ECP Walnut X.3 impressions thread

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by Erroneous, Oct 13, 2021.

  1. futbutts

    futbutts Friend

    Pyrate
    Joined:
    May 15, 2018
    Likes Received:
    469
    Trophy Points:
    63
    Location:
    Shermie Oaks, CA
    Unfortunately, even after downloading drivers manually and fudging parameters ad infinitum, I was never able to get my system to work with the Walnut x.3 free of the odd 'pop' here and there. Despite that I really enjoyed the DAC -- it's a pretty much impeccably balanced listen all around. When I first got into the hobby and heard the phrase 'audibly transparent' thrown around, usually in relation to Topping DACs (bllllurg), this is pretty much what I had hoped for/expected. Nothing feels overly accentuated or embellished at the expense of other areas of the soundscape -- it's all there, rich yet not overindulgent, incisive yet not bright or clinical.

    If I had to really dive deep for personality descriptors, I'd say that there's a slight etchiness to the upper mids, but not in an unpleasant way at all -- but certain transients (grit in male voices, and acoustic guitar in particular) have an extra visceral presence that is a little more prominent than most other DACs I've heard. To my ear this robbed it of the 'airiness' of Gungnir or Yggdrasil, but lends it great authority as well. I also enjoyed how it handles bass, there's some really nice composition to the sub-to-mid-bass range that lends those frequencies a deep-reaching throatiness while also lending plenty of texture and heft. At the risk of slightly contradicting myself this actually might be something it does better than other DACs in general -- one of my favorite go-to demo tracks is Steely Dan's 'Home At Last' which opens with a salvo of piano notes interladen with bursts of drums and bass guitar -- in those quiet moments of piano I noticed an odd sound -- and realized it was the mallets moving inside of the piano. I'd never noticed that before, as it's enmeshed pretty down low in the frequency range; I'd missed it even with the Iggy LIM, yet I heard it here fairly clearly.

    It's also a beautiful little DAC -- I love the small footprint, as well as the wood and pretty colored lighting

    For anyone keeping track I had quite the recent spate of DAC trials (BF2, Gungnir, 2541, Yggdrasil LIM) trying to find an upgrade to my Bifrost, and sound-wise this was my favorite of the lot, value-wise as well. My issues with the USB input would be my only hesitation in pulling the trigger on one were they to become available again.

    Thank you again @Erroneous for the loaner tour!
     
    Last edited: Nov 19, 2021
  2. dsavitsk

    dsavitsk Friend

    Pyrate
    Joined:
    Oct 2, 2015
    Likes Received:
    1,616
    Trophy Points:
    93
    Home Page:
    Not sure why the pops are there. My best guess sis that the Amanero USB board is a 32 bit board while the DACs internal circuitry is limited to 24. So you might try setting the output word length to 24 bits and see if that helps.
     
  3. Syzygy

    Syzygy Friend

    Pyrate
    Joined:
    Jun 13, 2018
    Likes Received:
    2,144
    Trophy Points:
    93
    Location:
    DFW, Texas
    That's generally good advice any way, when dealing with USB for transport. There's no reason to add 8 zeroes to each sample. You can't get 32-bit music, so why make the electronics work harder to add, transport, then remove, those bits?

    When I was using USB I always had Audirvana set to output 24-bit words.
     
  4. futbutts

    futbutts Friend

    Pyrate
    Joined:
    May 15, 2018
    Likes Received:
    469
    Trophy Points:
    63
    Location:
    Shermie Oaks, CA
    Indeed, the least-poppingest results seemed to be at 24/44100 or so.
     
  5. TomB

    TomB MOT: Beezar

    Joined:
    Dec 14, 2015
    Likes Received:
    1,753
    Trophy Points:
    93
    Location:
    Atlanta
    Home Page:
    I cannot get mine (or anyone else's) to start popping ever since I stopped using the clone USB boards (instead of genuine Amanero). That includes Dasman's Walnut, which I currently have connected in my system right now. I can not make it pop. With the clone USB boards, I could reproduce it every time when randomly switching between songs of 16-44.1 and high-res. It was if the drivers/clone board could not keep up with the resolution changes and hiccupped the popping noise. FYI, none of the clone boards exist in the wild anymore.

