Kyma & Its Alternatives

[ghettosynth]

There is a relatively old thread at Gearslutz that includes a nice comparison of Kyma, MAX/MSP and the Nord Modular. Post #4 is particularly detailed ...

http://www.gearslutz.com/board/electron ... dular.html

Peace,
Andy.

Funny you found that. I forgot about it but I read it some time ago. I think that I ran across it when I was considering upgrading to a G2. Changed my mind, went with Reaktor.

Funny that you found that. I met the guy who wrote that post... Anthony Bisset. I went over to his place one time and swapped my Nord Modular rack for his keyboard version. I later read that he claimed the one I gave him produced a horrible whining noise, and he ended up selling it. I certainly never heard that noise myself, nor did he notice it when we tested the units! He seemed like a very strange guy, always in search of something more rare, obscure, exotic. I'd be surprised if he's still using Kyma and hasn't moved on to something even more esoteric...

I'd love to get one myself, but I still just don't have the free time to make proper use of it.

That is funny, bet you money that we (you and I) know each other from some other internet venue. No need to try and find out, that's why I change my alias every few years or so.

As far as his scale of usability goes though, I would put reaktor north of max, but, based on what he's saying, south of kyma.

[whyterabbyt]

As far as his scale of usability goes though, I would put reaktor north of max, but, based on what he's saying, south of kyma.

Depends what you mean by 'usability' (eg ease-of-use versus flexibility-of-use) and what you're trying to use it for.
MAX/MSP has several significant areas where it is more fully featured, or more extensible, than either Kyma or Reaktor, although audio, in some ways, was one of its weakest areas up until v6. Even more than the Kyma, one pretty much needs to work from scratch with MAX, as there's minimal 'factory presets' for anything, but it gives you a far greater breadth of tools; its core has dedicated audio, graphics, video and MIDI processing intrinsics; it will even host VST plugins and ReWire. It can be extended via an internal Javascript interpreter, which also allows the creation of custom UI and processing modules, external modules can be written in native C++ or Java, and the optional GEN toolset even gives you an on-the-fly native compiler operating at much the same level (per-sample) as Reaktor Core but capable of working on video streams as well. Going further, MAX also has third-party externals which leverage other audio-processing engines like Supercollider, and CSound.
Unlike either of the others, you could probably implement your entire studio toolchain in MAX, from audio editors to audio processing to a custom DAW.
The obverse of its 'blank canvas' approach, though, is that, whilst you could conceivably build anything, you need to put a lot of work in to do so, and those that have done so dont tend to have the same 'regular music' focus as Reaktor people, so much of what was built wasnt shared the same way; there have always been lots of available libraries of prefabricated low-level components, but traditionally there was little that was comparitive to the Reaktor Library, ie built synths or effects. People were custom-building tools for themselves, so the 'sharing' was in information and
The change in that culture has been MAX4Live (which, if you look at it the other way round effectively adds prebuilt DAW functionality into MAX), and the M4L device library site, now the balance is changing, but to me MAX still needs an out-of-the -box library of medium-level components to lessen the learning curve.
Horses for courses though. Just as some people will prefer Reaktor to MAX/MSP because its more narrowly streamlined to a particular use case, so some people will prefer, say a G2, to Reaktor for even more of that that same streamlining, and some people again will prefer a Eurorack or somesuch to a G2. Personally, I'm equally comfortable with all of them, except for the Kyma which Ive never used. (Although I had a brief flirtation with Smalltalk in the late 80s or so when object orientatation was noticably The Big New Thing and it (SmallTak) looked like it might get some sort of footprint in industrial useage.)

[josematou]

i'm selling my pacarana on ebay

ebay id number : 221278817552

it's a good choice for european (the pacarana is almost new and you don't pay the taxes)

[secretkillerofnames]

but to me MAX still needs an out-of-the -box library of medium-level components to lessen the learning curve.

I'm of the opinion that Jamoma is trying to be this.

Max 6 made it a lot easier for me to get into building my own stuff with the added help and reference functionality - though it still isn't as newbie friendly as it could be.

[whyterabbyt]

but to me MAX still needs an out-of-the -box library of medium-level components to lessen the learning curve.

I'm of the opinion that Jamoma is trying to be this.

Max 6 made it a lot easier for me to get into building my own stuff with the added help and reference functionality - though it still isn't as newbie friendly as it could be.

possibly, but jamoma seems, to me, to be one of those projects that goes through constant refactoring rather than evolution. It started to look over engineered, for the sake of the engineering alone. I'm was also concerned that windows development trailed at various times, especially after the history of taptools, which jamoma basically started off as a framework for, and basically became mac only.

