Ayre USB DAC

[Hopstretch]

It's definitely on the plain side, innit?

Some further details here, will be interesting to see that white paper.

[Hopstretch]

Maybe. I kinda like having buttons to push or switches to flick, just to make myself feel a part of the process.

[Smeggy]

I don't like it, it doesn't go with anything I'm wearing. It needs moar plaid.

[Dusty Chalk]

Or perhaps you should just start wearing bright blue LEDs.

[Smeggy]

.......hmmmm, never thought about that, and I have the perfect place to keep the batteries....

[deepak]

TAS1020 I2S to FPGA

Don't know which dac chip they decided on.

Output is discrete, IV discrete no opamps

No upsampling, their own digital filters.

From AA. Interesting...

Edited January 25, 2009 by deepak

[Dreadhead]

From AA. Interesting...

Agreed. I'm looking forward to the reviews.

[deepak]

Since it accepts 24/96 I assume the DAC chip is still an upsampling DAC though?

[Dreadhead]

Since it accepts 24/96 I assume the DAC chip is still an upsampling DAC though?

I don't know enough what all the chips offer but it could just support each of the sample rates and work as a NOS system.

I'm assuming that it has optical inputs too which is what I'll be using since I need to pass the data through my DEQ anyway.

We shall see what comes out the other end soon enough. I'm hoping Stereophile does a review and measurement session soon.

[Hopstretch]

Sorry, it's USB-only I think.

[Dreadhead]

Sorry, it's USB-only I think.

Well screw that then.... Still interested to hear reviews

[Dreadhead]

Sorry, it's USB-only I think.

Well I've confirmed this with the article over at Computer Audiophile. I was excited for this product but it now looks like it's a lot of money to not include the capability to use it without a computer in the loop.

Oh well maybe they'll make a version with digital inputs but maybe they did not want to compete with their disc players.

[deepak]

I don't know enough what all the chips offer but it could just support each of the sample rates and work as a NOS system.

Anyone can correct me on this, but I don't think there is a NOS Dac chip that supports anything higher than 20/44.1

[Dreadhead]

Anyone can correct me on this, but I don't think there is a NOS Dac chip that supports anything higher than 20/44.1

I looked around a bit after posting that and it appears that you are correct that 24/96 seems to be all OS. Makes me wonder what the heck you are doing when you select the upsampling on some DACs, I guess that is the upsampling that the manufacturer controls but once it gets to the actual DAC chip all bets are off?

The 1794A seems to allow you to circumvent their digital filter (which is my understanding of what Ayre did) and maybe that will allow you to control the oversampling.

[cetoole]

Anyone can correct me on this, but I don't think there is a NOS Dac chip that supports anything higher than 20/44.1

What do you mean by "NOS DAC chip"? There are plenty of DACs with no internal upsampling/oversampling that can support higher than 20/44.1. The PCM1704 is probably the best example of this, and IMO, still the most interesting current production DAC out there.

[Icarium]

Mmmm PCM1704.

[cetoole]

Indeed. Certainly my first choice of current production DACs.

[cetoole]

Well I've confirmed this with the article over at Computer Audiophile. I was excited for this product but it now looks like it's a lot of money to not include the capability to use it without a computer in the loop.

Oh well maybe they'll make a version with digital inputs but maybe they did not want to compete with their disc players.

My guess, based off what I know about Ayre, is that they didnt include SPDIF input because of worries about jitter on that interface. I have read that they havnt made an outboard DAC before for this reason alone. Async USB basically solves the jitter problem for them, as the only jitter which should be present is the intrinsic jitter of the fixed XO they use to clock it and the jitter added on the board by parts like the FPGA (which will be handling the oversampling), but I wouldnt be surprised to find they are reclocking between the FPGA and DAC(s).

You can make a SPDIF interface with very low jitter with the application of a good secondary PLL reclocking the incoming data, either in analog with a VCXO or digital with FIFO buffering, but neither of these solutions are very simple or cheap. I am sure Ayre has people who could easily handle it, but suspect it was omitted because they were trying to meet a price point.

