Review: The Aqua Formula xHD Rev.2 D/A processor- What the broadband concept is for loudspeakers and the single-ended triode for amplifiers – that seems to be the so-called NOS ladder DAC for D / A converters. So: pretty rare and off the beaten track. But somehow also a bit old-fashioned. Because it’s not the case that an R2R resistor ladder for converting the digital signal is the whole new thing. On the contrary, it used to be the only way to change. In the meantime, the process – as popular as it is in some audiophile circles – is rather marginal. Delta-Sigma DACs have caught on. “For the wrong reasons,” the vendors of ladder DACs never tire of stressing. The Milan-based company Acoustic Quality, or Aqua for short, belongs to the Ladder DAC group. This review is about their flagship called Aqua Formula xHD Rev.2 (Price: 13,890 euros).
Of ladders and streams
When the CD was introduced in 1982, the converters in the players were of the ladder type. The term is derived from the ladder-shaped arrangement of the individual resistors that are switched to convert the data word. The value of the next resistor (R) in the ladder is double or – depending on the direction from which you look at the matter – half as large as the previous one, therefore also: R2R ladder.
The concept is basically pretty clear. Let’s start with the Redbook standard, i.e. a 16-bit word that needs to be converted: If it only consists of ones, all resistors are switched and the greatest current flows, corresponding to the highest amplitude of the original signal – on the other hand, it only exists from zeros, no current flows at all: absolute calm. In between there are 2 to the power of 16 (minus 2) intermediate values and the performance takes place 44,100 times per second. Afterwards, it is filtered, nowadays mostly digital, in order to remove so-called mirror spectra – these are multiples of the original useful signal in the ultrasonic range that are created by the digitization process.
One difficulty of the R2R concept: the resistors have to be very, very precise, because if they aren’t, there are amplitude deviations that can be annoying, especially in quiet passages in the music. Another: Ladder DACs cannot be easily integrated into ICs, the concept may be elegant, but it is complex in terms of production technology.
It is different with DAC chips of the delta-sigma type, which dominate the market today and are much cheaper to manufacture. Their concept is a little more complex: the multibit word is initially oversampled ( oversampling ) and converted into a bit stream that has a significantly higher frequency than the original sampling rateruns through the converter, but knows far fewer possible current / voltage values, namely not 2 to the power of 16, but in the case of a 1-bit converter only two: the zero and the one. The original information on the amplitude of the original signal is no longer contained in a varying level of voltage, but in the varying width of the constant pulses of the bit stream. Inherent in the concept is digital filtering of the signal, which is also necessary in order to move the higher quantization noise due to the reduced bit depth out of the audible range by means of noise shaping.
Avoiding digital filters is now the crux of the whole NOS ladder DAC idea. NOS stands for Non Oversamplingand means not only avoiding oversampling, but also doing without digital filtering. Digital filters may do a good job in the frequency domain, but in the time domain they smear the signal through post-oscillation and, above all, through artificial pre-ringing of pulses. And because music is made up of impulses and exact timing is psychoacoustically more important than maximum accuracy in the frequency range, especially since inaccuracies only arise here beyond the audio range, it is sonically crucial to avoid digital filters and thus ensure perfect impulse reproduction – at least that’s what fans of the say NOS ladder DAC concept. Alternative designs such as minimal phase or apodizing filters would only postpone the problem, but not tackle the root of the problem. In addition, non-oversampling DACs are more robust against jitter: the higher the work cycle, the greater the proportion of a given timing accuracy in the individual sample. Ergo, you successfully put obstacles in your own way with oversampling. So – analogously – also Aqua boss and mastermind Cristian Anelli. I said yes: off the beaten track.
Technology Aqua Formula xHD Rev.2
There are now four “ladders” in the Milanese flagship DAC – one for each half-wave and channel. These cascades, which consist of 4 x 63 individual resistors, are controlled using a proprietary software solution that is built into an FPGA module (a field programmable gate array is an IC into which an application-specific logic circuit can be loaded).
The connection between the digital world – the algorithms in the FPGA – and the resistor ladders, which are actually already analog, are made by optocouplers, which ensure galvanic and magnetic isolation. For this reason, Aqua speaks somewhat grandly of an “Optologic D / A conversion system”. On the format side, PCM signals up to 24 bit / 768 kHz and Direct Stream Digital up to DSD512 are accepted. MQA is left out, because this format implies data processing and filters – which Aqua rejects. Audio data is never manipulated, but always processed bit-perfect with the incoming sampling frequency, so the credo of Acoustic Quality. However, DSD is internally converted to PCM, the “natural” format for such a multi-bit converter.
