The word “Disk” in the full name of the CD-77 concentrated my attention immediately. This is job-related of course, no doubts about that, but I immediately got an association with the nomenclature used by the Scottish Linn. There “disk” is common, Unidisk, Akurate, Majik, Classik, etc, are further examples of distinguishing from the rest of the world (in case of Linn), and especially from the English.
Abbingdon Music Research, a UK company, coming from London, also wants to separate from something. As we will learn later, it is about separating from the stereotype of the Compact Disc, and from the belief that this technology is without soul, mechanical and futureless. I have seen the devices from that company this year, during the Munich High End 2008. And even earlier I heard the information, that there is a new, serious player on the market, whose player makes other whine. But if something sounds too good, then it cannot be true, so I took those “revelations” with interest, but without euphoria.
But one of the most important information was the one about the digital to analog conversion chip. In this role the company used one of the first DACs constructed, coming from the early 80., the TDA1541A from Philips. It was designed by Philips just for one task – convert the digital signal from the Compact disc to the analog form. It is a 16-bit chip (multi-bit), a completely different design than the currently used delta-sigma converters. In the latter the current sources needed to generate the subsequent levels are based on laser corrected resistors. This means, that those can be made very precise and the linearity is perfect for low levels with those DACs. But they have problems, among others high nose levels. But the production of multi-bit DACs was not put on ice for substantial reasons, but only due to the high manufacturing costs – higher resolutions turned out to be difficult to manage in production, while the delta-sigma units achieve theoretical resolutions of 24 bits without problems. Still… Twenty years after their introduction the TDA1541 is one of the most regarded converters for DIY and among the most hard-core audio companies.
So the presence of this “antiquated” technology on board of the CD-77 was a motivating element and inviting to listening sessions. The looks of the device was not. I do not want to hurt anybody, or to superimpose my point of view, but I like to express my opinion: the AMR player, although made splendidly, without any weak points, does not appeal to me – the design, the looks are not of my kind. Seemingly everything is OK – this is a very big block of aluminum, finished in champagne color, with blue accents in the form of a lit disc chamber, blue LEDs inside the unit, that shine through the openings in the top cover, etc. It is not in my taste. And that is all I will say. Because like I mentioned the manufacturing quality, the ideas behind it are top notch in this device. I mentioned the Philips DAC. Let us add tubes in the analog path, powered from a tube rectified power supply, splendid word clocks, expanded power supply, refined drive section, and we will see that this is not a product from “I can do it if I want to” series. On the company page there is also much talking about the company philosophy, and we find there that the listening sessions are superior to the measurements.
And this not by coincidence. Another element of the company identification is the resignation of any kind of filtering after the converter. As a standard, after the signal is read from the CD and converted to PCM it is subject to oversampling (and sometimes upsampling) and then filtered. The theory says, that at a given sampling frequency, after the half of that value (this is called the Nyquist frequency) distortion in the form of reflections is added to the signal (called aliasing). Hence the need to filter the signal – at 44.1 kHz we need to filter everything above 22.05 kHz, at 96 kHz above 48 kHz, and so on. In case of the CD, with its 44.1 kHz sampling rate we need to fit between 20 kHz (this is the upper frequency limit defined for this medium) and the 22.05 kHz (the Nyquist frequency). The first value is important from the psychoacoustic point of view, as it states that (young people) can hear frequencies up to 20 kHz, and the CD must be able to reach it. This means problems, because we need to cut the signal in a very narrow band. This requires very steep digital filters (brickwall filters). And those filters introduce significant distortion to the signal, so called ringing before and after the signal impulse. To deal with this problem a technique called oversampling was introduced. This means that the sampling frequency is increased – in case of the TDA1541 it can be done by a factor 4 or 8. This allows for using much less steep analog filters, that have much less pronounced distortion. But it is still there. AMR, like for example 47 Laboratory, went for the whole thing and resigned from oversampling at all. This results in a very linear conversion and much lower digital noise (oversampling is in fact a digital filter). The price to pay is in the form of the mentioned aliasing distortion. The choice is up to the designer, and later to the buyer. But if this is not convincing to somebody, he or she can activate all the “magic” filters – with oversampling (x2 or x4), upsampling (96 kHz or 192 kHz). This is free choice…
There is nothing like coincidence. After finishing the test I received from the Polish AMR distributor a letter from owner of the company, where I found information what devices were used as reference when “tuning” the player:
„(When) we developed the CD-77, our benchmark was the turntable and phono-source. We used several exotic turntables including Garrard 301 to define the CD-77's performance. This is why vinyl lovers will love the CD-77 as it is the only digital source they can get enjoyment from. This is the most important fact. ”
It turned out, that after listening to the AMR in comparison to a few digital sources (like the DP-700 Accuphase, 561SE from Wadia, Gryphon Mikado, Ancient Audio Lektor Prime) and having in memory the sound of the Ancient Audio Lektor Grand SE and the Jadis JD1 Mk II + JS1 Mk III system, for me the digital sources nr 1, I had to go a level up, to make the examination, which is the test, have sense. Not that the CD-77 is better than the mentioned players, it is not about that. It is just, that some elements of the sound are so good, that they surpassed the best constructions I ever heard. And this is something different. In some other fragments the AMR was worse than each of the mentioned players. Like it is common in audio it does not matter WHAT we mix (which elements) but HOW we do it. And the cooking in AMR resulted in a product that can only be compared to the Lektor Grand SE or the Jadis system in terms of pure listening pleasure. Or with analog.
