Denafrips ARES II USB R2R DAC Review

[Jimbob54]

Can anyone think of a good practical reason to explain this perceived huge difference in 2 DACs whose job is simply to convert digital signal to analog?

I can. But I'd like to hear from more Ares users who have also used other DACs in their set ups

 

[voodooless]

Can anyone think of a good practical reason to explain this perceived huge difference in 2 DACs whose job is simply to convert digital signal to analog?

Something is broken.. question is: what?

 

[Maki]

Can anyone think of a good practical reason to explain this perceived huge difference in 2 DACs whose job is simply to convert digital signal to analog?

I can. But I'd like to hear from more Ares users who have also used other DACs in their set ups

There isn't though my Ares occasionally freaks out and sends a massively distorted signal until I pause and restart the music.

 

[Jimbob54]

Something is broken.. question is: what?

Was one of my thoughts

Or poster has a very active imagination and is listening sighted.

Or the level matching undertaken is horrendously wrong.

 

[Jimbob54]

There isn't though my Ares occasionally freaks out and sends a massively distorted signal until I pause and restart the music.

Does it sound like it is coming from the neighbouring room though ?;-)

 

[voodooless]

Or poster has a very active imagination and is listening sighted.

That's what I would file under "broken" as well

@Teroz, any chance you can make a recording of the two devices? Frequency sweep would already be a first good step.

 

[MBL'er]

Can anyone think of a good practical reason to explain this perceived huge difference in 2 DACs whose job is simply to convert digital signal to analog?

I can. But I'd like to hear from more Ares users who have also used other DACs in their set ups

I had for almost two months both the D70S (my currently "resident" dac) and the Musician Pegasus, a close cousin of the Ares (but not identical).

In my 2 channel system they did not sound the same. Not even closely. Blind, sighted, level matched, with "audio friends", with my "industry guy" friend, whatever combination.

Transients attack, bass contour, microdetails, midrange presence, treble roll-off, imaging. Not any single FR area in which there was overlap between the two dacs.

In the end, the D70S resulted more engaging for 90% of the music I listen and I got tired of switching to the Pegasus dac for the residual 10%. Pegasus sold.

 

[Jimbob54]

I had for almost two months both the D70S (my currently "resident" dac) and the Musician Pegasus, a close cousin of the Ares (but not identical).

In my 2 channel system they did not sound the same. Not even closely. Blind, sighted, level matched, with "audio friends", with my "industry guy" friend, whatever combination.

Transients attack, bass contour, microdetails, midrange presence, treble roll-off, imaging. Not any single FR area in which there was overlap between the two dacs.

In the end, the D70S resulted more engaging for 90% of the music I listen and I got tired of switching to the Pegasus dac for the residual 10%. Pegasus sold.

Which is interesting. But I really want to understand what could cause that night and day differences to 2 devices which (if both are doing their job right) should be, if not identical, very hard to discern between them by ear. I dont even really mind , if differences are perceived, which is preferable- I just dont understand how they can sound so obviously different.

 

[Jimbob54]

That's what I would file under "broken" as well

@Teroz, any chance you can make a recording of the two devices? Frequency sweep would already be a first good step.

Just remembered. Teroz is comparing to SU9- which is affected by this issue.....

strange SMSL M500 high 3rd harmonic

I just got a brand new SMSL M500. it's the latest 1.7 firmware and 1.3 hardware version. (windows shows 2.06 in xmos utility). The following measurements are done with computer -> USB-> M500 XLR output-> ADI-2 Pro input -> computer. I use REW to generate the sine wave and plot the fft. If I...

www.audiosciencereview.com

So, possibly broken.

 

[MBL'er]

Which is interesting. But I really want to understand what could cause that night and day differences to 2 devices which (if both are doing their job right) should be, if not identical, very hard to discern between them by ear. I dont even really mind , if differences are perceived, which is preferable- I just dont understand how they can sound so obviously different.

I really do not know first hand.

A friend of mine, deep in electronics, told me that R2R dacs can have sample to sample differences in the resistor rails.

Based on measurements, in the audible zone both the D70S and the Pegasus should sound the same. But, then, the Pegasus somewhat smoothes out small details, which are less “present” than in D70S: rock tracks sound less engaging compared to the D70S, for example.

Or, another easy to spot issue, with some female voices, the Pegasus somewhat took out the “tone” part and left intact the “air” part of the voice emission, giving a sort of “disembodied”, or “sucked in” character to the singer. With some male voices, that worked fine, though.

Lower treble in the Pegasus had a perceptible roll-off, resulting in less sibilants compared to the D70S.

I do not know either if that is due to the high output impedance of the Pegasus (1250 Ohm according to various sources). My MBL pre has a 50.000 Ohm input impedance, which should work fine. But, still, it’s conjecture.

 

[gvl]

Peeps, are you running your Ares II in NOS mode? That would surely explain what you’re hearing.

 

[MBL'er]

Peeps, are you running your Ares II in NOS mode? That would surely explain what you’re hearing.

That could also be. The Pegasus in NOS mode sounded compressed on the top end and muddy in anywhere else.