On the input side, in addition to coaxial S / PDIF in cinch and BNC versions as well as AES / EBU via XLR, a USB-B interface is of course also offered – and here, too, there is nothing off the shelf. It is a “two-stage hybrid solution” with an XMOS xCore XE216 and an FPGA, which was programmed by Aqua and is intended to ensure completely isolated decoding of the USB and generation of the clock signal.
As good as this USB interface is, the so-called AQlink is still the best connection in terms of sound, says Anelli. Behind this is an I²S interface, which was implemented via RJ45 sockets and corresponding Ethernet cables. Aqua’s La Diva CD drive or the brand new network bridge called LinQ serve as possible sources. Finally, the customer can choose the fifth input himself, there is a choice of Toslink, S / PDIF via Cinch, AES / EBU – or, as in our test sample, an optical AT&T interface that works similarly to Toslink, but has a better sound. so the Italians. Now all you have to do is find a player …
As befits a top-tier DAC, the power supplies for the digital and analog circuits are strictly separated from each other. This is done consistently right up to the two C-core transformers .
A look under the hood of the Formula xHD Rev.2 also reveals the modular structure that is part of the basic concept at Aqua. In the event of future upgrades, all assemblies can be brought up to date by exchanging modules. This is of course worth its weight in gold, especially with such an expensive component, and the exact opposite of built-in obsolescence. So far there have been two such upgrades with the Formula DAC, each of which had to be paid around 1,000 euros. Finally, the output stage was revised, in which, among other things, the transformers for the symmetrical output were replaced. Quite right: The Aqua Formula has a transformer-balanced output – this is rarely found in the hi-fi sector, but it is common in the studio.
But back to the basic concept, the filterless NOS design. Is the Aqua Formula xHD Rev. 2 really completely filter-free? Then what do you do with the freely floating mirror spectra? Yes, there is no digital filtering, says Anelli. However, a “light” analog filter was installed at the output, and at least with the XLR outs, the transformers do the rest. Anelli does not spread what “light filter” means exactly – and instead explains that the whole subject of mirror spectra and the intermodulation distortions that may arise in connection with itfrom his point of view it is a very academic one. On the one hand, the proportion of very high frequencies in the useful signal is naturally low and therefore there is also the risk that they intermodulate to the relevant extent with the lower mirror spectra (which start at around 24.1 kHz in the Redbook standard). On the other hand, it does not have a negative effect on the sound, because most of them simply cannot hear it – and in general, a clean impulse reproduction is much more important psychoacoustically. With higher sampling frequencies from 88.2 kHz, even the “academic problem” is off the table, because the corresponding mirror spectra are too far away from the audio tape.
Aqua Formula xHD Rev.2: Sound impression & comparisons
First of all, some “expectation management”: Just because the Aqua costs almost three times as much as the converter, with which I spend most of the time, I don’t require it to sound “three times better”. That doesn’t work at all, as everyone knows, because in these spheres additional small sound gains are continuously becoming more expensive. And to be completely honest: We are beyond reason, or at least beyond normal price-performance considerations. But what I do expect is that Aqua’s flagship converter will not just offer a different flavor than my Luxman DA-06 or similarly priced DACs, but will move into a new sonic league. Otherwise none of this would make sense.
Speaking of flavor: As far as the tonal is concerned, the Aqua Formula xHD Rev.2 actually serves a different one than the Luxman. Both D / A converters are balanced, the Japanese actually play neutral, the Italian, on the other hand, with a little more oomph and the slightly warmer mids. Of course, this was applied so sensitively that a term like “emphasis” led on the wrong track, especially since the bass is presented in a contoured and differentiated manner and does not shine through sheer bulk – an Accustic Arts Tube Dac II Mk 2 (approx. 8,000 euros), for example warmer on the way.
And yet, with the Aqua, something is distributed more physically. The grand piano that Nick Cave plays on Idiot Prayer is much more believable in the room, with more physical presence and grounding, and the same applies to the voice of the Lord. On the beautiful country track “Ain’t got no money” from the last record of Justin Townes Earle, who died too earlythe bass line rumbles (it must rumble!) a lot more audibly than I usually get presented. From the middle to the highest heights, the Aqua appears almost linear, nothing tastes good, nothing is rounded. So all in all an honest skin – which gets its warmer touch from the substantial gait in the basement / basic tone and not from shyness around the top.
So far so beautiful. And that was it with the questions of taste. In all other areas he is clearly better. Let’s check out a few of the common sound criteria.
