Topping DX5II Balanced DAC and Headphone Amp Review

[ziggurcat]

I understand, but what I mean is that "always on" option for the Windows driver does the same thing basically right? So there is no limit for Topping to release such a software on Mac. Unless the driver does something else too on Windows?

It is arguably the same concept, yes.

There is absolutely no limit for Topping, and they could probably do it with the release of a MacOS version of the XMOS Audio Control Panel.

 

[CuriousAli]

I don't have the Topping DX5 II so can't try it, but I tried all the zero audio loop playback methods and eventually the delay/pop still happened with my Fiio KA17 on macOS 26.1.
I wonder if you guys with the Topping and Mac has had any success with playing zeros as a workaround?
This is an example of a silent wav file made with @JIW's script:

zeros-96.wav

drive.google.com

You can give it to your audio player of choice and set it to loop/repeat and see if still delays or pops happen.

 

[xlinknz]

Hi all, I'm considering one of these to modernise my vintage setup (2 channel pre+power amps with no digital inputs and no headphone socket) so I'm keen for feedback on my proposed setup so I know anything I need to consider plus a couple of questions

• Coax Input: Sony Blueray/CD/SACD that has no DAC
• Optical input: miniDSP PockDAC which will connect to RCA Tape output on the pre amp to enabling headphone listening to a TurnTable
• RCA output: to before mentioned pre amp (or maybe power amp see below)
• Unbalanced Headphone: Sennheiser HD560s (yet to be purchased and plan to extend to 4m length)
• USB: To play back from a laptop DSD files ripped from my SACD using the likes of Foobar2000

Questions:

1. Can the DX5 II be used as a pre amp i.e. direct to a power amp (noted that I would loose turn table)
2. Is there a new HW version of the DX5 II anticipated?
3. Are there other DAC+headphone in this price range that are compatible in function and cost but also have a display that can show VU, EQ etc (not I'm not a user of PEQ yet)

 

[ziggurcat]

I don't have the Topping DX5 II so can't try it, but I tried all the zero audio loop playback methods and eventually the delay/pop still happened with my Fiio KA17 on macOS 26.1.
I wonder if you guys with the Topping and Mac has had any success with playing zeros as a workaround?
This is an example of a silent wav file made with @JIW's script:

zeros-96.wav

drive.google.com

You can give it to your audio player of choice and set it to loop/repeat and see if still delays or pops happen.

The closest thing to that I have tried is Qobuz in the background with the DX5 II set as the output device. No pops whatsoever using the USB input on the DX5 II while the app is running on Tahoe 26.1.

Questions:

1. Can the DX5 II be used as a pre amp i.e. direct to a power amp (noted that I would loose turn table)
2. Is there a new HW version of the DX5 II anticipated?
3. Are there other DAC+headphone in this price range that are compatible in function and cost but also have a display that can show VU, EQ etc (not I'm not a user of PEQ yet)

1. To a certain extent, yes. It does have a Preamp mode in the settings, but it only has digital inputs (USB, TOSLINK, COAX).
2. No, there's been no need for a HW revision.
3. WiiM Ultra, I think.

 

[Berwhale]

It is arguably the same concept, yes.

There is absolutely no limit for Topping, and they could probably do it with the release of a MacOS version of the XMOS Audio Control Panel.

Topping didn't write the XMOS driver or control panel for Windows, neither did XMOS, it's written and maintained by Thesycon. The DAC manufacturers license the software from Thesycon who then provide a branded version for each manufacturer.

My understanding is that the driver model is very different for MacOS. It's not possible for a third party to write and distribute a low level driver like you can in Windows. Topping have already explained this in the correspondence you shared.

This whole topic relates to a long running* argument about the different philosophies of Unix and Windows operating systems. I won't go into that here, but you can Google 'kernel vs user space windows vs unix' if you are interested.

*I am reminded of a session I attended at Microsoft's TechEd Europe 2004 in Amsterdam. The presentation was given by Mark Russinovich about the differences between Windows and Linux. The presentation is on Youtube:

Mark had input to the presentation from both Dave Cutler, the architect of Windows NT, and Linus Torvolds, the architect of Linux. There was a picture of the three of them + Bill Gates posted in the internet a few months ago...

Sorry for the OT, back to DACs!

 

[CuriousAli]

The closest thing to that I have tried is Qobuz in the background with the DX5 II set as the output device. No pops whatsoever using the USB input on the DX5 II while the app is running on Tahoe 26.1.

I didn't do extensive testing but Qobuz seems to do the trick for my Fiio KA17 too, no more delays or pops in a few minutes of test.

What I have been trying to say for the past few posts is this:
1) If Qobuz or another user-space app can prevent this delay/pop, Topping and others can release a similar tool.
2) You *can* create system level drivers for Mac using kexts, or user-space level system extensions, but they may not even be necessary for this.

