Do we still need "audiophile" USB gadgets and bridges?


Recently I bought a Khadas Toneboard at around 89€!!.
There had been a review on AudioScienceReview which got me hooked. I was curious what the hype was all about.

It all turned out to be a really nice solution. Sideeffect: I got rid of IansReclocker, all kind of HAT DACs I had around, asf.

What a relief! :D

I'm now asking myself if we finally made it.

I always said. If the (USB) transport setup makes a difference on perceived soundquality, it's the DAC which is to blame.
Pretty much all USB DACs I had in the past made a difference, when changing stuff on the transport side. That's why I switched to the I2S-HATs and left the USB Audio world.

There seems to be progress!

We now get insane low noise,THD and jitter levels on USB DACs without using any kind of USB filter gadgets or any kind of audiophile USB transports.

Looking at e.g. the upcoming Soncoz products or products like a Topping D90 DAC all rather reasonably priced, the situation even gets better. Soncoz ( the Khadas Toneboard designer is behind that new company) plans to launch two USB DACs in February. The small one will hit the 200$ range if I'm not mistaken and supposedly outclasses the Khadas Toneboard.

To me it seems that USB DACs become commodity products soon. I'd love it.

Great times.

... Colour me sceptical ...

I'm a long time member of that club. And still am. ;)


The AudioScienceReview reviews - which are actually unified measurements - outline a very nice picture though.
And Amir - the guy running that blog/forum - uses top of the line AudioPrecision gear for the measurements.

I do not agree to all of his conclusion though.

He reviews the iFi USB filter. That one reduces noise by let say around 30dB - confirmed by his measurements.
Then he measures his high quality DAC with the iFi in and out and finds no impact.
He concludes the iFi filter is useless. You see the flaw in his logic?

The iFi 100% delivers to its promises - it filters noise. A 5* rating would be a fair verdict.
However. He uses a high quality DAC which copes with USB noise very well.
Basically the whole testcase leading to his conclusion doesn't make any sense.
The result just confirms the quality of the DAC he's been using and says pretty much nothing about the iFi device.


To understand his measurements I recommend to read his "understanding digital measurements" article.

And it's IMO amazing that my 89€ Khadas - as USB DAC - equals the
Allo Katana - both @ 110dB Sinad and sits right beside a Benchmark DAC3 with 112dB Sinad.

These are both excellent performances.

The big Q is:
Is there anything that all these measurements won't cover properly? ""call me sceptical""


As I said. That Khadas DAC sounds extreme clean and right to me.
From that perspective I am pretty much in line with the shown measurements.

And the Soncoz LA-QXD1 pre-production unit seems to better the Toneboard on several aspects already.

Again. These are great times. :D

One thing we need to keep in mind though.

Many DIY efforts on these kind of devices will usually lead to degradation. I think it was the Toneboard/Soncoz designer who mentioned that you can't achieve extreme quality levels by
manually soldering parts to the board ( a comment about his pre-production board).
For us DIY geeks tweaking gets tough. Especially since pretty much non of us can really do the required measurements to confirm the tweak result. And the risk that you rather make things worse, than doing any good is a real risk.

Last edited:
This post tries to answer my earlier question "Is there anything that all these measurements won't cover properly?

One thing that becomes more prominent though are the digital filters and upsampling.
I havn't really looked into how much impact these features that pretty much every DAC has inside
would have on these standard measurements.
I do know that they can make quite some audible differences, see e.g. FilterBrewingThread, I saw poor rolloff behavior of minimum phase filters, asf. asf.

Gustard and many others did realize that there's more than just great standard measurements.

The top of the line DACs from Gustard are marketed with 64bit DS processing, high quality custom filters and top of the line upsampling.

I am pretty sure that DAC devices using the DAC chips internal default filters might fall short on exactly that aspect.

IMO ASR falls short on looking more closely on that DSP aspect. (let me know if I simply missed it ;) ).
Last edited:
The USB bus is just a means of moving data around. USB power is notoriously dirty, so don't power your DAC with it, or be prepared to filter, filter, filter as part of the design. Once USB gets the data to its destination, the DAC can do its thing. Look for a DAC that has external power, or one that is designed well to begin with vis a vis power and noise. There is no need to go the "HAT" route per se, although there are now some excellent options with that kind of hardware. I find USB devices to be more flexible, since I can move them to e.g. a desktop computer with no I2S/GPIOs exposed. I should probably add here that I am not interested in DSD or very high samples rate playback.

Case in point, I just spent a few days getting the new MOTU M4 to play nice with Ubuntu. I had to do a couple of kernel mods but it wasn't anything more than tedious (recompiling takes a couple of hours). The M4 is a USB bus powered pro audio interface that run at up to 32bits/192k. It has noise floor and distortion specs that are on par with e.g. the Focusrite Clarett product lineup, and inside, the DACs are ESS Sabre32 Ultra types. The M4 costs only US$220 for four outputs, and there is an M2 (two outputs) for around $170. This is an incredibly good deal, and I bought a pair for left/right PC processing/crossover work of up to a 4-way stereo system.