    I use Windows 10 exclusively with the WASAPI Event driver. My go-to music interface is MusicBee, primarily because I love the interface and organization and I can use it in Exclusive Mode. I have never, ever, ever been able to get Foobar to work in Exclusive Mode. Yes, you can select the option and Foobar will play just fine. However, it's not really locked and you can operate any other software and access the sound in the OS. With MusicBee Exclusive mode, nothing else will work if you try to access sound outside of MusicBee. That tells me it works as it should, but Foobar does not. I do this because I have some cheap, rebuilt Dell Latitude laptops that I use throughout Beezar processes, including music playback. Unfortunately, I can sometimes get dropouts if I'm not using Exclusive Mode because the laptops are ssslllloooowwww.

    That's a digression, though. The point is, I cannot currently reproduce the popping, not matter what I try. In MusicBee, you set the WASAPI driver and then you can set for Exclusive Mode, but there is no selection for word-length or sampling rate.

    In Foobar, you can select word-length, but not sampling rate, when using WASAPI. I tried 32-bit at Dsavitsk's suggestion (I always had it set on 24-bit), and I still can't get the popping to manifest itself.

    I still believe it's something in the USB stream, but am investigating other causes with a couple of Walnut owners. Hopefully, Doug will weigh in as well again. I sent him an e-mail detailing all of the above and more this morning.
     
  6. ColdsnapBry

    ColdsnapBry Almost "Made"

    Joined:
    Feb 4, 2021
    Likes Received:
    640
    Trophy Points:
    93
    Location:
    Earth
    When I play out of exclusive mode on Foobar it locks everything else. Maybe it's a plugin you need?
     
  7. TomB

    TomB MOT: Beezar

    Joined:
    Dec 14, 2015
    Likes Received:
    1,753
    Trophy Points:
    93
    Location:
    Atlanta
    Home Page:
    From what I've read, the WASAPI driver disables Foobar's capability to run in Exclusive Mode. MusicBee doesn't work that way.
     
  8. TomB

    TomB MOT: Beezar

    Joined:
    Dec 14, 2015
    Likes Received:
    1,753
    Trophy Points:
    93
    Location:
    Atlanta
    Home Page:
    I should add that many think the reverse of that is true, but I got tired of fiddling with Foobar and re-sequencing files on my 4-terrabyte music collection whenever I simply added a new album.

    There are some reports that say some soundcards don't recognize exclusive mode with Foobar, some drivers, etc., etc. It worked the first time, every time, when I used MusicBee. So, I gave up on Foobar. Sorry.

    I'd only been using Foobar for about 12 years ... time to change to something better. ;)
     
  9. TomB

    TomB MOT: Beezar

    Joined:
    Dec 14, 2015
    Likes Received:
    1,753
    Trophy Points:
    93
    Location:
    Atlanta
    Home Page:
    FYI - Doug and I (mostly Doug, of course) have determined that's it's very unlikely any power issue could cause the ticking/popping some have reported. That was one of the avenues we were investigating about the ticking. The power supply is not going to dip below the voltage required for the chips that are responsible for the USB stream. The voltage would have to drop well below ~4-5V and the walwart would be in full failure mode at that point, anyway.

    Instead, we suspect some issue still going on with the drivers being used. I can't emphasize this enough, but if you are using a PC - the WASAPI "Event" driver is the one you should be using. Preferably, "exclusive mode" as well, if you can select it in your chosen music app.

    I have another customer who uses an Apple MacBook. He has noted similar USB stream ticking with other devices, so it's not exclusive to the Walnut/Amanero. When I was using the clone boards, I could reproduce the issue with great regularity. With the genuine Amanero, I cannot - but then I never change from the WASAPI Event Exclusive driver. The "Push" driver has been reported over the internet as a mode that can cause issues (this is not about the Amanero, but USB in general). As always, if this occurs, simply disconnecting the USB cable and re-connecting it will make it stop.

