Select Comparisons
Chord Electronics Hugo TT2
$5,495.00
Technical
This is the closest I have to the Bartok, at least budget-wise but also in terms of ‘doing their own thing’. The TT2 is similar in terms of being a bespoke DAC and an integrated preamp and headphone amplifier.
It is also pitched to the high-end and, like the Bartok, it can be expanded within a wider audio Chord audio eco-system. Modules such as the Hugo M Scaler and the TToby neatly stack above and below with connections aplenty for both headphones and traditional HiFi use.
WTA
Both stars of the show are the built-in DACs, which are not delta-sigma though very different from each other in their approach. The Hugo TT2 focuses on the tap limitation of delta-sigma blocks and hugely expands with their in-house FPGA-focused engineering to expand those ‘100s’ tap markers into nearly 100k taps using a Xilinx Artix 7 FPGA processor.
The processor allows for a much more complex level of filtering with no less than 86x 208MHz cores running in parallel. This allows the TT2 DAC to deliver an advanced 16FS WTA 1 filter with a ceiling of 98,304 taps.
Ring DAC
The Ring DAC is a discrete DAC that combines the proprietary FPGA control board with a DAC board using linking upgradeable software. The main purpose of the control board is to reduce noise, properly noise shape the stream, and ensure the clocking is ultra-precise, i.e. do away with digital jitter before moving it to the digital audio conversion board.
The DAC uses dCS’s proprietary mapping algorithm to take the sampled audio from the control side or decorrelates it from the signal thereby rendering mute any parts performance shift that can come from resistor-based topologies. In theory, the Ring DAC performance will not shift or denigrate over time as some ladder DACs can do.
Decoding
The Hugo TT2 does have an edge here in pure numbers with a decoding ceiling of DSD512 and PCM 32BIT/768kHz compared to DSD128 and PCM 24BIT/384kHz on the Ring DAC. Of course, how relevant that is depends on what you have on file and the Bartok is firmware upgradeable for decoding so that may not be fixed in stone for dCS.
The Bartok does have built-in upsampling capability via dCS Mosaic whereas the TT2 can be stacked with the M Scaler for a much more comprehensive upsampling capability. Unfair says you? A separate device yes but combined both are still around $5 less than the Bartok so there is that to think about.
Amplification
Both have competitive amplification outputs though, once again, different approaches. The Bartok primarily operates in the Class A domain though pushes into a Class AB operation sub 150mW with both unbalanced and balanced topologies.
Chord does not a huge amount about the topology of their amp design but we do know that is also a preamp and headphone amplifier with balanced outputs for pre-amping but unbalanced only for headphone output. We also know it uses a discreet output stage between the DAC and filter and before the amplifier to reduce potential distortion.
The TT2 offers a wider range of unbalanced outputs at 3.5mm and two 6.35mm sockets compared to the balanced 4-pin XLR and a single 6.35mm unbalanced output of the Bartok.
The TT2 offers a flexible voltage supply of 288 mW RMS into a 300Ω load right up to 2.7W into 32Ω on the high-end unbalanced.
The Bartok has a similar peaking rating of 2.7W but eases off on the voltage for lighter loads at 1.4W into 32Ω opting instead to maximize current demand. That 2.7W comes into play at a higher 66Ω level and levels out at a more powerful 600mW into 300Ω balanced.
Tuning
For this comparison, I wanted to initially test a high-end current-intensive modern headphone, the Hedd Audio HeDDphones (review coming soon). At 87dB SPL and just 48Ω, it certainly fits the bill.
For the TT2 I also hooked up the M Scaler for some upsampling comparisons as I consider it vastly superior to the stock TT2 sound. Considering both Chord units together are still a bit cheaper than the Bartok this is fair game.
HeDD Audio HeDDphones
With the HeDDphones what you want to hear is that signature subterranean bass it is capable of. Both units perform brilliantly in terms of pulling out an excellent low-end extension from this pairing.
However, the stock TT2 sounds more neutral to my ear in terms of quantity and forwardness. Not that the Bartok is a bass machine it just so happens the HeDDphone’s forte is incredible bass and that is where you hear a big difference between these two.
The Bartok sounds richer, slightly warmer, and more natural-sounding in its bass timbre, and well, to be honest, that liquid tone extends throughout the curve.
