I've got to weigh in on the Brass first impressions. I'll begin by claiming some qualifications. I'm a classically trained trumpet player, jazz composer and arranger, pianist, and I have several years experience working with Yamha's old wind physical modelling card-including using it with a breath controller/keyboard combo. Now, my own opinions after several hours on the Brass demo:
True, the saxophone sounds like a kazoo being fed through a harmonica microphone into a practice amp. It's marginally acceptable in riffs behind the trumpet and trombone, but I would never use it on a recording to emulate a real saxophone. No amount of mastering control of a physically modelled instrument is going to substantially improve this fundamental problem. It has to do with the tone quality that Arturia chose to model.
I think the trumpet and trombone models are another story though. I'm quite impressed with the authenticity of both the sound and the responsiveness of the models. I've even spent some time A/Bing the jazz trumpet solo preset versus my own trumpet playing. The model is close enough that I can use it for a convincing duet with a real trumpet. I did find the responsiveness to my Yamaha breath controller (run through a Motif) to be substantially different than other instruments that I'm used to. Even if you are used to breath controllers, plan on spending a chunk of time learning how to use one to control Brass On the up side, the contorl I could achieve through a keyboard with aftertouch is really excellent.
I find the trombone sound and responsiveness even more impressive than the trumpt's in many ways. However--and this seems totlally inexplicable--the bottom of the trombone range only extends to a concert c below middle c. Excluding pedal tones, a conventional tenor trombone goes down to e natural below that. As the bottom of a jazz/pop horn section, those notes are often badly needed in trombone scoring. I view this as another major shortcoming.
I have experimented with Cubase's built-in pitch shifting and will attest that the bottom of the Brass trombone range can be very convincilgly shifted as much as an octave down (for Bass trombone). However, this would be a pretty annoying workaround to have to use regularly.
(I'll also note that the sax range more than covers the tenor sax, but does not reach to the bottom of the baritone range or top of the alto range. However, the tone quality is so bad that this point seems secondary. I recommend a sample based program called Candy for good, detailed, jazz sax emulations.)
Curiously, I'm not noting a huge, work-stopping cpu overload with Brass, though clearly, more than 3 instances (read 3 sepearte horn parts) would overload my machine--a 3.2 GHz Sweetwater Pentium with 2 gb RAM, running Cubase.
Even with the above mixed review, based on the demo, I am feeling quite tempted to purchase the product in order to have a convincing source of solo jazz trumpet and trombone. I would also consider using these for the lead parts on top of a section otherwise comprised of instruments from the new Garritan Jazz Library. However, anyone hoping to be able to use this for a general horn solution would be well advised to consider the above limitations.
I'm also a fond user of all of Arturia's vintage synths and want to see the company succeed. I realize that getting Brass out the door turned into a big hassle and hopefully will not prove a major misstep. I'd be satisfied if they would just lower the trombone range an aditional half octave to octave.
Hope others find this fair and helpful.