    So please - use WASAPI Event Mode, and Exclusive, if your app allows it.

    Also, if anyone's Walnut walwart fails, I will replace it. The replacement walwart is a Triad 1.5A. Triad makes the finest walwarts available, IMHO. Plus, the 1.5A means it shouldn't even break a sweat. So, any walwart issues should not re-occur.

    That said, good practice is to always make certain your various pieces of equipment are turned OFF before you go switching around cables. Shorts can and do always occur.

    I've shipped out Dasman's Walnut again and it should be back in the loaner rotation soon.
     
    • Like Like x 4
    • Epic Epic x 1
    • List
  10. loadexfa

    loadexfa MOT: rhythmdevils audio

    Pyrate Contributor
    Joined:
    Dec 26, 2017
    Likes Received:
    2,544
    Trophy Points:
    93
    Location:
    SF Bay Area Peninsula
    Loaner Impressions

    Chain
    Qobuz -> Windows 10 Gaming Desktop -> Walnut -> Freya S (Passive mode) -> MJ1 -> LCD-4 (RD Modded)
    Roon (Qobuz/Redbook) -> Pi2AES -> Pavane -> Freya S (Passive mode) -> MJ1 -> LCD-4 (RD Modded)

    Listening was focused on the LCD-4 since they are my most resolving headphones and make hearing more subtle DAC differences easier. I never got over how cute and sexy the Walnut is, I guess I like real wood on my audio devices. Fortunately I had no issues with skipping, pops, or anything like that.

    Considering the price difference and my personal preferences, this comparison isn't fair. Perhaps the SFD would have been better but I've been holding off because I need to do some close listening to settle on which tubes I prefer. Until then, the Pavane is my comparison DAC.

    I liked the Walnut, it has a nice even, straightforward presentation. It doesn't try and stand out, it seems the goal is to present music accurately rather than "WOW look at what I can do!". I know "accurately" in this context is a loaded word similar to "neutral", hard to define clearly but it's the best word I can think of.

    Walnut vs Pavane
    • Walnut overall FR sounds more even
    • Walnut has less heft
    • Walnut has some etch, probably lower treble
    • Walnut less detail/resolution
    • Pavane has mid bass emphasis
    • Pavane has nicer percussion hits (heft)
    • When I go from Pavane to Walnut it sounds like something is missing. Not quite sure if it's detail, but there's "more" with the Pavane.
    • Pavane is more organic and alive. Not livelier per se, perhaps it has more soul. More noticeable with Stevie Wonder.
    • Mid bass emphasis with Pavane sounds off, I prefer the low end of the Walnut
    • Sometimes cymbals are too much on Walnut
    • Voices and mids are more forward on Pavane
    • With the Pavane, "interesting" voices have more bite (Natalie Merchant, Stevie Nicks)
    • Pavane has smoother treble
    • The Pavane is more focused and "there" vs the Walnut sounding more dispersed
    In many ways the Walnut is "just right". Not dry, thick, wet, etc and my only real complaints are less detail and soul. Almost "polite" but not reserved, definitely not boring while also not getting in your face. Overall an excellent DAC, I see why it's well-liked.

    USB could also be a disadvantage vs the Pavane's I2S, sometimes I would find my A2 Gungnir with Gen5 USB too shouty and fatiguing (depending on amp/tube pairing) which was cured by switching to the Pi2AES. The treble etch and harshness I heard could possibly be tamed but hard to be sure without actually comparing. At this price point I would prefer the option of an alternative board or something. I've built my best chain around a streamer and I doubt that's going to change.
     
  11. StageOne

    StageOne Friend

    Pyrate
    Joined:
    Sep 23, 2018
    Likes Received:
    1,121
    Trophy Points:
    93
    Location:
    US
    ECP Walnut DAC

    Chains
    RPi Pi2AES (USB for Walnut) > Gungnir A2 / ECP Walnut DAC> Freya S
    > ECP T4 > Utopias
    > Aegirs Monos > Spatial M3 Sapphires

    I know others have commented on size, but I was expecting it to be about half the size of the ECP T4 but it’s closer to 1/4 or 1/3 the size. Perfect for a desktop DAC!!

