Review and Measurements of Topping D90 4499DAC

[restorer-john]

His cycle is over.
You should send your APx 555 to @◘restorer-john, so he can measure the equipment that interests me, Vintage from the 60s / 70s.
Thanks.

Trust me, if I could justify buying an APX-555B, I'd be all over it like a cheap suit.

 

[trl]

I too appreciate Wolifie's contributions. The only issue is that when I promoted one of this reviews to the home page (the Audio-gd R8), people on another forum bashed me for not knowing how to measure. Someone corrected them at the end but the damage was done. I was crying all day, looking for a cliff to jump from....

But WolfX is providing the full chart from the AP equipment? What would they be worried about? Photoshop perhaps?
Someone could easily provide worse measurements by getting a ground loop or some mains hum due to the poor cheap cables, but getting better measurements than the reality is kinda hard to get (at least with the full chart/print screen provided).

Maybe WolfX should aggregate his own measurements into a different chart or on a measurements webpage within this forum and later will be up to the readers to appreciate his work (and money and time involved into) or not; I know I am and I will.

Later, time will tell more, just need to wait until Amir will measure few of the DACs WolfX already measured and will see if differences will be higher than 1dB in SINAD or not.

 

[Spocko]

Any idea what is the release date of the D90?

 

[bequietjk]

I'm gonna send it and say the 22nd of this month.

 

[WolfX-700]

But WolfX is providing the full chart from the AP equipment? What would they be worried about? Photoshop perhaps?
Someone could easily provide worse measurements by getting a ground loop or some mains hum due to the poor cheap cables, but getting better measurements than the reality is kinda hard to get (at least with the full chart/print screen provided).

Maybe WolfX should aggregate his own measurements into a different chart or on a measurements webpage within this forum and later will be up to the readers to appreciate his work (and money and time involved into) or not; I know I am and I will.

Later, time will tell more, just need to wait until Amir will measure few of the DACs WolfX already measured and will see if differences will be higher than 1dB in SINAD or not.

In fact, I have my own blog and chart. But it's Chinese after all

In the measurement of sgd1, the measurement results of Amir and I are basically the same. Previous gustard x26 measurements were also within the error range.

In any case, the advantage of measurement is that it's easy to verify.

 

[VeerK]

Any idea what is the release date of the D90?

December 12 according to their Facebook

 

[JohnYang1997]

It's the uneducated people from other forums' fault. Who gives a shit to them, honestly. You publish measurements for people who want to know the objective performance or learn the truth not those retards..

 

[da Choge]

December 12 according to their Facebook

Gives them very little time (none) to look into or even possibly correct any jitter or linearity issues; would it be possible through firmware? Anywhose, I'm interested to see the price . . .

- john

 

[Tks]

December 12 according to their Facebook

Closing out the the year 2019 and just in time to take the title with best performing DAC of the year? Topping you sly dogs ;P

 

[MC_RME]

@ Wolf-x700: the pic of the Dynamic Range measurement has an added note, 'A Weighting'. That seems wrong. This measurement uses a CCIR-2K weighting (AES17) when done as provided by AP default settings. Maybe you want to remove that note to prevent confusion. In case you intentionally exchanged the CCIR filter with standard A-weighting then the AP labelling (AES17) would be misleading.

 

[WolfX-700]

@ Wolf-x700: the pic of the Dynamic Range measurement has an added note, 'A Weighting'. That seems wrong. This measurement uses a CCIR-2K weighting (AES17) when done as provided by AP default settings. Maybe you want to remove that note to prevent confusion. In case you intentionally exchanged the CCIR filter with standard A-weighting then the AP labelling (AES17) would be misleading.

Thank you for your reminder.

Before that, I received some opinions about which weighting filter should be used for dynamic range measurement. It is true that the default weighting of AP is ccir-2k, but many DAC chip manufacturers use "A weighting"

So here I have some confusion.

I have also looked up some information, and it seems that there is no clear indication of which filter must be used.

So here I use A-weighting and make a conspicuous mark.

Maybe I need more information. I would appreciate it if you could.

 

[MC_RME]

As I wrote if you use A-weighting then you should remove the AES17 label, as CCIR-2k is defined in there.

