AUDIO over IP - REDNET 3 & 16 Review. AES67 Sets A New Standard for Computer Audio

[rb2013]

  To be a bit more precise…
It really isn't that the timing and signal data are separated, but that the ethernet protocol doesn't use the timing info from the data stream itself, rather it uses it's own clock to 're-time' the data itself and ignores the timing from the incoming data stream altogether.
This is then passed along to the digital audio converter (AES, SPDIF, USB Toslink, etc.) which then feeds the dac.
 
I chose the word separated in my previous post because it was 'easier' to try and explain and to help provide more contrast to these 2 critical aspects.
 
This method of handling these 2 portions (timing and the data itself) of the data stream, is in contrast to both SPDIF and USB which rely upon the 'embedded' timing of the data stream itself, which is a major source of additional jitter among other variables and are problematic.
 
And if the output of this data stream (from the ethernet data path) is passed along using either AES or I2s, then the data and timing info remain separate as they are used by the dac.
 
These 2 protocols (SPDIF & USB) are the 'consumer' versions of the method of passing the audio data to the dac, where as AES is a 'pro audio' (meaning more robust) and I2s is a variant of HDMI (or is it visa versa?) which has not been formalized into a set structure (pinout and signal definitions) that all who use it agree upon.
 
I2s is potentially 'better' but really is best for use with short cable runs and is mostly an internal digital signal pathway to the main digital signal bus.

Ok some other things that make AOIP different from USB 2.0 Async - that is error correction - AOIP is packetized as is USB  - but with a robust two way error correction protocol. Something USB 1.0, 2.0 and 3.0 have  - but was changed in the USB 2.0 Audio - maybe why MS never supported it?
 
So SPDIF and AES - are very different then USB audio - in so many ways wouldn't use them in the same context.
 
As for SPDIF and AES - they are base on  the same underlying protocol - and have many of the same limitations:

S/PDIF was developed at the same time as the main standard, AES3, used to interconnect professional audio equipment in the professional audio field. This resulted from the desire of the various standards committees to have at least sufficient similarities between the two interfaces to allow the use of the same, or very similar, designs for interfacing ICs.^[4] S/PDIF remained almost identical at the protocol level,^[a]but changed the physical connectors from XLR to either electrical coaxial cable (with RCA connectors) or optical fibre (TOSLINK; i.e., F05 or EIAJ Optical), both of which cost less than the XLR connection. The RCA connectors are typically colour-coded orange to differentiate from other RCA connector uses such as composite video. The cable was also changed from 110 Ω balanced twisted pair to 75 Ω coaxial cable, using RCA jacks. 

 
AOIP operates using RTP and is Level 3 compliant - that means is requires no special LAN switches to work  - at least in theory.
 

An example, if a Audinate brooklyn card (or similar) were located inside a dac and fed an ethernet signal and it's output fed the dac via I2s, it could be the ideal means to get the digital signal from the source (computer) directly to the dac with minimal signal processing in between.
And assuming the word clock in the dac was 'good enough' (or could be externally supplied) this would be a single solution dac instead of several digital boxes between the source and the dac itself.
 
JJ

 
I guess you haven't been following this thread much.  It's not an IF but a reality.  Need to bone up my friend.

 

[rb2013]

  Um, I use my 800's as the arbiter of what is better, not the speakers.
And my electronics are quite nice actually and I do rather enjoy this system.
 
But they are after all ≈$75 speakers so I don't expect them to keep up, and true to form they don't even come close.
 
But my point was that the degree of improvement that both of us have noticed is, as the system that feeds both the headphones and speakers steps up, the speakers reflect these changes and in surprising ways.
 
Despite the fact that they are ≈$75 speakers.
In fact they are connected to my ROK amp using $125 speaker cables.
 
This tells me that these Dayton $40/pair + my mods ($30 for tweeters and better caps) actually do have scaleability and far more than I would ever have suspected.
 
JJ

So your ref system  - in which you make your great prognostications - is based on listening to $75 speakers (but oh $125 speaker cables!)
 
Dude you have got to be kidding me?
 
When you said 800's thought you meant these B&W 800's - had a pair back in the day - nice speaks:

 

[rb2013]

  Ok this is getting interesting. Can I ask (having only enough money for one box at 1K, I have at the moment the EVO full stack and LPS x 2, one for the clock, one for the EVO. I also use a Mac Mini tweaked up with 12V DC supply board and using Audirvana+ Async / Direct mode into a 25K DAC via SPDIF.
 
I was thinking the Mutec MC-3 + USB would be better than my EVO stack on it's own, but I read 2 reviews and they are real close,. no big deal. And TBH I don't dig the fact the Mutec is using USB power to drive the XIMOS chip board.
 
So can I use the Rednet ethernet switch from a Mac Mini and feed my DAC from the Rednet with SPDIF out, and not have to buy the Mutec as well? If yes, would the Rednet on it's own beat the EVO stack by some margin?
 
BTW I use the TotalDAC USB cable/filter and it elevated the EVO still further i.e. smoother in the treble for example.
 
Thanks in advance.

I never did get the EVO clock and that does improve it considerably.  That said - I'd be shocked if JUST the Rednet 3 - that is MAC>RN3>DAC doesn't far exceed the EVO stack.
 
I completely agree on the Mutec XMOS issue - IME (In my experience) comparing the Mutec MC-3+ USB directly against a well fed Singxer F-1 - the XU208 F-1 was clearly better at a fraction of the cost.  The Mutec XMOS board is not much to write home about - still has the previous gen XMOS U8.  Higher phase noise clocks (looks like generic XO's) then the F-1 Crystek CCHD-575's

 
I'd imagine even the TotalDAC USB cable/filter would not get USB past the AOIP RN3.
 