    On Speakers
    Gungnir
    • Much more forward staging
    • Smooths sounds over
    • Wider, more expansive soundstage, bit more precise spatially
    • More gray sounding
    • Hits a little harder
    • Warmer and more relaxed
    • Slightly muddied mid bass.
    Walnut
    • Set back with more depth
    • Better “blackground”
    • Subtle bass notes that appear on the “Kinda blue” album
    • Same volume but sounds like there is more music
    • Sharper “blat” of the horn. Edgier
    • Smaller sound stage. More intimate


    Rodrigo Y Gabriella “Echoes”
    • Walnut has a crisper sound to the strings, edge/bite
    • Gungnir is softer and missing that little extra touch.
    Hans Zimmer in Prague “Pirates..”
    • Gungnir has better separation in the fast huge sections
    • Walnut feels slightly congested, mids bleed together
    Miles Davis “So What”
    • Walnut 5:30 in horns were a little shouty, glare
    • Gungnir more depth, cleaner sounding
    Jamie Woon “Lights”
    • Walnut good bass and weight, less space
    • Gungnir extends deeper, more depth, space. Notes seem to hang a little more.

    Headphones
    • Similar sound but headstage and spatial differences are less noticeable.
    Overall, I really like the size, excellent build quality and great sound.

    Thanks to @Erroneous and @dasman66 for loaning out their personal DACs!!
     
  12. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    ECP Walnut DAC X3 Loaner Impression

    Personally, I’ve been curious about this dac. When I was seeking for upgrades from Bifrost 2 this spring, I got to know the existence of the Walnut. My current dac (Soekris dac 2541) was purchased from an audio pal in this forum and he let it go BECAUSE he liked Walnut over 2541. I thought flavor and taste were the most dominant reasons, but even if so, going from r2r to d/s looked quite unusual for such a seasoned audiophile.

    Aside: The walnut referred to above was very personalized unit departing from regular products btw, utilizing very uber transformers.

    Motivated by his enthusiastic comments via pm, I briefly considered the walnut as a possible candidate but gave up soon. That time Walnut was still a uncertain option for me as I couldn't imagine whether/how its sound translated into my liking. On the contrary, I already knew how 2541 fitted my preference well through the loaner tour. Anyway since then, I’ve been keeping my eyes on this dac and waiting for opportunities to try out myself. Finally could get my hands on the Walnut X3. I’d thank @Erroneous for this opportunity (also appreciating @dasman66 for offering an alternative unit for the tour).

    During my time with walnut dac, I found 2541 and walnut sounded shockingly similar rather than different in most aspects. They both sound like something of a “hybrid” regardless of actual implementation. No kidding. Indeed I almost screwed up in the double-blind tests (not very scientifically done tho) scoring only approximately 60% accuracy on average. Hearing the difference between the two required a healthy amount of effort and concentration to say the least.

    That being said, here are my general thoughts on Walnut vs 2541.

    What I liked with Walnut over 2541
    • Denser tone/timbre: When I listened to naturally mic’d tracks of fewer instruments, I almost always found Walnut had denser and meatier sound. I heard some of these in Bifrost 2 (to me this was one of a few competitive advantages of BF2 over 2541), but Walnut was even better with clearer representation than Schiit.
    • Low-end slam / General macrodynamics: I believe these are highly associated if not identical. Walnut reminded me of Schiit Mjolnir to some extent. Quite often I found Walnut a little more engaging and inviting. Switching Walnut to 2541 feels somewhat analytic (not necessarily in a negative way tho) and I haven’t felt this way until this loaner tour.

