Topping D900 - teardown

[Deleted member 89542]

Thanks so basically all the topping and smsl products sound the same and there is no need for all the different models other than if some features appeal. I certainly could not hear a noticeable difference between topping, rme, ferrum and violectric on Sennheiser 600 series and susvara headphones.

 

[archerious]

Thanks so basically all the topping and smsl products sound the same and there is no need for all the different models other than if some features appeal. I certainly could not hear a noticeable difference between topping, rme, ferrum and violectric on Sennheiser 600 series and susvara headphones.

Exactly. All well measuring products for the most part sound similar. Only with really high output impedance could there to my understanding cause some strangeness but most SMSL and Topping products measure well enough and keep the output impedance under 1 ohm.

IMO, the most significant perks of something like RME over Topping would be made in Germany, better build quality (possibly leading to better reliability long term), and a longer warranty. Of course at a higher price tag.

The DX5 II is IMO a really appealing product. The PEQ is useful, the digital screen + remote control, and powerful headphone amp output. Really nothing about it is bad, especially at $299.

Really hoping SMSL releases a competitor to this.

 

[Deleted member 89542]

Ok thanks I buy a lot on line but when it comes to HiFi I prefer to listen and have access to service and advice. I have ended up with a lot of different dacs/amps over the years and they all sound very different but they vary from dongles, to HiFi chip dacs to FPGA so fundamentally different architecture and power. I can’t tell the difference between different chip dacs and nothing makes a difference on my modest speaker system. I can only tell a difference when I try high end headphones systems and there the difference is dramatic. The chip dacs/amps all sound the same but something like the feliks envy brings the susvara alive which I guess is due to power and tube coloration. Theoretical as I’m not buying these systems as non of the dac chip systems sounded better than what I have but just tried lots of different systems at high end Munich and dealers and just trying to understand better.

 

[RichB]

I have the D90 III discrete and D900 with a rig that lets me use a web page to instantly switch between DACs.
I find the D900 the more natural DAC where I can best compare them with the Revel M126, Benchmark AHB2, Roon Raspberry Pi5/Pi2AES.
The Coax and AES are indistinguishable on the D900 so I use these connections for comparison.

I am not sure that LAB measurements prove that some DACs perform better in the wild.
Some DACs have galvanic isolation, others do not. processing is different between inputs. and volume control performance is not measured on ASR.

I have a better rig than most to do instantaneous switching and own these products. That does not mean the ready should take stock in these impressions.
However, interpretation of good science includes understanding what is and what is not measured.

- Rich

 

[archerious]

I have the D90 III discrete and D900 with a rig that lets me use a web page to instantly switch between DACs.
I find the D900 the more natural DAC where I can best compare them with the Revel M126, Benchmark AHB2, Roon Raspberry Pi5/Pi2AES.
The Coax and AES are indistinguishable on the D900 so I use these connections for comparison.

I am not sure that LAB measurements prove that some DACs perform better in the wild.
Some DACs have galvanic isolation, others do not. processing is different between inputs. and volume control performance is not measured on ASR.

I have a better rig than most to do instantaneous switching and own these products. That does not mean the ready should take stock in these impressions.
However, interpretation of good science includes understanding what is and what is not measured.

- Rich

Did you volume match them? Same voltage etc?

 

[Deleted member 89542]

I am really just trying to understand.

1 Do all these topping and smsl and other off the shelf chip dac /amps measure and sound the same within audible range? If so does it make sense to just buy the cheapest with the features you want and it should be made clear that all these products from the same and different off the shelf chip manufacturers are basically the same.

Or 2 Do they measure different within audible range and sound different.

Or 3 do they measure the same within audible range and sound different in which case are there things that the human brain genuinely legitimately percieves differently which aren’t caught by measurements yet?

I spent a couple of days at the world of headphones trade fair without really knowing about what I was trying or the prices before I tried. Certainly not ideal listening conditions or A to B testing. Much sounded the same. The Topping models I tried with Susvara and Sennheiser were absolutely fine but did not move me but then neither did the likes of RME, Ferrum or Violectric or DCS Lina. So I get that people may not hear major differences despite the price.

The feliks envy on the other hand was in a complete different league from everything else. Personally I would not buy online just based on measurements but I would not buy any of these systems as i enjoyed them less than what I have and I would not buy the susvara either as magnetic planar headphones don’t move me emotionally.

I get the concept of nyquist Shannon and that you theoretically need an infinite calculation to reproduce the music. I get that digital music recording and reproduction is maths and engineering but listening to music is a subjective experience as we all have different music tastes. So ultimately I’m looking for the system that measures most on the musical enjoyment scale for my brain and taste for the least price and that is what I would really like to understand better. For me buying something online just based on measurements or subjective reviews which I don’t enjoy is very poor value for money as I think listening to music should be about enjoyment but I’m still trying to understand better how I find that.

