The first thing I did after unwrapping the Ti2 was download OS4 and load it up, then after remembering hearing some horror stories I backed off the beta and installed the last public release version (actually 3.3x.. dont recall version exactly). That experience in itself was troubling, because of course the back-level version (OS3) was not smart enough to clean up all pieces of the beta version (OS4). This sort of installer-snafu is very common in software shops -- beta these days means "try at your own risk" so some of the testing is short circuited in favor of getting it in the hands of the risk-taker user base. The difference these days is that there has been a trend of off-loading the quality assurance process to the customer under the premise of a "beta" release. I went to Access's site, and it seems this OS4 "beta" was the first release I was encouraged to download. Back in the good old days of higher quality software, beta software used to only go out to designated customers who were willing to test software in exchange for something in return; it would have been arrogant to release software into the mainstream without investing properly in alpha (inhouse) testing. Today there is a trend of "customer is our test bitch, why pay?".
For example, back then, if I tested Access' software and provide them useful feedback, I should then get a free Virus Ti2 in exchange for my time and experience with the product. If I examined at how many hours I spent dinking around with this synth to get the beta working right, then multiplied that number by the hourly rate I charge my clients, Access owes me a free Virus and about seven grand. So at that point I cannot justify giving them two grand for the priviledge of being their beta tester
All of that said, I want to make clear that I loved the sound of the Virus, and being a fan of subtractive synthesis above any other type (actual sound design and creation is important to me), I definately appreciated the knob interface on this desktop module. The polyphony was a bit weak, I loaded up some sort of D50 Digital Piano sound (I know, I know.... why the faerk would I want to emulate a 1980s digital synth with this thing??? Just out of curiosity)... But this sound was showing as a 5-bar patch! This means with only a slight amount of handi-work on the keyboard, I was able to rob it of polyphony fairly quickly. So for purposes of that sound, the Virus felt like a monophonic D50! This is absolutely laughable compared to what my PC (currently a Core-i7 965 running Windows 7 64-bit) can do with a mediocre soft-synth and achieve the same quality sound.
Now when you read the last few sentences there, please don't think that I am saying a Virus sounds no better than the Roland D50 did. I'm also not saying that softsynths sound better than the Virus. The Virus, by itself, sounds better than any softsynth out there. But the advantages it brought to my particular table did not justify the $2,200 US pricetag once I factored in the latency issue. And by adding high quality FX plugins to soft-synths, with certain sounds, I can match or exceed the virus sound (at the expense of CPU, but my PC's CPU resources 'runneth over' whereas Virus polyphony does not!... obviously they see the softsynth movement as a threat or they wouldn't bother with total integration to begin with, right? Let's face it, softsynths threaten hardware and the difference diminishes month over month). More importantly, in a complicated mix I am unlikely to be able to tell the difference between a virus and a good quality softsynth+fx.
I did realize that for pads and sounds with slow attack, I could probably find a way to work enough beautiful sounds into my mix that I could get some use from it even with latency problems, but the latency issue, lack of quality control in the software, and other issues simply made the pricetag not worth the returns for me.
One other bit of food for thought -- it seems that despite the "USB 2.0" marketing, the TI (and repackaged same TI2) actually run at the USB 1.1 max speed. So maybe the next generation of products will have a data bus that reduces the latency enough to be usable? It's just hard for me to believe it is purely a host issue -- latency should be a matter of the driver, the hardware capabilities (USB 2.0 in this case), and in my research the latency is not DAW specific (if it is, I'd suggest to the Access QA team that testing and certifcation with more DAWs translates to more sales). I've seen users complaining across many hosts, at the end of the day VST is VST, USB is USB, and if they aren't getting proper coverage on the platform (Windows) that holds more than 90% of marketshare, something might need looking into. Also some might consider my Windows 7 64-bit setup dicey, but that argument is no longer acceptable since Windows 7 has been the fastest selling OS in history, Microsoft is the second most profitable company in America this year kicking sand in Apple's face, and 64-bit adoption has been amazing over the last couple of years, even starting with the seemingly jinxed Vista.
I do hope they find a way to address the problems I had, and be careful to fund their own testing process rather than using the customer as a free tester. I will keep an eye on progress and perhaps be a future customer?