Berkeley Audio Design Announces The Alpha USB Asynchronous Interface

<cite>But actually, I think there is a lot more similarity between HDMI and data communciations networks than not.</cite><br />

<br />

Could you elaborate how?<br />

<br />

HDMI is NOT asynchronous, audio clocks are derived from video clock, typically requiring combined divide and multiply operation for 44.1/88.2/176.4 rates. Audio bandwidth availability depending on video resolution, no audio without video, etc, etc..<br />

<br />

<cite>Perhaps a bit more like ATM over ethernet. In any case, a lot of potential there.</cite><br />

<br />

Most jitter measurements for standard HDMI audio equipment have been roughly order of magnitude higher than for average S/PDIF, so at least a lot to do there to improve. There are couple of vendor-specific extensions to make HDMI audio asynchronous between devices from the same vendor (by Sony and Pioneer, any others?).<br />

<br />

And it's already irrelevant interface because it's spec is changing too frequently. Ethernet, USB or Firewire are way better for audio. And Ethernet at least is by default transformer-isolated...

I think HDMI will "evolve" to be nothing more than ethernet, with twisted pair or fibre becoming the standard interconnect between devices. And I expect it will be ATM or some derivative that takes center stage as the standard. (Goodbye jitter - ATM limits jitter as part of it's definition.) <br />

<br />

<br />

In other words, I agree with you. <br />

<br />

-Paul<br />

HDMI is OK'ish design for transferring video data to low resolution monitor. [End Quote]<br />

<br />

I guess I may be a little more understanding of the appeal of HDMI since the general consumer has been taught that HDMI offers the highest quality video and audio. And while home theater and video gaming systems are not generally audiophile level music systems, they have introduced the masses to high resolution video, multi-channel audio and subwoofers.<br />

<br />

So the Best Buy marketplace has succeeded in leading many consumers into the very impressive land of home theater. And now we sophisticated audiophiles try to tell these consumers that this stuff that just impressed them greatly is crap. <br />

<br />

Though it generally is crap by our standards, it is really suitable for the average consumer. I happen to be fond of an old Blue Jeans Cable article comparing DVI, HDMI and Component Video for high rez video:<br />

<br />

http://forum.ecoustics.com/bbs/messages/34579/122868.html<br />

<br />

And if anyone is interested in quality HDMI cables you should check out the Monoprice cables. The typical HDMI cables are crap.<br />

<br />

So even though we are audiophiles we need to know how to behave with non-audiophiles:<br />

<br />

http://www.tnt-audio.com/topics/audiophile_rules.html<br />

<br />

<br />

<cite>Though it generally is crap by our standards, it is really suitable for the average consumer.</cite><br />

<br />

I'm talking from purely engineering point of view...

I am not sure I would call all HDMI "crap," but I agree with the gist of this. And I think the Bue Jeans article hits it on the head: "it depends." I just got done setting up my Oppo BDP-95 recently in my HT, and it has a wide variety of connection options, and it is a unit in which the maker has invested a fair bit of attention to audio quality. HDMI connections, using color-coded short Monoprice cables [incredibly reasonably priced], have produced excellent video [at least as good as the component video outputs, maybe better], and even reasonably satisfactory audio. Now, I have yet to upgrade my old Denon AV receiver, and I won't be surprised if that makes a difference for me—I do all my critical listening over my headphone rig—but the HDMI signal quality is quite reasonable. And suitable for the application in which it is deployed.

HDMI in its standard form has very high levels of jitter in the audio signal as transmitted. Most SSP's just send the audio through an ASRC chip and - Presto! - the jitter's all gone. At least on paper. But not to a careful listener...<br />

<br />

With HDMI 1.3a they added an optional feature called Audio Rate Control that acts EXACTLY like asynchronous USB to essentially eliminate transmitted jitter entirely. But so far only three companies have implemented it:<br />

<br />

a) Sony -- in a proprietary form that only works with their equipment.<br />

<br />

b) Pioneer -- in a proprietary form that only works with their equipment.<br />

<br />

c) Ayre -- in an open format compliant with the 1.3a spec. But we only make a player. To date no SSP maker has stepped up to the plate to make a matching receiver...<br />

<br />

The other HORRIBLE thing about HDMI is that the signal doesn't belong to you. Even though you paid for the disc, it is encrypted via HDCP and only can be played back on "authorized" equipment.<br />

<br />

I just got a new computer and tried to play a DVD. It came with WinDVD 8 or something like that. But it wouldn't play the disc because I have a DVI monitor and not an HDMI monitor. It only gave me an error message. Thanks, guys...er, I mean @ssholes.