More info here:
MOTU M2 Info and Specs
MOTU M4 Info and Specs
Last edited:
...Once USB gets the data to its destination, the DAC can do its thing...

Look Charlie.

You really didn't get the point.

Since more than a decade the vast majority of USB-DACs "didn't do their things" properly. Noise, power, timing and whatever kind of flaws of the transport side
somehow made it to the audio output. That's been subject to discussion numerous times.
Some DACs were and are coping better then others. DACs did improve over the years.

But e.g. even in ""2019"" Allo decided to launch a USB interface that supposedly improves the more than disatisfactory USB situation.
Like e.g. the Uptone Regen or all the iFi gadgets and filters, these devices follow the logic to clean the mess before it hits the DAC.
Simply because many DACs were and are not ""doing their thing properly"".

And to quote analog_as, as another example " sensitive is a MSB Select II to upstream components..." . I think that's a DAC that sells @ insane close to 90000$. :eek:

That meant and still means, even with the same DAC at play, people can't rely on its performance.
Because its performance depends on and varies with the upstream environment.

If now even budget USB-DACs are starting to reach the measurement limitations of top-of-the-line measurement equipment under "normal" USB conditions,
I'm pretty optimistic that upstream chains/devices are potentially going to have a pretty low impact on the sound experience.

For a starter. Even my budget Khadas device sounds very clean already. ;)

But. Again. Let see how things develop.
Last edited:
I have found this to be true.

I used to blame USB for everything bad and avoided it.

No longer do I think this way.

My latest computer setup is pure USB - input, which is externally powered - not using motherboard power is better, plainly better. I also use an USB SD reader - SATA is disabled.

After many weeks of using this system both using the USB SD card and using a SATA drive with SATA engaged there is no question - the USB SD card is superior.

I do think there is something to the SD card and the fact it leaves the data where it is initially placed might be advantageous for audio but I think using only USB makes a difference.

I am using the ASROCK J3455 motherboard with the HDplex 200 watts supply. I am using the MATRIX X-Sabre dac and I am hearing clean high frequencies I had never experienced digitally before along with the wonderful absence of record noise. I am using external supplies for the SD drives and the input for the DAC. There are great little adapters that are easy to modify to use external power and use the motherboard header.

I did not hear a great improvement when using these to power the SD drives but i intend to build a SALAS L Adapter to substitute for the 19 volts SMPS I got with the HDPLEX and wanted to minimize the power needs of the motherboard. Other than the thumb drive for loading WTF player there is nothing using the supply but the basic motherboard.

The SMPS makes noises through an AM radio in the adjoining kitchen. I can hear it squealing with my ear close to it. Worth a try. I figure if the noise affects the AM radio it is doing more than that though, at the moment, I hear no effect on the system but it could be one of those things that one cannot hear it until it is not there an longer.

I have far more LPs than CDs but I can see that digital is getting to be very good. The number of advantages for digital is getting longer and longer and the list for LPs is getting smaller.

Also have purchased a few 24/192 and 24/96 recordings and have found these to be very interesting. Always thought the whole high resolution thing was ridiculous.

I love finding out I am wrong.

I think we're talking about two different subjects here.

What I'm saying is that all these (upstream) PC tweaks should NOT cause any audible differences anymore considering all these USB audio devices and related measurements
proving performances beyond any audible thresholds.

And therefore all your SD tweaks or power supply tweaks or whatever other tweaks basically won't (shouldn't) matter anymore.


HiRez audio files is a complete different story.

In most cases you don't know if you have a master in front of you.
Usually that's not the case. You'd basically listen to already resampled material.

Since most HiRez tracks are remastered tracks, you basically have no reference
to compare it to. How do you want to associate the changes you hear to HiRez.
It could very well be just a different remastered track that makes the difference.
We know that from remastered 44.1 CDs. Every release sounds vastly different.

Of course 24 vs 16 makes a difference. But that has nothing to do with HiRez samplerates.

The DAC HW and internal SW makes a difference as well. That will cause
different conclusions about HiRez depending on the DAC you're using.

And there's more.

IMO the whole HireZ, DSD arena is a jungle.

Last edited:
I did not intend, though i can see that i did, link high resolution with what you were saying. It was supposed to be an aside.

I was attempting to agree that DACs and maybe motherboards, too, have made the USB gadgets superfluous. And there is no question USB is a good way to send music to and from the computer.

Yes, I have found the high resolution jungle to be very bizarre. I have only purchased four recordings that I am sure came from real masters. I will never buy them without some research since the storage requirements are rather grand. One track as large as a whole CD!

The vast majority of my listening is to 16/44 from CDs to the SD. As you know there is great variability in what you get. Some of these recordings are incredible and some are tolerable to a music lover. Certainly there is far more it it than sample rate and number of samples, no question of that.

