beautypg.com

Sony G90 User Manual

Page 41

background image

again that this instrument was the first attempt
at a full-range home music center, a range
rarely captured or replicated by electronic
devices today.

All systems are biased in some way, and

the three principal biases are toward time coherence, phase
coherence, or tonal production. It is not possible to have all
three unless you also factor in the room, which acts as a filter
affecting all three. This Tukan-configured system seems to me
biased toward temporal alignment, as a result of which its
dynamic capabilities are enhanced. I would like to hear more
of the subtler textures of the tonal range, and so would you, if
you’d spent 15 years listening to Quads and BBC monitors.
For voice intelligibility, this dynamic bias has much to recom-
mend it; I was able to follow unfamiliar Sullivan (The Rose of
Persia

) with surgical precision.

Controlling the dynamic qualities of this system requires

discretion, but this is encouraged by the system’s ability to
disentangle and relax the sound sources across the full width
of the stage when so adjusted. The soundfield projected is
also surprisingly stable when heard at some distance from the
focal point of the loudspeaker array, a very refreshing change
for those who dislike being confined to the sweet spot.

Working the Environment

It seems to me that the THX concept is good for the rooms in
which we listen to films, for it mandates an observational fre-
quency response at the listening position. In other words, the
response of the system at the listening position is measured
and guaranteed to meet specifications laid down and used for
mixing guidelines on sound stages, ensuring that the mix heard
in the cinema or your room will bear some resemblance to
what the mix engineer intended! This is something that no sup-
plier of music reproduction equipment has dared to offer in the
past. This, however, merely addresses tonality. To arrive at the
best solution, we must also have control of time and phase.

These factors are routinely addressed in live, amplified

concerts, following pioneering efforts spurred by the giant
concerts of the Seventies, which necessitated time alignment
of their public-address towers to achieve a single pulse
response over the full audience area. That technology is now
used to improve the acoustic performance of sound stages
and recording studios, where at least a smooth response is a
prerequisite.

Now it’s time to take that mature technology and apply it to

the domestic environment. Our controllers already include dig-
ital-signal processing to achieve stadium effects; the next ingre-
dient is the processing algorithm that optimizes the response
from the loudspeaker-room interface. The hot contender here
is SigTech T1100, which, to date, is the only device that corrects
for non-uniformity in tonality, time, and phase (see the Revel
r e v i e w, this issue). In a perfect world, its proprietary algorithms
would become a de facto system licensed to other manufactur-
ers. Regardless of who does it first, all controller brands will be
galvanized into action by the first one to use a microphone to
optimize the signal for any room.

The Linn AV5103 controller incorporates a PC port on the

rear panel, and I assume this will be used to update the host
software. The larger point is that the system uses a separate
channel of amplification for each driver, and its dynamic
capabilities reflect that. It is implicit in such a system that
there is a crossover filter upstream of each amplifier. For

now, the crossover functions (except for the bass manage-
ment) are performed in plug-in modules that are installed in
the AV5105 amplifiers. It would be much better to move the
crossovers even further upstream, to the DSP engine where it
can be performed elegantly in the digital domain.

This system cries out for a finer control of the tonal qual-

ities of the room response. Absent DSP correction, we will
see how these qualities can be further optimized by moving
the acoustic sources within the room in the next article,
wherein we devote ourselves to extracting maximum tempo-
ral synchronization from the system.

The Remote Interface

I’ll look at two levels of the interface, the customer’s and the
installer’s. First, the customer interface is kept simple by link-
ing video and audio bus switching to the source controls, so
all the customer has to remember is to switch the system on
from standby, and then hit a source button if the system is
addressed to the wrong source.

The remote control itself is built upon a cast metal chas-

sis resembling a footprint in plan, a rather weighty slab that
tapers from three inches in width to two and a quarter inches
to provide a hand grip. The area of the control pad is divided
into four zones, one at front left for controlling the Linn 5103,
another at front right for the source transport controls
invoked by the source selector zone keys, which occupy the
center of the remote. There is also a number pad zone on the
heel of the remote. Altogether a very good piece of industrial
design modeling after the Brancusi school, though perhaps
not the best choice for those with small or arthritic hands.
One quirk the user must get used to is the slight delay built
into the response of the 5103 processor before visual feed-
back confirms receipt of the signal. Perhaps this is where
some more instant feedback might be given, even something
as simple as a repeater LED within the display area. This
would also encourage use of the system without video moni-
toring for those of such disposition. The manual makes a good
case for this delay; it enables a short-press option for control-
ling sources other than those currently selected without dis-
turbing the video and audio images you are currently follow-
ing. In other words, if you tap the relevant keys quickly, you
can cue up a CD and route it to a CDR and make a recording
without having to stop following the currently selected pro-
gram. But sans audible or visual feedback, I was more likely
to start repeatedly pressing the source key until the confirm-
ing caption appeared.

Switching on the system from standby compounds this

problem. Unless your video display is on, you know that the
system has received the command only via a rearrangement
of the typography of the Linn 5103 caption, from one line of
type to two, on the display, which is rather dimly lit to start
with. If you have been successful in switching the system on,
you can confirm this by pressing a source key, but there’s a
slight delay.

The remote worked well as a universal remote, soaking

up all the functions of the other remotes with the notable
exception of the one that operates the cable television box –
surely one of the most commonly found remote-control sys-
tems in America?

The interface itself needs a little more memory to allow

tagging of soundfield settings to sources. This happens with
AC3 decoding, as confirmed by the caption “As mix” that