I had one USB cable maker tell me that their new 1, 2, and 3 meter USB cables sounded great while the 5 meter one sounded horrible (relatively speaking). So they went out and bought all four lengths in a half a dozen different brands. Every brand did the same thing.<br />

<br />

Moral of the story -- don't go past 3 meters for USB unless you use a repeater. (There is a separate thread on this site about using a USB repeater to extend the reach of USB for audio.)

I have never thought about making a two-box async USB DAC. That is an interesting idea. I have no idea how much difference it would make sonically, but it's pretty easy to guess that it will double the cost of the design -- two chassis, two power supplies, extra connectors and interconnect cables. Normally this type of approach is only suitable for cost-no-object designs. So if we ever make a $20,000 USB DAC maybe we will try this idea.<br />

<br />

Regarding galvanic isolation, I have never done any direct tests with a computer. But I have had enough experiences with the switching power supplies in video displays to know that it is extremely important to keep them disconnected from your audio system if you want the best performance. Making high-performance DVD players is how we learned about implementing galvanic isolation, so it was trivial for us to do the same with computer audio.<br />

<br />

The switching power supplies used in computers are quite similar to those used in video displays. So it never even entered my mind to try building a computer-based DAC without galvanic isolation.

>> Yet again you are disparaging a competitors products. <<<br />

<br />

No.<br />

<br />

You are the one who mentioned Weiss, not me. I simply linked to a third-party objective website that claims that the only the Metric Halo is asynchronous and that some makers instead uses the Jet PLL DICE chip.<br />

<br />

Now I have no idea if this is true or if Daniel Weiss's seemingly evasive replies on this forum were true (could have just been a language barrier thing) -- I have never seen one of these devices, let alone opened one up and analyzed its circuitry.<br />

<br />

Furthermore, the underlying technology doesn't matter one little bit. All that matters is how it sounds. Chris Connaker gave the Weiss an all-out rave. So they must be doing something right. In the end, that's all that counts.<br />

<br />

We try to be open and honest about all of the technology in our products. Naturally there are secrets that we have developed. In those cases we simply don't talk about those features. If someone happens to ask we simply say, "I'd rather not say" or "That's one of our trade secrets." Different manufacturers have different approaches. And as I noted before, the language barrier presents an additional complication. If I had to give a technical interview about one of my products in German, I would probably end up saying something like "It's powered by carrots!" and think I had done a brilliant job!

<i>"And as I noted before, the language barrier presents an additional complication.<br />

</i><br />

<br />

Or an additional advantage. You can see it in every courthouse in this country.

Charles Hansen writes:<br />

<br />

<i>If I had to give a technical interview about one of my products in German, I would probably end up saying something like "It's powered by carrots!" and think I had done a brilliant job! </i><br />

<br />

Es ist durch Karotten angetrieben!<br />

<br />

Charlie, I think we need T-shirts.... ;-)

I understand that if you talk about what you consider a trade secret, that reduces your protection under the law. Unless you patent your special sauce, which means its published, you really aren't protected. There's nothing preventing someone from opening up one of your units and reverse engineering your "magic".

<i>Es ist durch Karotten angetrieben!</i><br />

<br />

<i>Charlie, I think we need T-shirts.... ;-)</i><br />

<br />

I'm in! Three size mediums. (I bet if we get the total up to twenty we can get a good price on them.)

Patents are pretty useless for a small company because it can cost millions to defend them. Sometimes a mid-size ($100 million to $1 billion) company can develop a large patent portfolio that is attractive to potential buyers.<br />

<br />

I'm not so much up on trade secrets. I think the main thing is to keep your employees from talking about them.<br />

<br />

<i>There's nothing preventing someone from opening up one of your units and reverse engineering your "magic".</i><br />

<br />

Absolutely. Sometimes a competitor will try to borrow it from a dealer. But they can always just buy it. But some "magic" is easier to cipher out than other "magic". In general, any programmable device creates some fairly big barriers, although they can usually be broken. Then there are other things that are so subtle that they are very easy to overlook. And even if they are noticed, they are not always easily understood...

