Synthesis Roma 69DC DAC (£2,499)​

Review by Martin Virgo



DACs are an interesting challenge to review, not because they all sound the same, although there are definitely family resemblances, but due to their impact being shaped by the system of which they are a part.

Is a DAC a source? Or is that the streamer AND the DAC? And what of the server? With SOME of these complexities in mind, this is intended to be part 1 of a two-part review. In this first part, I will be reviewing the Synthesis Roma DC69 DAC as a ‘stand-alone’ unit. In part two I will report on its performance when fed an upscaled signal.

I was drawn to the Synthesis Roma 69DC DAC when I saw it used valves in its output stage. My previous experiences with tubed DACs included the Border Patrol and the Audio Note 2.1 DACs; both of which I had bought. The former used a valve in its power supply, whilst the latter used them in its output stage. These were two very good, if very different, DACs.

The Synthesis Roma 69DC DAC had been built using an AK4495SEQ chip. Asahi Kasei Microdevices is a Japanese company which suffered a fire at its fabrication plant in October 2020; it lasted three days and caused a spike in DAC chip prices. This is the first time I have listened to a DAC using one of their chips.

The DAC itself is a half-width pressed metal box with a choice of fascia, the demo unit was black and suited my setup. On the rear are a series of connectors that would not look out of place on an AV amplifier, however, they functioned as required. The inputs include two Toslink, two coax and a USB type B. In addition to the analogue outputs, there is a Toslink digital output, something I don’t recall ever seeing before on a DAC.

The unit is a reassuringly weighty 5kg. The rest of the specs can be found HERE.

Removing the top was straightforward, with three screws per side, and revealed the two EC88s in all their glory.





A DAC is a summation of many parts and I was keen to hear what Synthesis had achieved by marrying valves to a modern chip. In order to get a bit more insight I asked some questions of the company’s founder, Mr Luigi Lorenzon.

Interview with Mr Luigi Lorenzon​

Why the AK4495SEQ DAC Chip?
The 4495SEQ was at that time the best 32bit DAC of AKM production in terms of performance and of course sound reproduction

I note that AKM mentions ‘Velvet Sound’ in their specification sheets, is this a brand name or a technology?
Velvet Sound is a brand name of AKM

What were the reasons for using the AK chip?
I loved the AKM from the first time used for the details able to reproduce, for its architecture and possibilities to set the external digital filter.

What is the impact of the valves on the output?
The input tubes on the device are the most important to the sound, they can change the flavour of the system. Of course, this depends also on the design of the circuit and from the type of tube chosen. On the ROMA 69DC the choice is the 6DJ8/ECC88 that, for me, is an excellent tube as sound and technical parameters are perfect for a differential amplifier. It has a low output impedance and it can be used on the output stage. The sound improvement is very high especially on the middle-high frequencies.

How long should the valves last?
The tube life is around 5000-6000 hours

Can the valves be swapped by the user?
Yes, there is no need to use any bias setting, but is extremely important to use a matched pair of tubes to avoid channel imbalance.

Any views on valve rolling?
After checking the different tube brands on the market, I decided to use the JJ Tesla from CZ Rep or Electro Harmonix from Russia, for production. This choice was made after evaluating tests of performance in terms of sound and technical parameters.

The input tubes on the device are crucially important to the sound as they can completely change the flavour of the system.

Is the digital volume control within the DAC chip?
Yes the AK4495 has an internal digital attenuator (255 levels and 0.5dB step)

Is there an impact on sound quality by NOT giving the user an internal volume control defeat option?
By default the level of the volume output is set at max. The attenuation (adjustable by selecting the mode from the front panel) is set by the ASK4495’s internal digital attenuator, controlled by a micro processor. There isn’t any quality degradation to the signal.

Is the headphone amp volume control analogue?
The headphone has a separate section made by a Hi-Performance OPA. The output level is variable by analogue volume control (potentiometer).

Are the valves used for the output of the DAC and the headphone section?
On the ROMA69DC the tubes 6DJ8/ECC88 are connected as differential amplifier configuration to the output of the DAC chip. The signal on the headphone section comes from the output of the tube, so the sound result is pure “tube sound”.

Is the internal PSU linear?
Yes, we use linear PSU. I prefer this design in a DAC to obtain better performance in terms of clean sound and noise floor.

Was there any bespoke development done on the design and build of the unit’s PSU?
Yes of course, the PSU has a great impact on the sound and we take care of this development in all projects, a dedicated design for each. On the ROMA69DC the filament supply of the 6DJ8/ECC88 is designed to heat in the cleanest way possible. Same for anodic power supply, the DC voltage is stabilised by a dedicated device for a clean and stable voltage, avoiding noise and interference that can have a negative impact on the sound.

