ifi micro iUSB3.0 USB Filter Review

[AudioSceptic]

In defense of @JohnYang1997 , he was pointing out something rather obvious. Just because certain people disagree with some of your testing procedures doesn't mean you are being lectured, Amir.



The linked "review" is just rubbish of course, but he was using a Ciunas D/A with it and we know that is broken by design. Maybe it actually helped!

View attachment 48917

If the difference was enough to change a trumpet from "pain inducing" to "gorgeous", that would be measurable.

 

[digicidal]

If the difference was enough to change a trumpet from "pain inducing" to "gorgeous", that would be measurable.

And IMO well worth almost any amount of money... in fact, if it just made brass instruments tolerable to me (sorry @DonH56 ) I'd pay for it. I try and try, but that's where my ears run for cover... which is funny because a wailing guitar solo - even heavily distorted - in that same range isn't nearly as jarring.

On the other side discussion going on here - I partially agree with the criticism @JohnYang1997 brings up... if I thought it was possible for it to make a meaningful difference in any worthwhile piece of gear... I'd like to see it proven. On the other hand, I'm as positive as @amirm that buying something which costs a few hundred dollars to slightly reduce the "brokenness" of bad gear isn't a recommendation I'd bother pursuing regardless.

Especially when there are multiple DAC and HP amp examples (can't come up with anything else it could help) already reviewed which have a lower retail cost than this "fix" does - but don't have any of the theoretical problems for which it might actually help.

No matter what, the bottom line is that ifi should have the burden of proving efficacy of their solution - @amirm should simply be able to confirm that by replicating their claims. As it is... I think they know best what it does and doesn't actually do... and this review confirms that as well.

 

[somebodyelse]

I wonder if it would clean up the problems with tick and pop noises from the RPi 3 USB when using high sample rates. This was particularly bad when feeding a Topping D10.

Of course, you wouldn't spend $400 to clean this up as it only costs $70 to put together an RPi 4 streamer that doesn't have the USB problem.

But, it's a cheap way to produce noise that this thing might be able to clean up.

Given that the cause of the clicks appears to be the Pi sending data slightly too late under certain circumstances when teh DAC asks for it, it's not the sort of thing an isolator can fix.

 

[LuckyLuke575]

This is a review and detailed measurements of the ifi micro iUSB3.0 USB regenerator, power supply and battery bank. It is on kind loan from a member. It costs US $399. ifi says it was discontinued but then came back to market.

The enclosure is just like everything else ifi makes:

View attachment 48803​

Which is to say it is confusing as heck. I have a full container of ifi products and every time I go looking for one, I have to spend 10 minutes reading the hardly legible print on the back to know which is which.

This thing has more USB outputs than you know what to do with. Power-only, power and data. Port 1. Port 2.

ifi positions this device for people with systems that cost more than $1,500:

View attachment 48804

So I decided to test it with a top of the line DAC, the Topping D90. To see if it can deal with ground loops and such, I decided to test with unbalanced RCA output of it.

USB Decrapifier Audio Measurements
I fired up the dashboard and as before, we see the fantastically low noise and distortion of the Topping D90 by itself, being powered by my very "dirty" desktop USB port:

View attachment 48805

If there is any dirt, it is not getting in the eye of this DAC!

Now let's run the USB through the ifi iUSB3.0 with the ground switch set to floating:

View attachment 48806

Hmmm. We lose half a dB of SINAD due to extra spike around 16 kHz that was not there before. If I flip the switch it goes away and performance remains the same (i.e. the device is doing nothing useful).

Zooming in further, let's look at jitter test:

View attachment 48807

We better see the impact of floating the ground here with the same 16 kHz spike and addition noise around 6 to 12 kHz (little red pulses).

If I ground it once more, then the additional noise goes away but we are left with the same performance as not using the ifi iUSB3.0.

Conclusions
I have measured so many USB regenerators and filters and the outcome is always the same: they either make things worse or do nothing useful. Same is true of micro ifi iUSB3.0. Any audible improvements would have to come from your brain manufacturing it, not due to electrical improvements of your DAC.

Please don't waste money on such things. Remember that not one of these companies shows a measurement of the output of the DAC improvement. Not one. ifi has Audio Precision analyzer. If there was an improvement there, they would have post it. Don't pay someone to excite your imagination for you.

------------
As always, questions, comments, recommendations, etc. are welcome.

The beheaded pink panther is complaining that his arm is getting tired posing this way all the time. He wants either extra pay or massage service. Can't afford either one so please donate what you can using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

These ifi devices are a scam and a robbery.

 

[solderdude]

I'm offended someone is abusing my name for profit..

