AOIP (Audio over Ethernet) or is USB dead?

[astrostar59]

Hi guys

Not sure of there is a thread here for this. Anyway, I was wanting to let folks know who maybe haven't tried this connection method that it rocks! Those that have are already ahead of the game.

I had a demo of the C1 CH precision 6 weeks back, and using USB connection is was just too bright and forward, making many tracks sound harsh. I talked to the dealers and they said it works best with the D1 or using Ehternet. That got me thinking. I didn't want to spend anything like the money on a C1 for that level of sound quality, and TBH it was not as good as my existing DAC.

 

USB Out AOIP In
I have had various USB chains, Offramp 5, M2Tech EVO stack, used various inbuilt boards with and without USB power, and heard many others and was never totally convinced that a DAC connected via USB was so great. I just sounded a bit course in the treble, too digital and lost something IMO. 4 weeks back I bought a RedNet 3 and this all changed for the better. Totally amazing the difference. So liquid, real and just better in every sense. I would say give it a try and see what you find.

Long thread here:

http://www.head-fi.org/t/806827/audio-over-ip-rednet-3-16-review-aes67-sets-a-new-standard-for-computer-audio

and another on www.computeraudiophile.com

It is not plug n play, but is not difficult to set up and TBH cost me less than my USB chain and sounds much better. RedNet comes from the studio recording sector, hence why it is not plug n play, but uses Audinate (Dante) software that is optimised for audio, not a standard DHCP. I am convinced a good server / PC and USB can beat most CDPs, but AOIP IMO has gone above that now.

 

Edited October 27, 2016 by astrostar59

[RudeWolf]

So technically what's so bad about USB audio that's solved by AOIP?

[Dusty Chalk]

8 hours ago, astrostar59 said:

...using USB connection is was just too bright and forward...

You lost me. 

[CarlSeibert]

I love that "what is aes67 and why should I care?"headline. Since the story didn't tell me either thing.

So. Somebody, what the heck is this protocol?

And if it's got for real error correction, I could be seeing the point, vis USB....

[Sechtdamon]

On 27.10.2016 at 2:57 PM, astrostar59 said:

I have had various USB chains, Offramp 5, M2Tech EVO stack, used various inbuilt boards with and without USB power, and heard many others and was never totally convinced that a DAC connected via USB was so great.

If you don't like using USB, then why don't you simply use a DDC? Less pain in the ass than your method, and you get to use outputs/inputs suit well for digital audio data transmission...

Edited October 28, 2016 by Sechtdamon

[DefQon]

That Focusrite looks pretty.

[Dusty Chalk]

It shore is purty.

[mtoc]

I rather go back to the IEEE1394 and save some money.

[stax fart]

Congratulations, you've discovered America.

My Burl B2 DAC has Dante too, and I don't need that interface, only an ethernet port

[astrostar59]

Don't know all the technicalities, but know what I hear. As I understand it, the data transmission and integrity of the data is much more intact. USB is full of noise regardless of the chain used it seems. Check out the various thread on AOIP for more responses from ones who switched. Pretty much 100% who have don't want to go back to USB.

[Dusty Chalk]

Do you really think with language like that, that you should be the standard bearer for such an interface if it's that much better?  I mean, even if it is better, you still sound like a snake oil salesman.  You people with your "my statistical sample of 1 is adequate" -- no it's not!  Not ever!  Sure, maybe it's good enough for you to buy-in, but it's not good enough to tell other people to buy-in, just based on your word, what if your previous system was so crappy that AOIP couldn't help but be better?

My first question in such a thing is, "if it's so great, why don't the pros use it?"  So I looked.  Seems to be just starting out and used almost exclusively for broadcast.  Long runs.  The standard for studio still seems to be AES3.  Sometimes they'll add an extra clock.  I'll wait until it matures somewhat.

[Sechtdamon]

2 hours ago, astrostar59 said:

Don't know all the technicalities, but know what I hear.

Did you make any blind test? Or what you hear is just based on honeymoon and placebo effects?

2 hours ago, astrostar59 said:

USB is full of noise regardless of the chain used it seems.

No it is not. If you use a transport pc just for running what you use for playback. And i use WYRD it was great.

 

2 hours ago, astrostar59 said:

Check out the various thread on AOIP for more responses from ones who switched. Pretty much 100% who have don't want to go back to USB.

At where? On head-fi? People swim in snakeoil at there.

[grawk]

I think the super audio best friends or whatever they're called is where you want to hold that particular debate, thanks.

[astrostar59]

1 hour ago, Sechtdamon said:

Did you make any blind test? Or what you hear is just based on honeymoon and placebo effects?

No it is not. If you use a transport pc just for running what you use for playback. And i use WYRD it was great.

 

At where? On head-fi? People swim in snakeoil at there.