FYI there is a used Rednet 3 in the classifieds.
 
Cheers!

 

[rb2013]

I should add on the HP side of the equation - I had Senn HD800 with Moon Black Dragon V2 cables.

Sold them -as the well sourced and amped Maggies produced an imaging experience that the HD800's could come no where close to.

 

[Tboooe]

   
I completely agree on the Mutec XMOS issue - IME (In my experience) comparing the Mutec MC-3+ USB directly against a well fed Singxer F-1 - the XU208 F-1 was clearly better at a fraction of the cost.  The Mutec XMOS board is not much to write home about - still has the previous gen XMOS U8.  Higher phase noise clocks (looks like generic XO's) then the F-1 Crystek CCHD-575's
 
 
 

Very interesting comments.  I've been toying with the idea of getting rid of my Mutec.  Frankly, I am not entirely sure it is better than my el cheapo Music Fidelity Vlink USB converter.  Curious about your comments on the Singxer F1.  I assume the F1 also needs power from the USB cable right?  What do you think is the theoretical reason why the F1 sounds better than the Mutec?  Have you also compared the SU-1?

 

[jabbr]

Very interesting comments.  I've been toying with the idea of getting rid of my Mutec.  Frankly, I am not entirely sure it is better than my el cheapo Music Fidelity Vlink USB converter.  Curious about your comments on the Singxer F1.  I assume the F1 also needs power from the USB cable right?  What do you think is the theoretical reason why the F1 sounds better than the Mutec?  Have you also compared the SU-1?

Comparisons and discussion of USB devices is best done on the USB thread and not an this AOIP thread.
The use of Mutec 3+ USB in the context of AOIP is for the SPDIF reclocking and not for its USB interface.

 

[Tboooe]

Comparisons and discussion of USB devices is best done on the USB thread and not an this AOIP thread.
The use of Mutec 3+ USB in the context of AOIP is for the SPDIF reclocking and not for its USB interface.

My bad...

 

[Tboooe]

Has anyone compared HQ Player + NAA to HQ Player + Rednet?  
 
I got a pretty big jump in SQ when I added an NAA into my system.  Has anyone been crazy enough to try HQ Player >> NAA >> Rednet just for kicks?

 

[johnjen]

 
So your ref system  - in which you make your great prognostications - is based on listening to $75 speakers (but oh $125 speaker cables!)
 
Dude you have got to be kidding me?
 
When you said 800's thought you meant these B&W 800's - had a pair back in the day - nice speaks:

Um, read my first sentence again, HD-800's.
This is after all Head Fi, not speaker fi.
 
JJ

 

[astrostar59]

So interested in this Rednet 3 product.
 
A question
Can I hook it up as:
 
Mac Mini playing Audirvana+ to Ethernet out (download the Rednet driver for a 'virtual' sound card) to Rednet 3 and SPDIF out to DAC.
 
i.e no switch box / router. I only need a dead simple one way network. I prefer to use only the Rednet 3, so no Mutec and / or external clock.
 
Has anyone with a Mac done this? If so how did it sound compared to say the Berkley Alpha USB or EVO full stack with LPS?
 
Also do you need a PCi card to keep latency right down below USB specs, or does it still perform really well using the 'virtual' sound card app from Focusright?

 

[rb2013]

Um, read my first sentence again, HD-800's.
This is after all Head Fi, not speaker fi.

JJ

Ok that makes more sense. But you didn't write HD800. What doesn't, to me, is if you have $1500 HP's, why use $75 speakers to evaluate anything? Whatever works for you...

 

[astrostar59]

Ok that makes more sense. But you didn't write HD800. What doesn't, to me, is if you have $1500 HP's, why use $75 speakers to evaluate anything? Whatever works for you...

Agree. Most in the audio industry recommend a similar budget for DAC/Amplifier/Speakers. Yes there are some great budget DACs, some great book shelf type speakers, but getting it too far out of wack is just bonkers. It also applies to HP as well IMO. I don't believe a very fast and revealing headphone will be relaxed listening with a 300 USD DAC for example.

 

[motberg]

...................., why use $75 speakers to evaluate anything? Whatever works for you...

sheeeze guys... the way I read it was JJ was just making a rather amusing comment that the AOIP investment was worthwhile even using entry-level speakers...
In any case, thanks to both of you for further defining some of the technical differences between the various transport protocols.

 

[Soundsgoodtome]

I think JJ means the difference is significant that even cheapo (albeit decent near field for the price) speakers benefit from the focusrite.

rb2013 how's the Burl B2 Dante stock vs Rednet /your dac? I'm thinking maybe I keep my rednet and up my amp/dac. But if the B2 Dante is good stock maybe I consider it more.

 

[johnjen]

Quote:Originally Posted by motberg /img/forum/go_quote.gif
sheeeze guys... the way I read it was JJ was just making a rather amusing comment that the AOIP investment was worthwhile even using entry-level speakers...
  In any case, thanks to both of you for further defining some of the technical differences between the various transport protocols.
 

Originally Posted by Soundsgoodtome /img/forum/go_quote.gif

I think JJ means the difference is significant that even cheapo (albeit decent near field for the price) speakers benefit from the focusrite.
 
 
Yes to both of these.
 
And that when a well executed source is delivered with precision, it's as important as the HP's or speakers, perhaps even more so.
 
As in GIGO!
 
JJ