Weiss Engineering DAC50x (DAC501 and DAC502) Review
Dec 7, 2022 at 9:00 AM Thread Starter Post #1 of 377
Jun 9, 2011
Head-Fi LA

NOTE: If you can't see the embedded video above, please CLICK HERE to see the video.​

While new flagship headphones are released fairly often, new reference DAC and amplifier releases are much fewer and farther between. And yet, in this day and age where digital audio is ubiquitous, they are no less important to a Head-Fier's signal chain.

On this episode of Head-Fi TV, we take a look at the DAC501 and the DAC502, the latest flagship DACs from Weiss Engineering's DAC50x series. Join us as we see, and hear, if it has what it takes to meet the needs of even the most demanding and discerning of headphone enthusiasts.

Weiss Engineering DAC50x (DAC501 and DAC502) Review - Head-Fi TV - produced by Brian Murphy and Joe Cwik
Last edited by a moderator:
Dec 7, 2022 at 9:01 AM Post #2 of 377
Weiss Engineering DAC501/DAC502 Review Supplement: Versus Chord Electronics's DAVE

Given the DAC502’s price range and usage case, the closest direct competitor is Chord Electronics’s Dave, and in head-to-head comparisons, I found that the DAC502 more than held its own.

In terms of frequency response, I found the DAC502 to be more linear, especially through the mid-range


Though it’s not the most pristine of recordings, Lagoya & Presti’s rendition of Bach’s English Suite No. 3 — as transcribed for two guitars — remains an unparalleled example of exquisite harmony.

It is here that the Dave’s upper mid-range emphasis actually worked against it. In certain contrapuntal passages, especially those with considerable differences in note pitch, the lower notes sounded noticeably recessed… so much so, that either Lagoya or Presti (I can't be sure which) sounded physically more distant within the stage.

The DAC502 didn’t exhibit this quirk at all. It’s even-keeled mid-range linearity kept Lagoya and Presti as the marvelous duet that they were, on-stage together, as they always should be.

When it came to detail, the DAC502 often out-resolved the Dave, particularly when it came to bass notes. In addition to flat-out hitting harder, the DAC502’s bass presentation felt more incisive, more visceral, being both tighter and more textural.

If you primarily listen to Pop, K-Pop, Hip-Hop, Rap, and other highly-produced genres, this won’t benefit you greatly.

But if you have a deep appreciation of sophisticated electronic artists, who place a great deal of emphasis on sound design, then the DAC502 definitely has something to offer over and above the Dave.



For example, with Four Tet’s continuing exploration of found sounds — as well as Amon Tobin’s experimentation with meticulously constructed sounds — the DAC502’s textural advantages allowed me to enjoy their artistry all the more.

And finally, the DAC502’s precise imaging characteristics rendered placement more accurately in live recordings; whereas the Dave tended to be more diffuse and dreamy in disposition.


Listening to Sarah Jarosz’s Mansineedof, live at The Troubadour, we can hear the DAC502 image that performance accurately. We’re right there, sitting on stage at Sarah’s feet, with Nat Smith a few feet stage left, and Alex Hargreaves a few feet stage right.

By contrast, the Dave presented this more expansively, with all three players approximately where they should, but not as sharply in focus. I should probably mention that I know Doug Weston’s Troubadour fairly well. It’s actually my favorite live venue back in Los Angeles. And I can tell you that neither the stage, nor the venue itself, is as big as the Dave would have us believe.

Now, none of this is to say that I found egregious faults with the Chord Dave. not at all. It remains a very competent DAC in its own right, so could be fair to say that the DAC502 simply renders music differently?

That said, with the DAC502 offering better linearity, enhanced detail, and superior imaging to my ears, I consider it to be the more accurate of the two — especially from a critical listening standpoint — which is why I prefer it over the Dave… and why I consider it to be more suitable as a reference DAC.
Dec 7, 2022 at 9:02 AM Post #3 of 377

Weiss Engineering DAC501/DAC502 Review Supplement: Digital Signal Processing (DSP) Features

If you’re perfectly fine with the idea of shaping sound to suit your preferences, then Weiss’s 50X series DACs offer you a wealth of DSP options – like a de-esser, a headphone EQ, and even vinyl emulation – all of which are descended from their professional studio products.

