[wolfstar76]

EX vs R2R DAC such as Musician Draco?

 

[xtiva]

EX vs R2R DAC such as Musician Draco?

For Muscian or Denafrips, I avoid them... sound quality is one thing but the fact that you cannot connect both RCA & XRL at same time, makes it very limiting... if you will only use one output or dont mind changing output manually, Draco or Denafrips Ares might be a good choice...

 

[JerryLeeds]

EX vs R2R DAC such as Musician Draco?

Denafrips new Ares 12th anniversary edition… all the way for me … now with I2S input

 

[AudioQuality777]

Here are some snapshots of the current modifications.

All the yellow caps are Rubycon UFG aka "FineGold". The originals looked similar, but are UFW, which is one grade lower. The FineGolds are very nice! They produce an airy, clear sound. Their weakness is the treble, which they present as rather fragmented. They need a matching bypass.

Generally, the SMSL designers thought along the lines of "more is better" and oversized a number of caps. That did not do any good. The sound became too thick and dark for my taste. I downsized the 4 black Nichicon UKZ (upper right) from 470 to 220. The 220 is an outstanding cap that sounds quite different from its UKZ brothers. This is a late discovery as I used to prefer the 100, which sounds grainy in comparison.
Also, the big yellow ones are 4700 originally. The AKM datasheet suggests 2200 and that's what I picked. The FineGolds are bigger and that size fits perfectly.



The most delicate operation was the IV stage, which I modified to get rid of the coupling caps. A 2.5V offset can be trimmed down to 0 by replacing one miniscule resistor and connecting a trimpot instead. The common mode voltage in the IV stage is very sensitive and Black Gates 33µ help to beef up the midrange. The parts and traces are very small there and a certain danger of breaking things has to be dealt with.
I am awaiting "boutique" Amtrans AMCH polystyrene caps, which will replace the cheap no-name red thingies in the passive filter.



Another new discovery is the Rubycon MU SMD filter cap. Those are very expensive (the 2.2µ is 5$ a piece), but man, do they clean up! Two MU are sitting under the aforementioned Nichicon KZ (12V supply) and one is behind the DCDC converter for the digital domain. See the following image. There, the SMSL designers ignored the recommendations from the datasheet and deployed too much capacity (144µ total instead of the 44 recommended). I reduced it to two BlackGates 33µ plus the MU 2.2µ.



And finally, another important spot is the 12V decoupling in front of the two 11V regulators that supply all op amps. FineGolds 100µ plus 22µ POSCAPs are my current favorites. That's not yet the final configuration, I think.



So, what has improved? First and foremost the clarity and contours. I complained about the missing attack and that the original sound was soft. I also criticized a lack of texture in the upper midrange. Well, the AKM is able to attack all right and it can produce sparkling treble, fast transients, and plenty of texture! However without the crispiness of the ESS DACs. Instead with its characteristic "velvet sound" as they call it. This thing pushed my system forward and not in a subtle way. The spread between bad and good recordings has widened. Old stuff sounds old as it should. It's not spiced up. More realistic and believable in total.

If you ask me, the AKM design has much more potential than the ESS. In the D400, the power supply is just so sufficient. With more effort - which is not feasible here - the AKMs could do even better, I think.

You haven't found a way to mute the sound color.It is always on at 5.6 MHz.How to disable.?

 

[HermannS]

No, that's baked into the software.

 

[AudioQuality777]

No, that's baked into the software.

It plays great with this setting.Noticeably more detailed, especially in the upper middle and with a denser sound than the DAC on the ES9038PRO.And the overall detail is noticeably higher.DAC for 2 pcs AK4499EX excellent weight.

 

[HermannS]

Amir has measured the EX and it is flawless. https://www.audiosciencereview.com/forum/index.php?threads/smsl-d400ex-dac-review.40116/
Unfortunately, he overlooked the sound color settings and we are still in the dark as to what they do.

This proves that measurements don't tell you how a DAC sounds. The D90SE and the D400EX measure objectively identical or the differences are far outside the audible range, for that matter. And yet, they are so utterly different animals, it's almost ridiculous.

 

[Zbell]

Amir has measured the EX and it is flawless. https://www.audiosciencereview.com/forum/index.php?threads/smsl-d400ex-dac-review.40116/
Unfortunately, he overlooked the sound color settings and we are still in the dark as to what they do.

This proves that measurements don't tell you how a DAC sounds. The D90SE and the D400EX measure objectively identical or the differences are far outside the audible range, for that matter. And yet, they are so utterly different animals, it's almost ridiculous.

Perform a blind test consistently differentiating the 2 DACs and all the greatest minds at ASR will collectively be blown to smithereens.

 

[AudioQuality777]

Amir has measured the EX and it is flawless. https://www.audiosciencereview.com/forum/index.php?threads/smsl-d400ex-dac-review.40116/
Unfortunately, he overlooked the sound color settings and we are still in the dark as to what they do.

This proves that measurements don't tell you how a DAC sounds. The D90SE and the D400EX measure objectively identical or the differences are far outside the audible range, for that matter. And yet, they are so utterly different animals, it's almost ridiculous.

The ASR forum has descended to empty squabbles DAC the topic of why it is needed for $ 1000 if the measurements are the same for $ 250(Complete nonsense.).
Amir himself destroyed the meaning of his forum by leading users to this idea with the slogan that all dac have the same sound(Complete nonsense.).
But this is not the case at all.

 

[HermannS]

To be honest, I bought the D90SE to learn what a DAC with squeaky clean measurements sounds like, after the disappointing results for the Musetec DA005. Measurements are just fine as a basic prerequisite. A well-engineered device measures well. On top of that, there are things that still cannot be measured, which is actually disappointing. The discussions around sound quality will never settle.

 

[AudioQuality777]

In the meantime I got rid of the big blue coupling caps behind the I/V stage by modifying the circuit. I can trim down the DC component in the signal to low one-digit mVolts. The more I improve (that's an assumption ) the device, the smoother and richer in texture, especially the mids, the sound becomes. I still miss the transient attack of my former ESS DACs. The AKM does everything so differently that it is puzzling sometimes. The rhythmic structures of some tracks don't feel the same, it's weird.

Well, making a smoothie-DAC sound brighter is potentially a simpler task than taming down an aggressive one. Let's see where this is taking me.

It's very interesting, did you connect 400ex to the i2s input and what are the results?

 

[HermannS]

I don't have a source with I2S output.

 

[AudioQuality777]

An interesting device with an i2c output of SMS RO100 PRO .Suitable for 400ex.

 

[Baten]

I never understood why people would buy an USB device that outputs another connection (like I2S) to feed into a DAC that you would otherwise use with.. USB
This is really just spending money just to spend money, your DDC is still connected over USB. At best you get super marginally better jitter, at worst nothing at all.

 

[HermannS]

I never understood why people would buy an -USB- device that outputs another connection (like I2S) to feed into a DAC that you would otherwise use with -USB-
This is really just spending money just to spend money, your DDC is still connected over USB. At best you get super marginally better jitter, at worst nothing at all.

That's right. I2S needs to be integrated into a network streamer if anything. And I will not replace my excellent dCS network bridge only to have i2s. I am interested in a comparison to AES, but that must wait for a couple of years, and who knows what the connection de jour is then.