Yes – analog. I do not want to step on anybody’s toe, or become harsh with anybody, but I cannot pretend not to hear something that I do. Because only that comparison showed what the AMR does well, and what not so well. Every other comparisons were not complete, because comparing for example the Prime and the CD-77 I could only tell that the timbre of the latter is much better than the one of the Polish product. But what and when – that was not so easy. It was helped a little by the phenomenal Accu DP-700, that is also better than my player in terms of timbre, but not fully. The response to my searches was the sound that I got from an analog system, based on the parallel tested turntable SME10A. I repeat – I do not claim that the CD-77 is equally good, or that the digital source can beat the analog – absolutely not! It is just that I needed something as close tonally as possible, a bit better (but not too much) to be able to describe such a comparison and that it would make sense.
The listening session of this extraordinary device started with the comparison of its sound with that what I got from the turntable SME 10A (test in this issue of HF), a Dynavector 23R Ruby cartridge and the RIAA preamplifier RCM Audio Sensor Prelude IC. The system is important, because stating that we compared something with a turntable is not a full statement, and thus a wrong one. An analog system is made from the elements that constitute its sound (and price): the turntable (the basis), the arm, the signal cable, cartridge and finally the phono preamplifier. Only such a complete system can be called a ‘source’. In this case, the reference system cost 30000zl. I started with the disc Love The Beatles (Apple/EMI, 379 808 1, 2 x 180 g LP; 379 810 2, DVD-A+CCD; review of the digital version HERE). For a moment I was deceived, that the CD-77 sounds exactly like this beefed up analog. Not that I wanted to be deceived, because I approached this experiment like every other. But the quick transition from the LP to the CCD showed, that in terms of timbre and sound culture the AMR player can only be equaled by the mentioned Lektor Grand SE Ancient Audio, or the Jadis JD1 MkII + JS1 Mk III, leaving my Lektor Prime far behind. And although this is not the same level of resolution like in the mentioned players (but about that later, this is the element where we pay for that what I describe now – the nature wants to achieve balance), but the timbre is overwhelming, and the coming from the LP to CD does not mean a shock like with 99% of other CD players. Interestingly, the CCD disc (it is a pity this is not a Red Book CD) from the AMR sounded better than the DVD-A 24/96 version from the cheaper, but still nice, universal player from Arcam FMJ DV139. In the last case there was more information, the clarity was better, but I would have to take a lot of exercise to not to lie saying that the sound was better as a whole, or that I like the music better from the ‘dense’ format than form a ‘cripple’ CCD played by the CD-77. It was completely different – the CD player sounded in a more natural, believable way. Coming back to the LP, with the naturalism of this medium , a certain ‘ease’ with the approach to the sound, that results in an absolutely credible event being shown in front of us, showed vinyl as the better source. But even the bare fact, that coming from LP to CD was not a change of gender, but only a class says much. Let’s add to that, that my Prime, trying to show what the Grand SE shows, but far from being so refined, shows the CCD’s from their worse side – the Beatles sound a bit bright and squeaky, that is why I reach for the DVD-A when I want to listen to that disc. The same was some time ago with the disc Thunderbird Cassandra Willson (Blue Note/EMI 58 762, Copy Controlled Disc; review HERE), that I could not listen to due to the mastering problems.
