creates huge sonic advantages for the AVR350, over all other surround processors, which suffer the many sonic degradations just described.
       The AVR350's sonic superiority, already so obvious in direct analog mode, also extends to all digital modes, thanks to the Stealth Mat. These digital modes include simple A/D conversion with straight channel playback, the various Dolby and DTS decoding modes, and reverb effects modes. Even the built-in digital FM tuner seems to sonically benefit from the Stealth Mat, with better quieting and a richer tonal balance than heard before in the AVR300.
       In its various digital modes, the AVR350 clearly reveals more about music, voices, sound effects, and space itself, than any other surround processor we have ever tested. Its quieter, blacker background simply lets more information through than we have ever heard before, from a surround processor. And the sonic improvements extend across the board, to cover virtually all sonic aspects (which is natural, much as a disc transport with much lower jitter likewise can provide across the board sonic improvements). The AVR350, in its various digital modes, is not only quieter and more transparently revealing, but also cleaner (with lower distortion), faster, more articulate, more natural, and with better spatial imaging (including an even more believable portrayal of surround space than the already superb AVR300).

Dolby Pro Logic IIx

       If the AVR350 sounds twice as good as before (the AVR300) in various digital modes, then it sounds ten times as good in one particular mode, Dolby Pro Logic IIx (as applied both to two channel input and also to Dolby Digital multichannel input). The AVR350's sonic transformation of Pro Logic IIx is so radical that it sounds like a whole new decoding algorithm, so amazing that it merits its own special discussion here.
       Dolby decoding in general sounds somewhat rounded and smooth, especially on transient attacks that should be sharp, fast, and a little hard. Then, the Dolby matrix extraction modes (EX, Pro Logic, Pro Logic IIx) go much farther, usually making the sound fuzzy soft, defocused, and indirect. These matrix modes do this because they subtract information from the main channels in order to derive the new channels added by the matrix. This subtraction occurs on an instantaneous phase basis, so the remainder signals that it leaves behind in the main channels sound phasey from this subtraction, as do the newly derived channels. This soft, fuzzy, defocused, indirect change to the sound quality was a sonic weakness of the Dolby matrix extraction modes.
       For playing film soundtracks, we felt that this softening change was worth enduring with Pro Logic IIx, since Pro Logic IIx provides far superior surround spatial imaging than any other mode. But, for enhancing two channel music into surround mode reproduction, we preferred and recommended a DTS system instead. DTS enhancement modes have the design advantage over Dolby enhancement modes in that theDTS modes leave the main channels alone, and do not subtract anything away from them when they derive the added channels, so the main channels are left with their fidelity substantially intact, with proper articulation, attack, speed, and directness. Now, DTS matrix modes are not sonically perfect either, having a tendency to reduce warmth (making the tonal balance leaner) and add some artificial glare in the upper midrange. But on the whole DTS provided better fidelity for music than Dolby matrix enhancement modes.
       That whole equation, that whole balance of pros and cons of Dolby vs. DTS for various applications, has now dramatically changed, thanks to the AVR350. The AVR350 provides a sonic quality from Dolby Pro Logic IIx totally unlike we have ever heard before. The AVR350's rendition of Dolby Pro Logic IIx is far superior to that in any other surround processor we have heard, from Arcam or any other brand, and by a huge margin. Indeed, the sonic difference is so dramatic that we asked Arcam if there was a new Dolby matrix extraction algorithm they had implemented.
       The "new" sound of Dolby Pro Logic IIx in the AVR350 is hugely more focused, articulate, fast, and direct, with nary a vestige of the fuzzy softening, defocus, phasiness, and indirectness heard in every other processor's implementation of Pro Logic IIx. Arcam told us that their implementation of Pro Logic IIx was the same in the AVR350. So what then might account for the huge sonic improvement we heard?
