Review Context
The main system consist of the CH Precision C1.2 DAC, CH Precision L1 preamplifier, CH Precision A1.5 power amp, and Magico S1 MkII speakers. Other digital sources include the Grimm MU1 and Antipodes G4 Oladra Music Servers.
All interlinks are Driade Flow Link Reference 808, both balanced and unbalanced. The speaker cables are the Jorma Unity. I will listen to the K22’s USB output using Final Touch Audio Sinope, Vermouth Reference, Mad Scientist Black Magic Ultra, and Black Magic Gold USB cables, as well as its AES/EBU output using a Jorma AES/EBU cable.
The CH system sits on Artesania Exoteryc and Aire racks while the digital sources and most review components sit on HRS EXR racks. To allow for comparison with the Oladra on fully equal grounds, the K22 is positioned on the same type of HRS EXR rack, on the same level height (not that this matters a lot with this rack, but still…), and connected with the same Belden 19364 power cable.
Running In
The G4 K22 I was sent was the exact same sample I reviewed earlier, now upgraded from G3 to G4 status. Nevertheless, many of its components were brand new, requiring some running in. Due to my busy schedule, I could allow it 2,5 months of bedding in. From extended exposure to the K50, I know that the server changes most in the first 3 weeks of use. But as I later found with that server, there was also a longer-term effect after 3 months of use. Presumably, the capacitors and perhaps other parts in its three power supplies just need a little longer. It stands to reason that all Antipodes servers with a linear power supply behave similarly. Interestingly, I found that the Oladra did not seem to change very much after its initial running in period. No doubt, this is due to its different power supply.
In any event, during the first weeks of use, the sound becomes noticeably smoother and more relaxed. The presentation becomes richer and more organic while retaining the refinement and overall character. The effects are most noticeable in the midrange, which can be a little hard and relatively forward when the server is new but which relaxes after 3 weeks and becomes fuller and warmer with continued use. The effect is also noticeable in the bass, which remains full-bodied but becomes rounder and more relaxed.
Incidentally, the G4 Oladra, to which the G4 K22 will be compared, has been in use for the exact same amount of time as the K22, which is 2,5 months.
Medium Power
In use, there are no obvious signs to give away that the K22 is “only” medium-powered. Just like the K50 and Oladra, the K22 responds immediately to play/next and other commands, and even when scrolling album overviews, I can’t tell a difference. Maybe, just maybe, the Oladra responds a tiny bit faster when tapping in the track waveform to jump playhead position, but I can’t be sure it is not imagined. So, I guess the only practical downside of its lower processing power only applies to huge libraries over 8TB or heavy DSP functions.
Listening – Foreword
I don’t think anyone will dispute that Roon is the current dominant server/player software. Its user interface is simply superb, and naturally, Antipodes supports it. Because of Roon’s market dominance, and I think the majority of people will use it that way, I will focus most of my listening on this software,
But besides Roon, there are various other server and player apps that also work very well. None of them match the convenience of Roon, but perhaps more importantly, there is quite a noticeable difference in sonic presentation between Roon on the one hand and Squeeze and MPD on the other, and one of the alternatives may suit your needs or preferences better.
Another thing to note about Roon is that its overall sonic presentation tends to waver with every update. When reviewing the K22, Roon is at Build 1365, which fortunately sounds really good again. Nevertheless, for me as a reviewer, the changes in sonic presentation are a concern, and that’s why I wholeheartedly welcome Antipodes covering every popular format out there.
I mentioned that the K22 is also ideal for those who predominantly stream from cloud services. As timing has it, I just finished my review of JPLAY for iOS, which works via UPnP. Because the K22 is also compatible with this format, I will also report how the K22 sounds with this software, streaming directly from Qobuz.
It’s worth noting that Antipodes recommend using Squeeze for the best sound quality or Roon server with Squeeze player for the best compromise between convenience and fidelity. As it follows, I will also listen to the K22 using Squeeze. Finally, I’ve compared some of the other apps in my original K50 review, so if interested, please read my impressions there.
Listening
When I reviewed the 3rd generation K22 in April last year, I was really surprised to find how great it sounded, even as an all-in-one server. I vividly recall it having a slightly dry but beautifully beefy, robust, and impactful sound, along with the Antipodes-typical smoothness, deep tonal saturation, and outright non-technical presentation.
Since the same actual G3 K22 sample was sent to Antipodes for upgrading to G4 status, I don’t have the older model for direct comparison, but as soon as I started listening to it, I immediately found myself back in the exact same state of mind.
Starting with Roon and the AES/EBU output, the G4 version is still recognizably punchy and full-blooded but very noticeably smoother and definitely sweeter. The G3 K22 could be a little dry but with the G4 K22, there’s not even a hint of dryness left. I don’t have the K50 available for direct comparison, but having owned it for quite some time and knowing it well, I would say it is indeed very close to the KALA range-topper’s performance.
