STREAMER - WHERE DO I GO FROM HERE?

 

@mdalton wrote "There are numerous other examples, but the answer is the same.  Just  having a USB connection doesn’t eliminate all sources of noise from a streamer to a DAC.  Not sure why you’ve dug in on this indefensible point.  Really weird."

RESPONSE:  Hee Haw!  I was hoping we could continue this discussion.  This is fun. . .and you brought receipts!   You’re a stubborn one, ain’t ya?  I hope you didn’t drop your mic.  I thought that we put this matter to bed.   I guess not. Ok, now, thank you for that article which happens to support my position.   I’ll try not to publicly embarrass you.  That would not be polite;   however, it may be necessary in the interest of the greater good.   So. I apologize in advance:  I’m sorry.   <smile>

(For the audience, sorry for the spacing, but I type these extended responses in MS Word and paste them in.  I try to m ake my posts easy to read an coherent.) 

Now, give me your hand (again) so I can walk you through it, but first, let me restate my position:

1.  We’re discussing USB-oriented streamers connected an asynchronous-DACs (Specifically, the BAT Rex 3 DAC) through a USB connection.
2. My claim is that  “SNR is one of the least relevant factors when comparing streamers in revealing systems” because Asynchronous USB breaks timing dependency. The DAC’s internal master clock controls the timing—not the streamers.  So.  the streamer’s SNR, jitter spec, or analog noise floor cannot influence timing accuracy. The DAC sets the system’s SNR.  That is, the DAC only receives. data packets (aka 1s an 0s) which do not convey an analog wave form; Therefore, a streamer’s analog‑domain SNR spec has no bearing on what the DAC receives.

That’s a mouthful, but let’s move forward and discuss the article any why it supports my position.  Here we go (4th attempt):

The article shows that DACs with asynchronous USB + reclocking ignore the streamer’s noise/SNR

Hi‑Fi News explicitly states that with DACs like the dCS Vivaldi and Mytek Brooklyn, the Rivo made almost no measurable difference:  “both the dCS and Mytek DACs provide full galvanic isolation/onboard reclocking, so very little difference in their inherent ~10psec jitter was detected.”

This is important because—and I’ll say this again for clarity: 

• If the DAC is asynchronous, it uses its own internal clock.
• The streamer’s SNR, noise floor, and jitter do not pass through.
• The DAC reclocks the USB packets (those little 1s and 0s) and regenerates timing internally
• Therefore, the streamer’s analog‑domain SNR is irrelevant.

(Just call me a broken record)

The article shows that only adaptive USB DACs care about the streamer’s noise

(n.b. an adaptive USB is defined as a DAC that uses the incoming USB data stream’s timing as its clock reference.)

Hi‑Fi News shows a dramatic jitter reduction when using the Rivo with the AudioQuest DragonFly, which is an adaptive USB DAC:

• Jitter drops from 300ps → 135ps because adaptive USB DACs use the incoming USB clock, so the streamer’s noise matters.

This is the opposite of asynchronous USB.

Adaptive USB Is Inferior because the DAC is not in control of timing which leads to higher jitter, more sensitivity to USB noise, audible differences between transports, and greater reliance on the streamer’s electrical purity

This is exactly why the Hi‑Fi News Rivo measurements shows 300ps jitter → 135ps jitter improvement on the DragonFly and almost no change on asynchronous DACs like dCS and Mytek.

Uh, I won’t drop my mic because you may have more questions, young Padawan, which I’m happy to explore with you for your benefit that of those sitting on the sidelines eating popcorn.

This is great.  I love learning new stuff.  (Is that why you called me weird?) Sticks and stones, my friend, lol.  I hope you do too. 

If you still feel like your position is justified, then attack mine by poking holes in my argument.  Don’t just say that I’m wrong or weird. Please explain why. For example, "You’re weird because . . ." or stop digging. It’s ok to be wrong as long as you recognize and admit it.  That’s how me learn.

Hope this helps. 

That's 56 minutes that I'll never get back.

 

Lord.  And you started out so politely (many posts ago).  The irony here is that I actually believe that a good streamer shouldn’t cost much, that the differences among streamers are very subtle, and that in many systems, streamers are largely irrelevant if your DAC effectively rejects jitter, and also does a good job of filtering other noise.  

