The free VSTi PM Saxophone : Sax-o-phunk (beta available)

[steponeloops]

I like both of them (MilesTone and Sax-O-Phunk); the breath noise in Sax seems to be a little bit to constant in my opinion; maybe an adsr-curve for breath (with a controller-to-amount-link) would help? Just an idea...

[Ben H]

I'd like to see Guido continue work on sax-o-phunk.

But heck, it already kicks Arturia's brASS

Ben H

[O.L.T]

I'd like to see Guido continue work on sax-o-phunk.

But heck, it already kicks Arturia's brASS

Ben H

"Pun of the day"

[4damind]

It doesnt sound like a real sax. Have you tried Saxlab from Linplug? Its miles better and sounds more realistic then this synthedit-stuff. Is a nice try and of course good work but not in the same ballpark as Saxlab. I think that Saxlab works with samples and is not a physical modelled thing, but this gives very realistic sounding results.

[Ben H]

Yeah, but you cannnot get samples to play as well as sax-o-phunk responds. If it can get over the last little hurdle of tone. It will rock.

BTW I have played saxlab, and own john rekevics bigga gigga sax set, vsl saxophones 1, yellow tools candy, and play real saxophone professionally.

[Tony Ostinato]

people who think samples work dont actually "speak" sax.

its kinda like how people who didnt speak japanese liked the first vocaloid demos in japanese, once they heard it in english they said arggh thats awful.

the phrasing is so artificial it doesnt matter that the tone might be right, that just makes it worse!

physical modeling isnt there yet, just check out arturias, even yamahas vl series arent up to a real sax.

i think physical modeling is the only way to eventually do it, theyll never incorporate realistic behavior in samples.

like i say, in the meantime ill just play the real thing.

[Ben H]

Hey Guido,

Just wondering if you have done any more work on this instrument lately, especially the attack?

I would love to see another beta released!

Kind regards,

Ben H

[rkgibbons]

People who think samples work dont actually "speak" sax.

COuldn't disagree more.

Tony, I know you're anti-samples, but the progress the industry has made in sampling (and scripting) in just the last couple of years makes it pretty obvious that we're approaching something really awesome.

Historically, memory and disk-space were always the barrier preventing truly realistic sampling and articulations -- but in the last few years, memory has become extremely affordable. In five more years, we'll look at a 40GB sample libraries as something tiny and useless.

I know you claim that you can always tell when samples have been used, but as a professional musician, I've been fooled more than a few times -- and I'll bet that you have too (despite your claim otherwise).

How many times has physical modelling of a brass or wind instrument fooled you? In my case, I can't remember a single time.

Physical modelling is great...and theoretically it seems like it should work...but we'll will never able to "program" all the little subtle variations in acoustics, atmosphere, and instrument apparatus.

I believe the ideal solution will eventually incorporate both samples and physical modelling.

[sonarman]

I don't want to turn this into a 'gang up on Tony' thread but I agree with rkgibbons. I don't see what the commercial value of PM is anyway. If you need to be a talented sax player with a breath controller in order to play a virtual sax, why not just play the real thing?
There's something about physical modelling which kinda reminds me of Virtual Reality helmets.

[runagate]

The thing is, Tony's completely and utterly right.

The things that fool your ear sample-wise, are the straightforward little melodies.

Any jazz playing, playing rapid changes and the like is completely obvious. I imagine the same is true for violin samples and violinists. If you stay within the technique that the sampler has emulated well, it sounds great. If you wig out and play like you're soloing it's immediately obvious that it's a sampler. The sax is a damned quirky instrument and is further complicated by the fact that it's overtone series is the most complex of standard instruments and for that matter different for every note played.

It's really much more expressive than most people appreciate and I'm sure it will be emulated but it will take a while longer it seems. There's just such a huge amount of complexity in trying to emulate it.

[AndrewSimon]

It doesnt sound like a real sax. Have you tried Saxlab from Linplug? Its miles better and sounds more realistic then this synthedit-stuff. Is a nice try and of course good work but not in the same ballpark as Saxlab. I think that Saxlab works with samples and is not a physical modelled thing, but this gives very realistic sounding results.

sax-o-phunk is much more fun then Saxlab.

[Yossarian]

Tony, I know you're anti-samples, but the progress the industry has made in sampling (and scripting) in just the last couple of years makes it pretty obvious that we're approaching something really awesome.

Historically, memory and disk-space were always the barrier preventing truly realistic sampling and articulations -- but in the last few years, memory has become extremely affordable. In five more years, we'll look at a 40GB sample libraries as something tiny and useless.

Sampling technology has certainly come a long way but recently I've actually been more impressed by apps like Synful Orchestra and the woodwind/brass synth that arwa is working on (check out the mp3 demo at www.f.kth.se/~arwa02/vsti_demo.mp3 - there are definitely instances that would have me fooled). PM is not the only synthesized game in town.

Add to this the hybrid "sound morphing" technology in Garritan's Stradivari (based on samples but with a lot smaller footprint than a traditional sample library) and I'm not all that sure that we'll see these super-sized sample libraries in the future. The problem with these multi-articulation sampled instruments is how to access all those samples in a musically useful way. Take the Chris Hein Horns library as an example - using a combination of keyswitches and mod wheel you have access to up to 44 articulations within one patch. Learning how to make use of all those articulations on the fly seems about as difficult as learning how to play the real instrument.

I realize that's not how most people would approach this kind of instrument, they'd probably use their sequencer's MIDI editor to create their instrumental parts but that's just it - it's editing, not playing. I think this is where Tony gets off the bus - he's a performing musician and he wants to play the instrument in realtime. With Guido's PM plugin he has an instrument that reacts in an expressive way to his playing. The instrument's tone may not be exactly that of a saxophone but that's not really important - it plays like a real instrument in a way that a set of samples can't.

For myself, I'm not much of a player - I'm just a hobbyist and I care more about the end result than whether I got there by playing or editing. I'd have to say, though that if I were to create an instrumental part by editing I'd rather use an app like Melodyne and a phrase library than a regular multi-sampled instrument. With a good source of phrases you can get results which will sound much more convincing than a multi-sampled library (IMHO, of course).

/Yoss

[rkgibbons]

And that's where the trade-off with each of the two approaches becomes obvious. With Guido's approach, we have an instrument that plays more accurately but doesn't sound anything at all like the real instrument. With sampling, we have an instrument that sounds identical to a real instrument (because it is) but isn't as easy or realistic to play.

Hopefully, the two disparate approaches will merge someday into an awesome end-product.

[Ben H]

Hooray for "Sonic Morphing".

I think that by taking the tone of the real samples and "morphing" them (an oversimplified way of explaining harmonic phase-alignment), the way that Garritan's strad does, will see the death of these huge multi-articualtion libraries.

BTW have you heard the demo of version 2 of the Stradivarious violin... Awesome

http://www.garritan.com/Strad/ForClaire.mp3

[krraqk]

BTW have you heard the demo of version 2 of the Stradivarious violin... Awesome

http://www.garritan.com/Strad/ForClaire.mp3

Incredibly real and BEATIFUL!!