USB Nervosa Thread Decrapifiers, pro interfaces, and bears oh my

If you look for ways to be insulted, you'll surely find them.

My "at the level you're at" comment was in relation to the potential return on investment. Spending $340 on tweaks to improve a $249 DAC doesn't make sense. Doing the same for a $2300 DAC is another matter entirely. Though, either way, you don't need the iFi devices with the F-1. It's an either/or situation simply down to the way those items work.

 

Please get over yourself. You're reading malice where none exists.

 

Ah looks like you did not listen to my advice. Like said, being friendly does not mean being kind.

 

Okay, the reason the signals getting to the F-1 matter are the same reasons why USB cables can make a difference in sq. We aren't talking about simple zeros and ones. USB audio is not a guaranteed delivery protocol like you get with files transfers or other normal computer operations. If something happens to the data sent to the F-1, it's not getting fixed and the F-1 cannot reconstitute it. It can only pass along what it gets. Dirty 5v power increases the likelihood of altered data being passed along the USB chain to the F-1. This is why providing extremely clean 5v power helps sq. Anything that can affect an analog cable can affect a USB cable. The digital signal is really an analog signal with certain voltages being zero and others being one. It's all about reducing the errors that cannot be fixed....

 

What does a dirty USB signal sound like to you? In your experience how specifically does it reduce SQ?

 

If you'd read this thread (and understood much of the technical content herein) you'd already know that most of us get how USB audio actually works. A good few of us (me included) have implemented it ourselves and/or been responsible for significant projects (e.g. implementations that other major stuff/projects depends on) that have had to deal with it's various issues (either for audio only or audio and other contexts).

You're correct that there's no ECC or retransmit capability for UAC 2.0. However, bit-level data corruption on an in-spec USB cable is not common. You can reliably send and verify data via a USB 2.0 Audio pipeline, to a USB->S/PDIF converter, and then into an S/PDIF->USB converter and read the data back all day and never see a bit-level error. They occur, rarely, but they're not a major factor nor much of a cause for concern with IN-SPEC cables (lots of "audiophile" cables are NOT in-spec and cause all sorts of issues). @atomicbob has a nice post here demonstrating that, somewhere.

No, actually it doesn't - that claim is completely incorrect and more the perspective of marketing than engineering.

Noise on the USB power-lines can DIRECTLY affect very sensitive analog electronics, like the output registers or buffers on a DAC chip. There is not NEARLY enough energy in that noise to shift levels enough to cause digital-misrpresentation of 0 and 1 on the USB data transmission line and confuse the discriminator on a USB PHY. This proves out in tests time and time again ... or we'd all be running power-filtering and super-quiet power-supplies on our USB interfaces. And the fact is, we don't. Nor do we see lots of calls for re-transmission when everything is in-spec, or, again, we'd do something about that.

The F-1 is perfectly capable of discriminating 0 from 1 no matter how noisy the power-lines are (if they get noisy enough for it to not be able to do that it will fail to operate all together). And with that taken care of, all that matters from a "noisy power" perspective is that the electrical output is isolated.. It is. Job done.

The iFi units, and similar devices, can be somewhat useful if the entire chain is USB.

They are fundamentally pointless with a competent, electrically isolating, DDC in the chain (as long as it's not USB-out).

...

No one here is trying to sell you anything ... in fact the reverse is true ... we'd rather you save money by not spending it on situationally pointless tat.

 

Introducing yourself as the tenet of this forum requests, lurking a bit and using google can be very helpful in getting off to a decent start here. I am going to address the "altered data" premise again. I have performed the null test many times and found that unless terribly deficient, this is not an issue. When it does become an issue dropouts are experienced. No subtle alteration of the data. Repeated here for your convenience though it may be found in the thread from which they are quoted:

 

atomicbob,

That is useful information!

What have you test results had to say about jitter introduced on the USB side of a DDC and how a quality DDC deals with it? Is the jitter greatly reduced by the DDC? What have you seen cause jitter on the USB side? How have you mitigated that, if at all?

Thanks!

 

You don't. USB audio is re-timed at the receiver end anyway, making USB reclockers superfluous snakeoil. If the USB is disturbed enough to cause underruns, that's a gross error from a faulty chain that can't be addressed with audiophile magical thinking.

Its not like SPDIF or similar where you are dependent on the quality of the clock used by the sender. All you can do with USB is clean up some of the analogue domain noise coming down the cable. Thereafter, you are completely at the mercy of the clock in the USB receiver, hoping that it isn't total shit. USB audio is asynchronous, being very slightly buffered and then clocked out after being received. Reclocking USB audio is like carefully sorting a load of letters before just dropping them in a mail sack- a waste of effort.

Your Singxter appears to cover both of these bases to an extent already, so all the the other tchotchkes that you are futzing with at best don't help here- and at worst could degrade your chain by simply putting more things that could fail in the signal path for no good reason.

People are being blunt, but actually they're trying untangle this mess for a somewhat cocky stranger that they've never met. They're weird. They're trying to stop you:

• Wasting money
  
• Degrading your chain
• Looking like an arse

Consider not fighting them tooth and nail while they are doing so. Alternatively, if all you want is a predatory shop window coupled to an ignorance-based echo chamber, have you considered Head-fi?