The situation was similar with the Stockfisch samplers Stockfisch Records. Vinyl Collection (SFR 357.8006.1, 180 g LP) and Closer To The Music. Vol. 2 (SFR 357.4006.2, SACD/CD; review HERE). The voice of Sara K. from the second piece on the vinyl Stars was bigger, warmer and closer from the AMR player. The rough part of her voice from the upper midrange, that is exposed by the digital versions of her discs (even in XRCD²4 – see the review Hell Or High Water), was far less annoying in the vinyl version. But this can be attributed to the way the disc was made (I mean the vinyl), actually Love was done in a similar way. The base material was on digital master tapes – high resolution PCM (in Love I am talking about the new master). In case of Stockfisch the Direct Metal Master technology was used, but this still leaves a feeling of something lacking – the urge of wanting to know, how this could sound, if the material would have been prepared in analog domain from the very beginning. Mixed emotions were left by the piece Caruso Christian Willisohn, that I know almost by heart. The analog version sounded in a more open way, the vocal was more natural, but the English horn was shown better, had a better timbre – and interestingly – a much better location, on a broader stage from the CD (the CD-77). Frankly speaking, the digital version was closer to what I hear from the SACD version of the disc Hold On from Willisohn, where this piece was taken from (Stockfisch SFR 357.4038.2, SACD/CD; review HERE). And the version was recorded directly on a DSD recorder and was not processed afterwards. Transferring it to PCM for CD and vinyl (yes, this is the version on LP) separated the voice and the piano a bit. But from a good SACD player, like the tested for “Audio” DP-700 Accuphase, this recording sounded very similar to the CD layer on the CD-77. Incredible!
To try to resolve the case I took the disc Time Electric Light Orchestra – original vinyl pressing from 1981 (Jet Records, JET LP 236, LP) and its best digital re-master made by Sony Music Direct (Japan) (Jet Records/Epic/Sony Music Direct (Japan), MHCP 1161, CD). And here finally the gramophone showed, that digital is digital, and some things are for now (and probably forever) closed for the CD technology. It is mostly about the differentiation of the dynamics and timbre. When Twilightbegins, we have a quick transition on the drums and percussion. On the vinyl it is heard in a slightly compressed way, with a darkened treble – as if this fragment was inserted later in place of another one (what could have happened regarding the recording techniques). The whole band enters a moment later, and then the percussion sounds more open and shinier way. On the CD this cannot be heard in a so clear way, probably by strong compression of the dynamics. This was confirmed by a disc, which analog and digital versions were recorded in a purist way, best for the given medium - The Bassface Trio Plays Gershwin (Stockfisch SFR 357.8045.1, Direct-To-Disc 180 g LP + SACD/CD; review HERE). Here it was clear, that the AMR is not as dynamic as vinyl, or that it cannot differentiate the sounds and timbres at low signal levels.
That’s it for the vinyl. When we compare the AMR to my Prime or even the Accuphase DP-700, then it will turn out, that the timbre is set a bit lower than in both the devices, and that it is richer than in my Lektor, locating itself somewhere in the region of the DP-700. Because the timbre is the number one element in this case, and the element for which we can fall in love with the CD-77 forever. It is better than in my player, and maybe even better saturated in the mid bass and lower midrange than in the Lektor Grand SE. the CD-77 beautifully distinguishes the recordings, and if the disc is a bit brighter (the voice recorded on it) like the Songbird Eva Cassidy (Didgeridoo, G2-10045, CD), then it will be shown higher. But even then, the whole will be a bit warmer than in other players. This gives an impossibly credible transmission. I did compare the CD-77 to vinyl on purpose, because the sound goes that way. And you need really an expensive analog source (in my case it is about well spent 30000zl) to better the AMR player in terms of presenting timbre, naturalness of the sound, etc.
Please remember one thing: those remarks are true for the Digital Master II (or Digital Master I) mode. Like I mentioned in the beginning, the CD-77 is actually a few devices in one. But the first two modes offered the most saturated three-dimensional and really natural sound. Switching to 2x oversampling immediately flattened the sound, and only the 96 kHz upsampling mode was worse, setting the worst point from all the modes. Changing this upsampling with its completely flat and lifeless sound to 4x oversampling bettered some aspects, and only upsampling to 192 kHz opened the whole and liberated from the internal tension. The return to Digital Master II was like returning home – long awaited and longed for.