The K22 has a sound that is remarkably meaty and dense – there is most definitely no paleness or leanness here! The flip side of this is that its focus does not appear as razor-sharp as it can be. On the other hand, servers that do sound highly focused tend to also be leaner and less saturated in balance. The midrange has bloom and richness, and there’s a touch of warmth, but less so than with the Oladra and also less than I recall from the G3 K50. Sometimes, dense-sounding audio components can have relatively narrow soundstaging, but this is not the case here – just like other Antipodes servers, the K22 sounds captivatingly spaciously, both in the width- and depth planes.
Until G4, there has always been a considerable difference in sound between USB and the other outputs in all Antipodes servers. Contrary to what one might expect, USB interfaces are tricky to implement, and to make matters worse, the success of the connection is dependent on the implementation on both ends (in the server and the DAC), and on how they interact. In practice, this can and does lead to large variances in the perceived quality or the subjective musicality of the connection.
USB
As mentioned in the introduction, Antipodes have lifted the USB sound quality to match that of the Reclocker S/PDIF, AES/EBU, and I2S outputs, as part of the G4 update. So, how does USB compare to AES, exactly, you ask? I won’t beat around the bush – USB now also sounds fantastic!
Listening with a Vermouth Reference USB cable, the K22 retains its punchy, planted, and meaty sound and expansive soundstage, including its impressive imaging in the depth plane. It adds to this slightly leaner, more articulate bass behavior with just a touch more control. On the other hand, AES/EBU is still more free-flowing and sounds a little bit more mellow and relaxed.
As mentioned, the sound/performance of USB ties in with the implementation of the interfaces on both ends. While I have no reason to feel that one of the CH’s input modules is “better” than the other, it is entirely possible that USB has a slightly different presentation than AES/EBU. The experience accumulated over many years using countless servers and DACs gives me a reasonable handle on these things, but I can never “prove” which interface is responsible for a perceived effect. Is it the server side, the DAC side, or an interaction between the two?
Naturally, the cables have a very large say in the matter. But, try as I might, I have not yet found a pair of AES/EBU and USB cables that truly sound the same, not even from the same manufacturer. A case in point is the Jorma USB cable, which offers a truly majestic and highly refined sound but is markedly warmer and more mellow than the Jorma AES/EBU cable.
The Vermouth Reference USB cable is a neutral yet refined performer with a gentle sweetness. So, how would the robust and meaty-sounding K22 work with the inherently also very robust and meaty- sounding Final Touch Audio Sinope USB cable? Indeed, as one might expect, this leads to a rather voluptuous bass. It’s pleasant and entertaining for sure, but a little bit too much off-balance for this server and my taste. But if your room or system needs it, or you just like Big Bass, it may be right up your alley.
Next up is the Mad Scientist Black Magic Ultra.
Blam! This cable immediately hit home with a sound that is still planted, meaty, and well-saturated but more articulate, open, and expressive. It’s less airy and refined than the Vermouth but more “Rock & Roll”. I also used the Mad Scientist Black Magic Gold as a final test. This cable adds the final touches in terms of refinement, flow, and air whilst offering a big slice of the Black Magic Ultra’s “Rock” capabilities. Between these cables, I feel the Vermouth and the Black Magic Ultra provided the most well-balanced presentations. The Vermouth will be relatively more ideal if you favor airiness and refinement, and the Mad Scientist will be more ideal if you favor incisiveness and expression. Of course, YMMV.
Squeeze
Using Squeeze instead of Roon is as simple as can be. Just select Squeeze as the Playback Method in the pull-down menu, and the software will make sure that both server and player are ideally set up.
While using Roon, you could also go to the Server App overview and enable Squeeze in addition to Roon. However, it must be noted that having Roon run simultaneously will impact the sound.
Good ‘n old iPeng (above) still works very well, offers all the basics, and doesn’t look too bad.
LMS (Logitech Media Server) with the Material Skin User Interface is a great alternative. You don’t need to download an app because it runs as a service within your internet browser. Simply tap “Manage Server Apps” in the AMS web interface and then tap the “Open” link.
This will open LMS in a new tab of the browser. iPeng may still work well enough, but to be honest, the LMS Material Skin sure makes it look quite dated.
So, how does the K22 sound with Squeeze? I won’t sugar-coat it; Squeeze sounds a lot better to me than Roon, certainly on Antipodes servers. Playing the same tracks as before, there is a new level of tightness, crispness, articulation, and overall precision to the sound. It’s a leaner presentation, for sure, but the server simply refuses ever to sound dry or anemic. The transparency and precision are definitely increased, but even so, I still hear the richness, saturation, and flow that characterize the server.