More than that the dynamic range on the Bartok is a bit more convincing. Spatial cues leap out at you, the staging is more complex yet at the same time, it holds onto that wonderfully natural tone not once sounding artificial in its delivery.
The TT2 on the other hand is the cleaner, lighter in delivery, and perhaps a ‘faster sound’ with very precise placement. You could make an argument that is the more linear of the two in terms of how it sounds with the HeDDphones with a neutral to slightly warm tone throughout the mids and into the treble. It is prim and proper with nothing out of place.
Yet, it lacks the ‘drive’ or PRaT of the more physical and punchy Bartok sound. The Bartok/HeDDphones just fleshes things out a bit more, offering more body throughout. Vocals have more authority, bass guitar plucks have tons of wonderful sustain or body without a hint of smear.
M Scaler Added
Now throw in the M Scaler and the gap tightens up. The M Scaler makes the stock TT2 sound rather muted and flat by comparison. The upsampling delivers a much more vivid performance, particularly with vocals that come further forward in presence, close to how the Bartok positions them.
Certainly, in terms of dynamic range and engagement, it is much closer to the Bartok. However, the overall tone doesn’t change and that may be to Bartok’s advantage. The TT2 is still has a cleaner, pacier tone and a little lighter in body compared to the Bartok with the Heddphones.
The Bartok still has more ‘character’ in its timbre, and by that, I mean nuanced details such as vocal breathing techniques, lisps, or short intakes. All have a bit more definition on the Bartok like you are standing in front of a live performance.
The TT2 seems to erase some of those vocal ‘quirks’ in favor of that perfect studio delivery and in doing so takes a little bit of ‘soul’ and realism out of the same sound.
Xi Audio Formula S
$3499
Technical
The Formula S is a pure analog amplifier so it will be compared directly to the Bartok amplifier using the Ring DAC pre-amp output (2V setting). That will mean both amps being serviced by the same DAC allowing for a fairly tight comparison.
The Formula S is a single-ended Class A fully discreet BJT amplification design and though weighty at nearly 5KG, it is quite a bit smaller and lighter compared to the 16kg Bartok. It does lack the dual mono-balanced architecture option of the Bartok, however, the numbers are very good in terms of output power from the Formula S.
The Formula S is capable of a single-ended output rating of 2.1 watts into 46 ohms which is a competitive rating compared to the balanced output peak rating of 2.7W into 66Ω from the Bartok. It may be more powerful given the sliding scale of the Bartok once you drop down to 32Ω loads given dCS’s preference for current priority overvoltage.
Distortion levels, however, are better on the Bartok at THD+N of 0.00025% compared to 0.0006% from the Formula S. As also the tested noise floor with up to a 6dB difference with the official spec of the Formula S at 110dB and the Bartok at 116dB (3rd party tested, A-weighted).
Of course, the supporting DAC will have a big influence on the Formula S performance but with the Ring DAC supplying both amp stages for this review it should stay fairly true to their numbers.
Tuning
For this pairing, I went with the Abyss Headphones Diana Phi as the Formula S tuning is referenced with Abyss headphones specifically in mind. Since both amps have excellent power and current draw the Diana Phi is also suited considering its SPL is a low 91dB.
Set at 2V the dB output of the variable pre-amp is fairly high at around 10dB for some headroom on the Formula S analog pot but your preferences may vary. The 6V line level voltage setting from the Bartok is more explosive sounding but you get less headroom and possibly more prone to distortion so 2V would be the norm here.
Dynamic Range
The star of the show in this contest is not so much the amp contest but the Ring DAC. With the Formula S connected to the Ring DAC the dynamic range and resolving capability take a huge jump up from my competing DACs such as the Chord Qutest which was my previous source in conjunction with the Hugo M Sampler.
The Diana Phi sounded brim-full of detail, punchy, and expansive sounding. And yet, the amps do paint the final picture in terms of timbre and you can hear a difference between these two Class A amps.
Timbre
The Formula S timbre is lighter but sweeter sounding and not as precise or dry as the TT2 amp stage. It has a slight inclination to draw you towards the upper mids and treble and of the two amps, it is airier and cleaner sounding with more treble forwardness.