Yes, but another massive thread on computer audiophile. Try it and then say is it snake oil. It is a game changer IMO, and TBH costs the same as many USB chains.

I ran my USB chain alongside the Rednet for a week solid. I could tell very quickly, but after a week I wasn't going back to USB no chance. The difference is not subtle. The removal of treble glaze and harsh leading edges, that impression of digital noise in the music. I think HP users are sensitive to this more than speaker guys IMO. 

On the Carbon and 009s plus my horn speakers it was real obvious. I would hazard a guess a lot of the 'brightness' issues in the 009 is from the failures upstream of USB audio. The detail and transparency is the same, but it is dead smooth, liquid I would call it, less 'hifi' and digital. For example I always suspected some female vocals had a on/ off type sound to it, like it was dropping out at an incredibly fast rate, like a chain saw type of sound to it. Anyway, that aspect is now GONE, it is as I say, fluid and super smooth.

I have no fixed opinion as to why 1000BaseT Ethernet is better for audio. I guess obvious reasons maybe latency and speed, accuracy of delivery. Possibly less noise in the data connection, possibly the way Dante sends that data. As I understand it USB was never designed for high end audio streaming. Anyway, give it a try, it would be good to get others on here who have tried it.

[Sechtdamon]

I shoud've make that clear, I did not use WYRD for sonic improvement and it did not (maybe a little, really little). I used it cuz my computer (designed for gaming) couldn't bring stable signal (too much latency) to my dac.

I've tried nearly all digital connections over year(s) with different dacs, or digital chains. The differences between connections and DDCs/usb cleaners were only "nuance" differences. And believe me I've tested them blind to avoid honeymoon and placebo effects for I'm extremely skeptical.

The thing we (I) don't like as I understand, if I understand correctly, you behave and talk like a vacuum cleaner salesman, who insist the vacuum cleaner he sold is way better than any others based on no useful basis. And you might be comapring your vacuum cleaner with broomstick.

There is no theorical basis, and as it seems there is no practical basis which can be acceptable too.

If you wanna make a point why don't you send "the interface" or whatever this is to people that can measure - theorical or hearing based (with blind test) -, and let him/her share foundings with head-case? Or send it to me, I have various digital inputs/outputs i can compare with it.

Edited November 6, 2016 by Sechtdamon

[Dusty Chalk]

17 hours ago, astrostar59 said:

I would hazard a guess a lot of the 'brightness' issues in the 009 is from the failures upstream of USB audio. The detail and transparency is the same, but it is dead smooth, liquid I would call it, less 'hifi' and digital. For example I always suspected some female vocals had a on/ off type sound to it, like it was dropping out at an incredibly fast rate, like a chain saw type of sound to it. Anyway, that aspect is now GONE, it is as I say, fluid and super smooth.

I have no fixed opinion as to why 1000BaseT Ethernet is better for audio. I guess obvious reasons maybe latency and speed, accuracy of delivery. Possibly less noise in the data connection, possibly the way Dante sends that data. As I understand it USB was never designed for high end audio streaming. Anyway, give it a try, it would be good to get others on here who have tried it.

Again, without knowing what was wrong with your previous system, just telling us it is better tells us it's better than broken, and that's not telling us much.  Most of our systems aren't broken. 

USB is fine for transferring data -- think about a program, or an encrypted archive:  it needs to be a perfect digital replica of itself after being transferred.  If one bit is off -- one bit! -- it is considered corrupt. 

So the only thing left is timing.  Jitter.  Jitter just isn't that audible, not as night and day as you are describing.  It's subtle, as Sechtdamon says.  Not unless something was really really wrong, and you should have been able to hear a difference just by trying a third system.

Look, I'm happy for you that it's so much better than your old system, but most of us wouldn't have put up with that shit. 

[MLA]

2 hours ago, Dusty Chalk said:

USB is fine for transferring data -- think about a program, or an encrypted archive:  it needs to be a perfect digital replica of itself after being transferred.  If one bit is off -- one bit! -- it is considered corrupt.

Well, if I read this correctly, the perfect replica USB protocol used for backups etc (i.e. bulk with checksums) is not what is used for USB audio streaming, so it would appear that packages at least theoretically can get lost or get screwed up when streaming audio over USB (as hinted by CarlSeibert above).

http://www.edn.com/design/consumer/4376143/Fundamentals-of-USB-Audio

That said, among explanations for why USB audio sound different in some contexts, I think Craig's observations have larger impact than transfer protocol type:

https://www.head-case.org/forums/topic/12477-usb-audio-tweaks/#comment-721196

And it happens to be measurable

[ironbut]

Pro audio has been fooling around with ethernet for a while.

High track count and long cable runs are some of the main advantages. 

MADI and Dante are the most common so far. 