While the DAC502’s web interface is much easier to work with, due to increased screen real estate, I've also come to appreciate that each and every DSP module is also fully addressable from the front panel as well… with the one exception that not all DSP modules are available in all output modes. For example, the Room EQ DSP module is not available in headphone out mode, nor is the Headphone EQ available in line out mode.

That said, let’s take a look at the ones available to us in headphone out mode… beginning with the first DSP module, and the one that I found to be the most useful, the DeEsser.


DeEssers are basic and essential components of an audio engineer’s toolkit, where they (as illustrated by their name) are used to de-esss audio… which is to say that they reduce or eliminate the sibilance commonly produced by the pronunciation of esss sounds, as well as a host of other fricative consonants. I found Weiss’s de-esser implementation to be particularly effective, owing to their use of real-time m/s-stereophony (or mid/side stereophony) encoding and decoding, instead of more conventional split-band or equalization based methods.

In taking a look at the DeEsser, we see two parameters available for adjustment: Mode and Amount.


With Mode, we’re able to choose between two settings: Surgical and Smooth. And while the DAC502’s DeEsser is not equalization based, it does help to draw an analogy between Mode and Q-value in this case, as Surgical and Smooth describe how narrowly or widely the DeEsser is to be applied. For me, Surgical was the clear choice, as I found Smooth to be far too broad a stroke… in that it removed not only sibilance but some detail as well.

For the Amount parameter, we’re offered a range with values. The most extreme of which is -33.0 dB at the bottom end of the range, and 0.0 dB (signifying no sibilance reduction) at the top of the range. Of course, the value you enter will definitely depend on the headphone you’re trying to compensate for.

In my case, the vast majority of my headphones are relatively neutral planars and warmer dynamics that aren’t particularly prone to sibilance. However, I just happen to have a Sennheiser HD 800 as well, and that is a headphone that many in our community would regard as the very pinnacle of sibilance.

Now, one might think that the HD 800 would require an extreme measure of reduction here - much as I thought before I experimented with the Amount parameter myself - but no. Surprisingly, I found -4.5 dB to be the ideal amount of reduction for me. That modicum of reduction was just enough to remove the sting, the sharpness, the involuntary wincing — without culling any appreciable amount of the HD 800’s sparkle and detail.

Knowing what I know now — and not being able to un-hear what I’ve heard — I’m not sure I can truly enjoy the HD 800 again without the DAC502’s DeEsser.

Vinyl Emulation​

For those of you who are enamored of vinyl, imagine the amazing convenience of having the vinyl experience available to you, at the push of a button. That’s exactly what I was hoping for when I first began this review, but did the DAC502’s Vinyl Emulation deliver on that promise?

Looking at the Vinyl Emulation DSP module, we find that there is only one adjustable parameter, that of Saturation. For those of you working in pro-audio, with plenty of experience using VST and AU plug-ins, this setting is similar to what you’ve worked with in the past.


For everybody else, don’t let the simplicity of a single parameter fool you into thinking there isn’t much going on here, because there is. Weiss’s Vinyl Emulation DSP applies RIAA Equalization and distortion, much as you’d expect. But there is also a Colour Stream block that consists of additional noise generators, resonance generators and an amplitude modulator.

Weiss-50X-Review_Supplemental_FP-Vinyl-Emulation .jpg

Luckily for us, all we need to do is adjust that one Saturation parameter between a range of -9.0 dB to 20.0 dB. And for me 3.0 dB was just right. Any less than 3.0 db only offered marginal “vinylization.” Any more than 3.0 dB, and I began to hear a hollow, flanging effect. But at 3.0 dB of saturation, the Weiss’s Vinyl Emulation was just wonderful. Synths were warm and analog sounding, the distortion was enjoyable and never offensive, and all hints of digital glare just vanished.

Incidentally, the Vinyl Emulation effect was relatively balanced tonally-speaking, which is to say that it was much more akin to, say, an Ortofon 2M Red (as opposed to a Grado Black cart or the like).

Having said all that, I have to admit that it couldn’t fully replicate all that I’ve come to love about listening to records. It didn’t replicate my childhood of laying down on shag carpet, staring at album covers and liner notes, while listening for hours on end without getting sick of my limited collection. Nor did it replicate the ritualistic cleaning, or self-denial of instant track selection, that has come to dominate my current vinyl listening sessions. And as strange as it might sound, it turns out that those aspects do actually matter. Sound alone was simply not convincing enough to preserve my suspension of disbelief.

The DAC502 did its part. The problem was me.

