posted 23 March 2006 03:36            

I am a 7 year Clavia Nord Modular fanatic, both G2 and G1, and have been using AudioUnits and other analogue and digital synths and programming langauges for more than 8 years now, and so yesterday was my first day with Kyma - X

I went through a few hours of serious doubt and scepticism, and then ended an 11-hour session with a few sounds, the like of which i've never made before... My own taste in source audio saved the day..

So, I want to clear up some of my doubts with the system...

My first big issue, is that the basic system ( which is all I can afford ) has its DSP limitations quite low I think - I was running out of realtime processing power quickly, and the demo sounds that interested me most , I couldn't even hear - the physical modelling, the Impulse Response cross-synths, Tau synth etc... At the level of experience I am currently clocking with electronic music, for me to hear something interesting and novel, seems like I have to already start thinking about the next DSP card or more... thats an expensive route! Is there a guide to optimising sounds/pathces before runtime, so that they can run at 48 or 44.1 without glitching? Is there a way to increase the buffer size even more, so that I can hear a sound? It seems a shame to have a dual processor G5 sitting there doing nothing, when the Capy could share that processing power?

When I started to build my own patches from the prototypes, I was frustrated to see that I can't seem to add a decent reverb to colour the sound without either dropping the sample rate to an unacceptable 32K or just making cutting down oscillator counts.. Maybe I'm not doing it right? On a typical G2 patch I can run 3 or 4 reverb modules in a patch without a problem.. My G5 can run 7 or 8 convolution reverbs and more... How can I warm things up if i need to, without plug-ins or reverbs available within a sound?

I'm used to patching on the modular or Max, where you drag in a module/prototype and place it in the chain - when I drag in a prototype in Kyma, it comes along with a load of other prototypes, always with a demo sample, or a signal flow already defined, which I find myself having to delete , and figure out how to try and shuffle things around so that i'm back to my patch with the module i was trying to patch into it, in the right place. This is frustrating, I can't seem to figure out how i'm supposed to build patches from basic buidling blocks, if the pre-defined modules are all patches within themselves - its seems difficult in Kyma to just say 'i want the output of this to go to the input of this' and plug things up quickly.

Also, I can't really figure out how voices work - how can I get polyphony working on Oscillators? How can I get say, 4 multiwaveOscillators voicing - can a basic system even manage that?

I really like the control mechanisms, the expressions library - these algorithmic control mechanisms give the system its authentic computer-music sound, in my opnion, and i'd like to start to write my own, for sure - is there any place where i can find more expressions that you can drag and drop? Like Swarm etc? These are very interesting for composition.

My biggest insecurity comes from the quality of sound itself - I've spent years keeping my digital signal path to the highest quality, investing in high-end convertors and clocks. My ears have heard the best and the worst, and I'm just not sure about the convertors on the Capy - Of course, I can't run 96K, cos I got only a basic system, and you can't do much at 96K with one card it seems. Maybe if I try clocking the Capy with a jitter-free clock? Or is it the nature of some of the sound-producing algorithms. I can hear aliasing at 48 / 44.1 , the sound stage doesn't seem very clear. I don't know. Can't figure it out.

Nevertheless, as far as really extreme electro-acoustic sound design goes, I feel it has great potential. This, of course is day One of a new relationship, so things will change i'm sure as I find out what I can and can't do, but I do feel that my imagination is being restricted by the DSP and RAM limitations with only one card. So far i conclude, that as a sound design environment, you need to be aware of the limits, and work within them - this is different to say, Metasynth or G2 , where you get a feeling of limitless creativity - when you run into a dsp overload on a G2, you are really patching hard - hundreds of modules, and you can still squeeze extra juice out of it if you know how. I know its not right to compare, they are different design philosophies, yet the feeling is important.

