INTERVIEW: BRIAN ZOLNER

Bricasti Design

Brian Zolner of Bricasti Design, talks to Jez Ford about his new M5, streaming, and why USB doesn’t work as well as it should!

Brian Zolner is the ‘Bri’ in Bricasti Design, a company founded by two former employees of Lexicon, joined subsequently by engineers from Madrigal Laboratories. Here, he talks to Jez Ford about the Bricasti M5 reviewed on page 50 of this issue…

Jez Ford: So tell us about your Bricasti M5 Network Player...

Brian Zolner: So we have the M5 which we announced a year ago, but we also launched before that the M12. And the idea of the M12 was what I call a source controller; it has analogue in, digital ins, so it’s a nice front-end for somebody, a preamp/DAC. And we developed the network rendering streaming interface for that product, and designed that card—the physical shape of it, how it mounts—so it could be retrofitted into any M1. So it fits in the M12, it can be put into an M1 or M1 DAC, so people can upgrade if they want, or they can order it with that.

Then I realised I could put that network rendering streaming interface in its own little box. So the M5 is this little box with the analogue power supply out of the M1 powering the network interface, and that drops down to a card we made that has a SHARC (processor) on it, and with AES and SPDIF out. Because the real idea of the product was not to use USB, the whole advantage is saying ‘let’s do this without USB’, OK? So this M5 can sit close to the system, or it could be further away because it’s going to go SPDIF or AES; it can go halfway around the room, who cares? And the computer is hidden back here or in the other room, another part of your house; who cares where, the network takes it.

JF: And that you see as important for convenience or for electrical separation?

BZ: Well the whole idea of this, from my perspective, from when I start to make these things, is that I make them for myself. And I said, you know, I don’t want that damned computer near me, and I don’t want to spend a pile of money on some server thing that’s just a computer in a box that sits there and you’re back to USB. Because you want to get that out, and if it’s USB you can’t, it’s got to be close. So the M5 allows you to take from wherever your source is—it happens to be this laptop for the show, or in my home it sits up on the second floor of my house—and I play it remotely from here. I use JRiver because I like JRemote [the JRiver remote app for tablet or phone] but you can use other apps—you can have an app just on your phone and then the phone becomes the library manager... Bubble or mConnect, you can do all these things. So the idea is to get this out of your listening environment, clean it all up, get that noisy drive, all that crap out of the way. And then the advantage comes also in that the M5 is a product that anybody can use, with the same features— DLNA, Roon Ready, UPnP device—and it comes out AES or SPDIF, or USB.

JF: So explain why you think USB has problems; a lot of us play USB from our computer. Can you explain the problems with that?

BZ: Well... this all gets very, sort-of, subjectively audiophile craziness. JF: Well, that’s fine, because I do that. BZ: [Laughs.] OK, you know, one would say it shouldn’t matter, right? And I think technically from the pure data point of view it probably doesn’t matter, depending on how things are done. But I believe the issue with USB is that first it’s not really made to do this. USB was really meant to be a thumb drive or just a point-to-point thing. It’s not meant for real-time transfer. What’s going on in the network now is not real-time. It’s pulling a drive, buffering it and playing it. So it’s not real-time streaming of audio. Now you take an interface and you’re trying to do that. Of course it can be done… and was done.

But also I think the issue resides in the fact that the interface needs to be powered. And that power has to come from the computer, because that’s how it’s supposed to work. So you have, you know, even in our M1 and all our products, we galvanically isolate that from the product. But nonetheless it’s all in there, right? So you have to say, OK what has it done on hearing.

And my experimentation and research tells me it has to do with power. Because I used to take my computer, plug it in, and if it ran it on the battery I could hear a difference. It’s better. Got rid of the hash. Not all of it, but a lot of it. OK. Then you have another level of hash which is the motherboard in any computer, I don’t care—any board you buy, unless you do it like we did in the M5, where we made a small card with an ARM processor running a Linux core. That’s what an ARM processor is for, to say ‘reduced instruction set’—it’s only going to do what you stick on it. In the computer world there’s things doing all kinds of shit, because it’s got to support it—not that it doesn’t work, it works fine. But you have all these different voltages that are being regulated because the power comes into a laptop or something at 19 volts, you’ve got to get five volts out for this, you’ve got to get all the one volts, three volts for the digital things inside the computer all the time.

So all of this has to be regulated, and it’s all done with digital regulation because you’re not going to do it in analogue—I can tell you that, because I know what I face with our products as we run everything analogue! If you run the digital part with an analogue supply, you know, suddenly “it doesn’t work, it crapped out on me, what happened?” Well the voltage dropped, sorry, that’s why you use a switch-mode

“... the best-looking, best-built, best-sounding speakers I have had in my listening room.” John Atkinson Editor/Stereophile

supply on digital stuff because you can’t afford to lose it. In analogue, analogue power goes down low, OK maybe distortion goes up a little bit but it still works—you don’t get as much power out of the amplifier or something. But with digital it’s dead. So getting back to the computer world, that’s why all the regulation on the board is digital. There is no way you’re going to create all these voltages linear. And so you didn’t really win anything. So that was the reason I found. But primarily it was the convenience of getting that thing out of my room and saying, well, OK, then you gotta have something like this.

JF: Back the 1990s you wouldn’t be seen dead with a computer in your hi-fi listening room. Now we live with them next to our hi-fi, inside our hi-fi.