       The ideal mathematical model of Dolby matrix extraction directs that signals will be subtracted and extracted from the main channels, and fed to the surround channels being enhanced, when certain exact phase relationships obtain among these signals. And the ideal mathematical model does this extraction in instantaneous, moment by moment fashion, so it depends upon and also affects the exact phase relationships among signals, on an instantaneous, moment by moment basis. Thus, the accuracy with which this Dolby extraction can be actually accomplished in practice depends crucially upon the circuits' ability to accurately assess the exact amplitude and phase relationships of the signals, to perform this assessment instantly, and to then perform the extraction with perfect (perfectly accurate and instantaneous) phase matrix subtraction. If the circuits doing the Dolby matrix extraction fail to achieve these ideals, then they will fail to truly replicate the ideal mathematical algorithm that the Dolby engineers intended. Obviously, circuits cannot be perfectly accurate and infinitely fast, so they will fall slightly short of the ideal mathematical algorithm. But let's suppose here that these circuits are at least basically capable of doing these things, if not perfectly, at least well enough to be audibly OK.
       Well then, what else could possibly intrude or go wrong, to make actual Dolby matrix extraction fail to achieve the ideal mathematical algorithm? One word. Noise. If noise is added to the signals and/or to the digital processing circuitry, then the purportedly exact phase extractions will become very inexact, occurring at the wrong amplitude, the wrong phase, the wrong effective time.
       What is worse, and sonically far worse, the error of this extraction is not constant, but rather is continually randomly modulated, in both nature and degree, by the random noise, which by definition is itself continually changing. In other words, the extracted derived signal will randomly vary from its correct amplitude, phase, and effective time, as determined by the continually varying random noise in the system. Likewise, the remainder signal left in the main channels by this extraction will randomly vary from its correct amplitude, phase, and effective time, again as determined by the continually varying random noise in the system.
       Thus, all channels, as you hear them from a Dolby matrix mode, are continually having their amplitude and phase and time (in short, everything about the signal) randomly modulated by any noise in the system, that noise being a randomly varying phenomenon that is continually swishing about. No wonder that, with all the signals from all loudspeakers continually having their amplitude and phase and time being swished about by random noise, the end result sounds soft, phasey, fuzzy, and indirect. Old timers may recall a similar sonic effect in Toni Fisher's hit, The Big Hurt, wherein flanging was used to deliberately make the sound phasey, swishy, and soft.
       The bottom line is simple. Noise added to the system makes a mockery of the ideal mathematical algorithm of Dolby matrix extraction modes.
       We saw above that the Stealth Mat in the AVR350 reduces the electromagnetic noise impinging on all digital circuits in the AVR350. That includes the digital circuits which perform the Dolby matrix extraction modes. Thus, it stands to reason that the AVR350 is drastically reducing the noise level in the Dolby matrix extraction circuitry, and is thereby allowing this circuitry, for the first time in any processor, to get close to the ideal mathematical algorithm that the Dolby engineers intended. And that's why we hear such a huge difference, such a huge improvement, in Dolby Pro Logic IIx reproduction in the AVR350, over any and every other processor.
       For the first time, Dolby Pro Logic IIx, as heard in the AVR350, need make no apologies for its sonic quality. It sounds admirably articulate, direct, and fast. It still sounds gentler than DTS Neo:6, but its error on the side of softness is now less than Neo:6's error on the side of hard glare, so Pro Logic IIx is now overall the more sonically accurate.
       Dolby surround enhancement modes always featured much richer enhancement of surround ambience than DTS surround enhancement modes, which is why we favored the Dolby modes for film soundtrack playback. Now, with Pro Logic IIx also becoming more sonically accurate than the DTS surround enhancement modes (ES and Neo:6), thanks to the AVR350, we would recommend Dolby Pro Logic IIx (music mode) over DTS Neo:6 for surround enhancement of music also.
       As we wrote before, one of the chief uses and joys of a surround processor with true high fidelity, like the Arcam series, is that they can bring enormous pleasure not just for film soundtrack playback, but also as the anchor in your main music system, for surround enhancement of your vast library of two channel recordings (as well as FM). Using the AVR350, we were able to re-create the surround soundfield of attending a live music concert, even from two channel recordings. The effect is so convincing with the AVR350 that you'd swear you're listening to a full surround multichannel recording, especially when the recording is an excellent one actually made at a live concert.