But the only question we’ve been debating is whether noise from a streamer can affect SNR at the DAC end.  And the answer remains “yes”.  I’m not saying it always does, I’m saying it can.  So for example the iFi dac tested by Miller has the same SNR -110 db - for both the PC as streamer and the Rivo.  That’s a case where you really don’t have to worry about SNR and your streamer.  But that’s not always the case, even with some really expensive DACs.  

Regarding adaptive USB, wrong again, the AQ Dragonfly uses asynchronous USB.  While typically that should result in negligible added jitter, that’s not an absolute.  

Finally, given that you seem to listen to AI more than actual people, here’s an AI response to the question: “Does using asynchronous USB from a server to a DAC eliminate all sources of noise from the server?”

 

“Short answer: No. An asynchronous USB connection does not completely eliminate the possibility of noise from the streamer reaching the DAC, but it reduces some types of timing-related issues (jitter). Other noise paths can still exist.

Let’s break it down. 🔍

What “Asynchronous USB” Actually Fixes:

In asynchronous USB audio, the DAC controls the timing (clock) instead of the streamer/computer.

• The DAC tells the source when to send more data.

• The DAC uses its own internal clock to convert digital → analog.

This helps reduce clock jitter that could occur if the source controlled timing.

So asynchronous USB mainly addresses:

• Timing accuracy

• Jitter from the source clock

It does not isolate electrical noise.

Noise Paths That Can Still Reach the DAC:

Even with asynchronous USB, several noise paths remain:

1. Power Noise via USB 5V

USB cables include a 5-volt power line. A noisy streamer (like a computer or cheap switching PSU) can send:

• switching noise

• RF noise

• ground ripple

into the DAC.

Good DACs often ignore this power line or regulate it internally, but not always perfectly.

2. Ground Noise

USB connects the grounds of both devices.

This can allow: • ground loops

• digital switching noise

• RF interference

to propagate into the DAC circuitry.

3. Radiated RF Noise

Even if the data is clean, the USB cable itself can act like an antenna carrying RF noise into sensitive analog sections of the DAC.

Why Many Modern DACs Handle This Well:

High-quality DACs often include:

• USB isolation chips

• galvanic isolation

• reclocking

• local linear regulation

 

✅ Bottom line:

• Asynchronous USB fixes timing/jitter issues.

• It does not guarantee electrical isolation.

• Noise from the streamer can still reach the DAC, though good DAC designs minimize it.

You can do the AI research yourself on how jitter can come from a streamer using asynchronous USB, but the short answer is:

“why some measurements show streamer differences even with asynchronous USB—the explanation involves RF noise coupling into DAC clocks, which surprises a lot of engineers.”

AI is a great tool when you know how to deploy it and learn how to filter though the information.  I spend my days working with it as part of my work.  When AI says something I don't understand, then I try to verify or ask for clarity.  I also use it to help me better convey a point to help other understand my position.  That's good AI's purpose it to be pimped. 

I'll be brief: 

 

 @mdalton wrote “Lord.  And you started out so politely (many posts ago).”

RESPONSE:  Lol, Lord, yes. I’m very, very sorry if I came across as less than polite.  It’s not my intent to be otherwise.  I hope you know that I was just poking fun. Then again, sometimes it take a few callouses to play in the forums:  

 

You wrote, "

You can do the AI research yourself on how jitter can come from a streamer using asynchronous USB, but the short answer is:

“why some measurements show streamer differences even with asynchronous USB—the explanation involves RF noise coupling into DAC clocks, which surprises a lot of engineers.”

RESPONSE:  FACT:  Jitter does not transfer from a USB‑oriented streamer to an asynchronous DAC.  Not timing jitter. Not clock jitter. Not transport jitter.
It is blocked by design. (MY FACE IS BLUE)

@mdalton wrote “ The irony here is that I actually believe that a good streamer shouldn’t cost much, that the differences among streamers are very subtle, and that in many systems, streamers are largely irrelevant if your DAC effectively rejects jitter, and also does a good job of filtering other noise.”

RESPONSE:  We agree.   A good streamer or DAC doesn't cost much.  I didn't breed.  So, I have no greedy munchkin's waiting for an inheritance.  My last check will bounce with my credit cards maxed out.  You only get so many rides around the sun. Max every day. Life's too short to listen to Schitt. 