Creative EQ​

I'm not going into any great detail on the DAC502’s Creative EQ here. It’s a perfectly competent EQ. And it’s very convenient to have a universal hardware EQ without having to mess with a computer-based software solution.

So, just briefly, you can apply up to three bands of equalization. And for each band, you can choose the type of equalization, the frequency to center around, the amount of boost or cut you’d like, and of course the Q value. I definitely appreciated the fact that I COULD apply equalization, and adjust parameters, right from the front panel of the unit if I had no other option...


...but in practice, this was far better controlled from the DAC502’s web interface. I simply found it far too complex to manage, through the front touch panel.

Headphone EQ​

This next DSP module, Headphone EQ, should probably be re-named Headphone Compensation as that would be more apropos to its function.

Using frequency response measurements supplied by headphone manufacturers, it applies compensation curves to bring any of the supported headphones to a pre-defined neutral target. Now, I don’t know exactly what that pre-defined neutral target curve looks like, but I can tell you that it definitely sounds neutral to my ears. And before you ask, no, it does not sound like the Harman curve.


Running through the list of currently supported headphones, we can see that Weiss has received compensation curves for a host of Audeze headphones:
  • LCD-1
  • LCD-2
  • LCD-2 Classic
  • LCD-2 Closed
  • LCD-3
  • LCD-4
  • LCD-4z
  • LCD i4
  • LCD-X
  • LCD-XC
  • LCD-MX4
  • Sine (Open-Backed)
  • Sine (Closed-Backed)
  • iSine 10
  • iSine LX
  • iSine 20
  • EL-8 (Open-Backed)
  • EL-8 (Closed-Backed) and
  • EL-8 Titanium

At the time of this review, the Headphone EQ module only compensates for Audeze headphones. So if you’re running other premium headphones - from Abyss, DCA, Focal, HifiMAN, or Sennheiser just to name a few - you won’t benefit from this unless their headphones get added through future firmware releases. Luckily for me, I’ve been listening to Audeze headphones for just about a decade now, and I happen to have several of them on hand, including their: LCD-X, LCD-XC, and even an EL-8 Open Back.

I’m happy to confirm that - in terms of frequency response - the DAC502’s Headphone EQ did a great job of bringing them all to a consistent neutral signature. Whether you prefer such a studio-monitor-like neutral signature or not, well that’s a different matter altogether. But to the extent that I could test here, the Headphone EQ does exactly what it says it does.

BTW, while I have an LCD-5 here - and while I’ve been told that the LCD-5’s compensation curve will be added soon - it was not available to test that at this time.

Loudness EQ​

If you’re not already familiar with Equal Loudness Contours — or the Fletcher-Munson curves that preceded them — I heavily recommend that you take a deeper dive to learn more about them, if and when time avails. Understanding equal loudness will shed some light on a fundamental psychoacoutic aspect of the human hearing system. It will also give you an understanding of how this Loudness EQ DSP module works, and why it might come in handy.


For the moment, it’s simply necessary to understand that what we hear changes based on how loud we listen. Specifically, the louder (or quieter) we listen, the louder (or quieter) certain frequencies will sound, relative to other frequencies. For example, let’s take a quick look at the Equal Loudness Contour chart above, and compare the 20 phon contour with the 60 phon contour. Music will seem to have a far more prominent mid-range, and sound more bass-light as a result, at 60 phons than it does at 20 phons.

Okay, back to the DAC502’s Loudness EQ! What Weiss has given us is the ability to alter tonal presentation, based on how loud we believe our music should be heard at. After enabling it, we can access the Level parameter, and choose between ten different settings ranging from 60 dB to 105 dB, in 5 dB increments. As you’ve probably surmised by now, choosing the right setting is going to depend largely on knowing how loud that music was when performed, or your best reasonable guess as to such.

A typical piano concerto would weigh in at 60 dB to 70dB SPL. If you favor recordings of chamber music in small venues, then you’re listening to music that would have been approximately 75 to 85 dB SPL when it was performed live. If you’re an aficionado of full orchestras in concert halls, that’s closer to 90 dB SPL, or even more, with the finale of Brahms’s 4th Symphony reaching as high as 105 dB SPL.

Personally, being a multigenralist, I left this module bypassed for the vast majority of my listening sessions, which were of studio recordings. But if you are an aficionado of classical or live jazz performances, and you have some knowledge of how those pieces were played and recorded, Weiss's Loudness EQ might just help bring you even closer to your music in terms of realism (or your perception thereof).