If there's any advice on how to reach a state where I'm not feeling restricted by the instrument i'm working with, whether psychologically or in DSP, it would be truly appreciated.

posted 23 March 2006 17:15            

I'm working on it , its obviously going to take more than a few days, I've read Carla's guide at the back of the book, and she mentions user's frustration , I guess i've just got a lot of habits to change or break - i have to change the way i think about creating electronic sound with this system - today I stumbled across a few prototypes that have inspired me a great deal, plus scripts controlling resonators to make a cheaper reverb

Still, advice is welcome

thnks

posted 23 March 2006 20:48            

quote:

---

Originally posted by cristian_vogel:
I haven't yet figured out how to take one prototype and move it to another position in the signal flow - i end up making a right mess

---

There isn't a good way to drag prototypes to new positions. What I usually do is put a new prototype of the same kind (doesn't really have to be the same kind, it could be anything that allows you to drag it into the place) in the place I want to move the prototype to, then I copy the prototype I want to move and paste over the temporary prototype. Then delete the old prototype. Seems a little awkward, but it works.

Some useful key commands to remember are Command-[ and Command-] (brackets) to make the prototype view smaller and larger (works on text too). Command-L enlarges the field to full screen (enter to close, rather than return). Double clicking in the background of the prototype display redraws the view after an edit. Double clicking in the background below with the fields makes the fields fill the sound display. And of course, be careful where you drop prototypes if you don't want to create a branch with a mixer.

After you've worked with Kyma awhile, you might think of doing something else with your G5 simultaneously - maybe you could free up cpu by running audio out of Kyma and into a VST host running a reverb. Or process midi through Max and send it into Kyma (maybe using the flame~ object from the Tweaky).

Eventually you'll want more processors, but you should be able to do quite a bit with the basic system. There are some limitations that seem strange at first, like not being able to hold that many samples in RAM, and limitations on disk streaming, but you'll figure out ways to get the most out of what you have.

I personally find the converters to sound better than my RME multiface...

posted 24 March 2006 06:15            

I now patch most of my Kyma Sounds into Logic in order to make use of my AU plugin library, including some nice reverbs, this frees up valuable Capy DSP and keeps the G5 in action. I too hit the DSP limit very early on, I worked with the basic system for about a couple of months before buying an expansion card. That improved things massively, and in combination with now using external reverbs etc. I rarely run out of DSP these days...

posted 24 March 2006 07:03            

I just want to clarify. You bought one expansion card, and it made that much of a difference for you? That's great news. Can you give some examples of where you hit the DSP wall before acquiring it, but now are able to do what you wish in real time?

posted 24 March 2006 08:26            

quote:

---

Originally posted by JackRosete:

I now patch most of my Kyma Sounds into Logic in order to make use of my AU plugin library, including some nice reverbs, this frees up valuable Capy DSP and keeps the G5 in action. I too hit the DSP limit very early on, I worked with the basic system for about a couple of months before buying an expansion card. That improved things massively, and in combination with now using external reverbs etc. I rarely run out of DSP these days..

---

And how do you patch out of the capy into an AU host and back into the Sound again? Like a send and return? or is there a more direct routing to an AU from within a Sound?

posted 24 March 2006 18:12            

I think that the problem is to think that the grey lines that interconnect prototypes they are representing physical signals instead of thinking that they are a logical junction between two prototypes or pieces of code.
In my opinion Kyma is very different to a classical modular system.
First thing is that you don't have cables to interconnect modules.
You have a signal flow where you INSERT pieces of code. That's not the same that interconnect output to inputs.
A typical modular system create the sound module after module. This adds different amouts of delay to different branches of processing. That's a big problem.
The signal flow in Kyma it's not temporal, the order of prototypes it's the order of operations but not the order of processing, because the processing it's done sample-by-sample.
Kyma does a purely mathematical process.
And the great cosequence of all this is that I call "Controland".
"Controland" it's the space that you can create in your mind when you use CapyTalk.
Like a neural process you can convert each fader of the VCS into a neuron controlling the flow between different parameters.
That's unique to Kyma.
I like this exercise: with only one hot variable, that is, only one fader in the VCS, how many differents things can I make with CapyTalk programming?

