SMSL VMV D3 Review (R2R DAC)

[DHT 845]

R2R sucks !!!

 

[Nango]

I expect an official from SMSL to intervene soon and give us kind of explanans.

 

[NYfan2]

I gave it the headless panther though it is not a broken product it is a poorly performing product.
For me it would be more correct if I could vote:

headless = broken product (for the broken/bad designed/snake oil products)
bad/poor = poor performing product
fine = OK performance but not great
Great = best in class product.

Just my 2 cents.

 

[bidn]

The well-known, practical pitfall which makes R2R DACs fail compared to Delta-Sigma DACs is the issue of very accurate resistance values, leading to unacceptable conversion errors, and this is shown so perfectly by the following two measurements by Amir:
- only 11 bits of linearity!
- the abnormally (unacceptable for me) jagged intermodulation curve, showing again the non-linearities = the practical impossibility to get each resistor to meet its expected resistance value within an acceptable error margin.

( Nothing against SMSL, it's only that the nice concept of R2R cannot be implemented as it could be on paper, unlike what is achievable in practice with ΔΣ).
Head-Fi and co, educate yourseIves!

 

[aj625]

R2R sucks !!!

try posting in some other forums and see how r2r brigade reacts.

 

[xaviescacs]

I voted headless because of this:

Distortion is below noise floor of 16 bit audio which people listen to. So there is no way that is what they are hearing.

If the distortion is not audible, what is the benefit of this device?

 

[Kegemusha]

and what a price, US $3,499 !?

 

[JSmith]

what is the benefit of this device?

Apart from parting with more money when there are multitudes of better performing DAC's... nothing, nothing at all.

But, you know... higher price = better, doesn't it?

I understand SMSL are serving a niche of a niche market here... but these are very very expensive for naught.


JSmith

 

[ZolaIII]

I mean, it's not like one would go down to the store to purchase a new refrigerator, then suggest to the salesman you want a 30 year old model

This worked for 50~70 years so far and still do in a storage rooms, garages and such burying 50 years younger Samsungs but that's another story.
For me CS43131 is a great DAC design to the extent we so far didn't find broken or bad design with it.
The VMV D3 isn't broken nor it performs bad but for two orders of magnitude less money you can get a DAC dongle that performs better (with mentioned CS43131) and for an order of magnitude a top performing one with lot of folklore with it.

 

[xaviescacs]

Apart from parting with more money when there are multitudes of better performing DAC's... nothing, nothing at all.

But, you know... higher price = better, doesn't it?

I understand SMSL are serving a niche of a niche market here... but these are very very expensive for naught.


JSmith

Yeah, I mean, if there is some audible distortion, one can justify a special sound signature, but if there isn't, then it's just marketing and design. Perhaps they have gone too far and it's too transparent for this market segment.

 

[Limopard]

R-2R love !

I also think R2-D2 is much better. At least in cuteness.

 

[Limopard]

To be serious: I'm also curious about this R2R stuff. But at best I consider buying an used CD player from the 90's for 50 - 80 EUR or a cheap DIY kit.

 

[bidn]

Yeah, I mean, if there is some audible distortion, one can justify a special sound signature, but if there isn't, then it's just marketing and design. Perhaps they have gone too far and it's too transparent for this market segment.

Even if their Signal to Noise Ratio may be fine, R2R, due to the conversion errors (see my post above #24, https://www.audiosciencereview.com/...smsl-vmv-d3-review-r2r-dac.30277/post-1063003)
R2R DACs will not properly convert the digitally recorded information into an analog signal, so I can imagine fake or deluded reviewers having a justification to pretend hearing "microdetails" (! I hate this word) they had never heard before...

This was actually the case with the owner of the TotalDAC DAC tested by Amir: The owner, who was very nice to have this super expensive DAC shipped from France, and honestly and humbly accepted the bad measurements by Amir (unlike the angry guy running the company) would say that he still liked it because with it he was able to hear things in the recordings which he would not hear with other DACs. I can imagine this to be real, but then as corresponding to those conversion errors typical of R2R DACs (due to the lack of accuracy of the resistance values), i.e. artefacts created by the R2R DAC and not conform to the recorded music.