[sadgod]

Since I have recently dove into this foray of resynth fft programming dsp... stuff, figured I would add what I've found. Pecking away at supercollider, CDP, and Read Kyma Revealed from cover to cover and use PD and Openmusic.
Every one of them can get as low level as you could possibly imagine. What I will say is, Kyma appears at least to me to have plenty of modules (sounds as they are called) The work flow in Kyma, though I've never used it hands on, is very ideal... iffff you got the cash. People have a hard time describing it, because it can't be described on a forum so quit giving people so much grief for not being able to describe it. All I can say is, after messing with all this stuff, you can get faster mostly realtime results on the kyma that for obvious reasons are going to be better than any realtime plugin that could be developed at this time. The dsp in those things are not small, the new ones at least from all the info I found on the net seem to equal something as powerful as an i7. You can't run those algs on a computer with anything else running at the same time. Anything else claiming to do this in realtime and not seriously drain your cpu is obviously going to be an economical means to that end. It won't be the same. On the other hand, Something like CDP will get you results fast, but far from realtime, and people are hung up on the interface. That being said, I'm no expert, but Kyma has its place, and it makes perfect sense to me as of its purpose, you just have to dive into into things and get your hands dirty with other things before you start to understand, unfortunately.
It's more visual than most, but I would call it a sound designers dream and expressive interface, reading the book and going through the tuts really left me wanting one. Something like Max and pd are somewhat similarly a higher level/visual type interface, but I personally use them for sequencing.
Open Music is all about algorithmic composition but can do anything else, if you can get past the french on some modules, similar to pd.
Metasynth and CDP are very similar in function, but I've never used Metasynth hands on.
Supercollider is all in script, and just the list of classes are 37 pages long, its very powerful, and I'm liking it now that my programming skills have improved. Csound, you are going low level stuff, very deep, all the time.
Any of these are going to be one hell of a commitment, the difference with Kyma is the Commitment scale weighs heavier on the monetary side while seeming easy to get going with the basics. Everything else is lots of time. It's all a mind opening experience for me. If you don't want to spend the time and money, something like reactor and Metasynth will do just fine, in a complimentary nice visual interface kinda way, if you want some fairly fast gratification.
Don't know if any of this helps, but I honestly tried.

[timbres]

I was like a lot of you, SO enamoured with everything I heard and read about Kyma and overwhelmed with all of the potential aspects of original timbre creation. I finally dove in and bought one. I'll never forget my wife and I going on vacation right after it arrived and choosing the manual as my "book" for the trip. She though I was crazy! (I had the same experience and extreme joy with Mark Vail's Vintage Synthesizers

When we got back, I took the phone off the hook (when that was possible) and connected everything up with a smile that couldn't fade. That is, until I had played though all 1,000+ presets. There wasn't a single sound that I could use in writing a song! I was so disappointed. As Syncretia mentioned, everything about the interface was completely alien and unfamiliar. The additive and resynthesys modules were very cool, so much so that they were almost enough to keep it, but I couldn't afford to just for those features. After 3 weeks, I sent it back and was charged $300 for the 'rental'

Carla kept telling me on the phone that any sound was possible if I just worked at it. I asked her, "Then, why hasn't anyone done it?". And there's my point. For all the theoretical talk about what it can do, I have yet to hear ANYTHING that is impressive musically or timbre wise. I kept remembering the blue demo album for New England Digital's Synclavier and how I couldn't sleep after hearing it! Nothing had ever been done like that before and the presets were breathtaking - as was the price! I remember seeing one live at a jazz club in Austin just after they were introduced and wondering how any musician playing in such a small venue could afford one. Then I found out that The Car's keyboardist had 2 of them for playing those little melodic ditties :/

Why didn't Symbolic Sound hire someone like Larry Fast of Synergy to program presets for them? Something we could all get excited about and use immediately. We would have delved into them to see how they worked and gone deeper from there. They still need serious help bridging the vast gulf between their academia perspective and average user.

What I came to realize is that it's not so much a sound generation tool but an expensive signal processor, manipulating preexisting audio and voice; which is why film studios are probably their biggest client; morphing and mangling imaginary creatures (think Gollum).

To be honest, I searched to see if they were even still in business and found this thread.
With everything available today, I honestly can't imagine putting out $4,000+ for this.

My experienced 2 cents...

