(Continued from previous page)
WSR Reber: So the volume control is either implemented in analog or digital?

Sinclair: That's right. With digital volume control you can only have attenuation, there's no such thing as digital gain. Think of amplitude in digital as a ladder of 16 or 20 or 24 bits-what you're doing to lower the volume is dropping from one rung to the next rung, to the next rung of the ladder. You're specifying that this 20-bit signal is now a 19 1/2-bit signal, and so on. The higher the bit rate, the higher the resolution. As you attenuate, you lose resolution. By using an analog volume control, we lower the volume, not the resolution. But it's hard for either of us to say, "The key thing is..." because our approach is essentially holistic. We're practicing this right now on a new DAC we're designing. We're listening to four different board materials.

SR Reber: Actual PC composite board materials?

Cardas: Yes. Same parts, same design, same everything. The actual board samples are made of different layered materials.

Sinclair: The board manufacturer thinks we're nuts.

Cardas: We should have just said we were going to test the boards, implying some-thing about heat or rigidity ... we theorize that the differences among the boards are due to small amounts of dielectric absorption.

Sinclair: There is little else we can really measure, but they do sound different. In our volume control, as Mary mentioned, we use separate metal film resistors. Most people who do it in analog just use motorized pots. But using individual resistors gives us the clearest sound.

Cardas: I try hard in my sales trainings and in discussions, even with consumers, not to sell our products by insulting some-one else's. But with the depth of questions you're asking, it's hard not to do this by comparison.

WSR Reber: My intent is to give our readers an idea of how to think about and evaluate products based on their design and component quality. I mean, one product can cost $300 and one can cost $5,000, yet both perform the same function and share similar features. How can consumers know they're getting the best performance for their money?

Cardas: Are Widescreen Review readers people who actually listen? I mean, even among our own customers, some want to listen, try stuff out, and compare before they buy---while some are just looking for the right person to build their theatre.

Sinclair: That's because audio/video has gotten so complex.

WSR Reber: Our reader polls indicate that over 92 percent of our readers set up their own systems. And that over 84 percent use the same system for both music and movie pleasure.

Sinclair: Wow.

WSR Reber: So how can people best evaluate different systems?

Cardas: I recommend to people that they shop using a disc they know well. It can be their reference disc or just music they enjoy. They should use this disc to test the system using two-channel music. If you're putting on a multichannel movie, there's no way to not be distracted by the video or by the center channel, and it's easier for a multi-channel system to make you go "wow" and think it was great. It's much harder to do this with two-channel. When it gets right down to comparing high-end equipment, my advice is don't listen to details. Rather, listen to the music and decide which system involves you. Which makes you want to sit and listen to the whole disc?

When I got interested in home theatre, part of what drew me to Theta was the quality Theta puts into its D-to-A converters. Other manufacturers focused on various kinds of components-they did this, they did that-but Theta is one of the few companies that puts DAC design first, and brings that sensibility into its processor design now. This makes a huge difference in quality, and in making someone want to listen to a system.

Sinclair: Our original mandate was simply to make a digital signal sound like music because the early digital was so bad. It's hard to remember because it was so long ago, but audiophiles recoiled in horror. What Mary's saying is that you can tell pretty quickly if a system will reach your emotions with music. When you have a product or a system at home for a few months, you'll have a good idea of what it will do. Mary's talking about when you're in the showroom, where you are trying to make a valid, pretty quick decision-because you are going to own the product for years. And it's easier to get a good idea quickly about a system from music than from movies. And what serves you well for music will serve you well with movies.

WSR Reber: That's our philosophy as well. When we test equipment, we always listen to music first for our impressions. If it works great on music, it's going to reproduce any signal well. This gets back to using the idea of "I want to hear what the boards sound like-does it move you?" That ultimately ends up as hearing differences in the output stage, which is the loudspeaker system. Loudspeaker quality varies widely. Speakers can distort a signal, even if it has been pristinely maintained through the source and amplification stages. What do you listen to in evaluating your designs?

Sinclair: There are a small number of really good speaker brands, and we have a good selection of the best-speakers from Martin Logan, Wilson, Arial, Thiel, and a few others.

WSR Reber: So, you use a variety of loudspeakers?

Cardas: We need to. When we're testing something, each of us also brings it into our home systems, which have very high-quality equipment, and are very different in each place. It's important for a product to be able to adapt to the end user's choice in speakers.

