Soekris DAC2541 Review And Measurements

To anyone else receiving their 2541 from Denmark, beware, let it warm a little. Cold from the box:
-veiled/etched, analytical sound. limp bass.

After some moderate warming-up time:
-much tighter/cleaner sound and much improved bass

I quickly preferred the filter light off (the soft knee butterworth one). The crossfeed on the HP out I don't like too much, prefer software alternatives to be honest. It's a nice-to-have I guess but don't think I'll be using it. As for the HP out, it's fine. A good amplifier will still improve things by a lot, though; yes even the BAL out (which sounds surprisingly similar to the SE out).

In any case I'll audition it for some months to come, but some warming up or breaking in or whatever you want to call it is absolutely mandatory before judging this DAC. Once there, it is quite resolving and involving. I concur with the OP review. Truly top-notch toe-tappingness, really the opposite of emotionless super-feedback dac/amps the like are trendy nowadays. Perhaps NOT the thing you want to listen to during maximum focus/work !!

I feel like my references, Audial (laid-back TDA1541) and PecanPi USB (high-resolution burr brown sound) are really quite different in presentation, so I'm thinking of simply keeping all three.

 

What the eff. I was chilling out listening to Spotify piano selection and this song fucked me up:
https://open.spotify.com/track/22N92htJMGsPVDLwh3PJuH?si=b7El_tT-QkuN1wJlJsMjPQ

From 07:00. I thought the soekris USB was being overloaded or hitting the clip limit. But no. The song is full of LOUD distortion? DON'T listen to it on expensive speakers. But for my sanity, can someone check this song? last 1.5 minute is super glitchy, right? Quite the scare I thought something was very wrong for a moment there "digital remaster", my ass!

 

Yeap that was awful, i didn't even, at the first moment, thought about the music itself being "broken" but maybe my PC was acting up.

 

Just to re-iterate, left the soekris 2541 off couple of days and once again it sounded pretty "meh" cold, left it on overnight and today it's back to its awesome self. I'm surprised I haven't seen much mention of this. Cold it's like a 6/10 to me, after 12-24hours of powering on I feel like it's closer to a 9/10 presentation (don't believe in perfection, haha).

A heads up, the soekris website mentioned the 2020 price (1100 excl VAT) was an introductory price, I see it has now officially increased to 1150 excl VAT. That's around €1460 including Danish 25% VAT and €20 shipping in Europe, was €1400 all-incl before. They're also all out of stock all those positive reviews must have paid off!

 

Wow, at the intro price I paid in US I really snagged a deal then.

 

Well, customers outside of the EU do not need to pay that value added tax which sweetens the deal by a lot..

 

I was lucky enough to be included in the loaner tour of this dac as a contributor. I received it today. It's probably been warming up for around three and a half hours. As I don't have my Pi2aes fully setup yet, I've been listening through the USB port of my computer, A/B-ing to the Bifrost 2 with SE connections to the ZMF Pendant and Auteur. These are preliminary, first thoughts:

In favor of the Soekris 2541:

-If there is one word to describe the sound here, it is "Clean". And not in an unnatural, artificial sense: everything just seems clearer and better defined compared to the Bifrost 2. Makes the Bifrost 2 sound a little blurry and grey in comparison. This may have to do with Unison USB.

-Soundstage is blown up in width compared to the Bifrost 2. Layering is really improved - the Bifrost 2 almost has a "Wall of sound" effect compared to the Soekris, something I never thought I would say!

-Microdynamics noticeably improved. Small nuances in volume level much more easily discerned.

-More coherent overall presentation. With the Bifrost 2, I sometimes find that the bass is a bit too prominent and hits -too- hard, especially with the Pendant. As a result, it sounds a little strained and unnatural.

-Better resolution, especially in the treble. Also, texture of treble is better discerned. Probably has to do with detail retrieval, but also leads it to have a more pleasing treble timbre. In addition, treble is both less rolled off yet also not any more fatiguing.

-Quicker transients yet without any noticeable "overshot" or "ringing". Very nice, and doesn't at all have that "edgy", "square" sound that some find unnatural.

In Favor of the Bifrost 2:

-Bass has better slam, texture, and dynamics. The distinct timbre of different bass instruments is better perceived. Bifrost 2's bass is also noticeably better extended to my ears.

-More visceral, in-your-face presentation, for those who like that. But it's combined with slower transients and a thicker overall sound - leads to a very different sonic presentation.

-Richer midrange. The Soekris sound is clean, and, for the most part, supremely natural, but sometimes one just wants a bit more warmth and body in the mid-range. The Bifrost 2 does not suffer from this problem at all.

-Better at rendering older recordings listenable and pleasurable.

-Vocals sound more realistic. This probably has to do with the mid-range richness.

