Topping D50 III Balanced DAC with EQ Review

[yys310]

(I'm glad that I'm not the only crazy one too measure channel balance right after measuring for DC and malfunctions right out of the box in any device I get)

I also get DC output on both channel. One about 0.37mV and the other about 0.14mV IIRC.
Cleaning the flux isn't helping I guess they're using no-clean flux and it's reasonable that they don't clean it.

I first discover this when I connect them to oscilloscope and try to do the stereophile -90dB test...and I realized the offset from their measurement from other devices is much much smaller...

 

[DLS79]

So let's face the problem instead of playing with words.

Anyone check the channel balance?
Mine is about 0.07dB off for L and R channel. My motu M4 is now under repair so I cannot measure the output from M4 now but IIRC the deviation from M4 is much smaller than this.
I measured this by MOTU 896MK3 from my friend.
As a comparison I measured the 1,3,5,7 output from MOTU 896MK3 and the loudest and quietest is 0.015dB off.
The D90III is 0.06dB off for L and R channel.

I also measured the Manhattan II and it's 0.001dB off but that's just too expensive so no need to compare with that.
View attachment 366456View attachment 366457

The often sited requirement for ABX testing is output matched to within 0.1dB, and every devise you mentioned meets that.

What's ironic is that's a lot of people here have complained about Topping chasing numbers in the past, but now that they switched to working on something else people are complaining about the numbers!

 

[zdykstra]

Is anyone having issues after updating to firmware 1.21? Now remote won't pair and firmware requires reinstalls and modes no longer accessible via D50 III right button control. Two emails to Topping unanswered.

I'm on 1.21 with out any issues (those you listed or other). I assume you've yanked power to the device and booted it back up?

 

[yys310]

Is anyone having issues after updating to firmware 1.21? Now remote won't pair and firmware requires reinstalls and modes no longer accessible via D50 III right button control. Two emails to Topping unanswered.

I don't have your problem either.
Did you turn the bluetooth off? Although the remote will work when you set them off but the pairing process requires that to be on.

 

[Robbo99999]

Here is a really good thread about the impact EQ can have.
(a partial fix was the result of this thread)

DSP Measurements and Rising Noise Floor

In several threads related to DSP platforms I've seen requests for measurements with filters engaged. I thought it unlikely that implementing filters in a DSP would have measurable effect on noise / distortion other than the obvious change in the fundamental, however @Sokel recently mentioned...

www.audiosciencereview.com

(note that these are pure digital measurements so no analog penalties involve )

What were the main conclusions from the testing & that thread for those that aren't reading it? Does it answer the question about potential inaccuracy of some of the PEQ filters?

 

[Sokel]

What were the main conclusions from the testing & that thread for those that aren't reading it? Does it answer the question about potential inaccuracy of some of the PEQ filters?

The conclusion for the tested devices was more about noise as inaccuracy was not observed so to be tested.
The results showed rising noise with filters enabled and they vary depending the device,I think one of them showed none.
It was strange to see that pure digital measurements was affected this way but people reported this as a known issue in some implementations.

The ones more affected was the Sharc based ones and miniDSP took action as a result and applied a partial fix.
There a lot more in that thread,I urge all who interested to read it.

 

[oscar_dziki]

That's in comparison to RME implementation. That assumes that RME implementation is rock solid, is it safe to make that assumption? Also is the measurement procedure bonafide & not open to any errors? It's true though that we see differences in the graph, not massive but there.

EDIT: assuming it's an error in Topping, do we know what is causing the error, as in with more filters combined & more complex filters combined are we likely to see increased error or is that not a safe assumption? I wonder what Topping have to say about this.

Which proves my point - we need a standardized eq frequency response measurement procedure.

 

[PeterWest]

I'm on 1.21 with out any issues (those you listed or other). I assume you've yanked power to the device and booted it back up?

Yup. I'm powering via the P50. I have it working but when I push the off button and hold it on to check if the 1.21 firmware loaded (which is said it did) the display goes DFU and I have to reload. When working now the remote works but I'm suspicious I'm on the original factory default firmware.

 

[mocenigo]

Unfortunately no Topping Tune for Mac, no EQ use with optical or coax and the software might be half-baked so far, however it's a good trend to have EQ in an affordable DAC.

This is silly.

