The TAS1020B isn't capable of USB 2.0 so if its _exactly_ the same board as his converters I don't think it will do 192khz anytime soon... although I guess he could just use a different USB receiver chip.

The chinese thingy is full of little compromises like a poorly generated clock and a reliance on USB 5V, so I have little doubt that Gordon could do much better than that. Hopefully for a lot less than $900 though.

You're using a North Star DAC right? Look for something that can do I2S... I recall reading that the ESI Juli@ could be very easily modded to output I2S...

--RME HDSPe AIO user

I don't know anything about modding though ._. Asked Steve N. about an I2S configuration for the North Star, and it comes out to $699 + $100 (5V I2S option) + $389 (Digital Cable) ... the last part was a deal breaker, so I'm giving up on the North Star in I2S config. Supposedly, the North Star's I2S input isn't "standard", and the digital cable needs to be wired differently :-/ Guess I shoulda done my research before buying the North Star.

There's no such thing as a standard 12S input to begin with, afaik.

Yeah, that is unfortunate.

I meant that it is possible to connect some I2S components together using standard RJ45 terminated cables. For example, I believe the North Star DAC and Transport can be used with a standard RJ45 cable; same with the MSB DAC and transports.

edit: It appears from show reports that the final version of the Wavelink will come with Nirvana Audio's new T0 BNC-to-BNC digital cable and BNC-to-RCA adapter.

The TAS1020B isn't capable of USB 2.0 so if its _exactly_ the same board as his converters I don't think it will do 192khz anytime soon... although I guess he could just use a different USB receiver chip.

It will need a new receiver. I think they're in the process of porting the code over now. Last I saw, Charles Hansen from Ayre was suggesting they might be ready to upgrade the QB-9s in December, at a cost of $250 or so. Although -- as he pointed out -- there really isn't a whole heap of 24/192 material out there unless you're creating it yourself.

From a relative value perspective, the new Weiss FW interface is $1,300 but has USB -> AES as well and does 24/192 put of the box.

Wow, quite the discussion I appear to have sparked.

I appreciate not all use cases are the same, but given the availability of things like, say Virtual Village - US | USB Mini Converter to SPDIF Output for $32, is that "good enough" to solve the problem at hand.

If jitter on the input S/PDIF is an issue, get something with a word clock input, and clock your audio bits, using the clock on your DAC. problem solved.

If your DAC uses ASRC, or reclocks the incoming bits, why are we having this conversation?

And why would you want your noisy, RMI/EMI emitting computer, within the maximum length of a usb cable, electrically coupled to your dac?

Just some thoughts to play with

Just some thoughts to play with

Good gravy, chap, how many cans of worms do you want to open at once?

Good gravy, chap, how many cans of worms do you want to open at once?

I must say, for one post that's quite an impressive feat Grahame just posted.

I was reminded of the quote

"SPDIF is so bad it has given birth to its own Snake Oil ecosystem. Here is a monument to it :

from The Extremist DAC

I'm not aware of any cheap (say under $500) USB/Firewire to SPDIF converters that have clock input. Wouldn't such a unit also need to be async?

You'd also need a DAC with clock output (not exactly common either), or a separate clock unit and a DAC with clock input as well (not common or cheap)...

And why would you want your noisy, RMI/EMI emitting computer, within the maximum length of a usb cable, electrically coupled to your dac?

That is what pulse transformers on the SPDIF lines are for - galvanic isolation.

You'd also need a DAC with clock output (not exactly common either), or a separate clock unit and a DAC with clock input as well (not common or cheap)...

I honestly don't understand why most megabuck (my defintion of megabuck being $1k and above ) DACs can't have that at least. Its easy to get a source with a clock input though - just get an RME or Lynx soundcard.

We did some tests with the AES16 with the word clock enabled and disabled ("Synchro Lock", AES16 as the master. I am interested in trying this card with a DAC that allows it to slave) and the difference with BNC connection on my Assemblage DAC is readily apparent. Maybe big bucks doesn't have to be spent to get a good digital transport, but this test certainly made me a believer in not always going for the cheapest sound card available. At ~ $700, the bump in resolution is worth it to me....maybe not with some mid level headphone setups, but with a TOTL speaker setup it's worth it (again for me).