He was a Gremlin but his name was Mogwai.
His name was (ab)used here
You probably can't connect that amplifier to mains after midnight I reckon, nor is it wise to poor water on it. In bright sunlight you can't see the glow of the tubes.

 

[youduoliang]

View attachment 48807

We better see the impact of floating the ground here with the same 16 kHz spike and addition noise around 6 to 12 kHz (little red pulses).

If I ground it once more, then the additional noise goes away but we are left with the same performance as not using the ifi iUSB3.0.

Conclusions
I have measured so many USB regenerators and filters and the outcome is always the same: they either make things worse or do nothing useful. Same is true of micro ifi iUSB3.0. Any audible improvements would have to come from your brain manufacturing it, not due to electrical improvements of your DAC.

Please don't waste money on such things. Remember that not one of these companies shows a measurement of the output of the DAC improvement. Not one. ifi has Audio Precision analyzer. If there was an improvement there, they would have post it. Don't pay someone to excite your imagination for you.
[/QUOTE]


The final comparison can't support your conclusions. Even you grounded ifi USB 3.0, your usb cable was still using the ifi low noise power supply and ReGen ReClock Rebalance. you should add "USB direct connection between PC and DAC" into that DAC to compare the performance

 

[Veri]

The final comparison can't support your conclusions. Even you grounded ifi USB 3.0, your usb cable was still using the ifi low noise power supply and ReGen ReClock Rebalance.

What? How so? Are you making this up? number of 'regen' mentions in this thread is literally 'zero'

 

[DeepSpace57]

I cannot understand ifi staffs. They are watching this forum, results and brutal comments on ifi products. However i see no reaction by them at all.

Himm! I therefore believe that these stupid expensive products could be scam.

On the other hand, i have an ipurifier 3 usb. It is impossible to say there is no improvement on Sabaj D5.

@amirm , can you use this device on Sabaj D5 if you still have them? We might be observering a different story.

 

[youduoliang]

What? How so? Are you making this up? number of 'regen' mentions in this thread is literally 'zero'

If you never have this product, at least read the spec before typing nonsense words.

 

[youduoliang]

I cannot understand ifi staffs. They are watching this forum, results and brutal comments on ifi products. However i see no reaction by them at all.

Himm! I therefore believe that these stupid expensive products could be scam.

On the other hand, i have an ipurifier 3 usb. It is impossible to say there is no improvement on Sabaj D5.

@amirm , can you use this device on Sabaj D5 if you still have them? We might be observering a different story.

I'm using ifi iUSB3.0 with WW Platinum USB and Shunyata Venom USB cables to connect my dCS Rossini DAC and NUC, the improvement is subtle but positive and noticeable.

 

[majingotan]

I'm using ifi iUSB3.0 with WW Platinum USB and Shunyata Venom USB cables to connect my dCS Rossini DAC and NUC, the improvement is subtle but positive and noticeable.

Maybe your Rossini DAC unit has USB Suspend issue with your NUC so you need the iFi just to prevent the USB module from losing connection due to Windows sleep state. I use the iFi iUSB 3.0 since it prevents the Unison USB module from "sleeping" thus preventing relay clicks from my Bifrost 2 DAC between stopping or pausing music. https://www.audiosciencereview.com/...o-iusb3-0-usb-filter-review.11284/post-321622

As for sonic difference, you have to provide objective evidence in terms of DBT volume matched 19/20 correct choice, and if you do manage to hear a difference under those conditions, then the USB module in your dCS is inherently broken by design just like the USB module in Schiit Modi 2 DAC

 

[BDWoody]

I'm using ifi iUSB3.0...the improvement is subtle but positive and noticeable.

Uh huh...

 

[amirm]

The final comparison can't support your conclusions. Even you grounded ifi USB 3.0, your usb cable was still using the ifi low noise power supply and ReGen ReClock Rebalance. you should add "USB direct connection between PC and DAC" into that DAC to compare the performance

What? There are three comparisons, not two:

1. Just a direct USB cable without ifi being in the loop at all.

2. ifi in the loop with ground floating.

3. ifi in the loop with ground not floating.

#2 made things worse. #3 just did nothing better than #1.

So at best, this device does nothing. At worse you get some extra interference you did not have before.

 

[amirm]

I'm using ifi iUSB3.0 with WW Platinum USB and Shunyata Venom USB cables to connect my dCS Rossini DAC and NUC, the improvement is subtle but positive and noticeable.

It always is even if nothing has changed in the sound waves coming out of the product. You seem to think your hearing is more reliable in remembering things better than an instrument. If so, what sound did you hear yesterday at 2:31 pm? And 2:32 pm?

