Topping D30Pro Review (Balanced DAC)

[DavidK442]

A very pleasant surprise to meet the man behind the meter. The video review was informative and concise with just enough personality. Thank you.

 

[AudioSceptic]

It's the apparent inconsistency too. Here's an example, Motu M2 review:

View attachment 111686
The Motu is the only D/A product reviewed so far (as far as I know) that has exhibited proper response in its LPF.

This Topping is a joke in that respect and should be called for what it is.

The Pro-Ject Pre Box S2 Digital <https://www.audiosciencereview.com/...ct-pre-box-s2-dac-headphone-amp-review.10085/> and <https://www.stereophile.com/content/pro-ject-pre-box-s2-digital-da-headphone-amplifier-measurements> has a choice of 3 filters that do this (plus 4 others that don't). I think these are mostly built into the ESS DACs but John Westlake added 1 or 2 of his own. I have one of these, and this is one of the reasons I bought it. Some of its other measurements are short of SOTA, but what is the point of a DAC with SOTA SINAD if it can't correctly reconstruct CD-quality audio?

 

[bennetng]

The fact these manufacturers insist on this kind of filtering has always been odd to me though. As in, 'Why'.
Velvet sound.....?

The only downside is the obvious filters and clock not being bang-on 1Khz (I get it, Topping ain't to blame, and you're not going to get an FPGA or some discrete solution at what I assume would be a very good price for this thing? Correct me if I'm wrong on this notion). I wonder if an equally costing ESS chip can be had, they have the proper filter selection (seems to be the only folks that do sadly).

As mentioned before, there is a price to pay for an "ideal" filter.
https://www.audiosciencereview.com/...mkii-distortion-at-44-1-khz.20142/post-665472
https://www.audiosciencereview.com/...ds/new-dac-chipset-from-akm.14391/post-450135
It is trivial to use something like SoX to get 160dB+ attenuation and full attenuation at fs/2 and 99% steepness so that frequency response is completely flat beyond 21.8kHz (for 44.1kHz sample rate), but I heard no difference, honestly.

What I said in 2018, don't know SMSL regrets or not at this moment

CS43198/43131 based devices please. I can predict there will be a lot of AKM implementations in 2019 and SMSL will face a lot of rivals. If you offer some high quality Cirrus Logic products you can probably have a firm position in that product line.

 

[bboris77]

Both are clearly faulty. The only reason they get away with this crap is there is so little "up there" in the first place. Digital pioneers would be turning in their graves.

And no, Topping are just lazy. No buck passing on LPFs back to the D/A manufacturer, what a joke.

I would not call them lazy exactly...they have been churning out DACs and amps like their lives depended on it.

They should not get a pass on it either though - they are the ones that selected that particular chip from Cirrus. If these forums are all about objective measurements, it is disingenuous not to call the product out on it, especially if other brands have been criticized for it.

 

[Max]

Just a question about features : does anyone know if it has bluetooth ? I would say yes with regard of other Topping's models but I don't see neither (external) antenna nor dedicated label.

 

[AudioSceptic]

Both are clearly faulty. The only reason they get away with this crap is there is so little "up there" in the first place. Digital pioneers would be turning in their graves.

And no, Topping are just lazy. No buck passing on LPFs back to the D/A manufacturer, what a joke.

1. Choose a DAC with proper filters.
2. If none available (clearly not the case but anyway), add your own filters. Add something to the value. Don't just rely on the DAC maker doing *all* the heavy lifting.

 

[AudioSceptic]

A very pleasant surprise to meet the man behind the meter. The video review was informative and concise with just enough personality. Thank you.

Yes, smooth and professional. And no annoying background muzak that afflicts so many videos.

 

[JohnYang1997]

Less than 10dB down at 22.05kHz (@44.1k) is not "very high attenuation" nor is it remotely "nice".

It should be 80-90dB down. You know that.

What I can say is, watch the video.

 

[TheTalbotHound]

I think the point is to point out this noise shaping or ultrasonic leaking. Maybe it's worthwhile to show two bandwidths though. WolfX does this too.

Yeah thats what i thought. I like the way Wolf does things.

 

[TheTalbotHound]

What I can say is, watch the video.

Also could just use HQplayer upsampling with one of the Sinc like filters to get a nice hard cutoff at 22.05.

 

[AudioSceptic]

The fact these manufacturers insist on this kind of filtering has always been odd to me though. As in, 'Why'.
Velvet sound.....?

It looks to me (no expert) as if they were designed for 48k sampling and they couldn't be bothered to design another set for 44k. Why else would 24k be such a common cutoff?

 

[JohnYang1997]

Also could just use HQplayer upsampling with one of the Sinc like filters to get a nice hard cutoff at 22.05.

Also the beauty of our new product is that we can add another filter if we want to. Everything in the machine can be upgraded (except for hardware of course).

 

[JohnYang1997]

Nobody knows.

A bit like the rear panel with no designation of inputs or outputs. An unfinished product if ever I saw one. How does this stuff ever see the light of day?

Outputs have white font black background, inputs have black font white background. If you are not sure read the god damn manual. lol

 

[AudioSceptic]

Also the beauty of our new product is that we can add another filter if we want to. Everything in the machine can be upgraded (except for hardware of course).

Nice to see you here, as always. Can you explain the naming system you use? Some of us are puzzled that the D30 Pro seems unrelated to the D30.

 

[JohnYang1997]

Nice to see you here, as always. Can you explain the naming system you use? Some of us are puzzled that the D30 Pro seems unrelated to the D30.

S is usually minor upgrades where pro can be completely different from previous product (if there is). And this will also happen in the future line up with 'pro'.

 

[AudioSceptic]

S is usually minor upgrades where pro can be completely different from previous product (if there is). And this will also happen in the future line up with 'pro'.

Thanks. So the number (30 in this case) refers to the market segment more than anything else?

 

[JohnYang1997]

I would not call them lazy exactly...they have been churning out DACs and amps like their lives depended on it.

They should not get a pass on it either though - they are the ones that selected that particular chip from Cirrus. If these forums are all about objective measurements, it is disingenuous not to call the product out on it, especially if other brands have been criticized for it.

Amir has to talk about everything every time from now on? Go watch his video, he talks in much more detail.
BTW I know what you did on other forums. So I hope everyone is automatically nicer and more logical on ASR forum. Thanks.

 

[AudioSceptic]

Also the beauty of our new product is that we can add another filter if we want to. Everything in the machine can be upgraded (except for hardware of course).

How likely is it that you would add your own filters in this case?

 

[PeteL]

1. Choose a DAC with proper filters.
2. If none available (clearly not the case but anyway), add your own filters. Add something to the value. Don't just rely on the DAC maker doing *all* the heavy lifting.

once your out of the chip, you're in the analog domain, so 2 is not really an option unless you design a FPGA dac. That said, we may be a bit fast on judging I believe, It's an oversampling architecture, that use interpolation, I see in the datasheet a "non oversampling mode" I also see a sentence that says " These filter has been designed to acomodate a wide variety of music taste and style". Yes, sound funny from a chip manufacturer, but. We can't fully assess just from these responses nowaday that it don't respect Nyquist. "Proper" is not as binary as it once was. Oversampling is precisely to be able to use less steep filter and still respect Nyquist, we don't really fully know what's behind design decisions, altough it may look wrong.

 

[JohnYang1997]

Thanks. So the number (30 in this case) refers to the market segment more than anything else?

In some way yes. But it's not exactly. There are so many different products that it's impossible to make a new series for a new idea.