The CD player CD-77 Abbingdon Music Research, a brand owned by the British Abbingdon Global Group, is an unusual device from any viewpoint. Its weight (28kgs) tells, that no-one has made any shortcuts regarding the enclosure. And yes – it is made from thick aluminum plates, with an even thicker front panel. In the front we see only a big, blue dot-matrix display, that displays the track number and time, or the chosen input (we can use the CD-77 as an integrated CD player or a DAC with an USB input) or the chosen digital filter. And this all by pushing a button on the remote. The display is covered by a big, smoked plastic plate, much bigger than the display. This is why this element dominates the front panel. Under the display we have big, lit blue when the function is chosen, sensors. There is no tray, as the CD-77 is a top loader, with a manually operated cover on top. On top we also see the ventilation cut-outs, covered partially by glass plates, with the electron tubes underneath. Like I mentioned, the CD-77 is an unusual project. The base for the drive is the Sony optics (the same like in the Ayon CD-3, the one in CD-77 has not the plastic cover) and the servo from Philips (CD-18). The disc the CD is placed on was exchanged to a bigger, aluminum one. From the top the disc is held in place by a medium sized puck with a place for the motor spindle, exactly like the one in Ayon. This allows centering it better than in most Philips CD-Pro2 drives. The element with the lens is made from heavy brass. On the axis there is also a small spirit level. Through the openings we see blue light – those come from the power supplies – there are SMD LEDs mounted. On the back we have one IEC socket with a mechanical power switch, a USB digital socket and placed on the sides analog RCA and XLR outputs. The enclosure is colored light champagne. The whole is supported by solid aluminum and rubber feet.
After unscrewing a few things scream for attention. One thing is regarding the drive. It turns out, that the carriage with the optics is mounted on a metal plate, and this is again mounted to a big, solid cast element, visible after opening the disc cover. This element is placed on long springs, and those further on soft supports. A similar solution, meaning heavy mass on an elastic element, was used by Electrocompaniet in the EMC 1UP player. But there were no springs; the mass was mounted on rubber feet. The second thing is about the enclosure – the bottom and the back are not only aluminum, but also a second, internal layer from thick copper sheets (this is good RF damping), where all elements inside were mounted to. The drive control section with the software written by AMR is placed on a PCB underneath the drive. There is also the place where the CPLD (Complex Programmable Logic Device) and DSP (Digital Signal Processing) units were placed, that handles the upsampling and oversampling. We have six modes at our disposal. The first two, Digital Master I and II, have no oversampling (oversampling=1), without the digital and analog filters. In DM I only an antialiasing filter was used, and in DM II not. Because there is no antialiasing filter we get the aliasing distortion, but at the same time we get perfect impulse processing, without any ringing. Near the nicely looking clocks with the AMR logo we see the heart of the device – a big, 16 bit chip from the early age of audio: the Philips TDA1541A D/A converter. The same chip is the base for another unusual player, the D/A converter Model 5000 from Zanden that is also renowned for its musicality.
Most circuits in the AMR nomenclature is named by adding the Opti- prefix. We have OptiDrive, OptiSignal, OptiSample and OptiClockLock. I will not go into details – please read the company web page. The audio circuitry is located on two PCBs on both sides of the device. They are tube based – low power double triodes NOS: in the first stage the Mullard ECC81/12AT7, and in the second stage the 5687/6900 from Philips selected military JAN tubes. Both receive power from a rectifying tube EZ80/6V4 Mullard. The tubes are equipped in rubber rings, that eliminate vibration. Between the stages big polypropylene capacitors were used – between the tubes a CAP-777 with AMR logo and before the output the Mullard KP. On the output we have two sockets – RCA coming from the American company CMC and the not gold plated XLR. A balanced cable from the RCA socket leads there, but single capacitors suggest this is an unbalanced circuit. The PCB houses also elements for the power supply of this section – Nichicon capacitors and chokes. The power is supplied by two, shielded transformers, with 12 secondary windings, among others for the word clocks. A separate transformer and power supply is dedicated to the display, that can be turned off on request. The whole can be controlled by a splendid metal remote, where the most used keys have a classic look, but the less common are just icons on a LCD touch screen. This remote can also control the AMR amplifier. This is a phenomenal device. And finally: the player comes in a brilliant metal box, with soft elements for all items (a interesting power cable, AMR interconnects, Telarc CD, etc) – a box like professional devices. This is how every hi-end device should be supplied. Bravo!
CDs FROM JAPAN
|© Copyright HIGH Fidelity 2008, Created by B|