Jump to content

EXS24 Alternate Sample-Round Robin Discussion


the sinner

Recommended Posts

Okay, see if this works for you. Let's say you want to string 3 samples so they trigger on three subsequent keystrokes.

 

1. Assign each sample to a different group. Give each group a unique name (or leave the default names given for each group if you prefer).

 

2. For group 1, make the first select by "Group: ---start---"

(Leave the second select by empty ("-----"))

 

3. For group 2, make the first select by "Group: name_of_1st_group"

(Leave 2nd select by empty)

 

4 For group 3, make the first select by "Group: name_of_2nd_group"

(Leave 2nd select by empty)

 

 

Does this make sense? You should be able to string together any number of groups in this manner.

 

Let me know if I need to clarify.

Link to comment
Share on other sites

I don't think I'm understanding....

 

Is the second and third sample played consecutively after one triggering (like trigger 1=guitar strum down, then up, then finger noise), are they each being triggered by the same key individually/consecutively (like trigger 1= guitar strum down, trigger 2 same key=guitar strum up), or something else entirely?

 

If I could grasp this, I'll brainstorm and throw some ideas out there for y'all, LOL!

 

X

Link to comment
Share on other sites

I think you have it right, X-man.

 

Here's an example:

 

You have three samples of spoken word: Help, Me, & Understand

 

                Lo/Hi

                 Key

Zone 1 = C3/C3 = "Help" and is assigned to Group 1 (named "A")

Zone 2 = C3/C3 = "Me" and is assigned to Group 2 (named "B")

Zone 3 = C3/C3 = "Understand" and is assigned to Group 3 (named "C")

 

Group 1 = ---start---

Group 2 = assigned to "A"

Group 3 = assigned to "B"

 

If you play C3 four times in a row, you'll hear, "Help, Me, Understand, Help"

 

I hope that helps you understand! :D

Edited by ski
Link to comment
Share on other sites

Thanks ski.

 

So if I'm getting it right, you could make an incredibly detailed, human sounding drum kit.

 

If each time you trigger a snare and it produces a hit that is the same velocity, but slightly different sounding (e.g. closer to the rim, left handed, a hint of rim contact, etc etc) that would be really cool.

 

X

Link to comment
Share on other sites

That's right. That's exactly the point of "round robin" (though we're looking for non-traditional uses for it as well).

 

But round robin does not produce random results. It does let you simulate randomness to a degree, which I'll explain...

 

If you have a snare sample -- just the one -- and you play a 16th note pattern with it, you'll very quickly hear the "machine gun effect". That's what it's called when the same sound is repeated verbatim over and over again very quickly. The ear can discern the attack as being the same every time.

 

If you have two snare samples which alternate (round robin in its simplest form), every other snare would be a different hit. That would improve the degree of "machine gunniness" but it would still sound like two samples. Better, but not great.

 

If you had three snare samples which alternated (1, 2, 3, 1, 2, 3, 1, 2, 3) the ear has more of a chance to get fooled by the subtle changes in timbre between the different snares. If you play a 16th note pattern (4 hits per quarter note) with three snares, there's a chance of it actually sounding more realistic than if you had four snares, because 3 doesn't divide evenly into 4.

 

So the way the theory goes, the more samples you have to alternate between, the more convincing the part will sound. That's the theory, and it's based on the fact that on acoustic instruments (as well as real analog synths) you can never hit a note the same way twice. The ear has a tendency to pick up and "tire" of repetition. But it will quickly discern -- and thus find musicality -- in the subtle variations that would normally occur with acoustic instruments, or even in environmental sounds.

 

But simulating randomness comes with a price: Time to prepare and program all those samples is one. Sample memory capacity is another potential issue (i.e., using up a ton of RAM to hold lots of variations of the same sound for any one instrument). And if you limit the number of sample variations to a low number, you'll start to hear patterns form in the timbre over the course of a bunch of repeated notes unless you resort to some programming tricks (like 3 samples for patterns of 4 notes. Or 5 or 7 samples for patterns of 4 notes).

 

Anyway, at this point I feel like I'm taking the joy out of this otherwise cool feature, so I'll stop.

Edited by ski
Link to comment
Share on other sites

Is it rude or helpful to answer a question directed to another? While that is pondered...