Before spending so much stuff on things that don't have merit, have a loved one test you with and without this gear 10 times. See if you get it 8 times right. If not, then you did not hear anything. Your brain assumed you did.

 

[headmania]

Hi Everyone. I believe that this product might be useful for some. I actually experienced a very annoying issue recently. If you don't have a good USB output from your computer, you might be getting some noise through it in your system (if you have a sensible dac or power amp). I experienced this recently (I might need to change my mini pc for audio purposes), as the usb seems to suck.... With microusb 3.0 I got rid of this issue.

 

[doitttt]

clean power, to usbport on a laptop
i work ,all ifi crap are not bad.
i use iDefender+ with ipower powersupply.
disconnects the usb port, from the laptop, uses the external ipower 5v, it can be heard
maybe amir testing with clean power

 

[snowsurfer]

clean power, to usbport on a laptop
i work ,all ifi crap are not bad.
i use iDefender+ with ipower powersupply.
disconnects the usb port, from the laptop, uses the external ipower 5v, it can be heard
maybe amir testing with clean power

I had a ground loop on my W10 PC, which I managed to isolate to my case with a multimeter and fix using a 15€ Adum based ground loop isolator. When it's inline, no groundloop. I take it out, it's horrendous again. I'm feeding it power from an external USB linear power supply (it was OK without this, I just had one around). This has a Talema toroidal 15w transformer and costs about 50-60€, but it was not necessary. A 15€ no-name, generic Adum based design did the trick.

I guess you need to have the problem before you can fix it, but I am 100% certain there is no need to spend serious cash on this. I see products like this and I immediately ignore the manufacturer forever. 400 bucks? Lol...just lol...

 

[Dnix]

I'd be curious to see what the measurements would be measuring what is going into the dac/out off the ifi device. The way this test is set up, you are also testing the filtering of the d90. (this would allow us to see if the device actually does something, as this current test shows doesn't show JUST the devices filtering, it does show that for dacs with already good filtering, this doesnt appear to help any)

Another thing that I have noticed "missing" from almost all of your measurements, is impedance. both on power and data signals. This is where "changes in sound" for certain things tend to happen, and its a measurement that you dont seem to include in your reviews. (I think this is where people "hear" differences in cables, I haven't really noticed this myself, but most of my cables are fairly similar in resistance and capacitance.)

This is just food for thought. maybe you would see a measurable difference with impedance. of course even if it does measure differently, the subjective test is really what matters to people and can only be done on their ears. As while measurements are a great way to get a basic idea of how good a product may be for your use case, it is not an end all be all, as some people prefer different sounds (I.E. I prefer the sound of the modi multibit over the sound of the tone2 pro, and the tone2 measures WAY better.)

 

[Veri]

Another thing that I have noticed "missing" from almost all of your measurements, is impedance. both on power and data signals. This is where "changes in sound" for certain things tend to happen

If the current set of measurements don't show that difference/those things that "happen", what would measuring the cable's impedance accomplish, exactly?

 

[Dnix]

If the current set of measurements don't show that difference/those things that "happen", what would measuring the cable's impedance accomplish, exactly?

Think along the lines of input/output impedance. Say you have a HP amp that has a high output impedance and are using low impedance headphones. On most headphone this produces a very audible change in the sound signature of the transducers. There is a headphone amplifier, that i cannot recall the name of currently, that has a low z and high z switch. There are reviews of this online showing the difference of sound signature between the two different impedance outputs. The cables example was just that, an example. Cables are usually all very close in impedance, so audible change *should* be minimal. Impedance on the power side also makes a difference as far a voltage stability. Higher impedance power outputs tend to have more stable voltages when current draw changes. I have tested that to some extent, testing power packs, SMPS, and a cheap 5v 1a LPS. I plan to test attempting to increase the impedance of battery power packs by use of an LC filter. This would also act as one more power filtering stage.

Objectively, the only way that I can see people hearing an actual change in sound from any decently made cables(again, just an example), is the impedance variations of the cables (other than conformation bias, etc). So testing the impedance of a circuit when measuring things, to me anyway, would seem like where you would see what the "audible" difference in certain products, and would be the most useful measurement for those things. Please keep in mind, I am not saying I am completely right here, but it is just what seems to be happening when things can objectively measure the same (I have not seen a measurement review outside of transducers themselves and input/output impedances of line level input/outputs that has impedance measurements), but produce subjective differences (again, ignoring conformation bias, etc).

EDIT: Although come to think of it, on the non-power side, impedance changes via cables, or high/low z switch, etc., should show up on frequency response measurements. Differences in cables would likely be below the margin of error of measurements unless the cable drastically changes the impedance.