       Essay recordings (essaycd.com) has a Concert Collection, comprising 11 CDs of stunningly recorded live concerts (classical music, from Haydn to modern, plus movie music). These amazing Essay CDs superbly demonstrate the magic that the AVR350 can achieve (even from two channel CD) in transporting you to a surrounding concert hall space, and giving you the exciting sense of participating in a live event (with the audience buzz around you), while at the same time portraying the music on stage so realistically, with such solid palpability and three-dimensional believability, that you feel you can reach out and touch it.
       Because the AVR350's implementation of Dolby Pro Logic IIx cures the fuzzy defocus and is far more articulate, all direct sounds have much more precise localization, all around the surround circle, for both film soundtracks and music recordings. And the same is true for indirect sounds from a recording, such as reflected ambience from the walls of the concert hall or other recording venue. These indirect, subtle imaging cues are reproduced more articulately and revealed more clearly by the AVR350, without being hidden by the veiling and obscuring from the fuzzy defocus of information that occurs in the Dolby Pro Logic IIx of all other surround processors we have heard to date. So, with the AVR350, the surround space itself becomes even more believable than ever before, since you can hear the boundaries of the large hall or other venue as never before. Arcam's AV8 made a big advance in the reproduction of space itself, as we discussed in its review, and the AVR300 subsequently advanced this rare achievement even further, and now the AVR350 advances this reproduction of space itself yet further, for the most believably convincing portrayal of surround space that we have ever heard.
       We did notice one interesting side effect from the AVR350's more accurate Pro Logic IIx extraction, namely that the back surrounds seem to be lower in volume level, compared to the conventional Pro Logic IIx extraction heard from all other surround processors. There might be several reasons for this sonic effect. First, a correctly instantaneous subtraction from the main channels (and addition to the back surround channels) might actually extract less energy than the conventional extraction, which is time smeared by noise, and therefore might well extract more energy from a given fast sonic transient, over a longer period of time. Second, the ear/brain generally perceives fast transients as sounding less loud than extended, time smeared transients, so the ear/brain might well perceive a correctly instantaneous derivation of a quick transient by the AVR350 as seeming less loud in the back surrounds than the conventional time smeared transient fed to the back surrounds by other processors. Third, the instantaneous, time correct extraction by the AVR350 might produce an derived transient in the back surrounds that is better correlated with the transient remainder heard from the main loudspeakers, than the decorrelated swishy noise extracted by conventional processors, and the ear/brain generally has a harder time hearing correlated sounds, due to the Haas effect, if these correlated sounds occur within the Haas time window.
       In any case, the solution is simple. After extensive experimentation, we found that we could optimize the surround space experience from the AVR350's more perfect Dolby Pro Logic IIx by simply boosting the levels fed to the surround loudspeakers, and also extending their time delays (to get them beyond the Haas time window). The optimal AVR350 settings we arrived at, for Dolby Pro Logic IIx, are as follows. Front left and right loudspeakers, 0 ms delay and -1 level. Center loudspeaker, 7 ms delay and +3 level. Side surround loudspeakers, 29 ms delay and +2 level. Back surround loudspeakers, 37 ms delay and +10 level. These optimum parameters assume seven identical full range loudspeakers all around, equidistant from listening seat (if your setup is different, use our suggestions here merely as starting points). The Pro Logic parameters are set for no effect: dimension 0, width 0, panorama off.
       These optimum settings provide the most convincing sense of surround space, from both film soundtracks (Pro Logic IIx movie mode) and music recordings (Pro Logic IIx music mode). When the most convincing sense of surround space is achieved, the hall (or other recording venue) seems to be radiating equal indirect, reverberant field energy or pressure at you from all sides, even though the majority of the direct music or sound might be spread across a stage that is only up front. Incidentally, the special settings for the center channel are important to free the center dialogue track from the small closet it usually seems to be trapped in, due to too close miking of the voices, and liberate it so that the voices sonically integrate with the rest of the huge surround soundfield.