 

No, the arrangement of groups won't do it. If you have three variations assigned to three groups, you'd set them as follows:

 

Bassoon staccato 1: select by: ---start---

Bassoon staccato 2: select by: Bassoon staccato 1

Bassoon staccato 3: select by: Bassoon staccato 2

 

This will round-robin infinitely through each group.

 

In theory, you can see this happening in the selected white zone boxes, which should travel up and down as the groups alternate. But sometimes the visual feedback falters even when the round-robin still occurs.

 

Keymap has an excellent way to visualize the process as you build your patch.

 

It's best to learn this method with very distinct groups, like a stac to a pluck to a bow, to get a sense of EXS' behavior. By contrast, the nuances of VSL variations can be so subtle, you keep hitting the key and asking yourself if it's working.

 

This method usually uses velocities determined by zone, then all velocities of a given iteration assigned to one group.

Link to comment
Share on other sites

Thanks guys for all the answers. It is great to know how to use RR in EXS24. I also understand how to group in EXS24, i am an ex DP man so it has taken time to get used to this.

 

Ski, as you can see I implied the keyboard was involved or else how would you get a long cymbal crash with no mod wheel?

 

 

 

 

see that I can use modulation to change pan balance, output and volume, but can i use it to trigger an entirely new sample? A muted cym crash with mod wheel up and a long one with no mod wheel.

 

Sinner,

 

A mod wheel at rest triggers sample A.

A mod wheet at full or anything triggers sample B.

 

So the mod wheel does not vibrato or anything it only triggers pre recorded samples.

 

After going through the trouble to illustrate everything on the previous page, and now seeing that this isn't what you wanted, it's time to get the nomenclature straight...

 

"Trigger" is synth terminology for "the thing that initiates a sound". So to say that "the mod wheel.....only triggers pre-recorded samples" indicates that you want modwheel movements to actually initiate the sound.

 

As Fred has pointed out, Orsanct has already provided you with the answer.

Link to comment
Share on other sites

But round robin does not produce random results.

Ski,

 

I've been thinking about this "problem" with round robin. Are there known work-arounds? The only thing I could think of would be to:

 

1. Set up different velocity Select Ranges for each group. (I suppose if one wanted avoid volume changes altogether, one could assign a single number for the range, e.g., Group 1: 80-80, Group 2: 81-81, etc.)

 

2. Notate the repeated note as desired.

 

3. In the matrix window, use the transformer to randomize the velocity of each of the note, keeping the ranges to those set in the group windows.

 

 

Of course this isn't truly random since the velocities are fixed once you apply the transformer. Might there be some other, more efficient, ways to do this?

 

 

By the way, I thought your "Help Me Understand" illustration was incredibly elegant! You are a natural teacher. :!:

Link to comment
Share on other sites

Of course this isn't truly random since the velocities are fixed once you apply the transformer. Might there be some other, more efficient, ways to do this?

 

Wiring a transformer object in with a random note velocity operation allows you to set a hi-lo range for randomness. A fader object set up with meta messages can control that range, thus the fader could be automated from the arrange or Hyperedit. So for example, for 10 bars the random velocity range could be 79-127, then for 3 bars it's 45-90.

 

You could also experiment with subtle use of the EXS24's random waveform LFO's set to a low rate and assigned to one or more timbre altering parameters. But definitely subtlety is key here! Velocity could be assigned to additional parameters too. Although I wish the EXS had more interesting mod destinations, you can still get pretty fancy with it.

 

And yes, ski is an excellent teacher. I just wish he'd stop sending us those tuition invoices every time he answers our questions!

Link to comment
Share on other sites

I did come up with an idea, but it would probably take more work than it's ultimately worth. Here goes....

 

What if you were to take one sample and beat the crap out of it a few times?

 

My suggestion would be to take a sample, drag it into the arrange, and add some odd effects/panning then bounce it however many times you want a different sound. Then you could layer the original sample, panned in the center, over the new audio file so the original sound is there with some cool sounds going on underneath.

 

for example:

Take a synth lead (my favorite is the Garageband preset, Ominous Dancefloor) and record it for however long you think is necessary. Then bounce. That is your "original sample." Drag a copy back into the Arrange and go to town.

