History of the PARIS DAW


Timeline of the PARIS platform

looking for volunteers to add to this section - specifics on designers of subcomponents etc.


Stephen St. Croix, interviewed early in the history of PARIS in Audio Technology magazine, gives a little of the historical background of PARIS in his own words: (contributed by Kim Webster - excerpt posted for educational purposes under "Fair Use")

PARISIAN ARCHITECT: Greg Simmons talks to PARIS designer Stephen St. Croix...


Greg Simmons: As a person who has designed quite a few digital audio workstations, what are the most exciting things you have done with PARIS?

Stephen St. Croix: The most exciting thing with PARIS is this: No Lies. ... Seriously, I am amazed at the amount of deception in digital audio workstations. I know how it happens, because of marketing pressures, and I know why it happens: because of real limitations. The original design crew says "Oh, we've got to do this and this" and everybody says "yeah, we can do that", but then they run out of horsepower or hardware bandwidth, and realize they can't do it. Meanwhile, the marketing department has made a big noise about it. I've been consistently agitated over the years by lies, and I wanted to design something that was honest.

GS: So where did you start?

SSC: I looked at some of the leading systems and asked myself five questions.

1) What do they do?
2) What don't they do?
3) What do they say they can do, but actually don't do?
4) Are they stable?
5) What does it really cost a client to live with one of those systems for five years?

The picture was appalling. These systems don't do much, but they do more than anybody else. That tells us that the industry hasn't started yet; it is not "real" yet. There are many things they don't do, but nothing stupid - they're not dumb products. And now, more important for me because this is a pet peeve of mine: what do they say they can do, but actually can't do? That list is impressive. There is a great deal of misrepresentation and it;s not just limited to the leading systems. There are lots of lies out there and I hate that sh*t. Who is going to tell the user the truth? Somebody has to.

GS: How did PARIS originate?

SSC: Ensoniq approached us, or we approached Ensoniq, I don't remember. They make hardware and we make software.

GS: When you say "we", are you referring to your company, Intelligent Devices?

SSC: Yes. Intelligent Devices was formed from a freelance team of people who had designed lots of other DAW packages. There are two other people who do the core of the design. I do the conceptual work, sitting there for endless hours while my friends are surfing, getting tans and being fed grapes by blondes on beaches. I spend my time going "no, I'd rather sit in this dark room and look at a computer". I do the graphics. I do the interface concepts. I make the decisions on what it should look like and feel like. Edmund Pirali is head of the coding team, and he's scary. Within two or three weeks after meeting him, I decided I was never going to code again, and I never have. It's like, you're a real hot rock guitar player and you're pretty proud of yourself, and then you bang up against Les Paul or Eddie Van Halen and you kind of think "Gee, maybe I shouldn't do this any more".

GS: Speaking of coding, you're creating and upgrading Mac and PC versions of PARIS simultaneously. That's rather unusual in this market. What software are you using?

SSC: There's a product by Microsoft that is a cross compiler. You write your code, and it gives you a Windows version that is pretty good. Then it gives you a Mac version that bites. It's horrible. Our team did something that at first I was very nervous about, but now it is the part of the project that I am happiest about. Edmund developed what he calls an abstraction layer. It knows about Ensoniq's ESP2 DSP engine, the Power PC engine and the Pentium engine. We compile on Code Warrior and our abstraction layer simultaneously kicks out Windows and Mac versions. They act the same and they look the same, and they appear within seconds of each other.

GS: Sounds like a great advantage for the future.

SSC: Upgrades. Updates. Add ons. It's all amazing.

GS: What if you want to take advantage of something special, such as the MMX features on a Pentium?

SSC: We just add it to the abstraction layer. We are already multiple processor supporting. There's nothing to it. If your OS can talk to it, we can talk to it.

GS: You were saying that you make software and Ensoniq make hardware...

SSC: We make really good software. Ensoniq make really good hardware, for very low money because they have silicon design facilities and they autoload and wave solder surface-mount boards, all in house. They're big. Ensoniq delivers 160,000 sound cards to Gateway every month.

GS: ... and they've got their own DSP chips.

SSC: Oh yeah, the ESP2 chip. We were lucky enough to talk to them before that chip was done and made architectural requests which were honoured. We got to say "oh, that's cool, but it really needs to be able to do this, and we need this width and this versatility". We wanted more DSP flexibility than the other systems. For example, with ProTools once you grab a DSP it never lets go, so you can't use if for anything else. To avoid that problem, we wanted dynamically allocated DSPs. We got them.

GS: The ESP2 chip is also in Ensonq's new DP Pro processor.