Next: JPLAY and Conclusion
Hi Christiaan,
another great review!
As you promised in the comments for your JPLAY for iOS review I would like to ask you about the comparison of K22 vs LinQ. I think it makes sense to compare them in the configuration where both sound best. Personally I am not interested in a configuration with Roon involved. Thank you very much.
Sorry, I forgot to mention please consider only streaming from Qobuz and not playback of local files. Thanks again.
Hi Matt, comparing the Antipodes K22 to the Aqua LinQ both as a streaming endpoint via UPnP controlled from JPLAY for iOS, providing signal via a Jorma AES/EBU cable to the CH C1.2 DAC, the K22 offers a tonally fuller, creamier, more voluptuous, darker rendition whereas the LinQ offers a tighter, leaner, more open, and expressive rendition. The K22 also feels more spacious whereas the LinQ feels more focused. The magnitude of the difference becomes smaller when using the LinQ’s HQP UPnP engine (instead of the standard UPnP input) but the relative differences remain. When using the K22 with USB, the magnitude of the difference is further reduced, but still, the typical Antipodes traits such as smoothness and tonal saturation remain evident. In a nutshell: the LinQ is more precise and transparent but can also be perceived as dry, lean, or analytical whereas the K22 is smoother and more forgiving but less transparent and revealing.
Hoi Christiaan,
Thank you for expressing your thoughts. I particularly value your feedback following our recent discussion about the Oladra G4. I’m pleased to see the incorporation of Squeeze to Squeeze in your review, and it comes as no surprise that opting for Squeeze for both player and server can significantly enhance the performance of any Antipodes server.
As I recently wrote on the Antipodes forum:
“ I have complete faith in Christiaan’s observations, and I’ve noticed and have let him know that lately he’s been approaching his writing more from the perspective of a Roon user. It’s not because he lacks knowledge or prefers Squeeze less; it seems to be a matter of convenience. However, his latest review truly highlights the strengths of Squeeze + Squeeze, which is also a testament to the efforts Antipodes has been making to propel Squeeze in the right direction.”
I also encourage Antipodes users to explore alternative combinations, such as SqueezePlayer with Roon server and the three MPD presets. However, it’s worth noting that there’s no wrong choice here; each option offers its own merits.
_ _ _
antipodes.support/t/audiophile-style-reviews-antipodes-audio-kala-k50/1962/60?u=paulw
(Links are not permitted here, please paste above in your browser)
Thank you for the feedback and support, Paul!
Hoi Christiaan, graag gedaan.
Thank you for this review. The Antipodes K22 along with the Playback Design MPS-X (which you reviewed last year) are the two streamers I am considering. Which one would you prefer? Would the PLINK connection to the Playback Design DAC nudge you in that direction? Of course there is nothing stopping one from connecting the K-22 to the MPD-6, but would it be as good as with a total Playback Designs setup?
With Playback Designs, you’ll still need a server section unless you use streaming services only. Other than this, it is a very personal matter. I found PLINK to be the best interconnect format for Playback Designs devices but I would not use this as the main deciding factor. These are two very different sounding solutions so it really depends on what sonic signature you are looking for. The Playback Designs MPS-X sounds leaner and more agile in the bass and is airier, more refined and more transparent, while the Antipodes K22 sounds fuller and more ponderous in the bass and is tonally more saturated, darker, and harmonically richer.
Hi Christiaan, thanks for the in-depth review. It is your Antipode reviews that triggered my attention for the brand. I have a K22 and K41. I sent the K22 for upgrade just before your review came out. Are you planning also to do a K41 G4 review? I am curious which improvements might be discovered there (and my K41 is a G3)
Thanks in advance for considering…
There are currently no other Antipodes reviews planned but one never knows what the future may bring:-) That said, I think you can get a good indication by extrapolating the noted differences between G3 K22 and G4 K22 to the K41.
Thanks for great review! I like that you compare sound to other brands/models. I would like to see an easy way to find your latest reviews, e.g. sorting the reviews by date. Most review sites have the latest review on top.
Thanks:-) If you scroll down on the main page, you see the “Reviews in Progress” and further down, you see “Latest Posts”, in order of appearance, with the latest ones on top. Were these perhaps missed or are you referring to something else? In the review categories, I did indeed opt for showing them alphabetically, as that is what I was missing on many other sites at the time. For me, it was not handy seeing them in order of appearance, as I was typically looking for a certain brand or model. Nevertheless, I am open to suggestions to making the site better. At some point, it will need an overhaul, and I will take such suggestions on board.
Hi Mikael, I’m happy to confirm I have now implemented sorting by date (latest articles on top) by default. I hope you like it!