You can put me down for 2 XXX Large please. <br />

<br />

-Paul

Actually Charles, you're exactly wrong about patents. The dream of many small companies is to be infringed by a large one. The only issue with patents is that it requires a discipline, writing things down in a bound, numbered notebook, having pages notarized if you think you've come up with something really special. Its date of discovery that counts, not the filing date. Filing isn't that expensive. If you've been infringed, its easy to find a patent troll lawyer to take the case on contingency. <br />

<br />

In some ways, programable devices are better protected. The code is copywritble, and once some one sees the code, the issue of derivitive rights and infringement becomes sticky.<br />

Totally heading off topic...<br />

<br />

But I think it's very unfair to class all patent lawyers as trolls. Patent trolls are only where a company buys dubious patents, does no development of own and just goes after legitimate products just because they wrote a paper "patenting" using wireless communication to collect email (or some such nonsense). Putting two existing technologies together in an abstract is not an invention (IMO). <br />

<br />

Eloise

<i>"If you've been infringed, its easy to find a patent troll lawyer to take the case on contingency."</i><br />

<br />

That's not exactly the quality of lawyer I'd want on my side. As you probably know Dave the quality of a lawyer can determine the outcome of a case. I wouldn't want to waste a ton of time and associated expenses pursuing a case with a patent attorney deemed a troll. I'm willing to be the best ones don't have to troll.<br />

<br />

Charles Hansen's point is a very valid one. I'd judge it very unlikely an attorney would take patent *defense* work on contingency, in case your patent registration leads to someone else suing you to have your patent declared invalid. And frankly, quality plaintiff's work in high technology intellectual property areas like patent or copyright, the only kind really worth having, is with few exceptions expensive and carried on by big firms, not lone "trolls" working on contingency.<br />

<br />

If you're the sort of company wanting to hire trolls working on contingency, you're a patent mill, not an audio company that needs to spend the vast majority of its people's time actually making things to stay in business.

Not really in the U.S. Eloise - they buy up patents that are not dodgy at all, and sit on them until someone uses the idea, or comes up with it independently. Then they descend like a avenging horde of crows. <br />

<br />

The big one that is going to hit us are the patents on audio/video formats. <br />

<br />

-Paul<br />

Chris, the reason that what I've been calling patent troll lawyers, really IP lawyers who work on contingency, are hated is because of their effectiveness. You've obviously never been involved in patent litigation. If you have a business plaintiff, they are always making a cost benefit analysis of multi hundred dollar an hour patent lawyer fees relative to the cost of the infringement. The incremental cost of yet another motion to a lawyer on contingency is pretty small. The fact is that many IP contigency lawyers can often run circles around anyone Wilson Sconcini can throw at a case.

There are (to my mind) some legitimate patents that are owned by "non-practicing entities" which are quite rightly defended. This is the (one of the) reason patents were created: so an individual could create an innovative solution to the problem and, even without the means of putting that into production, could gain some benefit from that idea by licensing that idea to others who could develop it into a real product.<br />

<br />

On the other hand there are other patents which should not rightly have been granted as they are not inventions but simply marrying of two exisiting technologies in a totally logical manner. These are kept until large companies "infringe" them and then the patent troll litigates: these have no benefit to anyone but the lawyers and cash benefit to the patent owners.<br />

<br />

A good example of the true patent troll is NTP Inc. who own a very sketchy patent for the "invention" of wireless email. So far they have successfully sued RIM (Blackberry), used the threat of litigation to force Nokia to pay licence fees for that patent and (at last count) were suiting countless others including Apple, HTC, Google, Microsoft, LG, Motorola and other. Many people contend this is a bad patent because it was only logical that people would marry Wireless communication (already invented) with email (already invented). My understanding is there is nothing in the NTP patents which describe any technical methods for transferring email via wireless. There is even doubtful if their patent is valid as there are claims of prior art predating the NTP patent; yet the lawyers won out and had RIMs demonstration of an earlier invented system thrown out on a technicality.<br />

<br />

Eloise<br />

<br />

PS. Chris, can I suggest that these posts about patents could be better transferred to their own thread?

Sorry, I couldn't quite read through all those comments (did about half) so this may have been answered.<br />

<br />

Is the 'best' system via Alpha DAC still to use the soundcard that used to be recommended, then AES from there to Alpha DAC, or does this new device do something that makes the sound better than that combo?<br />

<br />

Also, I'd be curious how the best of those two compares to the Perfect Wave DAC w/ethernet card as the buffering from that ethernet card is supposed to do something magical. Does the a-synch usb do anything magical as the PS Audio device does, or does it just provide low-jitter information to the DAC?

First, no one yet knows how the async USB solution compares to the previously recommended Lynx AES16 card for connection to the Alpha DAC.<br />

<br />

The async USB is just a method of supplying a low jitter, low noise SPDIF / AES signal for the DAC.<br />

<br />

Eloise