SSC: Correct. There are six of them on the card. Which is pretty amazing, when you consider there are only two of them in the rack mount DPPro. A single ESP2 chip is roughly the same as a good 56000 series DSP chip from Motorola, not one of its monster high speed Symphony versions, just the usual 56000 series. But the ESP2 is better suited to its purpose because it;s a purpose-designed chip. How many functions can it perform? That depends on the type of functions. You may be able to get 16 really high quality multi-band EQs on one chip but it might take one and a half chips to do a good reverb.

GS: Why six chips? Did you tell Ensoniq that's how many you needed, or did they tell you that's how many you were allowed?

SSC: They wanted to go less, but we said we needed 6 because the real world demands it. Everybody accepts that if you buy a DAW you'll be making certian compromises. There are certain things you can do with a razor blade and tape, and with patch bays and insert points, that you can't do with a DAW. They say the compromise is worth it, because you get all the benefits of working in the digital domain, yada yada. But the compromise is not worth it. So we told Ensoniq that with 4 DSPs it's going to be pretty good, but with six we can break new ground. No compromises. So they gave us 6.

GS: Does every channel have a set amount of processors dedicated to it?

SSC: No. The total power provided by the six ESP2 chips exsts as a common pool of DSP and is dynamically allocated. The only things that are permanently allocated are the default EQs - every channel has four 20Hz to 20kHz EQs, each with five different modes. With sixteen channels per I/O card, that's a total of 64 real-time EQs.

GS: PARSI can play different sample rates simultaneously. This level of real-time sample rate conversion requires extensive DSP power, doesn't it?

SSC: When you've got enough DSPs, you can do anything. PARIS has enough DSP power to gear box [sample rate convert] tracks on the fly, so if you want to keep them at a dozen different sample rates, you can. Or, if you decide to go to one sample rate for everything, PARIS will convert them permanently for you.

GS: Apart from the six ESP2 chips on the card, PARIS also takes advantage of the host computer's CPU. Is it doing a Cubase VST trick?

SSC: Exactly. You know, VST is amazing but there's nothing VST can do about the hardware limitation of the host computer, and when you hit them, you hit them. THe nice thing about PARIS is the intelligence. I tuses its own DSP chips on the card and when they're used up it checks the computer's CPU. Of you've got something good, it automatically grabs the CPU and adds it to the horsepower. SO at any given time, when you turn a knob or make an adjustment, you don;t even know; you may be using the card or you may be using the computer.

GS: You said earlier that PARIS "doesn't lie". Does that mean all its specified features are things it can do with just the DSP chips?

SSC: Yeah. Everything we spec refers to the DSPs only, so you'll always have at least that level of functionality. But because PARIS also uses your computer's CPU, you could get more functionality than specified, and its power automatically grows when you upgrade your computer. In fact, you can even change platforms after you upgrade, from PC to Mac or vice versa, because the PARIS package contains everything you need to run on either platform, including the software.

GS: Speaking of software, PARIS looks almost identical on either platform.

SSC: I am a Mac guy. I don't like Windows at all. I was at a recent trade show in New York, happily demonstrating PARIS to somebody at a kiosk that wasn't on my stage. And I was showing it and he was saying "What's it like on Windows?" and I said "It's the same, but Windows sucks so I won't show it to you in Windows". Then another guy points to the left hand side of the window frame and says "Isn't that the Windows bar?"... I had just done a fifteen minute demo on the Windows version of PARIS and didn't know it. That's how much the same they are.

GS: What about the user interface?

SSC: PARIS is an imitator of a physical world. Many of the functions are accessible in multiple ways, by using pull-down menus or key commands or buttons. THe idea here is, why make the customer learn again? Why not make everything work in as many ways as possible so that the first thing the customer tries is likely to get the job done. So you can cross-fade by slamming two pieces of audio together, or by putting them together and hitting the cross-fade button, or by pulling down the cross-fade menu. That's what I mean: if you push the play button, ot plays. If you hit the space bar, it plays. Whatever approach the customer is used to using on other systems and equipment will probably work with PARIS. Why make the poor guy learn the way you want to think?

GS: And the control surface?

SSC: PARIS is the only DAW I know of that has a real studio monitor with steeo solo-in-place. It's got a mix bus, and a monitor bus with mute, mono, dim and its own level control for studio monitoring. You use it like a real studio. There are individual hardware outputs on the back of the I/O interface.

GS: That's something sorely missing on many computer-based DAWs. You buy the system and then you have to buy a little mixing console to go with it...

SSC: ... and you still won't have solo-in-place. Everyone has given up features; they've grown up in a real studio just to get into a DAW. I don't want to do that any more.

(...snip...)

There are no comments on this page. [Add comment]

Valid XHTML 1.0 Transitional :: Valid CSS :: Powered by WikkaWiki