posted 26 March 2006 17:20            

1 - The quantitative difference between a standard system and one with an addition expansion card is about 50-55 % more power. However, in my case the subjective difference has been far greater because prior to adding the extra card most of my sounds were JUST peaking out the DSPs, and with the extra card they're now running smoothly. So the additional card provided just the amount of DSP that I needed (on the whole). I still run into the wall for sure, but much less often than before. I guess every case is different though, for someone else one expansion card might not make such a difference...

posted 27 March 2006 10:02            

quote:

---

Originally posted by cristian_vogel:
And how do you patch out of the capy into an AU host and back into the Sound again? Like a send and return? or is there a more direct routing to an AU from within a Sound?

---

It is via coreaudio input / outputs. the capy has a driver which advertises 6 IO into OSX coreaudio. you can then route audio to them like any other device except that instead of phyical IO you route to sounds (objects) you insert into you signal chain inside of Kyma.

In Logic you can use things like the IO plugin to insert into a given part of the mix or you can setup Bus and Aux routing. See the tweaky for additional files including stereo objects for kyma.

posted 27 March 2006 16:25            

A week in complete immersion, and I'm really feeling the power of Kyma X now, although I clearly have a lot more learning - the syntax of Smalltalk is a fiddly, and when you are experimenting in order to learn, mixing up dynamic Capytalk and Smalltalk and getting brackets and dots wrong can throw you.

posted 31 March 2006 05:59            

Looks like I am a classic case of 'interference' generataed from a complex system, confusing my focus! I am now feeling a sense of deeper understanding, and a sense that I have learnt a great deal more about where I want to be going with my music , and a really welcome feeling of having a powerful system at hand, that is already honing in on my personal need for depth in electronic music production, studio and compositional techniques. I think the key for me, was to find out how to integrate Kyma with everything else I have learnt and mastered over the years, rather than try to replace other tools,augment them with this wonderfully rich universe of sounds and new found ways of thinking about digital sound creation.

I've just picked up a Wacom tablet, and that has cured my onsetting tendinitis from using Kyma 10 hours a day, and also allows me to explore performance aspects of my sounds a lot more. Recommended.

Sorry for the panicky style of my first post in this thread, but when you are set in your ways , and something comes along that slaps you around the face, and shakes things up, no one said it was gonna be easy! Thanks again to Carla for being a guiding voice in the book.... and hopefully we can start getting things a bit more dynamic over at this forum, Kyma X needs alot of sharing, teaching and discussion, at all levels of ability and application with such a complex system.

posted 02 April 2006 00:25            

This is very true. It took me a long time to realize this. The sooner a new Kyma user realizes that things are done differently in Kyma, and more importantly they are done differently for a very good reason, the sooner that user can get on using Kyma in the best possible way.

This is also one of the hardest things to explain to new potential users. It`s like the expression, "a picture is worth a thousand words". I`m sure maybe Carla and a few other advanced users with a good journalistic sense could somehow write up a big long article that explains what it is about Kyma that is truely different, and most importantly useful. They could go on and on and even still one wouldn`t know for sure and would still be judging it somewhat on the traditional synth scale. Then you use it, and use it for awhile, practicing and not fully sure exactly what you`re doing. And the book is great, you follow along as Carla is the wise sensei that shows you the way. You just practice and practice, slowly working with it, and then things start to reveal themselves and then...BAM! You`ve got an understanding of it. And then keep practicing and then....BAM! An even deeper understanding of it. At least thats how it was for me. I really started to giggle at some points. I`m sure this doesn`t apply for everyone. I`m sure there are some that this was exactly what they thought it would be, but those few people probably couldnt find what it was they were looking for until they heard about and tried Kyma.