[Shy]

I can understand your feelings about this but I think it's more an issue of you not being the right user and not the hardware not being the right hardware. It does suck that the presets are not so good or interesting, that's a good point. I thought there are good ones published by users, though. The real reason to get that processor, though, is for the high quality implementations of the effects it offers and for making your own patches. The big problem is that people wouldn't usually know which of the algorithms in it are ones they're really interested in, and which could be useful for them. They also don't know whether they'd be capable of really customizing their own patches. Therefore, most people just have no way of knowing if it would be a good enough investment for them, if at all.

It may sound strange, but from how I see it, it should be easier to create your own unique patches on that system than on an Eventide DSP-series (and its descendants) which is "merely" a system that's very focused on pitch shifting and "basic", traditional delay-based effects. The software they give you is so great since it simplifies and speeds your workflow so much and enables you to patch things up in a way that's still "advanced" but which doesn't require anywhere near as much in-depth understanding of the modular structure of a patch/algorithm as in Eventide's interface (vsigfile) and doesn't frustrate you to death with a million things you have to remember and wade through every time you need to make a change, not to mention you could remove or rearrange things without breaking the entire patch with every little change, and you don't have to spend much time building the user interface yourself to enable parameter display and control for every different patch. It seems as good as modular effects interfaces get, along with Nord Modular's.

The question is, are you really interested in that? Are most people? The answer is definitely "no". Most people don't even know what they want except "great sounding stuff", they don't care much about how an effect is built and about the potential to seriously customize and tailor an effect of a certain type for themselves.

I'm coming to the really major points now: In fact, for me and I believe for most people, it's frustrating to have to learn how to patch an effect in a complex modular effects processing environment only to be able to create a few effects that I'm really interested in and from then on, not touch them anymore and not make others since the ones I already made are as good as it gets for my taste. The power and flexibility does allow me to create many more effects, but say I've learned and figured out all or most of what the processor has to offer and decided that there's no point to make stuff that's not as useful, then what is all that power good for, then? Then, you start to think about what a waste it is, but that's natural. It really isn't a waste, because apparently for me and some others there's no other way to get the effects you REALLY want unless you have the tools to make them yourself. No one else has made them for me, and literally, no one else could have, so I never would have gotten the sounds I wanted using anyone else's creation.

Almost all the presets even in an Eventide H8000 are totally useless for me and completely uninteresting. The only reason for me to have an Eventide unit is to use patches I customized myself or made pretty much from scratch. In each category I've covered with it, I get the best effects I've ever heard in my life, because I got the chance to make them myself on a system with hardly any signal path limitations and with essentially perfect sound quality. I never got a Kyma processor because even though I know the sound quality, flexibility and versatility of their spectral effects is unmatched, I know that it wouldn't be useful enough for me to justify the price. I just don't have very much use for that. The thing is, I know enough about effects to know what's useful for me and what isn't, and I had to learn some of it the hard way. I guess I could say Kyma processors are for people who know what they need and are very familiar with effects processors, and I don't mean it in any negative way, it's just a tough issue and there doesn't seem to be a way around it.

[timbres]

I agree with you, there's nothing wrong with their hardware. And I forgot to mention one of the strengths of Kyma is in processing live input.

It's really all about having wrong expectations, and that's what I had. I wanted something ground-breaking but with an intuitive UI that encouraged ans enabled creativity rather than something that was actually distracting and discouraging, which I found this to be in actual use. I still haven't heard anything made rather than merely processed that sounded truly original and worthwhile of the price of this unit. I chose poorly and made the immediate correction and sent it back after 3 weeks.

I just want interested people to understand from an Actual User that this isn't something like other synth platforms where you get large amounts of usable presets to start making music with. As Christian discovered over a longer period time of usage, the result is the opposite and creativity is basically killed.

You have to clearly understand what this system is and what it it not before investing so much money in it because if my experience still holds, they will charge you for the time you take to try it out and send it back. $300 for 3 weeks is not insignificant in my book.

Eddy Jobson recorded an entire album using only the Synclavier. That is what I'd expect from a Kyma user after all this time if it's even possible. Please point it out to me if anyone knows of one.

Like I said, i think post-production houses will find the best use for Kyma ulitizing the cross-synthesis and morphing features. Some musicians processing live input seem to like it as well. More power to them.

[Shy]

The thing is that anything that doesn't suit you could kill your creativity, like something you can't or are not willing to spend a long time to learn to get anything useful from it. A Synclavier would brutally murder your creativity if you wanted to do resynthesis with it. The FM and additive synthesis features are also complex to handle. Most things about Synclavier are complex and far from intuitive, especially compared to what's available today.

