Do USB Filters Make DACs Sound Better? (Video)

[grebg]

Give the facts some time to sink in. It’s hard to break the hypnotic trace of Marketing and Sales claims.

Well, a strict scientific approach would advice us to challenge what we're seeing. Regarding these USB filtering, so far we didn't prove that they're doing something but we didn't prove neither that they're doing nothing. We have no right to think that the measurements made here can cover every aspects of the audio experience, it's still experimental. Again and from a very scientific stand, we just don't know.

I would honestly suggest to continue the research on these device. Sure, it could be a very strong psychoacoustical bias, but also maybe very specific matching with some devices, or even maybe (maybe) a clue that these measurements are not enough. At this stage I wouldn't discard any of these hypothesis, many people (including me) can attest they're experiencing something different than what the measurement are telling

 

[solderdude]

We know very well that some of these devices 'clean' the USB signal or lower common mode currents in certain frequency bands.
This is very easy to measure in an EMC lab and objectively verifiable.
So the vast majority of these devices definitely 'do something'.

That doesn't mean it will always be audible nor that the things they do affect have any influence on the actual analog output voltage of the DAC.

This will depend on: The DAC itself, the circumstances it is used in, the amount of 'unwanted signals' and in what frequency band this is.

Amir's test show what in proper conditions (isn't necessarily everyone's home condition) changes in the actual output signal.
It does not show what it does about common mode issues.

The 'talk' about these devices is how they all are said to improve 'sound quality' which is a bogus claim. Treble will not be 'smoother' lows will not be deeper and sound will not be 'more analog alike'. When something is substantially wrong it will not 'degrade SQ' but you will hear weird noises, hum or have other issues.

This is not shown in Amirs tests as they tested gear may not suffer from a specific 'nasty' or the nasties are not applied. There can be many nasties.

What is shown is that in normal circumstances the actual analog output signal itself did not change. Signal fidelity is not improved yet that's what's being claimed.

The fact that you are experiencing an improvement is most likely caused by you knowing it has been applied.

Do yourself a favor. You have the device so can actually test if it does something in reality. When you really want to know (for yourself) I can explain how to find out. Most people, however, are simply glad it 'works' for them. If you want to continue to experience benefits you should not do these tests.

 

[grebg]

At least we agree that in a perfect environnement and with flawless hardware these devices shouldn’t provide any improvement at all to the sound.
But I think we disagree regarding what are the "normal circumstances" for most people and their audio setup.

In my case, despite decent hardware and a careful setup, the iPurifier filtering device improved significantly the SQ, letting me think that the integrity of the usb signal was affected beforehand. It doesn’t create any magic, but it seems my computer is just not able to properly transfer data to my dac and this specific filtering device seems to correct that.

Maybe there’s just 1% of people like me who have a computer which doesn’t properly communicate with the dac but based on feedbacks (even just on this specific thread) it seems I’m not alone.

Also, there is no glitch nor weird noise if I don’t use the filtering device but following the path is like saying that any digital source is the same because it’s "bit perfect" wether it’s a computer, a dedicated streamer or whatever spdif, usb, aes interface. I think it’s commonly accepted now that different digital source could have different performance regarding the audio restitution. But it’s an other topic in itself on which I hope we don’t disagree

Finally, and regarding your last sentence please don’t discard other personal experience just like that. Everybody has his own background in audio and in life. The "I know better what your experienced than you do" argument has probably no place in a scientific approach.

 

[tvrgeek]

Bought a LTM2884CY medical grade isolator demo board. Not cheap, but this is the real deal.
First just compared the noise on the output of my Focusrite IO 48V phantom.
It works, removing all sorts of hash exceeding 95mV. Down to the internal not very well filtered sawtooth power with 25mV on it. ( says I should make a better phantom supply eh? ) So the USB was injecting noise through to at least the DC-DC converter.
So, thought I would see if it had any effect on the Hanteck scope noise floor.
First catch-22 is it does not have the power. So one would need to build a very good external 5V supply, which it has provisions to feed.
Catch-22 coffin nail. The Hanteck has a residual noise floor of 5 mV no matter what and the lowest scale is 20 mV. The old analog is 2mV/div. So the Hanteck is really only good for digital, not for analog anyway. Bummer.
Now, knowing the isolator does do what it says, when the little boxes get here, I will try on my line powered DACs and see if there is any change either I or Rightmark can measure, but I doubt it as my IO box is only , get that only, -120 dB. I was testing my old laptop so if anything is noisy, it is. I wish I could find the $12 e-bay isolators to see if they have the same effect, but sending the Hanteck back so I'll be scopeless. Maybe I can fix my Tektronix.

 

[tvrgeek]

At least we agree that in a perfect environnement and with flawless hardware these devices shouldn’t provide any improvement at all to the sound.
But I think we disagree regarding what are the "normal circumstances" for most people and their audio setup.

In my case, despite decent hardware and a careful setup, the iPurifier filtering device improved significantly the SQ, letting me think that the integrity of the usb signal was affected beforehand. It doesn’t create any magic, but it seems my computer is just not able to properly transfer data to my dac and this specific filtering device seems to correct that.

Maybe there’s just 1% of people like me who have a computer which doesn’t properly communicate with the dac but based on feedbacks (even just on this specific thread) it seems I’m not alone.

Also, there is no glitch nor weird noise if I don’t use the filtering device but following the path is like saying that any digital source is the same because it’s "bit perfect" wether it’s a computer, a dedicated streamer or whatever spdif, usb, aes interface. I think it’s commonly accepted now that different digital source could have different performance regarding the audio restitution. But it’s an other topic in itself on which I hope we don’t disagree

Finally, and regarding your last sentence please don’t discard other personal experience just like that. Everybody has his own background in audio and in life. The "I know better what your experienced than you do" argument has probably no place in a scientific approach.

I am leaning to how each DAC handles the clock. Sure, USB or PCM are not reliable transport, as in spray and pray. But the bits are buffered and it is the internal clock that manages the jitter. I guess if ASIO, you can reduce the buffer to a very small one but not less than one bit. How bad does the noise have to be to flip a bit? Really bad I think. So we come down to noise bleeding into the analog. On a USB powered DAC, maybe. On a line powered DAC, it should not be a problem in the first place.

Conclusion? Buy a better DAC instead of hoping you can add on something externally to improve a crappy one. A thought at least.