Dynamics Adaption​

Weiss’s Dynamics Adaption module is actually two DSP functions in one.

First and foremost, and this is not at all obvious by looking at its name, this is a normalizer. As soon as you enable this DSP module, and without even touching the DynLevel parameter, you are in fact normalizing your levels from track-to-track. I mention this because, if you don’t want to normalize your output, make sure to bypass this rather than leave it on at 0.0 dB and then assuming that your output is not being altered, because it is.

Now I know what some of you are thinking. Why is this here? In this day and age of streaming, I can just normalize in Roon, or in TIDAL, or even in Apple Music. Well, first of all, not everybody streams. Secondly, there is no normalization feature in Qobuz. So this is actually a very useful option to have for some people.

The second DSP function found herein sounded to me like, well frankly, it seemed like audiophile heresy: the DynLevel parameter adjustment. It defaults to “0.0 dB” and swings from -15.0 dB to +15.0 dB at the bottom and top of its range. By adjusting this DynLevel parameter, to anything other than +15.0 dB, you are in fact REDUCING dynamic range. Yes, you heard me correctly, the DynLevel parameter governs the amount of dynamic range REDUCTION you want.

This graph from the manual will better explain what is happening:


With a positive DynLevel adjustment, the dynamic range of each track is preserved, and this entire DSP module functions as more of a limiter, where the peak of each track is set to the normalization point. This is equivalent to the gold line (Normalized B) in the figure above. And it's exactly what we’d want in order to avoid startling anybody with sudden crescendos — while preserving the emotional intent that is inherent to a wider dynamic range.

With a negative DynLevel adjustment, this module functions as more of a true normalizer — where dynamic range is compressed to bring both louder and quieter passages together — resulting in the reduction of low-level “drop-outs”. This is equivalent to the burnt sienna line (Normalized A) in the figure above, and it affords us a more relaxed and casual listening experience. That said, given that +15.0 dB is equivalent to no dynamic range reduction, I think a more intuitive way to label this range is to set 0.0 dB at the top of the range, signifying no reduction in dynamic range, and -30.0 dB at the bottom of the range.

So why would we want this? Why would we want to fight our own loudness war of sorts? Well, most of us probably wouldn’t, at least not for critical listening. But the DAC502’s manual suggests that this would be useful for dinner parties, where we don’t want background music to intrude upon conversations. I can see how that would be useful. And in discussing this feature with others, it was also suggested to me that normalized audio is ideal for mood music… specifically for romantic soires. Okay sure, I can get with that as well.

BTW, just as an aside, if you’re trying to woo somebody… and you’re anxious about that special moment being spoiled by a triumphant rendition of Handel’s Zadok The Priest… I’ma tell you right now, you’re doin’ it wrong.

Crossfeed HP​

Like other DACs of its calibre (for example Chord Electronic’s Dave or dCS’s Bartok) the Weiss DAC502 includes a crossfeed feature for dealing with certain tracks - typically older recordings) that were mixed and mastered for speakers and not headphones. This is because high-performance headphones weren’t a thing back then (because high-performance headphones didn’t exist back then).

Enabling the DAC502’s Crossfeed HP DSP gives us access to its one and only setting - that of the Amount of crossfeed we’d prefer - which ranges from 0 to 100. By the way, these are arbitrary values, and not percentages. So 100 does not mean 100%, which would of course be dual-mono, and 100 definitely wasn’t that.


“Somethin’ Stupid” - a classic duet by Frank and Nancy Sinatra - is one such track where I greatly appreciate having access to crossfeed. With guitar and percussion hard-panned to the left, and strings hard-panned to the right, this track can sound jarringly artificial with headphones. I love the song, but I hate the track. Thankfully, there is a silver lining here, in that “Somethin’ Stupid” has become my go-to track for evaluating crossfeed.

In the case of Somethin’ Stupid, as well as most other tracks that benefitted from some amount of crossfeed, I found that a value of approximately 65 seemed to strike a nice balance between: (a) the restoration of a more natural sound field; and (b) minimal soundstage collapse, which is a common trade-off when using crossfeed.


And with 65 being relatively close to the mid-point, it’s nice to know that I have quite a bit of play either way, should I want to apply more or less crossfeed, for more or less irritating tracks respectively. NOTE: Those of you who are familiar with the Chord Dave’s crossfeed feature, a setting of 60-65 in the DAC502 also roughly corresponds to the Dave’s Crossfeed 2 setting.

