[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Date Index][Thread Index][Author Index]

Re: Max/MSP vs. Kyma

Latency is a function of the sound hardware and drivers.  With ASIO and a 
MSP can get very low latency.  However, for any host-cpu-based solution,
you pay some varying amount when you use the short buffers needed for
low latency.  Different programs fare differently in that regard.  Thanks 
dedicated hardware, Kyma can get very low latency w/o a large penalty
in processing power.

I've found that Supercollider is much more efficent for processing
than MSP, OTOH, Supercollider is not graphical  and that really
bugs some folks out. (though, I think SC is  less complex, and better
designed than Max/MSP).

Native Instrument's Reaktor recently announced support for
Motu Audio System (thus smooth integration w/ DP). It already
supports VST. There's an online demo.  (BTW: MSP progs
can also run as MAS or VST plug-ins, via Pluggo).

Right now, if you want low latency audio for a powerbook, you either
buy Kyma, or a Magma rack unit and put a PCI card in it.  However,
mLan is imminent.  There are already 2 articles in the past two issues
of Sound on Sound (www.sospubs.co.uk).  I expect Yamaha and others
to announce mLan products at AES this Sept.  It is a good bet that
Yamaha's ASIO drivers for Mac (already demo'd at MacWorld)
have relatively low latency (otherwise, they would have a major
marketing problem).

Consider also the quality of algorithms available.  I've
heard pitch shifting on all these boxes, and nothing compares
with the algorithms from Wave Mechanics (used by Eventide
and TC Electronic).  Resynthesis with Kyma can be very good,
but also very expensive (2-4 DSP chips if I remember correctly).
Even simple algorithms, such as chorusing, can be tricky
to get right.

Study your needs carefully, these are all complex systems,
they will all have strengths and limitations that are not
initially apparent.


Ben Porter wrote:

> Well, after the whole Kyma vs. Orville fight a few weeks ago, I thought 
>I would ask about something that I've been thinking about. Lately, I've 
>been heavily researching Max/MSP. But, I'm wondering how much better a 
>Kyma system would be.
> I'm mostly interested in two aspects of the programs. One is live 
>performance. The ability to send MIDI in (using some alternate MIDI 
>controllers) and get audio out. I'm thinking mostly sampling, but maybe 
>some synthesis and possibly some sound morphing (thanks for letting me 
>know about that one!) and definitely some effects,  looping, and mixing. 
>The other would be sound design. Actually being able to create new and 
>interesting sounds as well as modify and/or mangle already sampled sounds.
> At the moment, I'm heavily biased towards Max/MSP because of the need 
>for minimal external hardware, but you give up alot of latency for it, 
>but I think I might be able to get to where I can stand it. With the Kyma 
>system, you need that large looking black box with the DSP's and memory 
>as well as a computer. Then there is the cost, but that might be made up 
>by getting a lower cost computer vs. buying a top of the line Powerbook.
> Any comments? Neither system is exactly what I would want, but I think 
>that what I really want will be possible in about a year. But, that is a 
>long ways away. I'm trying to decide if it would be worth getting 
>something similar now or if I should just wait.
> Thanks to all.
> Ben Porter.
> What are you N2?  Choose from 150 free e-mail addresses.
> http://www.n2mail.com