Review and Measurements of Topping D10 DAC

[Krunok]

Yes, it is a double blind study. I don't know how "modern" the opamps are. I just found the study and it is from 2011 so not ancient history.

Well, that sounds promising. But if they described the differences as something like "tighter bass with wider and taller sound stage and more forward and natural female vocals" you know we will be flaming you, right?

 

[Krunok]

Yes, it is a double blind study. I don't know how "modern" the opamps are. I just found the study and it is from 2011 so not ancient history.

Btw, I am also expecting that DAC was measured with all opamps to verify that neither of them is producing any of the nasty artifacts like increased distortion, noise, oscillations etc.

If it really turns out that there are people that can identify the difference between say THD 0.0008% and SNR 115dB (D10 with 3 OPA2134) vs THD 0.0005% and SNR 122dB (D50 with 3 OPA1612) on a blind test I would be truly impressed.

 

[amirm]

You are reading too much into my post and nature of the test. Let me post my summary of the paper and then we can discuss.

 

[Krunok]

Of course. And what is your opinion on the subject? Do you think anybody can hear the difference between say D10 and D50?

 

[amirm]

Do you think anybody can hear the difference between say D10 and D50?

With levels matched it is probably next to impossible. I like the additional functionality of the D50 though in terms of S/PDIF, Toslink, etc.

 

[Krunok]

With levels matched it is probably next to impossible.

I agree.

I like the additional functionality of the D50 though in terms of S/PDIF, Toslink, etc.

True, it seems to justify the price difference compared to D10. But hardly the case with DX7s as that one only brings XLR outputs to the playground, so in my eyes it should cost below $400.

 

[Stump909]

With levels matched it is probably next to impossible. I like the additional functionality of the D50 though in terms of S/PDIF, Toslink, etc.

Interesting. Would the same apply to the D30? It seems in some of you measurements the D10 outperforms it.

 

[amirm]

True, it seems to justify the price difference compared to D10. But hardly the case with DX7s as that one only brings XLR outputs to the playground, so in my eyes it should cost below $400.

Yes, the original DX7 was at that price and on sale even lower. Still, I don't know another device yet that has its full functionality including remote control, headphone output, XLR, nice beefy unit with internal power supply, etc. at its price.

 

[DonH56]

Ages ago when modding was very popular (not that it isn't today but I am no longer involved with that scene except vicariously through posts on sites like this) it was sometimes very easy to tell when opamps were swapped. Usually because the swap was undesirable and a misapplication of the device. For example, wideband video/RF amplifiers often have high noise at low (audio) frequencies and may not be happy with a high-capacitance load. MOSFET inputs have high noise (JFETs are preferred). Bipolar inputs are great except when the source is hi-Z (like a MM phono cartridge) so current noise dominates. Stability is always a concern, and how the opamp performs when overdriven. Etc. etc. etc. I saw way too many cases of people swapping components without understanding the impact and how it might affect other things. I am sure it is all different today... And of course there are times a positive difference was made, sometimes audible, and the swapper was hailed as a guru especially when he could dispute the "experts" (who could have explained why it worked and/or why that opamp was not originally chosen).

 

[Krunok]

Yes, the original DX7 was at that price and on sale even lower. Still, I don't know another device yet that has its full functionality including remote control, headphone output, XLR, nice beefy unit with internal power supply, etc. at its price.

Agree, original DX7 was right on the spot while DX7s only provided higher price with no real. value add on. Still better than most of the others.

I can only think of Matrix Mini-i-Pro 2 and Mini-i-Pro 2S, similar price with even the better specs. http://matrix-digi.com/en/products/122/index.html

Unlike with the DX7, Mini-i-Pro 2 can still be purchased.

 

[amirm]

I can only think of Matrix Mini-i-Pro 2 and Mini-i-Pro 2S, similar price with even the better specs. http://matrix-digi.com/en/products/122/index.html

Those two seem nice. Has anyone measured them?

 

[Krunok]

Those two seem nice. Has anyone measured them?

Not that I know of.

But Matrix uses LME49720 which are supposed to be even better than OPA1612 which are used in DX7s!

 

[Deleted member 65]

Agree, original DX7 was right on the spot while DX7s only provided higher price with no real. value add on. Still better than most of the others.

I can only think of Matrix Mini-i-Pro 2 and Mini-i-Pro 2S, similar price with even the better specs. http://matrix-digi.com/en/products/122/index.html

Unlike with the DX7, Mini-i-Pro 2 can still be purchased.

Seems as the Matrix is based on ESS SABRE Ultra ES9016S.

Here's two products that never saw the daylight, apparantly due to amongst others shortcomings in the ESS SABRE Ultra ES9016S according to designer JW. Creek eventually released new products using other hardware (Wolfson WM8742) & designers.

" Creek has developed a digital product, in collaboration with John Westlake and Dominik Peklo, offering a host of novel features and facilities. The EVOLUTION 50D is a CD player, high-end DAC and digital Pre-amp that sees Creek Audio entering into the world of digital connectivity and computer-based music storage. Primarily, this product should be seen as a very high-end DAC, with flexible digital pre-amp and built-in Slot Loading CD transport. At its heart, lies a ‘British-made’ XMOS event-driven multi-threaded 32 bit 8 core processor, programed to providing it with USB and SPDIF input, digital pre-amp with volume control, selectable filtering, sample-rate conversion, clock locking to external devices and several other novel features, including a ‘Bit Perfect’ test and analogue bit-depth enhancement.

