Review and Measurements of SMSL VMV D1 DAC

[amirm]

To be clear, this is a "hump" as it elevation, not a mains hum.

 

[Veri]

As with SU-8, you can choose the usual filter settings and special ESS DSP functionality to emulate tube sound and such with distortion products.

Hey amir can you elaborate on this special ESS DSP functionality? First time I hear about something like it

On the topic of these measurements I think SMSL has done quite good. While some Chinese builds on the 9038Pro like the HiFiME failed in a single chip set-up, this D1 DAC performs really well with a dual chip design! Lots of attention was paid to the powering of the device I'm sure.

Too bad about the unbalanced performance though, not everyone uses a balanced amplifier.. Still, quite respectable compared to the Benchmark as well as the Auralic Vega: both go around $2000 and the D1 comes very close, if not about equally performing in balanced.

 

[Krunok]

The latter. Problem is there at all sample rates.

But if you limit the bandwith to 22kHz while feeding 192kHz sample rate it wouldn't show, as it didn't show for 44.1kHz, right?

I'm aware that would kind of defeat the purpose of 192kHz, but I anyhow don't believe in those high sampling rates..

 

[oliveoilthansnakeoil]

Too bad about the unbalanced performance though, not everyone uses a balanced amplifier..

I agree. Most home audio amplifiers have only rca input. If someone uses active monitors, it would be good match.

 

[Krunok]

Too bad about the unbalanced performance though, not everyone uses a balanced amplifier.. Still, quite respectable compared to the Benchmark as well as the Auralic Vega: both go around $2000 and the D1 comes very close, if not about equally performing in balanced.

IMHO those spikes are well below hearing threshold and well above 20kHz so I wouldn't really worry about them, especiallly as 44.1 kHz performance looks excellent.

 

[bennetng]

@amirm
Are those spikes the same thing?
Could you unplug the USB cable and measure coax and toslink as well? Thanks!

 

[Veri]

Yes I agree I'm curious to see if there's any difference using optical/coaxical, if the 44/192Khz spurious tones and jitter problems persist.

 

[maxxevv]

I agree. Most home audio amplifiers have only rca input. If someone uses active monitors, it would be good match.

One could use XLR to RCA convertors in this case and get maximum performance perhaps ?

 

[Veri]

One could use XLR to RCA convertors in this case and get maximum performance perhaps ?

Pretty sure that would not go well? You can have a single end xlr output on an amp just for the connection but afaik you can't plug truly balanced into single end. Things might/will break.

 

[amirm]

@amirm
Are those spikes the same thing?
Could you unplug the USB cable and measure coax and toslink as well? Thanks!

Sure, but I am going on a little road trip so likely won't get to it until the end of this week.

 

[amirm]

Hey amir can you elaborate on this special ESS DSP functionality? First time I hear about something like it

See the review thread of SU-8 for more detail. Briefly, the ESS DACs have a distortion mitigation logic. This changes the shape of the transfer function of the DAC. SMSL has used this not to correct distortion, but to add it in different ways. Some of it for example attempts to emulate "tube sound" by increasing second harmonic distortion. See examples here: https://audiosciencereview.com/foru...measurements-of-smsl-su-8-dac.3778/post-91324

When I get back later in the week, I will capture the profile for all of them and post.

 

[Ron Texas]

Pricey for Chi-Fi. Noise in the balanced output underscores the need for independent measurements.

 

[helloworld]

Pretty sure that would not go well? You can have a single end xlr output on an amp just for the connection but afaik you can't plug truly balanced into single end. Things might/will break.

@amirm could you confirm that? Could you explain more detail if the xlr to rca or rca to xlr would affect the sound quality cause my monitor only have an xlr input but I am thinking of using a Pi DAC which only have rca out.

 

[RayDunzl]

you can't plug truly balanced into single end.

No, but you can pick the + signal from the balanced output and use that. Substitute the - output if you want to introduce an inversion in the signal chain.

Unbalanced has + signal and Ground.

Balanced has + and - (same signal but inverted), and signal and ground.

As an example, Rane's diagrams (figure 3) should cover it.

https://www.rane.com/note110.html

Pick your wiring according to the conditions of your application.

Things might/will break.

I would think no, unless the voltages are problematic

 

[restorer-john]

No, but you can pick the + signal from the balanced output and use that. Substitute the - output if you want to introduce an inversion in the signal chain.

If the balanced output is derived from a proper balanced line transformer, that won't work.

 

[RayDunzl]

If the balanced output is derived from a proper balanced line transformer, that won't work.

Then you ground the negative side of the transformer.

What goes wrong at that point?

 

[restorer-john]

Ground one side of the floating secondary on the transformer, it will work. Ground an opamp derived balanced output (most of them out there) and you might cook the device (driving into a short circuit).

Let's face it, most balanced outputs (on D/As CDPs etc) are simply twin opamps running one inverted, one non-inverting, off a single ended output. They are worse performing on the whole than the unbalanced RCAs. Even transformer balanced outputs come via single ended sources, so all the non-linearities of the transformer compromise the performance.

Some have come from balanced (twin or more) D/As, others come from paralleled D/As. There is some merit in balanced derived from twin D/As running at opposite polarity to one another, but the source itself, the digital signal, is not balanced- it is a ground referenced signal all the way through.

I have plenty of balanced (XLR) inputs and outputs on various bits of (HiFi) gear. Not one is anything other than 'faux balanced'- even the ones with expensive LCOFC canned transformers.

 

[Blumlein 88]

I'm thinking those spikes are common mode noise which is why they disappear balanced. If you go XLR to RCA, you'll lose the other phase to cancel out the common mode noise, and the spikes will be back. Even faux-balanced has some common mode noise rejection.

 

[gvl]

Is there a way to turn off the ASRC in the ESS chip, a.k.a. run it in "sync" mode?

 

[Deleted member 2944]

I'm thinking those spikes are common mode noise which is why they disappear balanced. If you go XLR to RCA, you'll lose the other phase to cancel out the common mode noise, and the spikes will be back. Even faux-balanced has some common mode noise rejection.

+1 on that. A very good observation.

Regarding grounding of the "-" output of an active balanced output scheme. Nearly all circuitry of this type would have a build-out resistor for protection and to decouple capacitive loads. All professional gear is built to tolerate this type of "conversion" adapting. Highly unlikely any active circuitry will be damaged. However, a distortion increase might be noted from something like a dual op-amp device.