I can smell from here the next step... an usb stuff that feed a sdCard with a lipo battery in between with a Bebee quantic thingy in between the in between...powered by the last small solar cell then advertised on a forum with number of clicks counter to feed the bank account of the posters.... The New World at the corner of Beethoven street...

Yes, there are progress, yes some of those progress are pure half facts as the layout and setuping are still important whatever the guy behind the last brandnew company and fall into the marketing talks as the laws of electronics stay the same as the EE schools...I mean the power is in the hand of the chips designers and dev labs.
As an illustration, lipo cells for instance are not especially better than a good power supply for all the stuffs around dacs but special rare case (go figure at the end when listening to) and so on. Plus the fact we are biased, also in the way we consider progress.

For me the true question is (as figured out already) what do you hear better? Is it the dac, the filter, the usb improvment, the applied filter ?? I mean each DAC is a whole experience when you listen to it but listenings kits as the Soekris' which don't fall at the end of the day as a good music stuff enough:eek:, i still remenber also the fifo and isolator cnips to be the ending story of all our digital pains...and no joke, the usb cable whatever Ifi gadgets were wtill hearable on a good hifi systems (btw for the worst or the best) or a little cap between the fifo and the the further clock improving it all while not being in the analog domain and playing on high frequencies of the digital side of the board....what a disapointment. So when a brand or the last Dac is enlighted in such threads, what is behind the better sound, is it really the usb subect, something else ??

At the end we still are rulled by the ADC chain quality in the studios, our listen rooms as main quality factors.
But yes there are progress, high rez, news stuffs...neverending story since I'm fond of music and hifi...37 years perhaps.
Oh listen, do you not hear the google quantic dac chip comming which will end the 0 and 1 and clocks nightmare....or will the zero and 1 reclocked by new crystals materials and datas stocked in the dac chips by revolutionnaries nano vegetables caps made by Monsanto (wait bio version will sound even better :clown: ).... doesn't matter, there is soon only rap and two weeks lasting artists soon:(

By the way give me the link for the last usb stuff, i want one if it solves usb stream problems...but I doubt :D
While one day I have no doubt we will suceed whatever the stream source, network, local stream from a close memory, wireless... that will mean also good electronics components at democratic price ... but hifi without marketing is it still hifi ;)... your true point imho is more about improving budget audio... which certainly is a good cause btw.
Well, duh,

There is far more differences from the recording itself than any hardware changes we can come up with. No question what makes a good sounding recording is full of mystery and good luck. If anyone could figure it out there would be nothing but good sounding recordings.

So with that out of the way we are left to do the best we can with what is available.

In my experience I find that soundcheck's speculation is true - but whatever means - whether the motherboards are getting better or the USB input DACs are getting better and more likely both that the gadgets are superfluous.

I am not using any USB gadget. I am simply using external power from a good supply instead of computer power for the XMOS input on the DAC. It makes a significant, anyone can hear it, difference.

Powering the SD card readers - either subtle or nonexistent - but retained since I want to use a linear supply to replace the smps brick to power the motherboard. I know that there are still a myriad of switching dc-dc converters to remain but the brick creates noise on a nearby AM radio and I figure that has to mean there is noise possibly interacting in a bad way with the rest of the system.

I need to minimize the load on that replacement brick. Standard HDPlex brick is 19 volts and 10 amps - I will use a SALAS L Adapter configured for 19 volts and 7 amps. Will be interesting to hear if it does anything.

I agree, the age of the USB improving gadgets has come to an end.

But I continue to think low power computing is very important for using the computer as a music source.
We agree each others and have the same experiences, but I still think as you do than power supply is still in it. I can't agree more about your conclusions from the sd card powering.

But I'm also sure than an old usb interli k with good optimization will sound better than a brandnew one but with a poor layout... it's still a whole as it was... the maybe only true things is it's becomming wallet democratic... good, good. I agree with Soundcheck than most of diyers are reaching a limit in the way they can tweak. Some parts of the cake are hard too coock with no professional expensive gears or standalone knowledge...EMI, etc.
Basicly, digital sounds better todays than my first 1987 cd player from Philips.... and much more affordable.

Some micro sd card are above 512 Go and even more for non micro sd card...something that maybe go towards software library optimization ? A little like what EC Design did with its usb transport with a library soft management ? Good by the rpi ? We still hear the RPI despite progress like the now old rpi fifo hat...
Last edited:
Btw, here a said very good devices according some observers and reviews (while I always doubt about the objectivity) : MP 8

This "transport" from the german T+A hifi is said to receive all the possible inputs and load an external DAC via USB and rj45 cable for dsd. reviewers noticed no differences whatever using analog or digital inputs whatever the cable and same when using the embeded cd-player...
I don't find any details about the chips used. It was always reviewed with its DAC sibbling for the reviews so a grain of salt would be carrefull...

I noticed some usb receiver chip has now up to 1024 k/b ram for their input buffer ??? Is it the major progress added with low current signal source (à la sd card) ??