Merging Technologies is using the AES67.

http://www.merging.com/products/pyramix/ravenna

Dante:
+allows 96 kHz without loosing any of 128 channels
+64 channels at 192 kHz
+redundancy (on most units)
+flexible routing
+cables are easy to get
+Dante Virtual Soundcard
MADI:
+soundcards with DSP available
+USB and ExpressCard interfaces available
+most units have optical interface (long connections and galvanic isolation)
+it isn't a proprietary protocol
+more second hand gear available

[Pars]

26 minutes ago, ironbut said:

+allows 96 kHz without losing any of 128 channels

Fixed that for you.

Just to interject, and I know most of you aren't jitter believers (beliebers? :)), but some theorize that clocking errors in the recording chain (i.e., pro audio) have ill-legitimized much (all?) of the digital recording legacy and library. Source of info is jockohomo, but I for one take note of what he might say on the subject, as obtuse and inferral as he may be. Scary thought, eh?

My point relevant to this discussion being that any opinion on an interface should be taken from a grounded reference of a known entity performing in a known acoustic space. If it sounds like the real deal, then there might be some merit in the technology, but keep in mind that there might be problems in the recording chain as well (blasphemy, I know). SPDIF has a multitude of problems, whether coax or optical. USB had a shitload of problems prior to asynchronous or whatever they call it now (Wavelength Audio IIRC). All of this shit (the clocking, and the transmission) is in the RF analog domain, and the implementation matters.

Edited November 10, 2016 by Pars

[Craig Sawyers]

Some interesting details of getting an SPDIF interface to work properly are here http://www.audiocircle.com/index.php?topic=45330.0

Although, even if you get the data transmission perfect with SPDIF you still have problems with clock recovery, the principles outlined can be applied to any digital link at all - as Pars said, a digital link of any sort is high bandwidth RF engineering, and many (most?) audio system designers do not understand this. Which would explain why different digital cables sound different on a particular system - if the impedances at high frequency are not well controlled, and significantly different from what they ought to be (50, 75 or 110 ohms say), then it is open turkey shoot for fiddling with digital cables to mitigate this.

Edited November 10, 2016 by Craig Sawyers

[astrostar59]

On 7 Nov 2016 at 6:30 AM, Dusty Chalk said:

Again, without knowing what was wrong with your previous system, just telling us it is better tells us it's better than broken, and that's not telling us much.  Most of our systems aren't broken. 

USB is fine for transferring data -- think about a program, or an encrypted archive:  it needs to be a perfect digital replica of itself after being transferred.  If one bit is off -- one bit! -- it is considered corrupt. 

So the only thing left is timing.  Jitter.  Jitter just isn't that audible, not as night and day as you are describing.  It's subtle, as Sechtdamon says.  Not unless something was really really wrong, and you should have been able to hear a difference just by trying a third system.

Look, I'm happy for you that it's so much better than your old system, but most of us wouldn't have put up with that shit. 

I am saying just try it. That is the best way to know for sure. I try to stay out of the technicalities. What set me on this path was the tech guy from CH precision saying the C1 sounded better on AOIP than USB. That the the 2 massive threads on computeraudiphile and head-fi. Pretty much 100% say it is better than USB. Quote like 'liquid' and 'super smooth, transparent'.

I would hazard a guess in my limited technical knowledge on this, that there is noise in the USB line, and the power sent down it doesn't help even if it is not used. Also the USB bus is shared on most PCs between other ports. Then almost zero latency may help with jitter? 1000 Base T is miles faster than USB. Dante in the case of the Rednet designed the system just for high quality audio, something USB was not designed for.

TBH my old system already sounded amazing, I thought I was done, until I found AOIP.

Edited November 10, 2016 by astrostar59

[Sechtdamon]

Dear astrostar59;

 

As I and some other Sirs said before computeraudiophile and head-fi are like pools filled with snakeoil. And you can be under honeymoon and placebo effect. I was a Pro gamer not so long ago, efff latency, do you know how much latency we (mankind) have? and efff fastness.

 

If you want to test us out. Then send the gear to any of us.

 

And last note, when you say you can see the difference between "amazing" and "moaaar amazin" it kinda loses the plausibility.

 

You sound like you're selling that product you hype.

 

Best wishes,

 

Özgür.

[Dusty Chalk]

4 hours ago, astrostar59 said:

I am saying just try it.

You lost me.  That's exactly what I'm telling you not to say.  -100 reading comprehension points.

[astrostar59]

Incredible. Ok so put your head in the sand, slag off what you haven't heard, and stop halfway up the hill, say you know everything. What is wrong with try it and see, don't believe my word for it. Use your ears, then you can come back and slag it off, not before you know zip about the effect. And there is a LOT of very good audio folk on other forums, yes really, there is a world out there.....

I am done.....

[grawk]

On 11/6/2016 at 5:58 AM, grawk said:

I think the super audio best friends or whatever they're called is where you want to hold that particular debate, thanks.

stand by this.