B&W 801 Matrix Series 3 

B&W 801 Matrix Series 3

vs. Audio Artistry CBT36 
y
Ground‐Plane Measurements
Measurements and Power Point by 
D B (Don) Keele Jr
D. B. (Don) Keele, Jr.
Audio Artistry 
www.AudioArtistry.com)
(And I take full responsibility for the content, warts and all!)

(With the able assistance of Marshall Kay of Audio Artistry and Rick Craig of Selah Audio, www.SelahAudio.com)

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 1
Measurements
• General Comments:
Introduction
– This presentation compares the B&W 801 loudspeaker system with the Audio
Artistry CBT36 curved‐line array by showing frequency response
measurements of both systems
y taken over an acousticallyy reflective gground
plane (the floor). Measurements were gathered over a concrete floor in the
workshop of Audio Artistry on Oct. 8, 2011. These measurements assess the
sound field generated by the systems when operated in a non‐anechoic
environment but over a reflective ground plane.
plane
• Test Conditions:
– All frequency response curves were gathered with custom software written by
Don Keele based on Angelo Farina
Farina’ss log‐sweep
log sweep technique.
technique Impulse responses
were windowed with a 50 ms half‐Hann window and all time‐of‐arrival offsets
were compensated before converting to the frequency domain. All frequency
response curves were smoothed with a 1/12th‐octave filter.
• Effect of Carpet:
– Measurements taken with a carpet on the floor (not shown) were materially
the same except for HF rolloff near the floor. The HF loss was a primarily a
result of the sound grazing the floor and was evident at heights of 10” and less
above the floor.
B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 2
Measurements
 Introduction: Cont.
The following measurements were taken:
• Frequency Response Versus Height:
– Data was gathered at five vertical locations in front of the systems: 0.0, 0.5, 1.0, 1.5. and
2.0 m; and four distances in front of the systems: 0.1, 1.0, 2.0, and 3.0 m. Measured for
both B&W 801 and CBT36.
• Frequency Response Versus Distance:
– Data was gathered at two heights: Sitting: 1 m (40”) and Standing: 1.7 m (68”); and seven
distances in front of the systems: 0.1, 0.5, 1.0, 1.5, 2.0, 2.5 and 3.0 m. Measured for
CBT36 only.
– Data was gathered at two heights: Sitting: 1 m (40”) and Standing: 1.7 m (68”); and four
distances in front of the systems:
y 0.1, 1.0, 2.0, and 3.0 m. Measured ffor both B&W 801
and CBT36.
• Off‐axis Frequency Response:
– Data was gathered at seven angles: 0°, 15°, 30°, 45°, 60°, 75°, and 90°; and two heights: 0
( the
(on h ground d plane)
l ) and d 1 m (40”).
( ”) Data was measured d on the
h sided with
h tweeters
closest to the edge of the enclosure. Measured for CBT36 only.
• Near‐field Frequency Responses:
– Data was gathered on the CBT36 at a height of 1 m; and distances of 2
2”, 4
4”, 8
8”, and 16
16” in
front of the enclosure. Measured for CBT36 only.
B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 3
Measurements
 B&W 801 M i S i 2
B&W 801 Matrix Series 2
Specifications

Note: The B&W 801 Matrix Series 2 systems were Don Keele’s
h & 80 i S i 2 l ’
reference systems for all his Audio Magazine loudspeaker reviews 
from 1991 to 2000.

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 4
Measurements
 (Circa 1988)

Oblique View Side View Front View

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 5
Measurements
 A di A i
Audio Artistry CBT36
CBT36
Specifications

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 6
Measurements
 CBT36 Specifications:
Description: A two‐way floor standing circular‐arc high‐
end loudspeaker line array. Each system contains 18 ea 3‐1/2”
full‐range
g drivers used as mid woofers and 72 ea 3/4” diameter
wide‐band tweeters that are crossed over at 1 kHz. The 3‐1/2”
drivers utilize a Neo‐Balanced 1” underhung motor that is
capable of 20 mm peak‐to‐peak excursion and features a
shorting ring to keep distortion under control.
The very‐small 3/4” wide‐range tweeters feature a 0.5” edge‐
driven inverted aluminum dome that utilizes a high‐energy
g gy
neodymium magnet for low distortion and high efficiency. The
drivers are mounted on a 36° circular‐arc front panel which
provides an extremely‐even wide‐band narrow vertical coverage
of 28°.
CBT which stands for Constant Beamwidth Transducer is a
loudspeaker line array theory that provides extremely even
coverage and flat frequency response at all locations in the
listening room with a system that has broadband constant‐
directivity (CD).
The drivers of the CBT36 are not equally driven but are passively
”shaded” in five banks that attenuates the upper drivers with
respect to the lower drivers.
drivers This is shown in the diagram on the
right. The curved front panel and shading vastly improves the
vertical coverage of the system as compared to a straight‐line
array.
The system must be bi‐amped and requires a DSP‐based speaker
EQ/processor along with two stereo power amplifiers. For
extended bass response below 60Hz, 60Hz one or two powerful
subwoofers are required. Note that the system can’t be used
without the EQ/processor.
Dimensions: H x W x D: 61.0 x 7.0 x 25.0 inches,
(154.9 x 17.8 x 63.5 cm)
Weight:
W i ht 55 lbs (25 kg)
Price: $2,999 USD per pair DIY Kit unfinished, $8,500
USD per pair assembled and finished.
B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 7
Measurements
 CBT36 Specifications, Cont.:
p p q
Complex Impedance vs. Frequencyy
Your amplifiers and cables are going to love these impedances!
LF (18 Woofs with Shading) HF (72 Tweets with Shading)