I think that since the come of ΔΣ, no one in search of high fidelity, true to the original recording, should buy a R2R DAC.

 

[fastfreddy666]

Delta Sigma converters are bad. Go cry to momma. But why is every DAC chip manufacturer using the technique?
Yes in the beginning there were 1 bit converters and that's a bad idea. I had a CD player once which was using it.

The Delta Sigma design is still being improved upon as newer chips (if implemented correctly) have shown better measurements. How?

- Oversampling (ratios of 64 or more are not uncommon in audio DACS)
- Multi bit quantizers
- Feedback loops (Some audiophiles are not going to like this. But it works. It's used in almost every amplifier)
- Cascade structures like multi-stage noise shaping (MASH)
- Quadrature mismatch-error-shaping
- Dynamic element randomization (mismatching)
- DWA (Data weighted averaging)
- Better reconstruction filter designs (LPF)

Of course there are a number of trade-offs (fact of life) to be made because you have to consider energy consumption (heat) and things like that (thermodynamics baby)

Scientists can simulate DACS in software like Matlab these days. So if they have some clever ideas it's easy to try them out before being implemented in the actual chip. For instance ESS has patented its HyperStream II algorithm. Hyperstreaming isn't exactly new. The technique is also being used in machine learning. Your ears are no measurement devices. Trust the science and enjoy the music. If you claim to hear a difference this is probably caused by confirmation bias. Scientists are using controlled double blind tests for a reason (The placebo effect is real)

Opinions are not fact, dear subjectivists. Now get off my lawn

 

[Lupin]

This DAC is bad because it is not a "true NOS" R2R DAC..
It's true, just check on Head-Fi. True NOS R2R DACs are the pinnacle of D/A conversion according to the members there, one can not even begin to hope to hear all the microdetails and airy instruments without a true NOS R2R DAC..

 

[xaviescacs]

Even if their Signal to Noise Ratio may be fine, R2R, due to the conversion errors (see my post above #24, https://www.audiosciencereview.com/...smsl-vmv-d3-review-r2r-dac.30277/post-1063003)
R2R DACs will not properly convert the digitally recorded information into an analog signal, so I can imagine fake or deluded reviewers having a justification to pretend hearing "microdetails" (! I hate this word) they had never heard before...

This was actually the case with the owner of the TotalDAC DAC tested by Amir: The owner, who was very nice to have this super expensive DAC shipped from France, and honestly and humbly accepted the bad measurements by Amir (unlike the angry guy running the company) would say that he still liked it because with it he was able to hear things in the recordings which he would not hear with other DACs. I can imagine this to be real, but then as corresponding to those conversion errors typical of R2R DACs.

I think that since the come of ΔΣ, no one in search of high fidelity, true to the original recording, should buy a R2R DAC.

This is very interesting. So it would still have a chance to be differentiated from a 50$ DAC in a demo. I guess the PS Audio guys wanted to be sure it can be differentiated and added a ton of distortion.

 

[aj625]

I also think R2-D2 is much better. At least in cuteness.

 

[xaviescacs]

very accurate resistance values

That explains why this people claim to have a very good resistors and such. Thanks for the insight.

 

[aj625]

some forums are full of theories like r2r sounds more natural, nos has no ringing so sounds more natural, measurements don't corelate with sound quality but they don't have nay explanation or science behind all these. sbaf is one such group which is mainly run by schiit guys. even "so called engineers" " proudly" advertise poor measuring version of schiit dac as better sounding as compared to better measuring version ( less is more yggy dac vs more is less yggy dac). going by their logic, may be in future with improvement to measurements say in next 20 years dacs may become totally unlistenable.

 

[Vict0r]

"Not terrible" incarnate.