|
Accuphase Basic Engineering Policy |
|
|
BASIC ENGINEERING POLICY
We seek, through audio technology, music that's exciting and inspiring,
that moves our hearts and enriches our lives. This is the idea behind
our slogan "Enrich Life through Technology". It is the basic guiding
principle behind all Accuphase audio components whose high grade
can even be considered to be extravagant. We intend, however, to
avoid non-audible frills, and concentrate our engineering on how
best to reproduce the musical contents of program sources. If "Accuphase
tone", as such, should come to be accepted, it would a manifestation
of our interpretation of sound as it should be, and its hallmark
would be its very special degree of transparency and clarity. To
this end, we are re-educating our engineers in music, to increase
further their knowledge and appreciation of this art which we are
dedicated to re-creating.
OUR VIEWS CONCERNING SPECIFICATIONS
It is our policy concerning specifications to avoid listing meaningless
figures. Rather we stress the characteristics most important to
best music reproduction, and consider each figure very carefully
as to what it means in terms of good music. Our specification figures
are then rated very conservatively, and represent what we can guarantee
the user for continuous, long-life performance. We acknowledge that
specifications are an important factor in determining the quality
of audio components, but believe that what lies behind the figures
is even more important.
FREQUENCY RESPONSE
Some ultra wide-band proponents maintain that frequency response
from DC to 100kHz is necessary from the standpoint of transient
response. However, the truth is that no program source goes beyond
20kHz. Moreover, if a square wave (a sine wave with countless harmonics)
is recorded on a highest quality 15-inch tape recorder, only the
fundamental sine wave can be recorded beyond 7kHz. Even at 7kHz,
the fundamental wave plus the 3rd harmonic (21kHz) forming a composite
wave with a single dip at the crest is all that can be recorded.
The range for the speakers is about the same. While the upper range
for human hearing is 20kHz, most people cannot distinguish the higher
harmonics beyond 7kHz. To them a 10kHz square waveform on an oscilloscope
would be meaningless from a hearing standpoint. We believe therefore
that a frequency response from 20Hz to 20kHz is sufficient. What
is more important as far as characteristics in frequencies beyond
human hearing are concerned is phase characteristic to prevent overshoot
distortion and vibration with input pulses within the audible range.
This problem extends up to several hundred kHz, so we cannot be
content with a flat frequency response characteristic that is only
wide.
NOISE LEVEL
The ratio between signal and noise levels is referred to as S/N
ratio or S/N. We regard at least a 50dB rating as a necessary minimum,
more than 60dB for good quality and over 70dB for high quality audio
equipment. Moreover, 10dB should be subtracted from S/N ratio ratings
for FM tuners since they are normally measured at 100% modulation
while the average broadcast signal level is about 10dB less. In
the case of power amplifiers, there should be absolutely no residual
noise under "no signal" conditions. This level is measured against
the output to determine its S/N ratio, so this ratio will vary at
different powers. Minimum S/N requirements for output powers should
be 10W: 79dB, 20W: 86dB, 100W: 89dB and 150W: 90,8dB. Noise can
be expected with anything less than these ratings.
DISTORTION
Different checking standards for total harmonic distortion (THD)
in relation to loudness, frequency, etc, have resulted in different
ratings which were often meaningless from a comparison standpoint
in the past. However, re-examination of methods is being regarded
as necessary by top experts today, with its improved equipment and
more discriminating listeners. Since we regard S/N 50dB as the bare
minimum requirement against noise, we believe that -50dB (about
0,2%) should be considered the maximum standard for THD. When two
signals enter an amplifier with THD distortion, intermodulation
distortion (IMD) results and sound becomes muddy. Both THD and IMD
are caused by non-linear amplifiers and have a mutual relationship,
IMD occurring at within 1 to 6 cycles of THD, and increasing sharply
past the clipping level. The tolerance level of IMD has been regarded
as to 2 to 5% for years, but our tests show that it should be held
from down to less than 0,5% in high quality amplifiers. |