 

For the first step in the round robin, I'd pan center and have the channel strip something like bitcrusher->AU Matrix Reverb->Stereo Spreader. Bounce that out.

 

Then the 2nd hit I'd pan hard left with a channel strip of Tape Delay->Autofilter->Flanger.

 

The 3rd would be panned hard right, channel strip would be a stutter effect->Guitar Amp Pro->Tremolo.

 

Etc etc etc etc.

 

Then for each round robin sample (I'd make a bunch, around 9 all panned in different areas...I'll explain in a minute) I'd bounce the original file (panned center) with the effects underneath so they don't take away from the original sound, just add interest. Then program the EXS.

 

When you play this instrument, you could have a rapid succession of key hits where the original sound would be dead center and the added effects would be bouncing all over the place.

 

Does that make sense?

 

X

Link to comment
Share on other sites

One thing about random effects is that they're fleeting. If you've got some kind of random anything set up and it sounds perfect as you're experimenting with it --- but you haven't recorded it --- that exact sound/part/effect is lost forever.

 

For me, when it comes to doing orchestral music, I don't leave anything to chance; I like to have the utmost in predictability. But for other styles of music, random stuff can be a fantastic inspiration, and remains that way as long as I've recorded it while the magic is happening.

Link to comment
Share on other sites

Wiring a transformer object in with a random note velocity operation allows you to set a hi-lo range for randomness. A fader object set up with meta messages can control that range, thus the fader could be automated from the arrange or Hyperedit. So for example, for 10 bars the random velocity range could be 79-127, then for 3 bars it's 45-90.

fader8,

 

Thanks for these ideas. This is my first sojourn into using the Transformer so this is all new to me. (Very cool. Kind of reminds me of several objects in Max/MSP.) So, a question. . .

 

1. I've set up the transformer object per your instructions. It works great.

 

2. I connected two fader objects to control the min/max of the random operation. They work fine as well.

 

3. Now, how does one automate the fader objects such that they can be controlled from the arrange/hyperedit windows? The manual says something about "using its In definition", but I just couldn't figure that out.

Link to comment
Share on other sites

For me, when it comes to doing orchestral music, I don't leave anything to chance; I like to have the utmost in predictability. But for other styles of music, random stuff can be a fantastic inspiration, and remains that way as long as I've recorded it while the magic is happening.

Indeed! That's why I really like the random button in Kyma. Sometimes, when I'm stumped for ideas, I'll just hit it a few times and, if I'm lucky, it will lead me to ideas I never would have thought up on my own. Same thing for the randomize button in Sculpture and (to a lesser extent) in Ultrabeat.

 

Generally, I work pretty methodically, but I do like those having those out-of-left-field ideas to help inspire me now and then.

 

Sidebar: Regarding your comments about orchestral music, it's been interesting to see how various composers handle "random/chaotic/aleatoric" passages. I know, as an interpreter, I usually prefer having some kind of notation to struggle with, even if the effect is intentionally chaotic. I have found that when composers give random notes or have long time-framed passages, performers often don't perform with the same intensity. (I am, of course, referring to the orchestration "technique" of creating random-sounding effects -- not to works that are based more "philosophically" on chance operations or processes such as those by Cage, Oliveros, etc.) It's all mostly psychological of course, but I've found it interesting to see how composers have dealt with this over the past 40 years.

Link to comment
Share on other sites

When you play this instrument, you could have a rapid succession of key hits where the original sound would be dead center and the added effects would be bouncing all over the place.

That sounds very interesting! Just to make sure I understand - each completed sample then has the same primary note in the foreground but a different "background" effect. Is that right? I could imagine some fascinating colors and flutter-effects resulting.

Link to comment
Share on other sites

3. Now, how does one automate the fader objects such that they can be controlled from the arrange/hyperedit windows?

 

That, Mr. O, is what the little icon is for in the fader parameter pane. Check the box next to it. Any environment object that has its icon enabled can be a target for a track.

 

In the arrange, create a track and click-hold near its name to reveal the destination menus. Since the names here reflect the environment layer names and then the object names those layers contain, you should see the two fader objects you created and hopefully gave unique names to, (if their icon is enabled). Pick one.