Sinclair: We're fortunate in that a large number of speaker designers use our stuff at shows, in their labs, and in their own homes. We ran an ad that appeared in your magazine about six months ago in which we used pictures of leading speaker designers. I can't remember them all at the moment, but most of the key speaker guys use our stuff. So there is a symbiosis.

Cardas: They're an extended reference for us. They have pushed us to offer all these crossover choices-three different types, 12 different crossover points for each type, and four slope choices. That's 144 possible crossovers per speaker set! That means that whatever speakers you have, they can be setup to perform at their best in every room, every situation.

Sinclair: Either you give the speaker designer all the flexibility he can possibly want, which is really the purpose for all these crossovers, or you need to advocate one right speaker, and I don't want to tell anyone that I know what speaker is best for him or her.

WSR Reber: Going back to the processor stage-we have a digital signal from a DVD player coming into the processor board, and that signal is going to be two--channel PCM, data-reduced Dolby Digital, or a DTS Digital Surround signal. What hap-pens in the signal path after that to deter-mine the sound quality?

Cardas: Every step of the way matters. What our input devices are. What clocks we're using. How things are detected as they're going through.

Sinclair: You can add a lot of jitter. You can do so as you're processing whichever system.

WSR Reber: How does jitter sound, and what impact does it have on sound quality?

Sinclair: Jitter is amplitude errors in the digital. When you get a jitter-rich signal into the analog you, get harmonic errors.

WSR Reber: So, these are added sounds that aren't on the source?

Sinclair: They're additive and take away from the richness of the music.

Cardas: It's the high-frequency interference that bugs me about digital.

Sinclair: It's not harmonically related to the music. There are wonderful-sounding tube amplifiers that have two, three, four percent distortion. Not point four, but four percent. These kinds of distortions, if you hear them at all, if they're setup right, if they're the right harmonics, can actually make music sound better. Not accurate, but better. But some kinds of digital distortions are so unrelated to the music that they make it obvious you're not listening to something created in nature. We take great pains to get rid of these. When Mary talks about clocks...

WSR Reber: Clock accuracy.

Sinclair: Yes. Using precise clocks, re--clocking, storing the digital signal, then clocking it out from a buffer where, for a very short period of time, the digital signal is in stasis in memory. While in memory it's perfect, time-wise; there is no jitter. As we clock it out from the buffer, we cannot totally avoid jitter creation, but we try to use as accurate a clock as possible and lessen the jitter as much as we can-but from starting anew in the buffer. We've gotten rid of all the jitter that accumulated before this point. So while we can't have a jitter-free signal, we work very hard to renew in a jitter-free state, then create as little jitter as possible.

Cardas: And we address this in both our transports and processors.

WSR Reber: So, as these clocks vary in quality, that's a major factor in expense?

Sinclair: Yes.

WSR Reber: Once it goes through the clock stages, and you re-clock...

Sinclair: It's clocked in several places.

WSR Reber: What's the next stage after the initial clocking?

Cardas: It gets identified as to type of signal, so it can be routed properly. For example, if it's identified as DTS, it gets routed to the DTS processor board, where we do some re-clocking and processing with our aesthetic sense, in as high-quality a way as we can.

WSR Reber: Do you do it in software or in a chip format?

Sinclair: We use Motorola processors with the DTS and Dolby decoding software fixed in them.

WSR Reber: Is that preferable to loading from outboard RAM?

Sinclair: It's more certain. I don't believe either company any longer approves new designs done any other way.

WSR Reber: What's the next step?

Cardas: Once the processing is complete, assuming it's a multichannel signal, channels are separated out. From here, still in the digital domain, we move into any crossovers being applied. This is where we have a huge number of choices, depending on which product we're talking about.

WSR Reber: Or you could bypass the whole path?

Cardas: Absolutely. You can do it full--range. If we're talking about an analog signal coming in, there's an analog direct mode, so we aren't doing anything to the signal, or it can be converted to digital, so that we can do all the crossover stuff and convert it back to analog. But, at any rate, after the crossover is done, we move into our D-to-A converter. All the quality things we've talked about-that Theta has learned over the years about D-to-A conversion- are done for each channel in exactly the range chosen for each speaker.

Sinclair: Then it goes to the volume controls.

Page Number: 1 | 2 | 3 | 4 | 5