The subtlest phase
variants characterizing the multitude of locations of individual
instruments or voices in a good stereo recording can be lost
by the slightest anomalies in the playback system. This is
why even speaker wire selection can have noticeable effects.
However,
assuming that high quality recording and playback electronics
are used, consumer (passive) speakers are inherently the
weakest link (that is, the least accurate hi-fi component)
in the audio chain. By comparison, other processes in the
audio chain, from microphones, recorders, mixers, through
distribution media, to our home electronics, are all capable
of maintaining close to original signal integrity with
good phase accuracy and low distortion.
Only when the amplified
signal arrives at the loudspeaker are we forced to accept
quite extreme corruption. For example, frequency response
errors vary through a typically 6dB range (i.e. plus/minus
3dB). Also, phase errors between drivers in a single speaker
enclosure, and also between multiple speaker enclosures,
tend to increase with higher frequencies, lowering resolution
and the experience of imaging and depth.
This is not surprising;
since in a consumer speaker's passive crossover, a single
amplifier channel output is expected to travel through
dozens of meters of fine wire coils and capacitors at high
power to inductive loudspeaker drivers, and yet reproduce
the original signal with high fidelity to the original performance.
Distortion
is caused by a passive speaker's crossover electronics
whose inductors use single strand wire that adds resistance
and impedance, interacting further as impedance varies with
frequency in loudspeaker drivers. This is one reason why
professional studio monitoring speakers tend to be active,
allowing amplifiers to be matched to each driver and avoiding
the need for any passive crossover components.
When the complex
interaction of multiple speaker drivers and their enclosures
is taken into account in addition to the problems caused
by the passive crossover, it is a tribute to decades of
perseverance and production skills that any speaker can even
get close to reproducing natural sound.
When we listen to
loudspeakers, the recognition that what we are hearing
is not real is largely subconscious; we immediately notice
that the sound arriving at our ears is unnaturally colored
due to inaccurate frequency and phase response and comb-filtering
distortion caused by time alignment errors between drivers
and even room effects. Other factors also help to give
the game away, for example, where just part of an instrument's
sound seem to come from the speakers themselves (beam)
whereas other frequencies behave omni-directionally and are
reflected in the room.
While some speaker designs approach
'being there' performance using heroic traditional analogue
technology done extremely well, the best way to fix such
a multitude of problems realistically is by using the digital
signal analysis and real-time correction FIR (Finite Impluse
Response) processing offered by DEQX Calibrated™ technology.
The
DEQX Calibrated™ PDC-2.6 processor is a stereo unit that
provides measurement and analysis features when connected
to a PC, combined with three-way, linear phase high-order
digital crossovers, and six channels of DEQX's proprietary
real-time phase and amplitude correction processing using
dual 32-bit floating-point precision digital signal processors.
Digital input to output (SPDIF or AES/EBU) introduces a
negligible minus 140dB of THD + Noise, with internal computational
dynamic range being 160dB. |