 

Now you have 2 choices. The fader objects input defaults to cc7 when it's created, so just enable automation view for that track and use volume to control the fader. Create another track for the other fader.

Or . . .

Create a region with the pencil tool for that track, open Matrix or Hyperedit and draw.

 

If you want to control both faders from the same track/region then create an Ornament in the environment, check its icon, cable it to both faders and assign a track to the ornament. Of course each fader will be set to respond to a different cc so you can have a Hyperedit lane for each.

Link to comment
Share on other sites

Another passing idea for "round robins". . .

 

In Kyma, I do a fair amount of manipulation of speech including chopping up and reordering words, syllables and phonemes. In Logic, I could see some use for placing syllables into separate groups (they could be "in order" or not) and triggering the assigned key to generate rhythmic structures that repeat the syllables over and over, but in different places in the rhythmic figures.

 

If we took a repeated short-long-short rhythmic figure, it might come out something like:

 

help me un

der stand help

me un der

stand help me

 

Or, applied to the palindromic figure used in Steve Reich's "Clapping Music":

 

help me un

der stand

help

me un

der stand help

me un

der

stand help

(etc.)

 

Furthermore, you could assign additional chopped up phrases to different notes to increase your number of "syllable instruments". With some filters, LFOs and other goodies, I could see some fun here. . .

Link to comment
Share on other sites

Just an additional footnote to this thread, for those of you with Kontakt2 there is not only a round robin feature but also a random round robin feature. This can be taken yet a step further for additional control using Kontakts excellent scripting features. Check out this gentlemans website for some "controlled", randomized round robin:

 

http://www.peteroregan.com/scripts.html

 

ski, you'll appreciate this as the mp3 examples are orchestral.

Link to comment
Share on other sites

Thanks, fader8. That actually works quite well. I think the only place that approach doesn't work is when you have double-tongued brass/flute passages. For that, I would recommend an alternated "tuh" and "kuh" sound. I suppose if you wanted to get really fussy, you could devise some way to randomize samples within the "tuh" and "kuh" -- but not mix them up willy-nilly.

 

At any rate, a terrific solution to some of the challenges posted here.

Link to comment
Share on other sites

In this example I've created a little environment chain where the mod wheel controls a cable switcher. The first route out of the cable switcher is directly to the EXS24. The second route goes through a transformer called "transpose", which could easily be set to transpose all your C#1 notes to C#2 (or some unused note), where you could put a choked cymbal. If you're not feeling terribly ambitious, you could as easily route this second chain to another EXS24.

 

Don't forget the parameters of the cable switcher (shown at left). The "input" should be set to "control", and the "-1-" parameter should be set to "1" (modulation wheel). All the "cymbalchoker" transformer is doing is scaling the mod wheel back to a maximum value of 1.

 

Note that the "icon" checkbox in the parameters for the "cymbalchoker" transformer should be ticked, which makes it possible to select this transformer as an instrument in the Arrange Window.

 

Rather than dick around with the EXS, that would be my method of choice.

cymbalchoker.thumb.jpg.dd0299ec16a7f24938b58b41a4196550.jpg

Link to comment
Share on other sites

(fyi: my most recent post, above, is a response to only seeing the first page of this thread ... sorry it's off-topic now)

Funny. So it happens that answering the original question seems 'off topic'...

Anyway, I think your solution has the advantage that it's independent of the instrument where the samples reside.

Link to comment
Share on other sites

  • 1 year later...
I think you have it right, X-man.

 

Here's an example:

 

                Lo/Hi

                Key

 

Zone 1 = C3/C3 = "Help" and is assigned to Group 1 (named "A")

Zone 2 = C3/C3 = "Me" and is assigned to Group 2 (named "B")

Zone 3 = C3/C3 = "Understand" and is assigned to Group 3 (named "C")

 

Group 1 = ---start---

Group 2 = "A"

Group 3 = "B"

 

If you play C3 four times in a row, you'll hear, "Help, Me, Understand, Help"

 

I hope that helps you understand! :D

 

I have followed these instructions and it works, sort of... But when I play really fast notes I still getting the machine gun effect!! Am I missing somthing???

Please listen to the attached audio file.

alternation.zip

Alternation

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...