I can easily see that the Aqua Formula xHD Rev.2 conveys a more spacious and, in particular, more deeply illuminated room impression and makes individual sounds more precise, more 3-D-like – and not only compared to my Luxman, but actually compared to all DACs and known to me Streamers. It might be interesting to compare it with a current Esoteric CD player in this league, but unfortunately none is available to me. I still remember the older models as extraordinary space talents. I almost suspect that an esoteric player would raster through the stage even more strictly, perhaps too strict for some tastes, while the large Aqua favors a flexible, organic stage design for all its imaging precision. But let’s leave the speculation – the adjectives wide, deep,
But there is more, even more crucial. For example with “Girl in Amber” from the Nick Cave live album mentioned. I experience the change to aqua as one from “excellent reproduction” to “now this is actually playing in the room”. Does that sound too poetic to you? I have come up with a little prosaic theory as to why I perceive it that way. It could be due to the fact that the Aqua Formula simply traces quietly and very quietly signal components much more accurately than other representatives of the guild.
Back to the music example: a piano string does one thing above all – fade away. So get quieter. And the reverb parts of the recording room – it is a “solo concert” in London’s Alexandra Palace, due to the corona without an audience: cave, piano, that’s it – are also one thing above all: quiet. Quiet in relation to the “main signal” that is lost in them. No, it doesn’t really sound more reverberant with the Aqua, and it’s also not the case that you can precisely measure the longer sustain / decay of the individual string with the stopwatch in hand. And yet it seems that the many, apparently negligible, quiet little things can be overheardwill. This is exactly what makes the subtle but decisive difference for me: the grand piano looks more harmoniously saturated, richer, with all the fibers “there” over the aqua – the illusion that the performance is taking place here in my room is frighteningly real. I wouldn’t have thought it possible either, but whenever I return to my standard setting, it looks a little bit gray, duller, duller – and less authentically depicted.
Does my “theory” carry you further? One of the strengths of the Aqua Formula xHD Rev.2 is the extraordinary wealth of timbres. And here the – again relatively quiet – spectral components of the overtones play a role. But I don’t want to overstrain that either, for the perception of timbres. Finally, other parameters are also decisive, such as the transient response. Perhaps the superior type of impulse processing postulated by the manufacturer is noticeable here by dispensing with digital filtering? Might be. In any case, the formula does not simply have an opaque brushstroke – the Aqua does not look like a watercolor at all – but a highly differentiated and sensitive one. The textures and color shades of the cello at the sporty start in the Sonata for solo cello, Op.8 by Zoltán Kodály (album: Grieg – Janáček – Kodály by Danjulo Ishizaka / Shai Wosner) worked out clearer and more meticulously than I’m used to – and since the instrument in turn appears to be more cleanly embedded in the recording room, the bottom line is that the performance looks more real.
The accuracy compared to low-level signal components naturally also pays off in terms of microdynamics. Sure, in terms of coarse dynamics, no eye stays dry either: When required, the Aqua can do the job. But is it to be expected that a DAC that is ten times cheaper would lag significantly behind? Rather not. Coarse dynamics do not seem to me to be a really relevant bottleneck in D / A converters, which they usually master. In terms of fine dynamics, it looks different. And this is exactly where the Aqua converter impresses, for example with the aforementioned Justin Townes Earle track “Ain’t got no money” (album: The Saint of Lost Causes), because the harmonica on the right is shown more precisely, not only in terms of space, but also in terms of volume. I find it even more noticeable with Julie Byrne’s voice and acoustic guitar (album: Not even happiness)
.
Limping comparison?
“With all the love and as good as it is – just under 14,000 euros for a DAC, who is so crazy?” It goes through my head. Then I look at the analog set-up in the rack, I add roughly … oh, look at someone. The formula converter in itself is still almost cheap. Apparently there is a certain market potential for crazy people. And while we’re at that kind of weirdness – let’s do a system comparison for a change! If I add my Innuos music server to the Aqua, digital and analog sources – SME turntables , transrotor pickups and BMC phono preamp – are at eye level.
What I hear here amazes me after all. At least in the tonal area, I would have thought that the differences would be more obvious – but they don’t. That may be a coincidence. I didn’t put together my analogue source as a “romantic alternative”, that’s pretty straight for vinyl playback and therefore possibly so similar to the Innuos / Aqua team. Which, by the way, ultimately offers a little more stability and assertiveness in the deep bass, as heard with the Poliça record United Crushers. But that’s a small difference and really only affects the lowest bass registers.