 

[Pdxwayne]

Both from RCA 2.0v. I did use Testo 815 Sound Level Meter from office so i can get 0.1Db accuracy from headphones BUT none of these did matter. Whatever the levels were, ARES is just so much better sounding. If you dont own it yourself, all this dont matter at all..
Oh yeah the filters. Well they too didnt matter at all. SMSL filters do about 0.01% change to the sound if at all. Sound coloring off ofc.
The difference is literally like you go from 100€ closed back gaming headset to HD 650, or even to HD800. It is HUGE.
This is not some little nuance change..

Hmm, the sound meter has ±1.0 dB accuracy. So the difference can be as big as 2db between DAC. It might seems small, but I have another big thread comparing a new DAC I got with an old DAC. With just 0.2db single channel imbalance, I already sensed a different. 1 to 2db different will be very obvious different between DAC, if you are sensitive like me!

It would be great if you can capture output using ADC to REW and show us the freq response from 20hz to 20khz for each channel of each dac. Using a good ADC to capture should give you more accurate dB information than the sound meter for the full freq range. You might find channel imbalance issue. You might find freq range issue, etc. I am curious to see the REW charts.

 

[Teroz]

Hmm, the sound meter got ±1.0 dB accuracy. So the difference can be as big as 2db between DAC. It might seems small, but I have another big thread comparing a new DAC I got with an old DAC. With just 0.2db single channel imbalance, I already sensed a different. 1 to 2db different will be very obvious different between DAC, if you are sensitive like me!

It would be great if you can capture output using ADC to REW and show us the freq response from 20hz to 20khz for each channel of each dac. Using a good ADC to capture should give you more accurate dB information than the sound meter for the full freq range.

Oh the seller said its 0.1db accuracy. Oh well it dont matter, even when SMSL is way louder the ARES is still way better. Freq response is same (due ARS tests.), its all about soundstage, resolution and imaging. Ofc the instrument placement is far superior in ARES II. Idk how you can those from recorded sweep...

 

[Pdxwayne]

Oh the seller said its 0.1db accuracy. Oh well it dont matter, even when SMSL is way louder the ARES is still way better. Freq response is same (due ARS tests.), its all about soundstage, resolution and imaging. Ofc the instrument placement is far superior in ARES II. Idk how you can those from recorded sweep...

In my other comparison, I was thinking that my old DAC was better than the new one in the highs, in the echo trails, etc. When I captured using ADC and diaplay the freq plots with REW, I found that my new DAC has channel imbalance. The right channel is lower in dB than the old DAC. Just a tiny imbalance is already causing a difference in what I heard.

So although ASR measurements look fine, the actual DAC you get might not function as good. You might get a DAC with large channel imbalance, etc...You can find that out using REW.

 

[kchap]

I really do not know first hand.

A friend of mine, deep in electronics, told me that R2R dacs can have sample to sample differences in the resistor rails.

Based on measurements, in the audible zone both the D70S and the Pegasus should sound the same. But, then, the Pegasus somewhat smoothes out small details, which are less “present” than in D70S: rock tracks sound less engaging compared to the D70S, for example.

Or, another easy to spot issue, with some female voices, the Pegasus somewhat took out the “tone” part and left intact the “air” part of the voice emission, giving a sort of “disembodied”, or “sucked in” character to the singer. With some male voices, that worked fine, though.

Lower treble in the Pegasus had a perceptible roll-off, resulting in less sibilants compared to the D70S.

I do not know either if that is due to the high output impedance of the Pegasus (1250 Ohm according to various sources). My MBL pre has a 50.000 Ohm input impedance, which should work fine. But, still, it’s conjecture.

What I've read on the web, a summary of that could possibly make R2R DACs sound different to over-sampling DACs:

• Non-linearity/mismatch of the resistor ladder. Still, the Denafrips has about 18 bits of resolution, reasonable.
• Most ladder DACS are Zero order sample and hold. Unless compensated by the anti-alias/LPF they have a natural frequency drop of over 3db at the Nyquist frequency. The Denafrips is one that doesn't, see next point.
• The Denfrips uses a cunning over-sample technique to implement a First order sample and hold. Less frequency drop I think, but still needs an anti-aliasing filter.
• To the current golden ear brigade NOS also means no anti-aliasing filter. HF noise and artifacts, possible frequency drop.
• Some R2R DACs do not use output buffers. The may have trouble driving some amps and maintaining a flat frequency response.
• In NOS mode and a lack of output buffer there behaviour might actually be better than oversampling DACs when dealing with intersample overs, the scourge of modern CDs.
• We are all easily duped by our own senses.

Take your pick.

 

[Snoopy]

I image the next more expensive thing suddenly sounds even "bigger".

Something is broken.. question is: what?

Sounds like for some people a DAC makes more of a difference than other Loudspeakers or room treatment, Dirac etc..

 

[Purité Audio]

Only in their imagination.
Keith

 

[zepplock]

An engineer in me is desperately demanding a solution to this puzzle.
One thing that could work is to do 1khz size and multitone, into a dac->amp->speakers, into a calibrated mic, into ADC.

It's also possible there's some sort of difference that we are not measuring, impulse? Cause I have no idea how to measure

soundstage, resolution and imaging

and other "feelings"

 

[Teroz]

Still funny how we got no answer how you can measure soundstage, imaging and instrument separation. Yeah some say they can but no actual measurement is shown..