Test Tracks: The Good, The Bad and the Bright​

The tracks were selected to allow:
  • Comparison of local and Qobuz sourced versions of the same tracks;
  • Comparison of standard and remastered versions of the same track;
  • Comparison of older and modern tracks, with their different mastering priorities;
  • How problem tracks were presented.
Qobuz Playlist:
Problem Files:

Headphone Listening​


The source was Roon, primarily Qobuz, with:
  • headroom enabled;
  • Sample rate NOT enabled;
  • Parametric EQ NOT Enabled.
Roon is hosted on a Vortexbox Audiostore Prestige 2 Server.

On powering the device the screen tells you that it is pre-heating the valves. It finishes its warm-up routines and then lets you know which input it is set to, these can be cycled through using a button next to the LCD screen. Two additional buttons allow you to adjust the chip-based volume control, this was always set to fifty, its maximum setting, as I was using the rotary manual volume control on the front with the headphones.

As a device, I have found that the Synthesis Roma 69DC takes about twenty minutes to come fully on song.

First up were my Quad ERA-1 planar headphones. I would say that these tend to lean towards a more detailed presentation, with slightly more emphasis on the higher frequencies. Here the top end was rather less spotlighted. Instruments such as cymbals were more inclined to shimmer and ring without being insistent. In fact, as I moved through the playlist I found that the tonality of the music via the Synthesis 60DC was that touch warmer and more inviting.

Luigi highlighted how the valves are used to tune the mid and upper frequencies, but there is no need to worry about the bass which I found to have both punch and articulation.

Having just bought a pair of Meze 109pro headphones I was wondering whether I had wasted my money. In the event, there was no need to worry. The Meze have an ability to deliver greater timbre, detail and dynamics; and the Synthesis Roma 69DC took full advantage of this.

Whatever the music the greater articulation of the instruments and the voice was evident, and drew me into the music. At no point did I find that vocals developed an edge, which will always pull me out of the music; this is not something you hear in any voice naturally.

Perhaps a flip side of this is that on balls-to-the-wall rock, such as Rise by Extreme and ‘For Those About to Rock You’ by AC/DC, the music is laid out cleanly allowing you to follow everything clearly, but it somehow doesn’t have that soupçon of excitement that gets your blood pumping.

I played one of my problem files, ‘Rubber Bullets’ by 10cc from their ‘Best of the Early Years’ album. This is full of problematic sibilance, but not through the Synthesis Roma 69DC. That is not to say that there was no sign of the problems, just that they were much reduced and tolerable.

While the Roma 69DC isn’t able to extract the same level of detail as the Hugo TT it is very musical and can highlight some dynamic expansiveness that escapes other DACs. The Greatest Showman contains some dynamic and highly produced music. In the opening number layer on layer of complexity is introduced as the track builds. At about 58 seconds the is an expansion of the soundstage as this reaches a climax. Many systems don’t reproduce this, the Synthesis at least acknowledges this effect is there, even if it is not as expansive as on some FAR more expensive DACs.

During my time with this DAC, I only had one issue, the manual headphone volume control. When adjusting the volume there was a swish that I would associate with a dirty contact in the potentiometer. I was using the demo model.

Time to move on to my main system.

Main System Listening​


The source was Roon, primarily Qobuz, with:
  • headroom enabled;
  • Sample rate NOT enabled;
  • Parametric EQ Enabled, deals with some minor bass issues.
Roon hosted on a Vortexbox Audiostore Prestige 2 Server.

Having confirmed that the unit was thoroughly run in I connected the DAC into my system I started to casually listen. I noted again that the unit took about twenty minutes to come fully on song.

As before the unit volume was set to its maximum of fifty, as I was playing this through my Icon 4 passive pre-amp.

Initially, I just enjoyed a wide range of music. It replaced my Chord Qutest, which I use with an SBooster LPSU and SBooster Ultra Mk II, raising the price to more in line with the Synthesis Roma 69DC. I made the following notes:

Flight Of The Cosmic Hippo, Béla Fleck And The Flecktones, Flight Of The Cosmic Hippo, Qobuz, 44.1/16
This is a track that I generally use to test BASS, and boy does this track have bass in bucketfuls. But, that bass is beautifully rendered and full of timbre. The Synthesis rendered this very well, powerful with lovely resonance. At the top end, the cymbals rang beautifully with easy differentiation between them and the way they were being struck or brushed.

Comparing this to the Chord Qutest, sans M-Scaler, I felt that whilst the bass was somewhat less profound it had greater resonance, I got a bit more feel that this was being produced by an acoustic instrument. Additionally, the sound stage was a tad more central with front-to-back depth.

Schéhérazade, op. 35 – Rimsky-Korsakov, Fritz Reiner - Chicago Symphony Orchestra, Schéhérazade, Qobuz, 24/88.2
A great favourite from the first time I heard it at age thirteen and fell in love with the piece. The Synthesis rendered the strings of the various instruments very well, giving them a touch of warmth and musical resonance the Qutest could sound a touch glassy and removed.