    What I liked with 2541 over Walnut
    • More comprehensive details: I’m confident that the two dacs do not differ in overall actual resolution (or they both perform far beyond my hearing ability). But information seemed to be presented a little differently. While 2541 was more spot on what I was looking after, I couldn’t identify the same amount of detail cues with Walnut by comparison. Note that I tend to hear more details with more analytic presentations.
    • Microdynamics: Same things went to dynamics, too. I tried to hear incremental loudness changes with both dacs. While Walnut nicely reproduces a perceivably smooth and fine gradation in volume, I found myself more comfortable to distinguish sonic amplitude delta at any level with 2541.

    Where I was rather indifferent
    • Tonality: I wasn’t very successful in differentiating two dacs in tonality. In the blind testing mentioned above, I first focused on tonality, which resulted in the poorest scores. Tonal balance was not a good indicator to infer.
    • Spatial presentation: Both dacs were decently holographic in speakers as well as headphones. Good balance between width and depth of staging. Sonic images were a little thin on both dacs compared to higher tier products (such as Yggdrasil or Holo Spring 3 among those auditioned recently). Not much to complain about on my end.

    Overall, I found the Walnut -- specifically X3 revision with amorphous cores -- a good alternative to DAC2541 if sonic performance only matters. Nevertheless, I still think 2541 fits better with my use cases because walnut comes with NEITHER balanced outs (amp synergy issue) nor remotable attenuation (convenience as a preamp).

    Associated chains below.
    • PC usb out (2x upsampling) → Walnut usb in rca out → 353X rca in → L700 Mk2
    • PC usb out (2x upsampling) → 2541 usb in bal out → 353X bal in → L700Mk2 (for comparisons)
    • I tested upsampling on and off with Walnut but they weren’t meaningfully different (note: 2541 with black filter benefits very obviously with the upsampling dsp). Thus, I didn’t change the DSP setting during the evaluation for convenience.
     
    Last edited: Dec 30, 2021
  13. Azimuth

    Azimuth FKA rtaylor76, Friend

    Pyrate
    Joined:
    Nov 9, 2015
    Likes Received:
    6,838
    Trophy Points:
    113
    Location:
    Nashville, TN
    Home Page:
    I just want to say first of all that this is a small box and kudos for fitting so much tech into a small box that is smaller than most hotel Bibles. The transformer outputs are something you do not see many DACs if at all. Combine that with the Amenero and other ECP tricks and you get one pretty amazing little DAC. Just plugging it in once and you can tell that this is a different kind of DAC.

    It needs to be said though that a lack of digital inputs is a serious limitation in the 2020 world with desktops with Pi’s, computer, and other digital sources. MORE digital sources should be the norm, not less. Amenero aside, I really think there should be other digital input options.

    The Walnut X.3 needs about a 15 minute warmup. This is kind of odd for a 1-bit DAC to me, and I don’t know if that is USB related, DAC chip related, capacitor related, transformer saturation related, but the change is certainly there.

    At first start, flat dynamics, light and soft on the low end, and just strained. Overall wimpy sound. After about 15 minutes, staging and SQ got better. And then improved again after about an hour and remained the same throughout. Go figure.

    Output is also lower than other unbalanced DAC’s you might be used to - 1.5V output instead of 2V. This is similar to the AMB Gamma 2 DAC, which uses the same Wolfson DAC chip. Quite a volume difference when comparing with other DACs.

    Overall presentation is soft, wide, deep, and just a bit tucked in on both sides of the spectrum. I liked @Vtory ‘s evaluation of a “denser and meatier sound.” No hard edges to speak of. If you are allergic to digitus, this is your DAC. The top end is not peaky at all, nor very grainy. Cymbals sound very natural and clean and not up front in the mix. The bottom in is textured, but not aggressive, albeit somewhat soft in dynamics. The mids are clean although somewhat recessed. Just like other Wolfson implementations, it is almost a U-shaped sound. The mids are there, just not up front and sometimes vocals can get lost in the mix. The lows are there, just with a kind of rubbery soft punch. Highs too are soft and quite non-fatiguing.

    The sense of space is quite surprising. It is a wide and deep presentation. A sheer feat for a 1-bit DAC. And really it is about the most impressive thing that this thing can do. And completely non-aggressive or offensive.