The Bartok is deeper, richer, and more natural-sounding or at least a denser presentation compared to the Formula S. You get drawn more to the bass and mids with a firm fundamental and a more rounded overtone that carries that weight into instrumental and vocal notes.
The treble is just slightly faded on the Diana Phi and that might suit a lot of people wondering what is the most natural pairing with the sometimes fussy Abyss creation. You get less of shimmer on percussion and more of a liquid attack with the better body.
It might not seem as energetic as the Formula S pairing but it also might be less fatiguing. The Formula S/Diana Phi pairing can be a shade lean at times through the upper mids and treble offering more of an ethereal vibe to the Bartok amp’s earthier tone.
ALO Audio Studio 6
$3900
Technical
Perhaps a bit left of field in terms of the pitch but one heck of an amplifier. The Studio 6 is a Class-A single-ended triode (SET) circuit design. This is a single-ended handmade amplifier. This contrasts sharply with the Class A discreet balanced and unbalanced design of Bartok’s amplification output.
The stock Studio 6 uses a 6SN7 input tube, two 6V6 output tubes, two OB2 gas regulators, and a 5AR4 rectifier. The stock JJ 6V6 tubes on my unit have been switched out for NOS JAN Philips 6V6GT’s which offer a fairly balanced tone.
Though it is balanced it has 3 dual RCA inputs and a very unique 4×6.35mm output all independent of each other rather than sharing the same output power supply. It sort of makes this amp the reviewer comparisons dream at times.
The Bartok only has two headphone outputs, however, it does also offer pre-amping line-out capability whereas the Studio 6 is a pure headphone amplifier despite its multiple analog inputs and outputs.
Performance
The Studio-Six was designed more for voltage than current performance and as such is more optimal for mid to high-sensitivity headphones such as my trust 120Ω AKG K501. It has less of an optimal current demand performance for demanding low-impedance headphones so SPL ceilings are more challenging than for the Bartok.
At 300Ω, the single-ended output of the ALO Audio is much more competitive at 500mW compared to the Bartok output in its balanced configuration at 600mW. It does drop, much like the Bartok, once you lower the load so at 32Ω it comes out at about 1W which is a little weaker than the 1.4W capability of the Bartok (balanced).
Do expect Studio 6 to be noisier, however, that is the nature of SET which is much more distortion, albeit a pleasing type of even harmonic distortion. THD+N from Studio 6 is way behind at 0.26% with minimal loads and SNR is also approximately 20dB lower.
Tuning
For this, we went with the classic AKG K501 which is fairly voltage-hungry at 120Ω as well as an inefficient 89dB SPL. It needs lots of everything to sound optimal but cleverly disguises it by sounding overly mid-centric when underpowered. It doesn’t sound terrible but you do not realize what you are missing from this smooth performer until you give it a lot more voltage.
With Studio 6, I gave it the same setup as Formula S with a 2V line out from the Bartok Ring line level to equalize the DAC side of the test. This is the first time I have heard the Studio 6 with the Ring DAC and it sounds super smooth, airy, and very resolving. The Philips tubes also give it a turn of pace and are much better than the stock JJ 6V6 pairing.
Timbre
I have always enjoyed pairing this with the thicker punchier ALO Audio CDM DAC line out and whilst I would still class the CDM coloration as warm and juicy and a lot of analog fun the level of resolution and separation it can offer from that old Wolfson delta-sigma chipset is not on the level of the Bartok Ring DAC.
In terms of amplification difference, there are a few between these two big amps. The Studio 6 is more in line with the Formula S with a sweet analog timbre and an airy staging quality.
The Bartok is meatier, more powerful, and a shade more intimate. It is also the more precise performer with a much blacker background. It is also a little more resolving in terms of how it fleshes out that texture and yes, denser sound compared to the Studio 6.
Vocals
Vocal resolution on the Bartok is incredible, especially in terms of mouth formation. I mentioned it before on the TT2 comparison but the realism in how a singer technically sings is uncanny in terms of understanding how you can almost pick out their breathing techniques.
What I did notice, however, is the width of the Studio 6 staging. It felt wider compared to the Bartok amp stage, particularly for vocals. Now, this might give you an initial wow factor but listen carefully and you will find that it comes with a slight loss of instrumental presence and separation.