The sound staging of the Roma 69DC was a touch more 3D than the bare Qutest.

The Greatest Show, Hugh Jackman, The Greatest Showman (Original Motion Picture Soundtrack), Qobuz, 44.1/16
This is not a great musical, it is too light a piece to claim to be more than what it is, which is a very entertaining bit of musical theatre. This is the opening number and is highly produced, having many layers added over time in a rather Hornesque style. However, the engineering of the album hasn’t led to the most free and dynamic representation of the music. But I do enjoy the album.

I recently had the chance to listen to this track through George’s system, via his Grimm MU1 and Mola Mola Tambaqui. Now this was playing through some significant other gear but I was taken with the way the music expanded away from the speakers, it didn’t just crescendo; at 58 seconds for instance. In my system, I was NOT hearing this, and I put this down to my systems and room acoustics. However, when I played the track through the Synthesis Roma 69DC I did hear a trace of this effect.

Picture This (Remastered), Blondie, Parallel Lines (Deluxe Collector's Edition), Qobuz, 44.1/16
A remastered album that sounds good via both the Roma 69DC and the Chord Qutest, but for subtly different reasons. The Qutest wins in the detail it extracts, the Roma for its presentation of Debbie Harry’s voice.

Here Comes The Sun / I Want You (She's So Heavy) (Album Version), George Benson, The Other Side Of Abbey Road, Qobuz, 44.1/16
This is an interesting track for a few reasons, including the complete change of pace in the middle. George Benson has a great set of pipes and the Roma renders these well. The piano is a bit weedy, but the 69DC has a bit more body than the Qutest. Again, the Synthesis Roma has a good verisimilitude with respect to the strings, in this case, the cello had a very nice rasp.

Rise, Extreme, Qobuz, 44.1/16
This track has one of the best guitar solos I know of, real seat of the pants stuff. This is where the Roma 69DC shows its slight Achilles heel, while it is resolving and informative it is also a tad polite.

The Magic Flute: Act 1 – O Zitte Nicht, Mein Lieber Sohn, Philharmonia Orchestra - Klemperer, Qobuz, 96/24
Beautifully presented soprano in a large acoustic, you are aware of her voice bouncing off the boundary of the space. While the Qutest is damned close I definitely give the nod to the Synthesis Roma 69DC DAC.

Playing my usual suspects of problematic files the Synthesis Roma 69DC is certainly no cure. On the whole, it was a tad better than the Chord Qutest with most tracks, although it ‘cured’ El Condor Pasa.


Reviewing DACs is never straightforward as there are so many variables, over and above the system within which it is being asked to perform.

Any component is always a small piece of a larger recipe. You will note that I did not report on the Synthesis Roma 69DC in the context of my nearfield system, this is due to the fact that I am now using an amp and speaker pairing that are not synergistic with the Roma. In order to achieve this I would need to buy an alternative pair of speakers, such as the Harbeth P3s, and no one is paying me for this gig!

There are definite and consistent differences between the Synthesis Roma 69DC and my usual Chord Qutest. Picking a winner almost feels like choosing what sort of music you prefer, I came to realise that there is a range of music that I love that I have been shying away from since I bought the Chord Qutest, but I returned to with pleasure via the Synthesis Roma 69DC.

I could write that the Qutest is honest about poor recordings, and there will be a degree of truth to this. However, the strengths of the Synthesis Roma 69DC include the slight extra body and harmonics it gives to acoustic instruments and the human voice. Additionally, where a recording fails the Roma does not thrust this in your face, it allows it to fail gently. I truly appreciate this as so much music I love is not well recorded.

The area where this DAC’s strengths slightly weaken my enthusiasm is harder rock and metal. It reproduces the music well, but somehow the blood doesn’t quicken as much, the music is more something to be observed rather than experienced.

With other genres, I worked my way through old and new music with great pleasure. This is a DAC that I have thoroughly enjoyed spending time with and I can happily recommend the Synthesis Roma 69DC DAC to include on your audition list.

Part II

When I first slid the Synthesis Roma 69DC DAC into my system I did so on the end of my Chord M-Scaler. This has a consistent effect on resolving data and expanding the sound stage.

Over the course of the next three hours three things struck me:
  • The higher frequencies are not as highlighted through the Synthesis Roma 69DC as through my usual Qutest;
  • The sound field generated by my speakers was a tad deeper than usual; and
  • I was more relaxed.
This was a very positive impression. However, not many people are going to buy a £3.5k upscaler to feed a £2k DAC! The combined price of £5.5k will have the Synthesis Roma 69DC competing with a different bracket of competition.

Putting on my thinking cap, ‘What if I use HQplayer to do the upscaling?’. This would require some changes to my digital chain so I have been talking to a company with whom I have done business for a fair few years, Vortexbox.

In association with Vortexbox I will be putting together an alternative front end to test this question.