How To Connect! The Absolute Best Way!

[DSD_mastering]

MADI2 optical is running now at 100 Mbps which is the equivilent of 64 channels of 24/44.1 or 16 channels of DSD! How much more do you want?

 

 

Regards,

 

[Renerator]

I am using a CA DacMagic with USB interface regarding my connection from computer to DAC. I just want to make sure that I'm not missing out with what I currently have. So the type of DAC determines the interface. Correct? One that comes to mind is the Weiss Minerva with firewire interface. In this case, can you say that firewire is the better of the two interfaces?(between USB and firewire)

 

Thanks Eloise

 

[manisandher]

Eloise, you stated, "... the AFI1 is the equivalent of the Lynx card - but connected externally via FireWire rather than internally to PCI bus."

 

This was pretty much my understanding... until today. I've just learned that the Lynx conforms to the dual-wire standard and doubles the clock frequency when in slave mode. The AFI1 cannot do this.

 

Now, this is purely academic for the vast majority of people who are using so-called 'legacy' DACs. But for those of us who use the DAC as Master and interface as Slave, the AFI1 prevents the use of files >96KHz. The Lynx (and the RME equivalent) have the dual-wire standard implemented correctly. It seems the Weiss does not... which is surprising from such a company!

 

Mani.

 

[idiot_savant]

I've never even seen a Weiss, but from what I've read I'd be surprised if it was wrong with respect to dual AES...

 

What clock is doubled?

 

your friendly neighbourhood idiot

 

[manisandher]

Well, I'm as surprised as you are i_s.

 

The clock to which I'm referring is the clock being sent back from the DAC to the interface (via BNC) to slave it (the interface) to the DAC. Now apparently, part of the dual-wire specification is that the word-clock rate must be half of the sample rate. This clock must then be doubled by the interface. The Lynx will double the clock frequency automatically, once switched to dual-wire mode. The AFI1 doesn't and, from my exchanges with Daniel, I understand that it simply hasn't been designed to.

 

DACs like the Pacific Model Two can only output clock frequencies of 1x/2x 44.1KHz or 48KHz and rely totally on the interface to double this frequency when in dual-wire mode [EDIT: and rates >96KHz] thus enabling the interface to be slaved to the DAC.

 

As I said, this is all news to me (courtesy of www.goodsoundclub.com).

 

Mani.

 

[idiot_savant]

The way I see it: The clock back ( or forward ) should be an integer multiple/submultiple of the audio - the clock could be half/double the rate of the audio with no real technical consequence - this means, for instance, if the soundcard did the same thing, the Weiss could output a 44.1k clock, and the PC could output 44.1,88.2,176.4 on 1, 2 or 4 wires without anything having to lose lock....

 

Having said that, I'm not convinced there is a spec for dual wire clocking, so you will get these situations where one manufacturer does one thing, and another something else.

 

My feeling is Weiss is actually in the right here ( from an engineering viewpoint ), and Lynx/RME are not being as flexible as they could be

 

your friendly neighbourhood idiot

 

[manisandher]

Daniel was kind enough to send me the AES3-2003 specs and a cursory look through them does suggest that you (i_s) and Daniel are right - there is no spec for the wordclock!

 

But it would be nice if the AFI1 had the option to double the clock frequency to accomodate DACs like the Pacific (in Master mode at rates >96KHz). I can't believe this would be difficult to do.

 

Mani.

 

[idiot_savant]

doubling the clock isn't that difficult....

 

but I suspect you'll be waiting some time - anything to do with manufacturers disagreeing over a spec will be fairly low down their list of priorities ( as each believes they are doing the right thing ). Like I say, I actually think Weiss are correct here, but without a spec, who can tell?

 

your friendly neighbourhood idiot

 

[DSD_mastering]

Well DSD is clocked at 44.1 You shouldn't have to "double" or quadruple the clock to match the SR.

 

 

Regards,

 

[cfmsp]

 

"So the type of DAC determines the interface."

 

I'd say that the choice of DAC, the interface, and computer are all inter-related, for reasons that I mention above. There are very few DACs, if any, that will support all 4 major types of interfaces - Firewire, USB, Toslink and S/PDIF (Coax or AES/EBU). Even LightPeak might not support all 4 protocols, enticing as that might be to audiophiles.

 

And there are no known (to me) computers that provide all four outputs natively, although most Macs now feature Firewire, USB and Toslink (albeit mini-TOslink), and the tower Macs have the additional ability to support AES/EBU via the addition of a Lynx card. Given that the Lynx card costs more than the ubiquitous Mac Mini and also the second hand G5s that are recommended, it's more than just a passing decision to add the Lynx card, but it does offer those who have chosen (or already own) a 'legacy' DAC a chance to get great results from a computer audio setup.

 

If you don't already own such a DAC, or have your heart set, I'd suggest comparing Firewire / Async USB DACS and increasing the budget the extra $700-800, plus not insignificant cable expense (easily an extra $200-300) for DACs of that persuasion, since neither the Lynx card nor special/expensive cables are required, IMO.

 

enjoy,

clay

 

 

 

 

 

 

 

 

[Audio_ELF]

Sorry you lost me a little on the thing about external clocks using Dual Wire...

 

Am I reading write that one device (the Weiss) expects the clock at the total sample rate (176.2kHz) where as the Lynx uses the clock at the rate of each wire (88.2kHz) when receiving the clock from an external source?

 

Eloise

 

[manisandher]

Eloise,

 

Yes, you're right: in dual-wire mode, the Weiss expects the clock at the total sample rate (176.2KHz or 192 KHz) where as the Lynx expects the clock at the rate of each wire (88.2KHz or 96KHz). Some people believe that the latter meets the 'true' dual-wire specifications.

