-
- This article was actually a listserve post on the ChurchSoundCheck
Discussion Group on March 12, 1999. Ray Rayburn has given
us his permission to post his message here. If you're not
familier with ChurchSoundCheck
Magazine, visit it here!
In the early days of sound reinforcement system EQ, practitioners
such as the late Dr. Boner would sweep a sound system's response
and either chart or manually write down the measured response.
They would then calculate on their slide rules the broadband
filters needed to re-shape the response to something much flatter.
They would carry a stock of inductors (often made by White)
capacitors and resistors, andsolder together one or more bands
of EQ. After this they would measure again, and see if they
needed to add more filter sections. After the broadband EQ was
to their satisfaction, they would then do the same thing with
very narrow band filters to notch out some feedback frequencies.
Overall the EQ process often took a week for a church sound
system.
Don Davis was trained by Dr. Boner in this process. After he
had used it in the field for some time, he approached an electrical
engineer at Altec (where they both worked) and asked for help
in designing a better, faster, way of doing this. Don saw that
what took all the time was calculating and hand building all
those broadband filter sections. He knew from studies of human
hearing that EQ differences of less than 1/3 octave at mid frequencies
were not heard as tonal differences. So he asked the engineer
if it would be possible to design a set of factory built filters
set at 1/3 octave intervals whose response would combine into
a close approximation of the response that would have been gotten
from those hand built broadband filters. The engineer did this
and they came up with a modular set of plug in filters, so you
could buy just those needed for a given job. The contractor
would come in with a master set of all the filters, and after
tuning the system, unplug those filters not used. They were
granted a patent on this system, marketed by Altec as Acousta-Voicing.
In real world use, Don soon found that by minor adjustments
of the EQ bands he could often eliminate the need for adding
the notch filters to get rid of feedback.
Don then realized that what now took all the time was the measurements,
since you still had to hand sweep your analyzer across the frequency
range and plot the results. If he could only look at all the
frequencies at once! Since he had EQ filters at 1/3 octave intervals,
he felt that it would be best for such a "real-time"
analyzer as well. This time he approached HP to ask if they
could design such an instrument. They said they would look into
it. Some time later they called him to say the instrument was
ready and cost $20,000! He had not realized they would interpret
his request as a firm order! Altec refused to pay for it, so
he mortgaged his house and paid for the instrument. Two years
later after seeing the improvement this made, Altec bought it
from Don. They then went to HP and worked with them to develop
an inexpensive $2,000 version.
After leaving Altec, Don Davis founded Syn-Aud-Con, the training
organization many of us have been raving about.
I gave this background to explain how we got the 1/3 octave
equalizer, and explain how it was designed to be used. It was
designed to replace broadband filters such as one can get from
a parametric EQ today. The problem with many 1/3 octave EQ's
today is that they are too narrow in frequency in each band.
Many manufacturers try to make each band act in an independent
way and minimize the interaction between bands. Original 1/3
octave EQ's had each band on a separate rotary knob. After the
slide adjustment became popular, users started to ask why the
ends of the sliders did not describe the frequency response.
Many manufacturers responded with EQ's using very narrow band
filters that gave a "what you see is what you get"
response when viewed on that 1/3 octave RTA. The problem is
that the arrow band filters do not combine, and the result is
a ripple in the frequency response between each 1/3 octave band
that you can't see on a 1/3 octave RTA since it is between each
band. No longer were these EQ's trying to simulate the response
of a broadband parametric type EQ. The narrow band ripples,
while not audible as a tonal difference, do significantly degrade
the sound quality. They make speakers far less pleasant to listen
to.
A quick test you can do on an EQ is to drive it from a sweepable
tone generator and meter the output. Set the EQ flat except
for three adjacent bands set to -3 dB each. Note the response
as you sweep the tone across the range of the bands you have
pulled down. You should NOT see three dips. You should see only
one smooth broad dip, that goes down to around -6 dB in the
center, and smoothly rises to each side. This is what the original
EQ's did, and if your EQ does not do this it may make things
sound worse instead of better.
I hope this long explanation helps in everyone's understanding
of how we got to where we are today in EQ's and why there are
sonic differences between EQ's. It should also explain why many
professionals today prefer to use parametrics now that digital
has made them less expensive than a good 1/3 octave EQ. |