If you spent some more time with Kyma focusing on only a few features, I think it's likely that you would have gotten some great and inspiring results with it. Whether the process was fun enough or not, that's something else. Personally I don't like spending so much effort and time learning, modifying / building and finally tweaking things to death, but I absolutely love the result, so there's always that balance that keeps me inspired.

I can also blame Eventide for killing creativity by not offering a much better modular interface for so many years, but there's no point. I think it's sad that most people only know the included preset algorithms and at most, Italo de Angelis's preset libraries. Then there are a few unimportant presets from others that are almost impossible to find and that's it. Even Italo told me recently that he doesn't know of anyone else who makes presets / algorithms for Eventide units. That's sad but totally understandable, because hardly anyone would bother with vsigfile, so everyone is left with an amazingly powerful effects processor that's largely misrepresented by the presets / algorithms included with it, and no intuitive way to make your own original effects. I forced myself to learn it and by doing so, I got the best effects I ever did, and the only way I could be happier is if there was a greatly improved interface for it. The bottom line is that neither Eventide nor Symbolic Sound have done their best to improve the interface and their presets, far from it, but they both got the basics right, so it IS possible to get over many of the issues, but not everyone can or cares to do that.

[Winstontaneous]

When I started playing upright bass years ago, several of the method books I used said something to the effect "You need to have a concept of the sound in your head before you make it on the bass." It's an unwieldy instrument that requires proper technique--you can easily destroy your tendons if you don't know what you're doing.

I've found the same to be true in my explorations of Pure Data, CSound, and SuperCollider (minus the injury). There are so many possibilities that unless I have a clear goal in mind and an inkling of the technique necessary to achieve it, it's nearly impossible to get anything useful, and the vast majority of presets are irrelevant.

I'd personally want to figure out one of these free DSP systems before spending money on hardware. But people like Ben Burtt can make a Kyma sing, and folks who actually figured out the Eventides have gotten great sounds that created and defined whole genres (Brian Eno, Daniel Lanois, Squarepusher, Bill Laswell, Laurie Anderson).

[jackmazzotti]

The best thing to do with kyma is to experiment with your own creations and then tame them inside of metasynth, pitchmap or melodyne. Then feed those samples back into kyma or reaktor. Trying to make traditional playing synths like plugins is a complete waste of time in kyma. Having it process external synths on the other hand can produce very interesting and sometimes beautiful noise and textures. I have a paca and I wish a got the pacarana instead because I run out of dsp so easily.

[Scot Solida]

A lot of people think of Kyma as purely an esoteric sound design/mangling tool. And, really, the preset factory "Sounds" don't dissuade this impression. Sure, there are a lot of "synth" patches using just about any form of synthesis imaginable (and one I'd never have thought of), but most of them are "single-purpose" designs, meant to perform one task only (say trigger a pair of sync'ed sawtooth waves). There are few full-blown, fully realized instruments in there, like you find in, say, the Reaktor library.

With that in mind, and primarily as a learning exercise, I recently spent a few weeks dedicating myself to building up some fully-loaded synthesizers in Kyma. The best of these were a PPG-style wavetable synth (using wavetables I made on my Waveterm), a Wavestation-like dual wavesequencing synthesizer, and a synth inspired by my VCS3 (complete with pin matrix - that was a chore, I can tell you! Also the looping trapezoid style envelope - I got some help from more advanced Kyma users for that bit, as well as other things that left me scratching my head).

Here's what they sound like in action (this is all Kyma - primarily the three synths described, plus some other things I've made - though one sound is my own version of Camille Troillard's JX3P clone):
http://soundcloud.com/scot-solida/phosphene-beacons

Here are the interfaces for each - they are Kyma, so they're pretty ugly. However, they do the job!




If you use Kyma and want the Sounds, you can get them here:
http://www.theelectronicgarden.com/Scot/WaveState.zip
http://www.theelectronicgarden.com/Scot/WaveSynth.zip
http://www.theelectronicgarden.com/Scot/Synthi.zip

[secretkillerofnames]

Thanks for the example Scott. I presume this was recorded into a DAW and edited into shape? Or is this a live recording?

[Scot Solida]

It was recorded just as you would any MIDI synth, soft or hard. I composed the MIDI sequence in Cubase 7.5 and tracked Kyma's playback of each MIDI track onto a new track. The more atmospheric and abstract things and the lead sound at the end were played in live using Kyma Control on the iPad.