It depends what modules you have set up.

You could build a standard subractive synth and have multiple pages of knob allocations. Making it possible to program it as you would (in theory this is - haven't played with one myself) a normal subractive VA.

Would love to have a go with one in the studio.

I've just had a go on the CPU demo of the G2 and it gives you a good idea of what it can do - I found it easier than expected building patches, it's quite an intuitive editor and I like the way you can select a drop down menu on individual modules and change them for alternative modules.

I hope the actual hardware doesn't sound like the CPU demo though - it can't be the same sound surely?

no,its several times better.
dsp is supirior over cpu for now when it comes to synthesis and such.

Cheers Tomer, that's what I thought, but I haven't been able to play a G2 at the stores I've been to so wouldn't know. Hmmm, it sounds like my type of thing to be honest. I'm going to hang fire and see how the Ti saga unfolds before I make my mind up.

yeah its cool ive been wanting a NM1 for quit sometime now.
got to get some "must-have" basic stuff first before i can treat myself with this kind of toy.

What?

First of all the software G2 depends on the quality of your soundcard's convertors. If that's a simple on-board thing with tons of latency and noise of course the hardware version will sound better.

Second, there's no difference in sound between a DSP and a CPU. A DSP -is- a CPU. It's specialized though, opposed to the generic purpose models which are used in computers.

It does a particular kind of calculations pretty well. Since it does not have to bother with nice pictures or e-mail or whatever crap the efficiency is much higher - and this allows a DSP to run on a lower clock speed than a generic-purpose CPU.

The bits are the same. You can write the software for a DSP in a higher-level language such as C/C++ instead of pure assembler. Eventually the zeroes and ones that come out of it are exactly the same; it's just that the computer version doesn't have the dedicated DSPs running so it has to emulate one in a way. Compare the G2 software with the G2x (preferably via a digital out - but the G2 doesn't have that :( ) and there should be no difference.

Also, for physical modelling a G2 user named Chet Singer has written several tutorials.

http://chet.getchwood.com/G2-Tutorial/Index.htm

sorry i was trying to speak out my mind in a simple non sientific way.
eventualy,dsp can do what cpus do in the field of synthesis in much less cycles this is why synths dont require high amount of proccessing power dsps in order to work\generate a rather high amount of voices and effects.