DAC manufacturers use A-weighting for many years. Measurement values would no longer be comparable if they now change to AES17 DR, so they don't. And they have no reason to change, using CCIR-2k does not bring any real-world advantage. It just gives numbers that are different (confusing). I always measure Dynamic Range without any weighting filter, so it equals Signal to Noise but with excitation by the low level signal. If a unit works as expected the SNR and DR values are identical (no relevant noise floor modulation), so it is easy to see if something is wrong or suspicious. The moment you add a weighting filter the values are different. But that's just me...

 

[WolfX-700]

As I wrote if you use A-weighting then you should remove the AES17 label, as CCIR-2k is defined in there.

DAC manufacturers use A-weighting for many years. Measurement values would no longer be comparable if they now change to AES17 DR, so they don't. And they have no reason to change, using CCIR-2k does not bring any real-world advantage. It just gives numbers that are different (confusing). I always measure Dynamic Range without any weighting filter, so it equals Signal to Noise but with excitation by the low level signal. If a unit works as expected the SNR and DR values are identical (no relevant noise floor modulation), so it is easy to see if something is wrong or suspicious. The moment you add a weighting filter the values are different. But that's just me...

I'll check the AES17-2015 later. At present, AES-17 1998 does define CCIR weighting for D-R. If CCIR-2K is also defined in the latest standard. In the future, I will delete the aes17 label when I use a filter to measure D-R

As you said, the measurement of DR is really very confusing in the whole industry. I even provide the measurement results of two kinds of weightings on my own blog.

I don't like denormalization, but I find the definition of industry is different

It's a real headache.

Anyway, thank you for your reminders and suggestions.

 

[MC_RME]

You're welcome. The point here might be that Amir uses the AP default (IMHO, he might want to confirm this), and as you use the same APx readers might expect identical/comparable measurement results. So it could be useful to use the same setting for DR, no matter if that makes sense or is the same in data sheets etc.

 

[WolfX-700]

You're welcome. The point here might be that Amir uses the AP default (IMHO, he might want to confirm this), and as you use the same APx readers might expect identical/comparable measurement results. So it could be useful to use the same setting for DR, no matter if that makes sense or is the same in data sheets etc.

This is also the problem I encountered. When I use the default settings --- I actually prefer this---it will conflict with the DAC manufacturer's statement.
And if A weighting is used, from my current AES17-1998, it does not meet the description in section 9.3.

The question you mentioned about the AES17 label was something I hadn't considered before. From my point of view the whole thing is really confusing-it has been tangling me for some time.

Like I said, I respect standards. But I still think it is necessary to check the latest AES17 standard. I will finish this soon.

 

[WolfX-700]

good news:
The manufacturer is likely to find a solution to the AES3 interface J-Test. They will perform engineering verification and will send a modified machine for me to re-measure.

I have recorded the serial number of this machine now (and return it to the manufacturer).

Looking forward to their repaired version!

 

[mt196]

good news:
The manufacturer is likely to find a solution to the AES3 interface J-Test. They will perform engineering verification and will send a modified machine for me to re-measure.

I have recorded the serial number of this machine now (and return it to the manufacturer).

Looking forward to their repaired version!

Now the question will be, will they modify the produced unit or will they do it just for you in order to solve the issue in the reviewers' jitter plots? I mean, probably no one will ever notice that jitter even if they lie about modifying all the produced units and instead they sell them as is, so maybe in the future we will need a retest of a random retail unit

 

[WolfX-700]

Now the question will be, will they modify the produced unit or will they do it just for you in order to solve the issue in the reviewers' jitter plots? I mean, probably no one will ever notice that jitter even if they lie about modifying all the produced units and instead they sell them as is, so maybe in the future we will need a retest of a random retail unit

I will talk to the manufacturer to see if this change is long-term or only one or two evaluation samples.

If the manufacturer's statement is long-term (as far as I know, the cost of modification is not high)

I will tell them that I will have a random sample (on the market) at some point in the future

Test results will be updated here.

 

[777]

I

Now the question will be, will they modify the produced unit or will they do it just for you in order to solve the issue in the reviewers' jitter plots? I mean, probably no one will ever notice that jitter even if they lie about modifying all the produced units and instead they sell them as is, so maybe in the future we will need a retest of a random retail unit

Be serious.... they didn't selling yet. They are not crazy to do that They will repair all the units because all the units are in the factory now. It is the easiest thing to do now.

 

[mt196]

I


Be serious.... they didn't selling yet. They are not crazy to do that They will repair all the units because all the units are in the factory now. It is the easiest thing to do now.

Yeah I was just pointing out one possible scenario, I feel Topping is a serious company, but a double check in the future will be always appreciated