Overall, I found the DAC502’s crossfeed to be very well done. It’s more refined and subtle than others that I’ve come across in years past, with the result being more of a delicate correction than a blunt compensation.
Dec 7, 2022 at 7:39 PM Post #7 of 377
Doesn't this just use standard DS converters?
Dec 7, 2022 at 10:47 PM Post #10 of 377
"Why so expensive? Because it sounds good."
Daniel Weiss
haven't heard this quote in ages! LOL

That's so classically Daniel Weiss, and must be said in his straightforwardly deadpan manner for maximum effect. :laughing:

Great review @warrenpchi

I've owned the Weiss DAC501-4Ch since late August 2022 and it's been everything I hoped it would be and more. A true reference DAC with incredibly powerful noise shaping tools and algorithms.

Thank you and well done.


Thanks @RobertSM! I have to admit, I was surprised by its performance when it arrived. I don't mean this as criticism, but let's be honest here, its spartan façade doesn't exactly scream sexy in any way.

Doesn't this just use standard DS converters?
Yes, the chips are ESS. As with anything, the implementation is everything.

As @RobertSM pointed out, yes, it does. That said, it doesn't sound like an ESS implementation. Had I not known it ahead of time, I doubt that I'd be able to audibly discern the SABRE DACs found therein.

I was lucky enough to borrow the Weiss 502 for three weeks and I really should dig up the review that is about finished and release it. It was truly something unique sound wise. Shame I’ll more than likely never be able own the 502 in my lifetime haha

I'd be interested in reading it for sure! If it's not too much trouble, would you mind dropping a link to it here in this thread when you post it? :relaxed:
Dec 8, 2022 at 1:15 AM Post #12 of 377
I was very skeptical when looking inside the chassis, but there is simply no extra nor redundant parts built in.

It looks empty inside but definitely doesn’t sound empty, just as Grimm MU1. Size and weight should not be the factors in the judgement of a source component.

Kimber copper PC is revolutionarily the perfect match to Weiss for me, as it adds musicalities, meat, bone, body and soul at the cost a bit spatial accuracy. Personally I wouldn’t consider keeping it without Kimber PC as it sounds a tad thin and lean with its stock cable to me.

Due to its history, Weiss is big on making interface, so I have no worries or doubts about its clean sound. It is interesting that I have difficulty connecting to USB input with AQ entrance USB cord Forest that I love and believe will be good match sonically.


  • 531B1110-F383-431A-9C25-AD101FD437E4.jpeg
    2.5 MB · Views: 0
  • C751510B-2270-43BA-9074-64DF189F0A47.jpeg
    1.9 MB · Views: 0
Last edited:
Dec 8, 2022 at 3:59 AM Post #14 of 377
DAVE was the ''greatest'' until this launch. I guess when DAVE 2 is launched for $20k it will regain the throne.

Looking forward to hype free impressions from my fellow headfiers
You certainly won't find it here :p
People's current favorites will always be the "greatest" from their perspective. I've just accepted that and moved on to deciding for myself. This helps to bring visibility over lesser known dacs. Like this one.
I wonder how much of an improvement in SQ this $10K Weiss is from a $1K RME ADI-2 DAC?
At the end of the day It was too much hassle to audition and while the proof is in the pudding, the only thing it seems to have going for it is adjustability (don't need or want it, that's most of what I'm paying for, tuning it for me) and Weiss' name. I can see paying a premium for it (Maybe 6k tops), but not 10x, especially when there's a lot of products out there better built and "originally" engineered.
It seems like a great product for a sound engineer that wants and knows how to dial in DSP for his speaker system "just right". People who pay this much for a dac usually expect something that works out of the box without the hassle. I think the reason why they're not more popular is this mismatch between who they are targeting with their pricing and their feature set.
Last edited:
Dec 8, 2022 at 4:52 AM Post #15 of 377
I wonder how much of an improvement in SQ this $10K Weiss is from a $1K RME ADI-2 DAC?
Take out expectation bias, pre conception, diminishing returns and measurements that only a bat could hear on a good night.
What Hi Fi reviewed and said it was “Possibly” (nothing like accuracy!) “the best DAC they've heard, and yet it’s relatively affordable”.

Users who are viewing this thread