BZ: Well everybody’s been making records for the last 20 years with a computer. Some people maybe make token tape recordings. But that’s just token. JF: So are we cursed at source with this? BZ: Well no, in professional use, you know, you aren’t using USB most of the time. But there are issues like that throughout the chain. But nonetheless, all we can deal with in this situation is what we’re playing back. But that’s what I noticed, that it was done initially because of convenience, but then once we got it going I said, wow that’s better. It’s a lot better. Because that data can be anywhere on the network, it can be on my phone, can be on a NAS, could be from the cloud, wherever the server’s going to pick it up from. So I certainly use this as my ‘server’, I use that word because it’s where I let the app reside—here. I find I prefer to just do that. People think they want to get around that and they say, “Oh, I just want to play from the NAS”, and you can kind-of do that, but then there’s a processor in the NAS. There’s a computer in there. People think that “Oh I don’t need that computer”, but they have one in the NAS and then they’re kind of hamstrung because you can’t do anything.

JF: And there’s the variable server software, where the DLNA won’t order albums in the right way and so on…

BZ: All that stuff—let’s call it library management, you’re going to have to flip your computer open and straighten this out. Right. So my life and the way I do it, I gave up. I just love this little laptop I’ve been using for years. It’s just running Windows 7 and all I ever use it for is audio. When I get home next week, I’ll go upstairs, plug it into the network, turn it on, change the library, take that M21 that I’m carrying home with me, turn it on, and ten minutes later I’m playing. And it just sits there, I leave it on, it’s got a solid-state drive, it’s just on. So I never turn my system off and on—it’s stable, it’s all IP addressed, all that stuff gets sorted out. I come here today, I turn it on, bang, spot on, it’s stable.

So basically we’ve made our own Linux player. We could, with a bit more work, turn it into a server. But why do that? Because then I’m in the app business. I don’t want to be in the app business. We could stick it on there, license an app from somebody or put jRiver on there, license it from those guys. But then you’re supporting that. And it doesn’t have to live there, that’s the point. There’s no real advantage for it to live there because you’re still back to the same problem—I’ve got to keep my database sorted out, how am I going to look at it, how am going to rip a CD, all this stuff, all this maintenance. However you do it, it’s not magic—it’s like organising your albums in the old days, you have to be able to find it, get it, and play it. People make it more complicated than it is.

If you use Roon, that’s what Roon tells you to do, Roon says you use a core. That’s how you do it. Well that’s what I’m doing, this is my core. It’s the same thing, I’m using this as my core.

JF: Nobody’s explained to me what the criterion for ‘Roon Ready’ actually is.

BZ: All it is saying… OK, all it is… is it’s their version of DLNA. So they took that type of a protocol and made little tweaks. Then they said “OK now you have to make it work our way, and comply with us, and we’re going to test that it does…” yada, yada… the whole thing.

JF: And they go through a testing procedure?

BZ: Yes, and then we have to give them a unit and all that crap and, arrrgh, it knocks on to all kinds of things, because they want to do all kinds of shit that you don’t really want to do.

JF: So that’s a tough decision, which you choose to license...

BZ: Very expensive and, for us, costly time-wise in development. It’s not Roon, it’s us implementing what they want.

JF: Do they change criteria or is it a pretty fixed protocol?

BZ: It’s fixed, but the point is we had to do a lot of work, we had to re-invent stuff to do it. Is it worth it? I don’t know. I tell guys, use what you want. I find that the most important criteria for users is that once they get used to an app, they don’t want change. That’s far more important here than some twiddly little—maybe does it sound a little different? It shouldn’t really matter. You’re just sending the files over. And the rendering is done here. We do the rendering. So the app isn’t doing it. In our testing, I got my guy in L.A., he’s got a Mac and a PC, I let him have all the apps—I don’t want to waste my time with that, but somebody has to do it. But he’s got Roon, he’s played with Audirvana; he says it doesn’t matter.

JF: So preferred connection between the M5 and the DAC? Not USB then?

BZ: That’s another point of discussion, because the idea was not to use USB. And I tested it in a way that no user can—well a user could, but no user typically can—because everything with that is very speculative and very subjective. You have no grounding point in anything, so to change something—sure it changed, but is it better or worse? Theoretically you’re sort-of after the truth; that’s the whole idea, right? Well it’s supposed to be, but it’s not. The problem is, how do you ground yourself?

So in this case I take the same file, same cables, simultaneously playing to the M5 and say in this case to the M21, same streamer, same everything right, in both products. I go into the M12 or the M1 or the M21, the same streamer card’s there and it drops I2S right to the DSP. And then out to get converted. In the M5, I2S right to the DSP—AES, S/PDIF, USB. So you can play the same track at the same time and switch. And of course the level is perfect because it’s digital, you don’t have that issue, and then you start listening, and the conclusion I come to is that the truth is very close between the AES and the SPDIF. The USB is not, and you hear why—you hear the noise and the hash. And the USB sounds more exciting, more lively, because of the noise. Noise does that; distortion does that. Various types of distortion will create thickness, loudness, apparent loudness, apparent width of the stage, you know, because it’s kind of ‘shaking guys up’. If it’s still, what you’ll have is a more clear deeper soundstage, looking at it that way. And if it’s bit noisy it’s going to go wider and excite things more—maybe you think it’s detail, but it’s not resolution.

The best resolution is hearing nothing, it’s transparency. But most people don’t understand that, they think something’s missing. But my listening says that it’s the closest to the original, OK, and that’s the only thing you can hang it on.