GPO Brass

[Michael Davis]

Is there anyway to get rips with the brass, like a loud sfz.

[Alan Lastufka]

Tom and Francesco were able to get rips, perhaps we could get them to do a tutorial or at least a short post on this technique...

[Michael Davis]

OK lets get that set up.

[Tom Hopkins]

Michael,

OK, I’ll put a tutorial on the list of things to do - there's a long list of things to do, as you can imagine. In the meantime, the trick is to use knowledge of the way brass instruments’ resonant modes work so that you can simulate the natural characteristics of a rapidly ascending rip. As an example: Take the case of a trumpet player doing a natural rip up to a “G” at the top of the staff. He will usually start the rip at a point at least a fifth below the target “G.” He will finger this “open” (no valves depressed). Only certain notes will sound as he rapidly tightens the embouchure. If the first note he plays is a “C” then the next note to sound will be the “E,” and finally the target note “G.” Those are the only notes that sound because of the resonant modes of the tubing for that valve combination. Second point: The first two notes (C and E) are very rapidly played. Third point: The first note is tongued, the other two are slurred. Fourth point: The addition of a small amount of portamento can help the illusion, especially with the last interval because a real player will have a small amount of “slide” as the note pops into the next resonant mode. Fifth point: The transition note(s) should usually be softer than the first and last notes.

If the player were to start the rip from the next possible lower note, that would be the lower “G.” So, the rip would rapidly touch the G, C, E, and finally land on the upper G. As the destination note gets higher the partials get closer together. Starting at the “C” below the example I just gave the ascending series would look like this: C, G, C, E, G, Bb, C, D, E, etc. When simulating the rip you must take into account which notes would actually sound leading up to the destination note and, as you can see, that depends on which note it is.

All valve combinations have the same pattern to the series, just a different starting note. These rules apply to all tubular brass instruments. That should give you something to experiment with until a tutorial is created. I would strongly recommend picking up a book that goes into detail on each brass instrument and its natural harmonic series.

What I’ve described is one way (perhaps the most common way) brass players execute a natural rip but there are others that involve using the valves in a rapid, rather random fashion. That would be much more difficult to simulate and is best ignored for simulation purposes.

Tom

[Tom Hopkins]

One other thing: As with most simulation tasks one of the best procedures is to locate a recording of a real player playing the type of effect you wish to simulate. Use audio tools like "time stretch" to slow the process down for analysis. That way you can see (hear) clearly what is actually happening when the real player is performing. Then you can take steps to duplicate the elements of what you observe.

Tom

[Jeff Turner]

Tom,

Interesting reading your comments here. I'm wondering in the upcoming Big Band library, if there is a way to implement any "half-valving" used in conjunction with portamento to further achieve this effect. That was the technique I used most often in playing rips. Don't forget fall-offs. They're kind of like rips in reverse, but not exactly.

Jeff

[Tom Hopkins]

Jeff,

I didn’t mention half-valve techniques because they are even more difficult to simulate, requiring targeted EQ for specific notes and greater amounts of portamento. Best not to go there.

On the general subject of the Jazz Add-on: I’m getting the impression that some are misinterpreting what this project is intended to be. Remember, it will be consistent with the design philosophy of GPO: as few samples as possible using programming techniques in place of sample count; easy to use; expressive control; single, all-purpose instruments rather than separate articulations; small footprint; as CPU efficient as possible; etc. Also, consistent with GPO, I expect it to perform well beyond its sample count but it is not intended to be an exhaustive exercise in sampling various techniques. Among other things, it will be valuable as a tool to check arrangements before you turn them over to real players. There are so many subtle characteristics for each instrument in Jazz performance that chasing that particular will-o-the-wisp is a losing battle – and is not our intention. I will integrate as many techniques as I can without violating the basic design philosophy.

Tom

[Markleford]

Also, consistent with GPO, I expect it to perform well beyond its sample count but it is not intended to be an exhaustive exercise in sampling various techniques. [...] There are so many subtle characteristics for each instrument in Jazz performance that chasing that particular will-o-the-wisp is a losing battle – and is not our intention. I will integrate as many techniques as I can without violating the basic design philosophy.
Tom, I love you. You're turning me on.

In a music software developer peer sort of way. ;)

Now, talk more of this design-philosophy sexy-talk for me... :D

- m

[dewdman42]

no doubt

[Tom Hopkins]

I can see that I need to take extreme care in the future to avoid incendiary phrases like “stacked parametrics” and “peaking filters.” Never mind “real-time manipulation” and “flex envelopes.” Some people appear to be too susceptible to the power of certain combinations of words, so I’ll need to practice “safe syntax” from now on. I’m just glad I found this out before something truly regrettable happened. “Assignable sliders.” I’m sorry, I don’t know how that slipped out (if you'll pardon the expression).

Tom

[Markleford]

Knob-tweaker. :p

- m

[dancase]

the trick is to use knowledge of the way brass instruments’ resonant modes work so that you can simulate the natural characteristics of a rapidly ascending rip.


Tom, this is an excellent example of an important principle in MIDI music production -- to make an instrument sound "like a real human playing," you have to understand the instrument. I'm a brass player... trumpet, flugel, cornet, and pretty much anything with three valves are things I understand because I play them. If I want to emulate a particular sound, effect, or texture, all I have to do is think through how I would produce that sound, effect, or texture when playing the horn. I also understand the limitations of the horn, and don't waste time trying to make it do something it can't do under the very best of conditions. If you balance that with an understanding of the limitations of your software tools, you can produce some incredibly real-sounding brass with GPO. When I listen to other people's GPO brass compositions I find it easy to identify pieces done by artists who either play or intimately understand brass instruments. There's nothing that raises a red flag quite like hearing something that can be done with GPO and a midi sequencer but can't be played on a real horn. There are also some playing techniques that can't be reproduced adequately by electronic means, and always sound phony. By understanding those limitations, I can avoid wasting time trying to produce something I'll never be able to produce.

OTOH, when it comes to the details, nuances, and limitations of strings I'm lost as a goose. :o


D.

[Tom Hopkins]

Dan,

Nail on the head!

Tom