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 8
Measurements
 On‐Axis Measurements
B&W 801
B&W 801 The curves below illustrate that the
This was published in my Nov. 1990 Audio Magazine Review CBT36’s frequency response is much
flatter than the B&W 801 when
measured over the ground plane.

All the following CBT36 measurements were taken with the

system crossed over and equalized with the DEQX HDP
Express DSP processor.

The B&W measurements were done with no processing g

except for constant voltage 2.83 Vrms input. The level of the
CBT36 was adjusted so that its level roughly equaled the
B&W at 1m away and 1m high.

B&W 801 CBT36
From Oct. 2011 Ground‐Plane Measurements From Oct. 2011 Ground‐Plane Measurements

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 9
Measurements
Measurement Photo Gallery
That crazy bushy-headed
bushy headed grey
grey-haired
haired old guy in the white “T”
T shirt is Don Keele!
The bald-headed guy in shorts, and blue “T” shirt is Rick Craig of Selah Audio.
B&W 801
B

Note: This CBT36

was loaded with
black-dome
tweeters. The
response with
silver-dome
36

tweeters is
CBT3

exactly the same

same.

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 10
Measurements
Perfect Sound Everywhere!
Perfect Sound Everywhere!
Even on
the floor!

Even this
Even way up
close! Even at 1 m away
here! Even up here!
and 1 m up!

Wow, great
imaging!

Even A perfect
Even way off directly on near-field
to the side! the side! monitor!

Even this Even listening laying on the

close! floor! Wow! Killer sound,,
knocked me to the floor!

Who’s this
Sounds
crazy looking
Great!
guy!

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 11
Measurements
 B&W 801
B&W 801 vs. CBT36
CBT36
Response vs Height
Response vs. Height

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 12
Measurements
Response vs. Height at 0.1 m (4”) in Front of System
Upp close,, the CBT36 frequency
q y
response is quite flat and well
The B&W 801 frequency behaved at all these heights!
response is quite erratic at all Highest level is close to floor
these heights at this close level. Note that response is fairly
distance! Severe HF rolloffs at all flat even at 2 meters high directly
but 1 m high (tweeter axis)
axis). in front and above the array!

B&W 801 CBT36

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 13
Measurements
 Response vs. Height at 1 m in Front of System
The B&W 801 frequency The CBT36 frequency
response is still quite response is quite flat
erratic at all these and well behaved at
heights at 1 meter away! all these heights!

B&W 801 CBT36

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 14
Measurements
 Response vs. Height at 2 m in Front of System
The B&W 801 In contrast, the
frequency response is CBT36 frequency
still quite erratic at all response is quite flat
these heights at 2 and well behaved at
meters away! all these heights at 2
meters away!

B&W 801 CBT36

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 15
Measurements
 Response vs. Height at 3 m in Front of System
At 3 meters over the ground
Even at 3 m over the plane , the CBT36 frequency
ground plane, The response is still fairly flat
B&W 801 frequency and well behaved at all
response is still quite these heights! Note how
erratic at all these close all the curves are.
are The
heights! level hardly changes from
floor level to a point above
the array!

B&W 801 CBT36

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 16
Measurements
 B&W 801 vs. CBT36
Ceiling Coverage
Based on Frequency Responses 
d
Measured at 2 m High

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 17
Measurements
 Ceiling Spectrum at +82°
Ceiling Spectrum at  82
The CBT36’s energy
spectrum aimed at the
The B&W 801 energy ceiling is relatively flat and
spectrum aimed at the the level is significantly
ceiling is very erratic less than the B&W 801!
and rough! This means less interaction
with room acoustics!

B&W 801 CBT36

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 18
Measurements
 Ceiling Spectrum at +45°
Ceiling Spectrum at  45
Rough! Smooth, flat,
and lower in
level than
B&W 801!

B&W 801 CBT36

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 19
Measurements
 Ceiling Spectrum at +27°
Ceiling Spectrum at  27
Rough! Smooth, flat,
and lower in
level than
B&W 801!