Vocals out wide and forward diminished the presence a little of instruments behind whereas the Bartok pulled it in and organized the imaging a bit better, no doubt benefitting from that black background. I suspect the richer timbre of the Bartok amp helps also for lower register notes to make their presence felt.
Our Verdict
The dCS Bartok is probably the best-sounding integrated DAC and headphone amplifier I have reviewed to date in the 10 years we have been operating this website.
That is some statement but I will not future-proof it because, well, the Bartok is modular, it is firmware upgradeable, and has plenty of legs in it to go on for a few years more and still stay relevant.
The Ring DAC may well be the star of the show but the Class A amp is no slouch either. Right now, the Bartok delivers a rich and powerful sounding component with a smooth delivery and tons of dynamic range with just about every headphone I tested it with.
Where other systems refine and distill to give you that perfect sound, the Bartok opts to give the rawest most realistic sound possible. Throw in all the mod cons of networked streaming, save for BT and built-in WiFi, and a very useable free app, and it is perfectly poised to cope with the digital streaming era.
Yes, the Bartok is huge, weighty, and oh so very expensive. However, it is likely all you could ever need for a high-end headphone setup and honestly, it could well be all downhill from here unless there is a Bartok 2 in the pipeline.
Please do not do that dCS, stick to the firmware upgrades and people will appreciate this beautiful example of engineering a lot more in the long run.
dCS Bartok Specifications
- Type: Upsampling Network DAC with Headphone Amplifier
- Color: Silver or Black
- Dimensions: 444mm / 17.5” x 430mm / 17.0” x 115mm / 4.6“. Allow extra depth for cable connectors.
- Weight: 16.7kg / 36.8lbs
- DAC: dCS proprietary Ring DAC™ topology
- Power Supply: Factory set to either 100, 115/120, 220 or 230/240V AC 50/60Hz
- Power Consumption: 30 Watts typical / 50 Watts maximum
- Software Updates: Download and update functionality available via Bartók App
- Local Control: dCS Bartók app for unit configuration and playback. RS232 interface (controlled by a 3rd party automation system). dCS Universal IR remote control is available as an optional extra.
Analog Outputs
- Output levels: 0.2, 0.6, 2 or 6 Vrms for full-scale input, set in the menu.
- Balanced outputs: 1 stereo pair on 2x 3-pin XLR male connectors.
- Output impedance is 3Ω, the maximum load is 600Ω (10k-100kΩ is recommended).
- Unbalanced outputs: 1 stereo pair on 2x RCA phono connectors. Output impedance is 52Ω, the maximum load is 600Ω (10k-100kΩ is recommended).
Headphone Outputs
- 1 x 4-way male XLR connector, 1 unbalanced pair on 1 x 6.35mm (1/4”) 3-pole jack.
- 1.4W rms into 33Ω per channel
- 0.15W rms into 300Ω per channel
- Output gain levels: 0, -10, -20, -30dB, set in the menu.
Digital Inputs
- 2x AES/EBU on 3-pin female XLR connectors accepting PCM at up to 24-bit 192kS/s or DSD/128 in DoP format individually and PCM at up to 384kS/s, DSD/64 & DSD/128 in DoP format or dCS-encrypted DSD combined.
- 2x SPDIF, PCM at up to 24-bit 192kS/s or DSD/64 in DoP format.
- SPDIF optical on a Toslink connector will accept PCM at up to 24-bit 96kS/s
- BNC connectors PCM at up to 24-bit 192kS/s or DSD/64 in DoP format.
- Dual-RCA unbalanced output
- Supports Apple AirPlay at 44.1 or 48kS/s
- Network Loop Out connector on a second RJ45 connector
- USB 2.0 interface on a B-type connector operating in Asynchronous mode, will accept up to 24-bit PCM at up to 384kS/s plus DSD/64 & DSD/128 in DoP format.
- USB- OTG capable of up to 24-bit 384kS/s plus DSD/64
Codec Formats
- FLAC, WAV & AIFF at up to 24-bit 384kS/s native sample rate
- DSD/64 & DSD/128 in DFF/DSF format
- WMA, ALAC, MP3, AAC & OGG