What is really happening in here in my opinion is Thesycon knows what they are doing to prevent this issue, but they are only releasing their software for Windows, and Topping doesn't know how to do it for other platforms.

 

[Ahmonge]

Optical input: miniDSP PockDAC which will connect to RCA Tape output on the pre amp to enabling headphone listening to a TurnTable

If you choose a configuration with two devices (DAC and headphone amplifier) that have XLR and RCA outputs (DAC) and similar inputs (amplifier), you can connect the DAC to the headphone amplifier via XLR and the preamplifier's tape output to the RCA input, allowing you to listen to the turntable on headphones without an analog > digital > analog conversion that will degrade the sound.

At Topping, you can choose from several combinations, such as E50 + L50 or E70 + L70.

 

[Berwhale]

I didn't do extensive testing but Qobuz seems to do the trick for my Fiio KA17 too, no more delays or pops in a few minutes of test.

What I have been trying to say for the past few posts is this:
1) If Qobuz or another user-space app can prevent this delay/pop, Topping and others can release a similar tool.
2) You *can* create system level drivers for Mac using kexts, or user-space level system extensions, but they may not even be necessary for this.

What is really happening in here in my opinion is Thesycon knows what they are doing to prevent this issue, but they are only releasing their software for Windows, and Topping doesn't know how to do it for other platforms.

I don't think we can blame Thesycon. The XMOS chip in these DACs is supposed to be UAC 2.0 compliant and should work with the default OS UAC 2.0 drivers. The Thesycon Windows driver is only required to add support for ASIO and provide some other extras over and above what is handled by UAC 2.0.

So why do DACs with XMOS chips exhibit this behaviour? And why is it only present (or audibly present) on only a subset of these DACS?

 

[CuriousAli]

I don't think we can blame Thesycon.

I agree, it is not up to them to provide such workarounds/fixes, definitely not on platforms other than Windows that they support.
I think it is either on the XMOS side or it should be handled by the device manufacturers like Topping.

 

[Berwhale]

I agree, it is not up to them to provide such workarounds/fixes, definitely not on platforms other than Windows that they support.
I think it is either on the XMOS side or it should be handled by the device manufacturers like Topping.

This situation feels somewhat similar to the 'ESS Hump' debacle, where OEMs were left to workout the 'fix' for behaviour that appeared to be inherent in the DAC IC (or behaviour that was very likely to occur unless the DAC IC was implemented in a very specific way).

I hope this issue doesn't follow the same timeline with Mac users having to wait 18 months before all the OEMs 'get it'.

 

[Mutex]

I ordered one. Let's see if it pops on Linux.

If it does, one can probably come up with something that periodically sends a couple of 0's to the unit as keep-alive ping.

 

[wavetrade]

Mine just arrived about an hour ago. I have it hooked up to a Mac Mini M2 and it seems to be working well. Sounds fantastic and no popping experienced yet with Apple Music or Quboz. I did get a soft pop of sorts when fumbling through the menu settings. I imported a couple of filters an got them loaded to the device. Wish that topping had skipped the first 5 built in EQ modes, but not a big deal.

The firmware on here is 1.72. I will try reading through the thread as I wonder if there is any consensus on the best/most stable firmware at this point? I am not using headphones.

 

[ziggurcat]

Mine just arrived about an hour ago. I have it hooked up to a Mac Mini M2 and it seems to be working well. Sounds fantastic and no popping experienced yet with Apple Music or Quboz.

The firmware on here is 1.72. I will try reading through the thread as I wonder if there is any consensus on the best/most stable firmware at this point? I am not using headphones.

You won't hear any popping with Qobuz (and probably not Apple Music, either) since we know for sure that Qobuz sends continuous data to the DAC preventing it from effectively going to sleep (therefore no pop will occur). To test that, shut down any of those apps, play some music from your local files in the Music app (not streaming from Apple Music) or even just go into Settings > Sound > Alert Sound or go to Bandcamp online, and just play some music there. Do note that if you're on Tahoe 26.1, I have noticed that there's a 5+ second delay between the time music has stopped and when the pop happens (no such delay with Sequoia 15.6.1, it will happen within a second after pausing/stopping). It is immediate when you play an alert sound, though, so that would be the better test to conduct. And if you're not using headphones, put your ear within 6 inches of the speaker drivers -- in my experience, it is more noticeable with the woofer vs. the tweeter.

I would also update the FW to v1.78.

 

[somebodyelse]

I don't think we can blame Thesycon. The XMOS chip in these DACs is supposed to be UAC 2.0 compliant

The XMOS chip is just a processor. The UAC2 compliance and general good behaviour are up to Topping's firmware developer. They're probably closely following the XMOS example code, but that's just a generic example. It's up to the implementer to ensure that muting of the DAC chip happens when necessary etc. for the hardware design they're using.