    I still feel like there are details lost there in the mix. It is not quite as wide or detailed as my modded Modi MB, (although the Modi with the op amp changes gives it a bit harder edge). Dynamically it feels like a soft knee compressor. It kind of sounds like a representation rather than the real thing (more on that in a minute).

    Things got a little more even with the Etir feeding the Modi MB - so I can now rule out the source. Modi MB + Eitr had the flatter soundstage and made the Walnut sound even more natural.

    Overall, there is just something keeping me from being fully engaged on the Walnut. I don’t feel like I am there in the room. It sounds good and presents the sound well, but things sound like a good recording rather than a totally live feel with real musicians playing. Something feels a bit disconnected somehow for me. Things did sound distanced, and there was space, but also kind of distant and not immediate clear sound. Maybe I am just used to feeling close and harder edge of things. I will say that acoustic music had a richness and pleasantness and some aggressive electronic recordings just did not come through in the same way and felt disjointed. My only guess is that the softer edge of things were just sometimes too soft.

    I DO think it is better than the AMB Gamma 2. Less grain in the mids, smoother, deeper, wider, more natural sound. And those that want that Wolfson magic, I think it is here. For myself, I am just find the realism factor down a bit and transients kind of losing definition, and an almost lost sense of immediacy and details.All the other stuff considered, it is still an astounding little 1-bit DAC that could.
     
    Last edited: Jan 4, 2022
  14. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    It's interesting seeing this sentiment come up again, in the same week we've had lots of talk about improving USB implementations.

    I'd bet the Walnut's USB implementation is as good as, but probably even better than, any commercial product. A dedicated modern USB receiver, galvanically isolated, feeding I2S directly into an ASRC, is just about as ideal a digital input implementation as you can hope for. Adding basic input switching is going to be pretty hacky, and it will create those real or perceived differences between digital inputs. To mitigate that you're installing FIFO I2S buffers, messing round with master clock distribution, etc. Then, the Walnut is no longer a cute little box of unique awesome, it's just another big DAC in a sea of other big DACs.

    I seem to be doing this a lot now, but I'm not even sure what my point is? I guess, if the singular purpose doesn't fit someone's needs, that's fine, I obviously can't dismiss that. But when that singular purpose is best-practice, there's a lot of power in that for the right consumer. Multi-purposing something just for the sake of it isn't always the best.

    Incidentally, I just looked at Tomb's initial build thread, and I bet it would be really easy to mod the Walnut with a different input. There are 4 lines coming off the USB isolator chip that are surely just I2S data lines. Install a transformer-coupled SPDIF/AES to I2S converter and tap it directly into those I2S lines, and you're golden.
     
  15. Azimuth

    Azimuth FKA rtaylor76, Friend

    Pyrate
    Joined:
    Nov 9, 2015
    Likes Received:
    6,838
    Trophy Points:
    113
    Location:
    Nashville, TN
    Home Page:
    I see what you are thinking, and hacky or not, there are other implementations that use Amenero or Unison (similar) that have some kind of switching. Obviously they have figured it out and should not hinder a DAC in any way. Although I can see that source switching poses a particular set if issues that is much easier and "purer" if you will with managing only one I2S source.

    Source switching has just been thing for many years on DACs and USB may not be quite as popular because of the popularity of Pi devices, and it is even hard for me to go back to USB and Windows because I am steeped in Pi since 2016 or so. So mostly for me it is a convenience thing and a preference thing.
     
  16. dasman66

    dasman66 Self proclaimed lazy ass - friend

    Pyrate Contributor
    Joined:
    Mar 13, 2018
    Likes Received:
    2,462
    Trophy Points:
    113
    Location:
    NW Pennsylvania
    agreed... and while I love my (3) Pi2AES based setups, I use a USbridge Signature Player with the walnut. Since it's a Pi3 Compute based device, the only thing it can't do relative to Pi4/Hat solution is run HQplayer (and USbridge Sig is less $$ than a Pi4 and hat).
     