Your call.

 

Somebody que the @MuppetFace neighborhood bar analogy for our new friend.

It reminds me of a live chicken wandering into a den full of bobcats. Lots of squawking from the chicken about the incivility of bobcats. @winders , your plan to teach the bobcats and change their culture on your first day has been oft-repeated with pretty uniform outcomes. Here's a clue... The chicken loses.

 

Having read the last several posts this thread came immediately to mind...
http://www.superbestaudiofriends.org/index.php?threads/gear-you-wish-existed.3909/

The piece of gear I wish existed after having read all this is a USB audio analyzer that you could place between your dac and the rest of the chain to measure "signal quality" based on playing some reference track. Obviously there is no single measure for "quality" but even some way to observe whatever measurable changes are occurring as you change equipment, cables, PC/MAC settings, software etc. would be interesting... think REW for the USB signal. Anything objective to add a bit of real science to support or refute subjective observations would be great.

I have an Uptone Audio Regen that I quite like but it is fair to say that my PC is far from optimal and the Regen is doing an ok job of addressing "low lying fruit". Still, it would be nice to know what is really happening via measurable differences.

 

There it is in a nutshell, and with that plainly laid out I will have to admit that I have started to fall out of love with the variety of DDC/decrapifier units I have been using over the past while (including the Singxer).

For the most part I have found that these are all capable of subtly changing elements of the sound, but I can't say that it is any better than the built in USB of a well made DAC. Often it is not, and I wonder why I would think that SPDIF would get me much farther on DACs that are doing internal conversions to avoid jitter anyways.

If anyone is truly having audible issues with USB I would look elsewhere for solutions. By far the best change in my setup I have made in recent times was to add a low ripple and silent PSU to my PC.

 

There's no reason why a good USB to SPDIF converter can't appreciably improve things. If it has good clocks and electrical isolation, and works reliably, then that's most of the battle. The USB input of a lot of DACs is a bit of an off the shelf afterthought, after all.

The only caveat is that if you find yourself spending more than the DAC's worth on decrapifiers or converters, you have to ask yourself if you chose wisely. If you find yourself buying a high-end converter for a cheap DAC, you should probably have put that money towards a better DAC.

(Given that you can build a surprisingly good streamer for about half the cost of a Modi Multibit, which will feed it clean, low-jitter SPDIF which will noticably improve how it sounds, this isn't a hard bullet to dodge.)

 

@winders , the following is the very first post on this thread.

 

I've always looked at audio accessories like this: is there noise? Is there interference? If not, f**k the bullshit! I run a Belkin USB 2.0 cable from my pc to my Gungnir Multibit, into my Kenzie Amp and Mjolnir 2. There's no interference or drop outs. Don't buy accessories in search of a problem; solve actual issues.

 

I used to have a Singxer SU-1. At the time I thought it gave a decent improvement for the money. Then I realized putting a Wyrd in front of it made it better, even though it probably shouldn't. At that point I said screw this usb stuff and got a rednet. The improvement in sound quality going from SU-1 to rednet was fairly substantial, particularly in clarity. Before I was listening though a usb haze and didn't even know it.

It's not worth spending any more than a minimal amount of money trying to fix or convert usb. Instead put that money towards a rednet or lynx and ditch usb altogether.

(this is of course assuming that the rest of your system is at the point where spending money on this makes sense)

I've come to the conclusion that source is an important part of the chain and not just an accessory. Spending money to improve sound quality is as valid a reason as fixing obvious problems. I mean if that weren't the case, why bother buying amps like the Kenzie or Mjolnir2? A Magni is only $99 and won't have drop outs.

 

I concur.
if the source is hampered and so is unable to deliver the SQ it is fully capable of, then that IS what you'll hear.
And much like what 800's do, in that they will deliver the signal fed them with precision and alacrity, delivering the digital audio stream 'cleanly' seems to be a significant factor in all of this.

JJ

 

Tried the iSilencer with my iPhone 6s + DragonFly Red + Sony XBA-Z5 set up. In a nutshell, the iSilencer makes the Jitterbug a thing (or critter) of the past. Lyrics becomes legible, midrange is more organised and focused, dynamic range increased. Amazing little widget.

It used to be the Jitterbug, now it is iFi Audio's iSilencer. Get it.

 

"Fighter jet technology now cornerstone"
​

Does this madness ever stop?! But wait there's more (insert drum roll here):

"More than one iSilencer3.0® brings exponential benefit"

followed by ...

"Multiple ANC® circuits will simply double the noise reduction each time the number of ANC® units is doubled. So 2 x iSilencer3.0® will kill double the noise of a single-one and four for example will kill four times the noise of a single one; if they are plugged into the same block of USB Ports and even though only one iSilencer3.0® has one device (e.g. DAC) connected into it."

I'd say that the only thing that doubles is the ka-tching in iFi's piggy bank each time an iSilencer is sold....

I give up. Good night.​

 

That is not exponential, it is linear. I think we need a Thread Decrapifier here.