Edited October 4, 200916 yr by deepak

We did some tests with the AES16 with the word clock enabled and disabled ("Synchro Lock", AES16 as the master. I am interested in trying this card with a DAC that allows it to slave) and the difference with BNC connection on my Assemblage DAC is readily apparent. Maybe big bucks doesn't have to be spent to get a good digital transport, but this test certainly made me a believer in not always going for the cheapest sound card available. At ~ $700, the bump in resolution is worth it to me....maybe not with some mid level headphone setups, but with a TOTL speaker setup it's a no brainer (again for me).

Ideally the master and slave relationship should be the other way around... shouldn't it?

Edit: Ah, nvm. People have reported that the AES16 isn't necessarily "the best" soundcard for that purpose - but with the variance in computer setups it probably doesn't matter.

Ideally the master and slave relationship should be the other way around... shouldn't it?

Nope can't work in slave mode.

Just one thing to add, an async USB DAC works exactly like a DAC with a wordclock output slaving its transport, which is why we're all harping for the async USB protocol. The argument for having an async USB transport is that if you can't have the DAC directly slaving the PC, the next best thing is to have a transport that slaves the PC (for computer audio).

... At which point we've moved from a push based protocol (which was reasonable when you we're reading pits on a spinning disk, with a laser, in real time) (with no re transmit / error correction) to a classic pull based Producer-consumer problem - Wikipedia, the free encyclopedia. The sad part is that its taken this long.

Well I have to weigh in here as this also seems to be excessively priced for what you get. I am very happy with all of my squeezeboxes (2), transporter, boom and duet. All run just fine and there is no lag when running squeezecenter (Now Squeezebox Server 7.4 although I am still currently running 7.3.4) off of the iMac (2GB RAM all music on one NAS). I also run my home office SB3 through a Musiland SRC-10 upsampler and then into the external DAC. This to me sounds better and has the advantage of using BNC for the output to the DAC (although that is not an option right now as my Buffalo32 is still partially [mostly] in pieces). Right now the only advantage the transporter has is that it can process native 24/96 files but the SB3 cannot.

Stereophile is saying "priced under $1000," which I assume is at least $900.

Ideally the master and slave relationship should be the other way around... shouldn't it?

Edit: Ah, nvm. People have reported that the AES16 isn't necessarily "the best" soundcard for that purpose - but with the variance in computer setups it probably doesn't matter.

What are better soundcards? It seems the choices are RME and Lynx. I prefer the minimal GUI of the Lynx and it has BNC. And a Lynx rep has guaranteed Windows 7 (x86 and x64) drivers.

Stereophile is saying "priced under $1000," which I assume is at least $900.

Priced under $1000, each WaveLink 24/96 Asynchronous USB-to-S/PDIF converter will come complete with Nirvana Audio's new T0 BNC-to-BNC digital cable and BNC-to-RCA adapter.

Suck it hard, Eejit!

That's a stupid amount of money for something that can't possibly cost more than $100 to build - it looks to be one board in a Hammond case with some ultra-simple milled front panels. An utter ripoff.

I'm working (slowly) on my own asynchronous USB implementation and may just open source it once it's done.

What are better soundcards? It seems the choices are RME and Lynx. I prefer the minimal GUI of the Lynx and it has BNC. And a Lynx rep has guaranteed Windows 7 (x86 and x64) drivers.

Actually, I should have said that I've heard _one_ person say the RME 9632 is better (there aren't many who have owned both cards) but like I said, given that every computer is its own environment, its hard to tell if you can reproduce that result.

... but like I said, given that every computer is its own environment, its hard to tell if you can reproduce that result.

Isn't this just a variant of the "we all hear things different" arguement? NTTAWWT.

Suck it hard, Eejit!

shit, nice!

*gets on his knees*