  17. loadexfa

    loadexfa MOT: rhythmdevils audio

    Pyrate Contributor
    Joined:
    Dec 26, 2017
    Likes Received:
    2,544
    Trophy Points:
    93
    Location:
    SF Bay Area Peninsula
    Even if this is 100% true, that doesn't address how awesome/garbage the source is. I saw the other thread you're referencing but that requires a desktop/tower where you can add an additional usb card. A Pi2AES setup is a lot cheaper, more power friendly, and in some ways more flexible than purchasing a desktop computer.

    I understand the desire to keep the Walnut tiny and as simple as possible, as I said in my impressions, a swap-able input card would have been much better, especially at that price point (even if additional inputs cost more). Considering the small run, I can understand the lack of desire to implement this. But labeling this as an "ideal digital input" ignores a large part of the equation which could make it sound a lot crappier.
     
  18. rlow

    rlow A happy woofer

    Pyrate Contributor
    Joined:
    Jul 18, 2017
    Likes Received:
    7,738
    Trophy Points:
    113
    Location:
    Canada
    Bottom line, USB-only will limit the market for this DAC to a certain degree. But depending how many they want to make/sell, this may not be a problem for them. Comes down to a business decision as well as potentially a design decision/tradeoff. There's lots of good USB streamer sources out there for relatively low cost - PI2AES isn't the only game in town, even though it's a darn good one for the price. If that's your main transport however, and you don't want to switch, this may not be the DAC for you.
     
  19. TomB

    TomB MOT: Beezar

    Joined:
    Dec 14, 2015
    Likes Received:
    1,753
    Trophy Points:
    93
    Location:
    Atlanta
    Home Page:
    Nothing like a loaner excursion to open one up to criticism. All of that is fair game, though, with the exception of one comparison up there that I simply can't let go by without offering an opinion and hopefully, a little bit of insight.

    AMB's Gamma 2? Seriously?

    It was fine 12 years ago, but has only an opamp output used to tie the balanced outputs from the Wolfson into single-ended .... and there are bypass-paired coupling capacitors - including an electrolytic(!) - between that opamp and the final output. (If in a signal path, a small fast film cap is often paired with an electrolytic in an attempt to keep the electrolytic from trashing the sound too much.) In fact, Beezar's PupDAC does the same thing, albeit with a TI DAC (PCM1794), and uses an ultra-fast opamp (OPA2836) - without coupling capacitors on the output.

    The Walnut, on the other hand, has nothing short of a low-voltage version of the DSHA for an output - all the way to the output transformers' inputs. The transformers themselves combine the balanced outputs into single-ended. Opamps are used in the Walnut (2 of them), but only in the power supply - they are not in the signal path. Moreover, an electrolytic or film capacitor is nowhere in the signal path.

    Finally, for those that are curious, the Walnut X was really designed by Doug - at my request - for one reason: to be paired with the T4. In that role, I couldn't ask for more and will be ever grateful to Doug. As it turned out, others were interested in it for other amp pairings, too, but that was never the purpose in the beginning. Hopefully, that explains the lack of features and connectivity that seems to be mentioned often in this thread.

    OK - nuff said. Back to it, guys! I'm enjoying the reading!
     
  20. Azimuth

    Azimuth FKA rtaylor76, Friend

    Pyrate
    Joined:
    Nov 9, 2015
    Likes Received:
    6,838
    Trophy Points:
    113
    Location:
    Nashville, TN
    Home Page:
    Tom, I mostly bring that up because it is the same DAC chip and there was a recent loaner tour out of that DAC last year that I also reviewed. Part of that Wolfson sound is still in there somewhat.

    The Walnut X.3 is a much more refined version and the best implementation of said Wolfson chip that I have heard. The Gamma 2 was much more slow/syrupy by memory. The X.3 beats it every way.

    The Wolfson chip is just one of those things. Some may like it, some may not, it is a preference thing. The imaging was just a bit too synthetic for me, just like other Wolfson DACs, including the old iPad. So I think that has more do with the DAC chip itself than all of the other things you guys are doing around it.
     

Share This Page