depresses a key on the keyboard, and after this initial attack the breath control will change the volume of the sample. This generally results in an attack that is uncharacteristic of a wind instrument, but does allow good control of expression while the note is playing. Keep in mind this attack problem doesn\'t exist for wind controllers because the sample begins to play when a breath is detected by the controller, and the keyboard is not part of the picture. Now what a software midi utility that I could program would do to fix the problem for breath controllers is to detect the note-ons from the keyboard without sending the note ons. This way it can \"know\" what note will be playing, and send the note on when the breath controller begins the breath - with the initial velociy determined by the breath (most likely). I haven\'t investigated it at all aside from this, but I suspect the end result is that it would give functionality exactly the same as a wind controller (aside from the fact that the notes to play are determined by keyboard).

[This message has been edited by Jeff Hurchalla (edited 02-12-2002).]<HR></BLOCKQUOTE>


Hi Jeff,

Yes, you are correct--this is exactly the behavior that would be required. It would be great if you built it into a standalone \"Maestro Tools.\" I have been furiously pitching the idea to controller and hardware manufacturers and DAW app manufacturers for a couple of years, and they have no interest at all. They consider us too much of a niche, and of course, Nemesys and TASCAM have not made the best of friends with the DAW guys--they have steadfastly refused to \"plugin-ize\" GigaStudio for performance reasons, and that pisses the DAW companies off (babies that they are on the issue--GigaStudio stomps ANY DAW app\'s disk-streaming capabilities, so the DAW manufacturers should rightly be trying to emulate GIGA, not the other way around!!)

The MIDI utility would need to \"hold\" the depressed key-generated notes, and \"wait\" to get CC2 information before actually sending the note on to GigaStudio, along with the breath information.

However, one problem I\'ve encountered with any of the BC-1, 2, 3 series breath controllers is that they\'re not really very sensitive, and don\'t respond anything like a wind instrument in practice. This is true in spite of me adjusting their sensitivities, responses, etc., in every way available. In contrast, the WX5 I use now feels almost exactly like a wind instrument, and responds that way.

I just bring that up, because what I wonder is what other issues and forces might be weighing in to create that phenomenon. Also, I wonder what effect CC2 \"leakage\" or jitter might have on a software-emulated \"note gate.\"

Another issue would be how such an emulation would handle polyphony, although I suspect it would be a fairly minor issue. A secondary feature might be a \"randomizer\" which would introduce some delays within chords that would otherwise trigger at the exact same instant (when CC2 information begins). Of course, one couldn\'t randomize \"earlier\" so this would be somewhat in need of post-manipulation to really get the feeling of an ensemble attack. Still, it would get partly there, and would sound better in a live playing context than a fistful of perfectly synchronized attacks. Even a few milliseconds would take off some of the \"curse.\"


Best regards,
Bruce

[Duncan Brinsmead]

I did a little more investigation and discovered that one can\'t use CAL for real time midi effects at all.

Help us, Obi-Wan., er..I mean Jeff. You are
our only hope!

Duncan