@mdalton wrote “But the only question we’ve been debating is whether noise from a streamer can affect SNR at the DAC end.”

(Uggh) Well, not exactly—that’s only part of it.   I believe this started when I challenged helomech’s positing that tube-based systems as incapable of differentiating between streamers.  You chimed in with a thought and it grew from there.  My responses have always been in the context of my DAC—it’s all that interests me because I’m committed to it as my endgame.  It’s the reason that I started the thread and I try to stay on topic. With that said, I believe that much of what I say applies to asynchronous DACs.

(BTW, what you describe in your “Asychronous USB” definition is exactly what I’ve been saying.)

I’m curious, what part of the article that you provided supports your point of view particularly given that the streamer when using asynchronous sends no analog waveform and only data packs containing no analog wave form.  You continue to challenge my position without supporting yours.

@mdalton wrote, that the DAC “. . .does not isolate electrical noise.”

Not true.  I’ve addressed this.  The Rex 3 DAC isolates noise in 3 ways:

1. Asynchronous USB Architecture (DAC is the master clock) - This is the primary noise‑rejection mechanism. (it’s only 1s and 0s)
2. Galvanic (i.e., electrical) through massive power‑supply Isolation (dual‑mono, multi‑stage, high‑current)
3. Fully Differential, Balanced, Tube‑Hybrid Analog Stage
   1. Fully Differential, Balanced, Tube‑Hybrid Analog Stage which reject common, ground-borne, RF and PS ripple noise.
   2. Plus,  the tube hybrid technology natuarally filters HF noise.

Any noise that survives the first two layers is canceled or absorbed in the analog stage.

And finally (yes, finally) I don’t think I ever said that noise was eliminated in other parts of the system (e.g., Amps, Preamp, Cables) even though I have a fine high-current power conditioner, with isolated inputs.  There's always noise in part of othe system.   I acknowledged that “Tubes are noisier than solid state.”  So, I understand that there's noise down the chain.  

BUT, Tell me how noise is getting from the streamer through the USB into an asynchronous DAC?

With that said, no need to answer that.  I already know that you agree with me.  You don't have to say it, because I know this topic is getting old and the gallery is likely tired of me repeating myself.  

Again, I really enjoyed this. 

 

“Sigh”.  When I was growing up, dancing in the end zone was considered bush league behavior.  And dancing before you get to the end zone, just before you fumble the ball thru the end zone is just embarrassing. 

When I respond to someone’s post, I really do try to give them the courtesy of reading and carefully considering everything they’ve written.  I wish you would do the same.  But here goes, one last time:

1) Re analog waveform - I never even responded to this point, because it’s not relevant to the points I was making.

2) Re Rex 3 DAC - I never opined on your specific DAC’s effectiveness at rejecting noise from a streamer.  My responses were always conceptual.  I would note however, that you might want to temper your confidence in your DAC if your views on this topic are informed primarily by the manufacturer’s marketing materials. That’s why, in the absence of effective measurements of such claims, I tend to maintain healthy skepticism, and suggest to other members that they may want to be safe, rather than sorry, and invest in a steamer that’s quieter than a Radpberry Pi, for instance.

3) How is noise getting from a streamer using asynchronous usb into the DAC.  Here, I’m just gonna cut and paste from my prior post in hopes that you’ll read it this time.

“Even with asynchronous USB, several noise paths remain:

1. Power Noise via USB 5V

USB cables include a 5-volt power line. A noisy streamer (like a computer or cheap switching PSU) can send:

• switching noise

• RF noise

• ground ripple

into the DAC.

Good DACs often ignore this power line or regulate it internally, but not always perfectly.

2. Ground Noise

USB connects the grounds of both devices.

This can allow: • ground loops

• digital switching noise

• RF interference

to propagate into the DAC circuitry.

3. Radiated RF Noise

Even if the data is clean, the USB cable itself can act like an antenna carrying RF noise into sensitive analog sections of the DAC.

Why Many Modern DACs Handle This Well:

High-quality DACs often include:

include:

• USB isolation chips

• galvanic isolation

• reclocking

• local linear regulation

 

✅ Bottom line:

• Asynchronous USB fixes timing/jitter issues.

• It does not guarantee electrical isolation.

• Noise from the streamer can still reach the DAC, though good DAC designs minimize it.”