 

[Berwhale]

The XMOS chip is just a processor. The UAC2 compliance and general good behaviour are up to Topping's firmware developer. They're probably closely following the XMOS example code, but that's just a generic example. It's up to the implementer to ensure that muting of the DAC chip happens when necessary etc. for the hardware design they're using.

What if the muting and unmuting is what is causing the audible pops on certain setups?

 

[somebodyelse]

What if the muting and unmuting is what is causing the audible pops on certain setups?

Then it's still Topping's work that's the problem, not anything to do with XMOS. It's Topping's responsibility to design hardware that can be muted without popping, and to write the code that makes it happen. Either they messed up the code that implements the UAC2 to I2S and control signals to the DAC chip, or they messed up the hardware implementation (or ESS messed up the chip - but then we'd expect to see the same issue in finished products from everyone using the same chip - but AFAIK that's not the case.)

Edit: muting and unmuting was just an example - the problem appears to be in starting and stopping a stream of zeros on the USB side without a pop, and muting the DAC is commonly used with noisier DACs to hide the noise when there's nothing playing. I think people have reported a similar noise issue when some SMSL DACs change sample rate, something that doesn't happen to other manufacturers using the same DAC chips, so is either SMSL's hardware design or their code. As Windows used to require manufacturers to supply a driver for UAC2 devices it wasn't uncommon for them to work around firmware issues by modifying the driver instead of fixing the firmware - firmware changes are riskier than driver changes. As a result we see a large number of quirks needed in the linux drivers to make these devices work properly - once someone has worked out exactly what the non-standard behaviour is.

 

[Scrivs]

Are there any gut-shots/ internal pictures, or tear down pictures posted. I’d like a peak under the hood for consideration before further.
Thanks!

 

[Berwhale]

Then it's still Topping's work that's the problem, not anything to do with XMOS. It's Topping's responsibility to design hardware that can be muted without popping, and to write the code that makes it happen. Either they messed up the code that implements the UAC2 to I2S and control signals to the DAC chip, or they messed up the hardware implementation (or ESS messed up the chip - but then we'd expect to see the same issue in finished products from everyone using the same chip - but AFAIK that's not the case.)

Edit: muting and unmuting was just an example - the problem appears to be in starting and stopping a stream of zeros on the USB side without a pop, and muting the DAC is commonly used with noisier DACs to hide the noise when there's nothing playing. I think people have reported a similar noise issue when some SMSL DACs change sample rate, something that doesn't happen to other manufacturers using the same DAC chips, so is either SMSL's hardware design or their code. As Windows used to require manufacturers to supply a driver for UAC2 devices it wasn't uncommon for them to work around firmware issues by modifying the driver instead of fixing the firmware - firmware changes are riskier than driver changes. As a result we see a large number of quirks needed in the linux drivers to make these devices work properly - once someone has worked out exactly what the non-standard behaviour is.

But this issue appears to affect Mac users much more than Windows users. The pop was very subtle when I thought i'd reproduced the issue on Windows at the week-end, but i'm struggling to replicate the issue again now.

I have streaming set to 'On when needed' in Topping Control Panel (so the streamed zero 'fix' is turned off) and I'm using WASAPI (Exclusive) in MusicBee. I can start and stop music without any pops, I can switch between albums with different bit depths and sample rates without any pops and I can randomly play test tones from Control Panel (when MusicBee is stopped) without any pops. I'm wondering if I really heard any pops at the week-end!

 

[Berwhale]

Are there any gut-shots/ internal pictures, or tear down pictures posted. I’d like a peak under the hood for consideration before further.
Thanks!

Yes, in 'Topping DX5II teardown' thread: https://www.audiosciencereview.com/forum/index.php?threads/topping-dx5ii-teardown.64527/

 

[ziggurcat]

But this issue appears to affect Mac users much more than Windows users. The pop was very subtle when I thought i'd reproduced the issue on Windows at the week-end, but i'm struggling to replicate the issue again now.

I have streaming set to 'On when needed' in Topping Control Panel (so the streamed zero 'fix' is turned off) and I'm using WASAPI (Exclusive) in MusicBee. I can start and stop music without any pops, I can switch between albums with different bit depths and sample rates without any pops and I can randomly play test tones from Control Panel (when MusicBee is stopped) without any pops. I'm wondering if I really heard any pops at the week-end!

What happens if you play music through Bandcamp online, and not through MusicBee (and is it possible to play anything on Windows without WASAPI)?

I think part of the reason some people don't encounter it is because they're running applications that effectively prevent the DAC streaming from going to sleep.