 

[zdykstra]

Yup. I'm powering via the P50. I have it working but when I push the off button and hold it on to check if the 1.21 firmware loaded (which is said it did) the display goes DFU and I have to reload. When working now the remote works but I'm suspicious I'm on the original factory default firmware.

When I press and hold the power button, mine also displays DFU (which I imagine means Device Firmware Upgrade). I can safely yank the power and it boots back up normally.

 

[PeterWest]

I don't have your problem either.
Did you turn the bluetooth off? Although the remote will work when you set them off but the pairing process requires that to be on.

YES that's the answer. Duh. Thank you for the answer.

 

[PeterWest]

I'm still mystified. Is anyone running their D50 III optically from a BlueSound Node 2i. The only way I can the DAC to run is if it is also connected via USB to my MacMini. This doesn't feel right. Is the DAC firmware residing in the computer and not in the DAC itself? And now I am questioning whether my P50 power supply is actually powering the DAC. I'd really appreciate some help LOL

 

[yys310]

I'm still mystified. Is anyone running their D50 III optically from a BlueSound Node 2i. The only way I can the DAC to run is if it is also connected via USB to my MacMini. This doesn't feel right. Is the DAC firmware residing in the computer and not in the DAC itself? And now I am questioning whether my P50 power supply is actually powering the DAC. I'd really appreciate some help LOL

You face the signal drop off issue? If so probably you didn't successfully flush the device. Not sure how to check the FW version on mac.
On windows you can see the FW version in the topping ASIO control panel. You should see 1.21 after you flash your device.
You just unplug the other typeC input to see if the device is still on. I don't see why the P50 isn't powering the DAC.

 

[howard416]

I guess they'll need to do this in an E series in order for it to come with a volume knob?

 

[DLS79]

I guess they'll need to do this in an E series in order for it to come with a volume knob?

The only thing in the E series with a volume knob is the E70 & E70 Velvet. Depending on your needs the DX or MX line mut be a better fit assuming they get PEQ at some point.

 

[Steve96]

Is there a difference in (sound) performance/advantage to using built in PEQ vs using external software like Peace?

 

[DLS79]

Is there a difference in (sound) performance/advantage to using built in PEQ vs using external software like Peace?

The D50 iii is limited to the same eq settings for both the left and the right channels, and usb input only.

Speaking generally, the main benefits of none pc based eq, is support for multiple types of inputs, and not needing to worry about OS or driver related issues.

 

[GaryH]

That's in comparison to RME implementation. That assumes that RME implementation is rock solid, is it safe to make that assumption? Also is the measurement procedure bonafide & not open to any errors? It's true though that we see differences in the graph, not massive but there.

EDIT: assuming it's an error in Topping, do we know what is causing the error, as in with more filters combined & more complex filters combined are we likely to see increased error or is that not a safe assumption? I wonder what Topping have to say about this.

Well we don't need to assume the RME's implementation is correct (I would think it is though) to know the Topping's isn't - the first filter was set at 100 Hz but the Topping produced a peak at 105 Hz. This is a significant error within audibility (unlike the difference between audibly transparent distortion and really audibly transparent distortion ). This post gave a possible cause of the Topping's PEQ errors (which may mean errors accumulating with more filters, but who knows). If correct that's just poor design.

Is there a difference in (sound) performance/advantage to using built in PEQ vs using external software like Peace?

In this case yes, the difference in sound performance is the latter actually produces the correct PEQ curve.

 

[DLS79]

@GaryH amir didn't have any issues with his peq tests!

 

[Robbo99999]

Well we don't need to assume the RME's implementation is correct (I would think it is though) to know the Topping's isn't - the first filter was set at 100 Hz but the Topping produced a peak at 105 Hz. This is a significant error within audibility (unlike the difference between audibly transparent distortion and really audibly transparent distortion ). This post gave a possible cause of the Topping's PEQ errors (which may mean errors accumulating with more filters, but who knows). If correct that's just poor design.


In this case yes, the difference in sound performance is the latter actually produces the correct PEQ curve.

Ha, well that's true re the 5Hz offset, that can be quite easily seen when I look now, I only gave it a quick scan last time! Yes, I agree, the Topping is slightly off for some reason - and according to the post you linked then that sounds like something akin to "rounding errors" if I understand his post correctly?