 

[RichB]

Did you volume match them? Same voltage etc?

Yes at my normal listening level -27.5, the D900 III Discrete outputs 1.543, the D900 outputs 1.530.
This is essentially identical.

D90 II Discrete

D900

- Rich

 

[Deleted member 89542]

I’m confused as some people say DACs make no difference and DAC technology has been solved long ago. In which case all the measurements and multitude of regular new models of eg topping and smsl imodels are unnecessary and basically just another form of marketing. But then I understand from reading Amir that power, filters and analogue output implementation make a difference and then they are different chips from different manufacturers and different technologies such as FGPA and R2R. So are all these different new models and more expensive models of Topping and SMSL just marketing with no real sound impact or not?

 

[Ahmonge]

So are all these different new models and more expensive models of Topping and SMSL just marketing with no real sound impact or not?

In a market like this, crowded with brands, there are two key factors from a marketing perspective: differentiation and the halo effect. Exotic technologies and exorbitant prices at the high end are their tools.

That is why it is necessary to continue measuring products to see if they really solve “the problem” or make it worse.

 

[Purité Audio]

If two dacs both do not add audible distortion then they will sound the same, the only possible difference the choice of reconstruction filter.
Keith

 

[audiofun]

I actually bet this will have more ultra sonic distortion / noise when measured than a regular off the shelf tech (such as ES9039Q2M) DAC.
while DSD256 is much better than 64, the noise shaping may not be as good as an ES9039Q2M.
Shouldn't be audible though.

 

[Deleted member 89542]

Thanks I’m just trying to educate myself as I build a new HiFi system and want to get best system possible without spending money unnecessarily. I very much doubt that for instance a 20k streamer dac sounds fundamentally better than a 500 DAC with same chip or all the different models are more than just consumerism or if slight differences in inaudible measurements matter.

I get that it is mathematically proven that band width limited 44khz is sufficient to capture all the information required to reconstruct the sound wave perfectly but in theory you need an infinite sinc function to perfectly reconstruct the sound wave. So the reconstruction filter impact makes sense to extent audible. The filters on chord mojo 2 make a subtle but noticeable difference for instance. I prefer a neutral filter.

I’m not sure to what extent measurements can fully capture the accuracy of the frequency and timing of the reconstruction filters in a complex dynamic context of music and to what extent these differences are audible given the limited understanding of human physcoaccoustics. But I understand that we can hear tiny timing differences in order to locate sound and these may impact our perception of music. For instance how do you measure soundstage,imagery, timbre , rhythm and essentially human musical enjoyment. Many systems sound the same but sometimes some systems just sound more enjoyable or the sound position sounds different and I don’t understand why.

I also get human perception is important. I understand for instance that the harman curve is not neutral but based on the average of human preferences but humans have different preferences. The AirPod Max seems very popular at the moment so clearly human preferences matter over objective performance. I get you can measure frequency response and SINAD but not sure to what extent these relate to human enjoyment.

 

[Svend P]

All of your questions are answered here. But you probably have to read more than the first page

'Master Thread: Are Measurements Everything or Nothing?'

Nov 18, 2020

Seems like so many review threads get challenged with:

1. Measurements are not everything.

2. You all never listen.

3. I trust my ears, not graphs.

4. I don't listen to graphs. I listen to music.

5. You all must not listen to music at all.

6. Why don't you all buy the best SINAD gear?

7. I have heard your best SINAD gear and they sound terrible. I don't like any of this Chinese stuff.

8. You don't trust your ears. I/we do.

9. All these reviewers/youtubers/audophiles say these amps, DACs, etc. sound different and you say they don't. They can't all be wrong.

10. Surely designers...

• amirm
• Replies: 16,652
• Forum: General Audio Discussions

• 

 

[Deleted member 89542]

An entirely fair response thanks! But 771 pages! I completely get the point and read Amir’s guide to measurements before posting but then I keep reading all DACs sound the same. So why even bother to measure loads of Topping models or produce all the multitude of new models when they all sound the same. And then I read Amir writing implementation matters. Which is it?

I studied cognitive biases. Should I get a dCS Lina stack just because it is expensive sounds fine but it didn’t do anything for me. I don’t think so. Should I get a Topping which I’ve heard with Sennheiser and Susvara headphones because it measures well and sounds fine but leaves me completely non-plussed. I’m not convinced either.

I get that no DAC is going to make a difference on my old modest speaker system. I get that I couldn’t hear substantial differences between dCS Lina, Topping, Violectric, RME, Ferrum DAC/Amps. However, I’ve heard significant differences between some dac/amp combos on high end headphones which I guess are more down to the ability to drive the headphone. I’m just trying to improve my knowledge and understanding which is difficult when an objective site like ASR seems to contradicts itself. So does the Topping D900 measure and sound better than other Topping DACs or is it just marketing and all the Topping and SMSL DACs sound the same and there is no point measuring them or with all the loads of different models every year. I don’t listen to features so I just need input and output connections and a volume control.