[justin]

My guess, based off what I know about Ayre, is that they didnt include SPDIF input because of worries about jitter on that interface. I have read that they havnt made an outboard DAC before for this reason alone. Async USB basically solves the jitter problem for them, as the only jitter which should be present is the intrinsic jitter of the fixed XO they use to clock it and the jitter added on the board by parts like the FPGA (which will be handling the oversampling), but I wouldnt be surprised to find they are reclocking between the FPGA and DAC(s).

You can make a SPDIF interface with very low jitter with the application of a good secondary PLL reclocking the incoming data, either in analog with a VCXO or digital with FIFO buffering, but neither of these solutions are very simple or cheap. I am sure Ayre has people who could easily handle it, but suspect it was omitted because they were trying to meet a price point.

Is it entirely USB powered? Lack of power supply which would be necessary for a SPDIF input is a major $$ saver

[cetoole]

Is it entirely USB powered? Lack of power supply which would be necessary for a SPDIF input is a major $$ saver

I doubt it, and I think I see a standard power cable coming out of the rear on the first pic posted here. I have no more information on this than anyone though. I do think I remember hearing that CH likes current mirror based I/V converters though, but this could be old data, and I dont even know that they use a current out DAC. The Ayre D-1 does (4x PCM1704) as does the CX-7 (PCM1738), but who knows? Gordon mentions IV discrete, and he would certainly know the difference, but as he doesnt know what DAC is used, I guess this could be an assumption on his part.

[Dreadhead]

My guess, based off what I know about Ayre, is that they didnt include SPDIF input because of worries about jitter on that interface. I have read that they havnt made an outboard DAC before for this reason alone. Async USB basically solves the jitter problem for them, as the only jitter which should be present is the intrinsic jitter of the fixed XO they use to clock it and the jitter added on the board by parts like the FPGA (which will be handling the oversampling), but I wouldnt be surprised to find they are reclocking between the FPGA and DAC(s).

You can make a SPDIF interface with very low jitter with the application of a good secondary PLL reclocking the incoming data, either in analog with a VCXO or digital with FIFO buffering, but neither of these solutions are very simple or cheap. I am sure Ayre has people who could easily handle it, but suspect it was omitted because they were trying to meet a price point.

Yeah I see your point and they said as much. To be honest though if you have the async USB going why not pass the TOSLINK data to the same ASYNC chip and you're off and running. Anyway they may do that in the future and I hope they do.

[cetoole]

Yeah I see your point and they said as much. To be honest though if you have the async USB going why not pass the TOSLINK data to the same ASYNC chip and you're off and running. Anyway they may do that in the future and I hope they do.

Its not nearly that simple. The reason Async USB is (practially) immune to jitter is because it contains the master clock and requests more data from the computer to keep it's buffer partially full. The computer's clock is slaved to the DAC clock, and all incoming data is buffered and reclocked by the DAC clock. You can modify a SPDIF source to act like this, though it adds a second wire to the interface and is completely nonstandard, so you wont see it in commerical devices, except for Lessloss, I believe. To keep SPDIF jitter down would require a completely separate reclocking system from the USB.

[Dusty Chalk]

Consumer commercial devices -- word clock is quite standard in the professional community.

[Dreadhead]

Its not nearly that simple. The reason Async USB is (practially) immune to jitter is because it contains the master clock and requests more data from the computer to keep it's buffer partially full. The computer's clock is slaved to the DAC clock, and all incoming data is buffered and reclocked by the DAC clock. You can modify a SPDIF source to act like this, though it adds a second wire to the interface and is completely nonstandard, so you wont see it in commerical devices, except for Lessloss, I believe. To keep SPDIF jitter down would require a completely separate reclocking system from the USB.

The Chord DAC64 for example just inserts a delay in the signal train and does the reclocking and doesn't require any extra outputs. I don't see much of an issue with this but I can see that it may require another implementation.

[deepak]

What do you mean by "NOS DAC chip"? There are plenty of DACs with no internal upsampling/oversampling that can support higher than 20/44.1. The PCM1704 is probably the best example of this, and IMO, still the most interesting current production DAC out there.

So the PCM1704 doesn't take a 16/44.1 input and convert it to 24/96 in the DAC chip? If it doesn't, then it was a poor assumption on my part (I was under the impression only the TDA chips did no upsampling internally).

I like the PCM DACs as well, I haven't read anything posted by Charles Hansen on his thoughts on digital on AA. But hopefully he is sticking to the PCM family