 

In any event, the Pacific can only output a clock at 1x/2x 44.8KHz or 48KHz at its BNC clock output, and will therefore not work with the Weiss in dual-wire mode AND Master clock mode.

 

However, the Pacific will work with the Weiss perfectly happily at all SRs in PLL or Slave mode.

 

Putting all of this into perspective, I believe only a handful of people on the planet would be at all perturbed by this...

 

On a final note, Daniel Weiss is looking into this issue for me to see if there is a 'simple' solution.

 

Mani.

 

[SeVeReD]

I'm using a Fireface 800 to go out to a Stello DAC while I wait for PeterSt to develop his new DAC. The Stello DAC has a USB interface, but SQ varied too much with type of usb cable/usb port out from the computer. Also the Stello accepts 24/96 into its coax, but not usb. Using firewire out to the Fireface has been very stable for me. I like that it's outside of the case; I like that it's firewire; I like that it passes hires material untouched.

 

 

[Nicholas.Bedworth]

Actually, some higher-end laptops from mainstream vendors such as HP and Toshiba definitely have S/PDIF available as a standard feature.

 

It is available, on a Toshiba Qosmio for example, through a mini TOSLINK connector, and on the HP HDX/DV series, widely sold at Costco, etc., on the output of an inexpensive docking panel that actually has an RCA coax connector on it. What service!

 

http://www.shopping.hp.com/product/notebook/notebook_hp/dock_solutions/4/accessories/KG461AA%2523ABA

 

Haven't tested them for audio quality let, but will be doing so shortly.

 

Of course almost all Windows-based laptops and desktops sold at retail as "digital media" oriented machines, which is most of them, have FireWire as well.

 

[Nicholas.Bedworth]

Good background material... just Google or Bing IEEE 1394 for more in this vein.

 

IEEE 1394 defines a single isochronous interconnection bus that serves many purposes and user scenarios. In addition to its adoption by the consumer electronics industry, PC vendors including Compaq, Dell, HP, IBM, Fujitsu, Toshiba, Sony, NEC, and Gateway are now shipping Windows-based PCs with IEEE 1394 buses.

 

• The IEEE 1394 bus complements the Universal Serial Bus (USB) and is particularly optimized for connecting digital media devices and high-speed storage devices to a PC.

 

• IEEE 1394 is a peer-to-peer bus. IEEE 1394 devices have more built-in intelligence than USB devices, and they run independently of the processor, resulting in better performance.

 

• The 100-Mbps, 200-Mbps, and 400-Mbps transfer rates currently specified in the IEEE 1394a standard and the proposed enhancements in IEEE 1394b are well suited to meeting the throughput requirements of multiple streaming input/output devices connected to a single PC.

 

• The licensing fee for use of patented IEEE 1394 technology has been established at US $0.25 per system.

 

With connectivity for storage, scanners, printers, and other types of consumer A/V devices, IEEE 1394 gives users all the benefits of a great legacy-free connectora true Plug and Play experience and hassle-free PC connectivity.

 

 

[cfmsp]

 

"Actually, some higher-end laptops from mainstream vendors such as HP and Toshiba definitely have S/PDIF available as a standard feature."

 

Thanks Nicholas for the clarification - that indeed some computers do have native Coax S/PDIF. I have a disturbing habit of dropping the 'Coax' when I mean 'Coax S/PDIF'.

 

And like many 'audiophiles' I only consider Toslink for 'convenience' listening (or wireless via Airport Express, or similar). Most if not all Mac computers support Toslink via the mini-toslink output.

 

Thanks especially for the reminder to folks here about the strengths of Firewire, as I'm perhaps the single most annoying Firewire user on the list. Glad to have company from a fellow Firewire supporter, in indeed you are such.

 

clay

 

 

[manisandher]

The AFI1 can now work in dual-wire mode at wordclock I/O rates which are either half or full audiorate...

 

... Perfect for legacy DACs like the PMII!

 

A big thanks to Daniel and Rolf at Weiss for listening to their customers (well, me at least!) and making things happen.

 

Mani.

 

[cfmsp]

 

 

"... Perfect for legacy DACs like the PMII!"

 

Well, if you're gonna have a legacy DAC, this would be the one!

 

Mani, I thought you said there might only be a handful of people in the world that would suffer your problem. That's quite generous of Daniel if that was the case. Kudos to Weiss.

 

clay

 

 

 

 

 

 

[manisandher]

Yeah, I really can't believe that there would be many people with an AFI1 and a Model Two, who insist on playing 176.4/192KHz SR files in MASTER mode (i.e. slaving AFI1 to PMII). But this didn't stop Weiss from looking into this for me.

 

Now that's what I call customer service!

 

Mani.

 

[The Computer Audiophile]

That's great to hear Mani! I've spent many hours listening to PMIIs and it's hands down the best DAC I've ever heard. Too bad they are like rare coins now, hard to find and very expensive.

 

[manisandher]

According to Dave Peck at Euphonix, my PMII is an early unit and was used as an in-house demo unit for a couple of years. I can't help wondering who might have sat there listening to my very machine...

 

Mani.

 

[Renerator]

Hello Eloise. Can you elaborate on which Weiss interface you spoke of in your comment?

 

[Audio_ELF]

Weiss currently sell two interfaces to connect AES (or SPDIF) to FireWire, with a third on the way.

 

The AFI1 is a pro-orientated device with 8 channel input and 8 channel output.

The Vesta is more consumer orientated, built in the same case as the Minerva DAC it supports 2 channel in and 2 channel out.

 

Finally there is a new device - the INT202. This is a pure 2 channel output device.

 

All three will support dual-wire interface which splits a 24/192k signal across 2 interconnects.

 

Eloise