Output Impedance Switch

       In our AVR300 review, we had recommended setting the rear panel impedance selector switch to 8 ohms, in order to obtain the maximum voltage swing for the power amplifier output stage. However, the new performance strengths in the AVR350 provide cause for reversing this advice.
       The AVR350 is so dramatically superior in quieting background noise, and provides such important ancillary sonic advantages therefrom (as noted above), that it becomes important to optimize and maximize these new sonic virtues. We found by experimentation that the 4 ohm setting, of the rear panel impedance switch, improves quieting even further, and improves all the other sonic advantages of the AVR350 as well (even curing a slight artificial solid state glare that is evident in the 8 ohm setting).
       This is probably because the 4 ohm switch setting draws from a lower voltage secondary tap on the power transformer, which would provide a lower source impedance for your loudspeaker from the output stage (via its rails voltage), and also would provide a lower source impedance for refilling the capacitor reservoir of the power supply. A lower source impedance provides sonic benefits because it can supply energy at higher currents more quickly, hence more accurately.
       A lower source impedance is also important because the model we all use, for obtaining accurate reproduction via a universal amplifier to loudspeaker interface, assumes that the power amplifier can act as an ideal voltage source, which means having as low a source impedance as possible (at all frequencies). A higher source impedance from the power amplifier means that the signal reproduction will not be as accurate through this interface, perhaps more veiled and distorted. It's a subtle effect, not obviously audible with most audio equipment. But the AVR350 is so amazingly revealing that it does clearly reveal even this subtle effect, making the 4 ohm tap, with its lower source impedance, clearly sonically superior to the 8 ohm tap with its higher source impedance. It's a testament to the superior sonics of the AVR350 that it is so revealing, of even subtle effects like this.
       Speaking of low source impedance naturally also raises the subject of power cordsets. With the AVR350 obviously deriving some of its sonic superiority from its virtue of having a very low internal throughput source impedance (especially in the 4 ohm mode), it makes obvious sense to make sure that the external connections fulfill this potential by giving the AVR350 the lowest possible source impedance connection to the mother of al low impedance sources, the powerline. The AVR350 is so revealing of sonic information on recordings that it is naturally also very revealing of the sonic quality of all other associated links in your system, so it is intrinsically more revealing of power cordset quality than lesser processors and receivers would be, and thus more sonically rewarding of your investment in the best power cordset. In addition, the fact that the AVR350 has such an admirably low internal throughput source impedance means that it is more revealing of the difference that a low source impedance energy source (hence a low impedance connection to the powerline) makes, more revealing than another processor or receiver with poorer (higher) internal source impedance would be.
       The best sounding power cordset we have yet found, at moderate cost, for the Arcam products is the 12 gauge (!!) Wan Lung black (available in the USA from the service department at Parasound in San Francisco). It provides a low source impedance connection to the powerline, and allows the AVR350 to sing at its full glory.
       Because the AVR350 is modest in price, yet sonically outperforms far more expensive surround processors and receivers, you'll get the very best sound that money can buy, and you'll save a lot of money buying the AVR350 instead of the more expensive competition. Here's some advice. Put this money you save to good use, by investing in the finest associated components you can afford, for the rest of your system chain. The AVR350 is so extraordinary sonically that it will happily reveal the better sound of all associated system links, and will thereby reward you with even better system sound. For example, if you want to partner the AV350 with an Arcam disc player, you might be wise to invest in the premium new DV139, which is part of Arcam's top FMJ line and includes a Stealth Mat (just like the AVR350), rather than the less expensive DV137. Likewise, you'll be sonically rewarded by partnering the AVR350 with the very best, most revealing cables and loudspeakers you can manage.



(Continued on page 141)