To enable it to provide its shattering audio performance, the EVOLUTION 50D‘s contains 2x ESS SABRE DACs running in a ‘Dual Differential’ configuration, for increased dynamic range and lower noise. The balanced audio output of each DAC feeds a discrete transistor ‘Class A’ current regulated ‘CROSSII MOSFET’ buffer circuit, that also acts as a ‘Class A’ headphone amplifier output, controlled with a digital volume control. Expected availability is from September/October, depending on region. Priced in line with the Evolution 50A at £550 in the UK, including sales tax.

Finally, Creek Audio will offer a new DAC for the first time since its famous OBH-14 went out of production. This half-width DAC/Pre-amp will share all the features of the EVOLUTION 50D, except the CD transport. However, it will be fitted with a built-in FM/AM analogue tuner, as standard, designed to work automatically in all regions."
https://www.creekaudio.com/news/munich-high-end-review/

 

[Krunok]

Seems as the Matrix is based on ESS SABRE Ultra ES9016S.

Here's two products that never saw the daylight, apparantly due to amongst others shortcomings in the ESS SABRE Ultra ES9016S according to designer JW. Creek eventually released new products using other hardware (Wolfson WM8742) & designers.

There are no technical facts in that article about any "shortcoming" of ES9016S, so I wouldn't recommend jumping into conclusion there is something wrong with that DAC chip.

In 2013 that chip was probably the best DAC chip money can buy and was used by many respectable manufacturers without any issues (see links for Pioneer and Yamaha as an example).

https://www.businesswire.com/news/h...ioneer-Chooses-ESS’s-ES9016-SABRE32-Ultra-DAC
https://www.businesswire.com/news/h...aha-Selects-ESS’s-SABRE32-Ultra-SABRE-Premier

ES9016s DAC chip was eventually replaced by 9028 and 9038 series as customers always expect "cutting edge" technology although it sometimes doesn't really provide better results, as we have seen from Amir's measurements of DX/ vs DX7s.

 

[Deleted member 65]

There are no technical facts in that article about any "shortcoming" of ES9016S, so I wouldn't recommend jumping into conclusion there is something wrong with that DAC chip.

In 2013 that chip was probably the best DAC chip money can buy and was used by many respectable manufacturers without any issues (see links for Pioneer and Yamaha as an example).

https://www.businesswire.com/news/home/20130916005230/en/Pioneer-Chooses-ESS’s-ES9016-SABRE32-Ultra-DAC
https://www.businesswire.com/news/home/20130916005246/en/Yamaha-Selects-ESS’s-SABRE32-Ultra-SABRE-Premier

ES9016s DAC chip was eventually replaced by 9028 and 9038 series as customers always expect "cutting edge" technology although it sometimes doesn't really provide better results, as we have seen from Amir's measurements of DX/ vs DX7s.

Not trying to start an argument about the 9016 DAC chip. The designer JW had lengthy discussions on the 9016 shortcomings some five years back when he was released from the Creek project. I'm not enough motivated to search for them (and too lazy). If anyone is interested the discussion took place here https://www.pinkfishmedia.net/forum/ around 2013.

 

[Krunok]

Not trying to start an argument about the 9016 DAC chip. The designer JW had lengthy discussions on the 9016 shortcomings some five years back when he was released from the Creek project.

You didn't provide any argument. Maybe it was easier for him to blame the chip than himself for his release..?

Anyway, unless there is a proven fact that ES9016S has shortcomings I don't really think there is anything to talk about on that subject.

 

[Krunok]

Just to add some light to ESS family of chips: their DACs are basically divided into 2 groups, mobile DACs and Audiphile DACs, main difference being that mobile DACs are 2 channel while Audiphile DACs are 8 channel to be able to accommodate multichannel formats.

ES9016 family has slightly lower specs than ES9018. The chip used in D10 (ES9018K2M) is mobile DAC while the chip used in Matrix Mini-i Pro 2 is (ES9016S) Audiphile DAC.

ES9038Q2M used in DX7s is also a mobile DAC, while ES9028PRO used in Matrix Mini-i Pro 2S is Audiophile DAC, obviously from the 9028 family.
ES9028Pro is pin compatible with previous generation ES9018S and the same is the case with ES9026PRO and ES9016S.

As far as I know all these DACs work extremely well and are used in a broad range of products without any issues.

Generally speaking, each next generation brings better SNR, THD and ability to handle higher bitrate formats, so yeah, soon we can expect DSD2048 to be possible. What implications will that have on real life listening I can't possibly imagine.

 

[Arnold Krueger]

True but why would an amp send transients through its input connection. Output, I could understand. But input???

Those transients can damage audio gear as they do not expect their output to be driven by anything.

I like to duplicate that problem. Is the AMP reasonably priced so that I can purchase it?

If a device has a single supply or strong variations in input circuit voltages during power up it will pass a signal out through its input coupling capacitor at that time. The source impedance of this voltage pulse is likely to be high.

 

[Krunok]

Speaking about THD, here is an interesting figure: Bowers & Wilkins specifies for their 803 D3 loudspeaker (which costs app $8500, so it is not meant to be a child toy) these figures for THD:

2nd and 3rd harmonics (90dB, on axis):
<1% - 70Hz - 20kHz
<0,3% - 100Hz - 20kHz

http://www.bowers-wilkins.com/Speakers/Home_Audio/800_Series_Diamond/803-D3.html?c=set

In that context D10's THD of 0.0008% sounds ridiculously small. That is why I think it is wise to save money for decent loudspeakers, as there THD really matters.

 

[amirm]

Here is the paper digest on audibility of op-amps under double blind test condition: https://www.audiosciencereview.com/...digest-do-audio-op-amps-sound-different.2625/