Keep in mind these impressions are preliminary. I will give a more detailed review once I have assembled the Pi2aes and given both a more extended listen. I do think that the Soekris 2541 is, overall, probably a superior dac (though definitely not 2x as good!), but I'm not sure it works better in my system. With the Soekris 2541, I am jamming; with the Bifrost 2, I am contentedly smiling.

 

That's a pretty damn good summary

 

@dematted Any chance you can also share some impressions of using headphones directly with the 2541?

 

You should try your hand at journalism, well written.

 

Does anyone have any insight into why a switching power supply was chosen? Most other R2R-type DACs seem to use linear power supplies with big toroidal transformers, which seems more consistent with "traditional" audiophile design. Just curious as to why Soren decided to go in a different direction than his competitors, surely there's a good reason that I'm not aware of.

 

Think he said he did it so he didn't have to spend a fortune on shipping his product globally

 

I think he also just wanted to streamline his DAC line-up as far as components go. In any case the 2541 filtering and power reserves seem to be fully redesigned from the ground up, so I don't believe the input switching supply is much of a bottleneck, if it is one at all, looking at the reception of this DAC so far. I'm sure atomicbob's upcoming measurement suite will shed further light on how good the internal PSU is, I'm expecting no issues but we'll see

 

I might have mentioned it before, but there are a number of reasons:

1) Smaller and Lighter, which of course mean lower cost shipping worldwide.
2) Worldwide powering, most use a switch to select 115V/230V, or even worse, have two different versions (like Schiit). Mine just need the correct Power Cable.
3) Less Noise, yes a correct designed and used switcher have less noise, especially not the 50/60 hz and its harmonics.
4) No power wasting, heating generating large Linear Regulators. Although I do have small local regulators.

And I don't really care much for "traditional" Audiophile designs, I'll rather do smart modern designs....

 

Well, this was unexpected.

The conventional wisdom tends to be that Unison USB is one of the best implementations around. I was expecting the 2541's small lead over the Bifrost 2 to only expand as the inherent limitations of its USB transport were lifted by Pi2aes. So whereas I initially wrote Soekris 2541>Bifrost 2, now I'd be confidently reporting Soekris 2541 >>> Bifrost 2.

But something else happened. Don't get me wrong: the Pi2aes expands the abilities of the Soekris 2541. It makes it even faster, even blacker, and even more dynamic. But plugging in the Pi2aes into the Bifrost 2 seemed to bring its technicalities almost up to the level of the Soekris 2541 through USB, while keeping the warmth, immersiveness, and overall sense of realism that the Bifrost 2 excels at. It's still a slower, thicker sounding dac, but with the Pi2aes, I no longer feel like the background to the Bifrost 2 is overly "Grey" and the microdynamics "muted" in comparison to the Soekris 2541. It -is- still greyer (But significantly less so), but this greyness comes with additional plankton, texture, and sense of tonal weight and density.

So, two conclusions.

Firstly, I don't think Unison USB is all that great. I think that whatever USB transport is in the Soekris 2541 is ahead of Unison! So, grats to Soekris on that one.

Secondly, with the Pi2aes, I don't think either of these dacs are clearly superior to one another. I might buy a Soekris 2541 down the line because it fits my sonic taste a bit more and I think it may synergize a bit better with something like the Auteur, but I can't really say that I feel like I'm -losing- out on anything too significant when listening to the Bifrost 2 with the Pi2aes. At the very least, the losses in some areas are made up by gains in others; something I couldn't say when using USB with these dacs. In short, I think whether one picks the Soekris 2541 or Bifrost 2 with the Pi2aes should be generally a matter of system synergy rather than anything else. If I was rocking a Focal right now I'm pretty sure I'd want to keep my Bifrost 2; with the Auteur, the Soekris 2541 looks very tempting, and I am definitely going to be quietly resentful of @dasman66, who I will be sending this off to shortly...

 

Another teaser, supporting Søren's statement:

dac2541 Residual Noise - AES input Balanced output

Note this scale has been expanded down to -180 dBFS as much of the dac2541 residual noise is below -160 dBFS where my standard measurement suite displays.

*edit*
I should point out the numbers are dBu, absolute while the FFT is dBFS. dac2541 0dBFS is +13.6 dBu.
Now think about total noise at -111 dBu, approximately 2.09 uV.
And how about the power line components at -133 dBu, 170 nV!!!!

 

A couple of related supporting observations: 1) When I got Unison on my Yggdrasil (I was a tester), combined with an Allo USBridge with LPS, it improved on several other Yggdrasil sources. However, Pi2AES was even better, with the kind of improvement you describe but less marked. 2) I own a Soekris dac1541, which among all R2R DACs I've tried is the only one where USB is competitive with other inputs.

 

How does this dac compare with the chord qutest in terms of sound when comparing rca.

Thanks in advance