The similarities predominate, yes, they are so big that you can easily get confused. Okay, if it crackles, it was probably the record player, but that’s banal. What is more decisive is: Especially where the strengths of vinyl lie – in a differentiated, detailed, color-fast and lively presentation of the midrange and treble bands – the Aqua is damn similar. With SME & Co I experience the vocal area a bit “closer”, in a positive sense more raw and immediate, finely dynamically minimally differentiated. But these are minor nuances, which is why I want that to be understood as extra big praise for the Aqua converter. Otherwise the differences are greater.
What the comparison underlines again: The way the Aqua puts the music in the room is simply great. Nothing remains in the rough, fuzzy, two-dimensional. But it’s also not this “hyper-realistic holographic show” that high-end digital sources sometimes offer. It is undoubtedly fascinating to begin with. But does it also look like it is made of flesh and blood? Isn’t there a slightly artificial note involved? Doesn’t a first-class turntable appear bigger, rounder, more organic, only apparently more casual in the stage structure – and that is precisely why: more real? At least that’s how I feel, especially with voices, guitars, strings and the piano. And that’s exactly what the large Aqua converter can do.
Conclusion: Aqua Formula xHD Rev.2
A D / A converter like the Aqua Formula is quite an exclusive story. One in which tough price-performance considerations are irrelevant. Glam factor and feature richness are also left out: there is no volume control, neither headphone nor digital output, no streaming module, let alone something as unworthy as a Bluetooth interface. For those who find this strange, there is good news: You can buy it all for a fraction of the money elsewhere and become happy with it.
Formula is all about what is possible in terms of sound when a concept is fully exhausted – in his case that of a digital filter-free, non-oversampling ladder DAC. And damn it, a lot is possible here! I don’t even want to claim that the sound image appeals to every audiophile. The Aqua sounds slightly warmer and does not belong to the genus of digital sources with a “show off” sound signature: it leaves it to others to use concrete-hard bass, extra-airy highs and a staggered stage graduation like on graph paper. A clear sign that he’s closer to the live impression if you ask me. But in the end it remains a matter of taste. After this test, however, it is clear to me: The Aqua Formula is the D / A converter with the most authentic sound character that has come before my ears so far.
Profile : The Aqua Formula xHD Rev.2:
- Overall balanced tonality with perfect broadband. A little more sonorous in the bass and fundamental.
- Very deep, somewhat stronger, always structured, but never “over-regulated-dry” bass range. The Aqua can make the deep tones swing with color and drawing, as is often an advantage with acoustic instruments – and if required, crack down on hard and screw synth bass surfaces into the room.
- The midrange and treble bands are a coherent unit, nothing appears attached or indistinct – the fundamental tone that is well placed in the lining makes the mix a little warmer.
- A clear strength of the Aqua is its extremely good resolution – it is so high that it appears “softer” with it (because it is less “coarse-pixelated”). Even the smallest signal components are tracked accurately, a virtue that is said to be of top analog equipment. This has an impact:
- Sound textures and colors appear extremely real, authentic. Opaque too, yes. But above all, very differentiated.
- The fading and fading of tones is presented in a strikingly real way. A major reason why, for example, piano music comes across so impressive with the Aqua.
- Quiet reverberation and reflections from the room do not crumble away porously in the background – but are first-class “long” traced.
- Speaking of spatial representation: The Aqua adheres to the specification and lets the stage start where the recording is intended, there is no preference forwards or backwards. The situation is similar with the dimensions of the room, spacious recordings are rolled out uncut, and chamber music is presented intimately. The representation of the individual voices turns out to be precise and exceptionally plastic. The sounds seem like they are made of “flesh and blood”: round, organic, natural.
- The coarse dynamics are mastered, but DACs usually offer much cheaper DACs as well. In contrast to this finesse in the fine dynamic differentiation. With the Aqua you feel closer.
- The Aqua Formula is a very expensive device. All the more commendable that it has been consistently structured in a modular manner so that future developments can easily be retrofitted. That makes the investment in him future-proof.
Facts:
- Product: Aqua Formula xHD Rev.2
- Category: D / A converter (NOS ladder DAC)
- Price: 13,890 euros
- Dimensions & weight: 450 x 370 x 100 mm (WxDxH), 9 kg
- Colors: front in black or silver, body Nextel dark gray
- Inputs: 1 x RJ45 AQlink (I2S interface), 2 x S / PDIF coaxial (BNC, Cinch), 1 x AES / EBU (XLR), 1 x USB-B, an additional input according to customer requirements: AES / EBU, S. / PDIF coaxial (Cinch), AT&T ST Fiber or Toslink
- Outputs: 1 x XLR / balanced, 1 x RCA / unbalanced
- Data rates: maximum 24 bit / 768 kHz (PCM) and DSD512, no MQA
- Other: Mute and phase reversal switch on the front
- Guarantee: 5 years