 

[Purité Audio]

No dac is going to make a difference to any speaker system no matter how expensive unless it is extremely poorly designed.
Amps need to be able to drive loudspeakers/headphones without clipping with that caveat they will all sound the same unless again they are poorly designed and are adding audible distortion.
Decide perhaps on performance, features and aesthetic appeal certainly not price.
Keith

 

[Deleted member 89542]

Thanks, yes I certainly struggle logically or intuitively to understand how there would be any noticeable impact on my current and proposed speaker system between decent DACs. My listening room situation is so compromised I’m only seriously interested in headphones. The dealers I use also told me they were not interested in selling me a different DAC and suggested cheaper active speakers with room DSP. I do perceive differences on highly resolving headphones with dac amp combos where power and reconstructive filters do seem to make a noticeable difference.

 

[Purité Audio]

Acoustic measurement with REW and a microphone should be your first step and then EQ where and if necessary.
Yes contemporary active cardioid designs offer tangible benefits in my opinion.
Keith

 

[Deleted member 89542]

Thanks I suspect something like Dutch and Dutch or Genelec that Amir recommends would give me a better speaker solution. But current circumstances dictate speakers are just furniture most of the time and it makes more practical sense to get a good headphone set and add a modest power amp and aesthetic pleasing passive speaker for occasional listening. Good knowledge for future when circumstances change though as not considered before.

 

[welwynnick]

I've suspected for a while that the DAC was a bottleneck in ambitious systems, and thought there might be a potential improvement in system gain staging by increasing the analogue output level beyond the usual 2V or 4V up to the higher levels used in modern pro studio equipment (typically 24dBu / 12.2V).

I'm a happy Topping E70/L70 user; occasionally thought about getting a D70/A70; and the D900/A900 got me curious.
The D900 offered 13V pre-amp output, but the dynamic range was no better than many recent low cost DACs.

Looking closer at their website, the D900 has a noise floor of 1.5 µV at 5.2V, which equates to their claimed 131dB SNR.
If they could maintain that noise at 13V, it could give a decent boost to the DNR (not that it actually needs to be any higher, before you reach for your keyboard).
However, although Topping usually wring the neck out of their gear and extract every last ounce of performance in their testing, the 13V output was no better.

Could it be that the noise is higher at 13V output than 5V, so the SNR is no better at all?
However, Topping's figure for 13V (1.6µV) is different to 5V (1.5µV), which suggests they have actually measured it, and not used the 5V figure by default.
But 13V is 138 dB higher than 1.6µV - so why didn't they quote that figure instead? That would be something to make a song and dance about.

In theory, that would be enough to drive a Hypex or Purifi power amp directly, without any input buffer,
It wouldn't compromise either the max power or the noise floor of these remarkable amps, which are normally limited by the upstream electronics.
In any other system, no matter how you arrange the gain staging, the bottleneck is upstream. Here are some calcs:

 

[whydan]

I've suspected for a while that the DAC was a bottleneck in ambitious systems, and thought there might be a potential improvement in system gain staging by increasing the analogue output level beyond the usual 2V or 4V up to the higher levels used in modern pro studio equipment (typically 24dBu / 12.2V).

I'm a happy Topping E70/L70 user; occasionally thought about getting a D70/A70; and the D900/A900 got me curious.
The D900 offered 13V pre-amp output, but the dynamic range was no better than many recent low cost DACs.

Looking closer at their website, the D900 has a noise floor of 1.5 µV at 5.2V, which equates to their claimed 131dB SNR.
If they could maintain that noise at 13V, it could give a decent boost to the DNR (not that it actually needs to be any higher, before you reach for your keyboard).
However, although Topping usually wring the neck out of their gear and extract every last ounce of performance in their testing, the 13V output was no better.

Could it be that the noise is higher at 13V output than 5V, so the SNR is no better at all?
However, Topping's figure for 13V (1.6µV) is different to 5V (1.5µV), which suggests they have actually measured it, and not used the 5V figure by default.
But 13V is 138 dB higher than 1.6µV - so why didn't they quote that figure instead? That would be something to make a song and dance about.

In theory, that would be enough to drive a Hypex or Purifi power amp directly, without any input buffer,
It wouldn't compromise either the max power or the noise floor of these remarkable amps, which are normally limited by the upstream electronics.
In any other system, no matter how you arrange the gain staging, the bottleneck is upstream. Here are some calcs:

View attachment 476583

I do have one 6525 under development progress, and I tried to input from D900 without buffer stage.
To be honestly, I didn't feel better with buffer stage and I prefer to add OPA buffer stage by my ear.

Sorry I can't provide any measurement since lack of toolkit, I am still working on it.
I believe it may get some benefit on measurement Theoretically but not by subjective feeling.