Thoughts on a bunch of DACs (and why delta-sigma kinda sucks, just to get you to think about stuff)

May 26, 2015 at 4:47 AM Post #5,551 of 6,500
 
Ethernet is inherently asynchronous.  Data is sent in packets (e.g. TCP/IP).  As the receiver consumes them it sends acknowledgements which signal to the transmitter that it can send more data.  The receiver will maintain a buffer of received data from which it will send audio samples (16 or 24 bit) to the digital filter and DAC chips at its own stable clock rate.  There is no need for the data stream (over ethernet) to carry accurate audio timing information.

 
Thanks; I'm here to learn.
 
So, is this 'streaming' the audio, or something else? Also, does this mean that a usb dac, with an asynchronous receiver chip, is much the same?
 
May 26, 2015 at 12:42 PM Post #5,552 of 6,500
   
Ethernet is inherently asynchronous.  Data is sent in packets (e.g. TCP/IP).  As the receiver consumes them it sends acknowledgements which signal to the transmitter that it can send more data.  The receiver will maintain a buffer of received data from which it will send audio samples (16 or 24 bit) to the digital filter and DAC chips at its own stable clock rate.  There is no need for the data stream (over ethernet) to carry accurate audio timing information.

99% right, the as you stated Ethernet is technically async.  However depending on how they implemented it the receiver may or may not send acknowledgements to the sender - this depends on exactly how they have implemented their communications in(see TCP vs UDP).  I don't think anybody wants me to get deep into the way networking protocols are layered here
 
Just to add a bit more color on why this could be super awesome
 
USB audio as it stands right now is NOT error corrected due to the real time nature of audio.  USB in general is error correct This is because USB Audio was designed to have low latency meaning you get a sound as soon as you do something on your computer.  Waiting for error correction is not useful in this case as is correcting samples who's temporal need was already passed.
 
This is where it remains to be seen if this is potentially any better than USB.
 
As their implementation can be completely proprietary they have an opportunity to prioritize data reliability vs latency as I think any of us can say "hey if there is a 500ms delay when I hit play to when it plays, im okay".   Doing this would require anything as simple as a ring buffer as Murrays said but allow error correction to take place.
 
At $8900 USD this is not in the same price range as the Yggy for most buyers however so any qualifications of "better" need to be weighed against that.
 
May 26, 2015 at 1:06 PM Post #5,553 of 6,500
Thanks; I'm here to learn.

So, is this 'streaming' the audio, or something else? Also, does this mean that a usb dac, with an asynchronous receiver chip, is much the same?


The def for 'streaming' is a bit fuzzy but in your quote seems to mean sending clocked audio signal (data bits + clock). A network Dac will just get the data bits over ethernet and the clocking happens inside the Dac. An Usb Dac does pretty much the same. It's not 100% technically correct but your comparison is good enough: a network Dac is ~same as an Usb Dac but with ethernet cable.
Hope that helps you wrap your mind around the concept.
 
May 26, 2015 at 10:48 PM Post #5,556 of 6,500
The def for 'streaming' is a bit fuzzy but in your quote seems to mean sending clocked audio signal (data bits + clock). A network Dac will just get the data bits over ethernet and the clocking happens inside the Dac. An Usb Dac does pretty much the same. It's not 100% technically correct but your comparison is good enough: a network Dac is ~same as an Usb Dac but with ethernet cable.
Hope that helps you wrap your mind around the concept.

 
I'm still trying to get my head around it (and many other digital and electrical related topics!), but I'm lacking info as well as intelligence.
 
I wonder how much it matters that the computer isn't just 'sending data', via usb or ethernet, but is actually playing the file. If the asynchronous dac 'sees' the feed as just pure data (the same as moving a file), where the clock information is superfluous, it should be ok.
I guess it just must respond when a signal is fed to it, and start the 'playing' process, with a slight delay.
 
Are you saying that most usb dacs are asynchronous?
 
May 26, 2015 at 10:56 PM Post #5,557 of 6,500
Another Theta Gen V on ebay
http://www.ebay.com/itm/121654818125
No relation to the seller, just curious to see how big the hype really is
biggrin.gif

 
Yeah; already watching; shipping not a problem. Thanks for your 'help'!
 
May 27, 2015 at 4:04 AM Post #5,559 of 6,500
I'm still trying to get my head around it (and many other digital and electrical related topics!), but I'm lacking info as well as intelligence.

I wonder how much it matters that the computer isn't just 'sending data', via usb or ethernet, but is actually playing the file. If the asynchronous dac 'sees' the feed as just pure data (the same as moving a file), where the clock information is superfluous, it should be ok.
I guess it just must respond when a signal is fed to it, and start the 'playing' process, with a slight delay.

Are you saying that most usb dacs are asynchronous?


Yes almost all Usb Dacs nowadays are async. Another somewhat fuzzy term but in this case it means that the Dac is controlling the data transfer and asking for new data packets as needed (instead of the PC sending them as soon as they are ready).

And what do you think 'playing the file' means ?! The PC-music-player 'grabs' a file from the hdd (flac, waw, mp3, etc), converts the data bits into a format understood by the Dac (e.g. PCM) and sends those data bits to the DAC through the Dac driver. For the PC it doesnt matter much whether it's printing a doc or playing an audio file, it's just processing of data bits as usual...actually printing a doc works exactly as playing an mp3: e.g. microsoft Word grabs the doc, converts it to a printing format and send the converted data bits to the printer through the printer driver.
There are some diffs of course but only in the details...also they mostly concern the dac/printer side, not the PC itself.

@the theta people
yap the hype is going strong. There were actually 33 bids even before I posted the link so I would expect this to be the most expensive theta ever. Btw, there's one in the classifieds here too.
 
May 27, 2015 at 4:32 AM Post #5,560 of 6,500
Anybody heard the Audiolab 8200 and Yggy recently and can post impressions/comparisons? Not looking to bash Yggy, honestly want how and what the differences are.
 
May 27, 2015 at 10:06 AM Post #5,562 of 6,500
Purrin, can you comment on wyrd+odac?
 
May 27, 2015 at 12:32 PM Post #5,563 of 6,500
@purrin, et. al., seeking information on the DCs "Ring Technology", a hybrid of Delta Sigma and R2R what are the differences in DCs's approach, what are the sonic attributes versus R2R/DS?
 
Listening to my "mothballed" Arcam FMJ CD23, wow Stevie Ray sounds really good, deep powerful bass, clean, clear, but not even close to etched, very sweet highs.
 

Users who are viewing this thread

Back
Top