B&W 801 CBT36

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 20
Measurements
 Ceiling Spectrum at +18°
Ceiling Spectrum at  18

At every angle the

Rough! CBT36 provides a flatter
spectrum with lower
level than the B&W 801!

B&W 801 CBT36

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 21
Measurements
 CBT36
Response vs Distance
Response vs. Distance

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 22
Measurements
CBT36 Response vs. Distance at Seated Height (1 m)

7
0m 0.5m 1m 1.5m 2m 2.5m 3m
6

At seated height, 3 1m (40 in) Test Mic

the level only 2
Locations
decreases 10 dB
1

from directly in
front to 10 ft
away and stays
relatively flat!

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 23
Measurements
CBT36 Response vs. Distance at Standing Height (68”)
0m 0.5m 1m 1.5m 2m 2.5m 3m

6
Test Mic
System
y defies 5
1 6m (64 in)
1.6m Locations
inverse square 4

law!
3

No level change
from directly in
front to 10 ft
away and stays
flat!

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 24
Measurements
 B&W 801
B&W 801 vs. CBT36
CBT36
Response vs Distance
Response vs. Distance

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 25
Measurements
 Response vs. Distance at Seated Height (1 m)
B&W 801 CBT36
At seated height, the frequency response
envelope spans nearly 30 dB from directly in At seated height , the level changes only
front of the system to 10 ft away and is quite about 10 dB from directly in front of the
rough and uneven! Note that at a distance of array to 10 ft away and stays relatively flat!
0.1 m, the mic is directly in front of the tweeter
and only 4” away (top black curve)!

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 26
Measurements
 Response vs. Distance at Standing Height (68”)
B&W 801 CBT36
At standing height, the response is Hardly any level change from directly in
quite erratic with distance and barely front to 10 ft away and stays quite flat!
fits within a large 20 dB envelope! System defies inverse square law!

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 27
Measurements
 CBT36
Nearfield Response
Nearfield Response

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 28
Measurements
 CBT36 Near‐field Response vs. Distance
2, 4, 8, 16 Inches Away at 1 m High Wow, this 
is pretty 
neat!

Note uniformity of very-

close response! Is a perfect
near-field monitor!

Mic Locations

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 29
Measurements
 CBT36
Horizontal Off‐Axis Response
Horizontal Off‐Axis Response

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 30
Measurements
 CBT36 Horizontal Off‐Axis Response at
2 m on Ground Plane
2m

0°
15°
8

6
Test Mic
Location
5

30° 4

45°
2

1
2m

60°
2m

75°

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 31
Measurements
 CBT36 Horizontal Off‐Axis Response at
2 m, 1 m Above Ground Plane
2m

0°
15°
8

6 1m Test Mic
Location
5

30° 4

45°
2

Frequency 1
2m

response isi 60°

2m

fairly uniform 75°

out to ±90° off
axis!

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 32
Measurements
 B&W 801 vs. CBT36 Ground‐Plane Measurements
Observations:
• Note the extremely even frequency response over the floor as 
compared to the B&W 801 No floor bounce!
compared to the B&W 801. No floor bounce!
• At standing height, the volume and frequency response of the CBT36’s 
hardly changes over a very wide distance range of  0.1 m to 3 m (4 in 
t 10 ft)! It d
to 10 ft)! It doesn’t blast you out when you are up close!
’t bl t t h l !
• Note that in every case, for equal 1 m axial levels, the CBT36 
illuminates the ceiling with less energy than the B&W 801! In addition, 
the CBT36 energy that is directed up towards the ceiling (or the side 
walls for that matter, these measurements are soon to come) has a 
much flatter spectrum than that provided by the B&W 801. 
• Also notice the very well behaved near‐field curves. It is a perfect 
near‐field monitor!
• The horizontal off‐axis frequency response of the CBT36 is quite well 
The horizontal off axis frequency response of the CBT36 is quite well
behaved and extended out to ±90° from on axis!
B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 33
Measurements
Great Sound Everywhere!

Hey, I had fun with this slide!

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 34
Measurements
The p
Vintage Photographs
photo on the right
g was taken on Oct. 27, 2004 at the Harman Tech Meeting g at the Renaissance Parc 55 Hotel which
occurred during the San Francisco AES Convention. It shows Floyd Toole (with wine glass in hand) and Don Keele standing
on chairs listening to the first CBT ground-plane circular-arc prototype, the CBT45. We were standing on chairs to illustrate
that the sound coverage was excellent at points high and close to the array. Note grins on both faces! The prototype was a
one-way system that utilized 48 headphone transducers and was constructed of clear plastic PVC pipe. The photo on the left
was taken in 2004 of the CBT45 prototype at Harman/Becker Automotive System’s workshop in Martinsville, Indiana.

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 35
Measurements
 The End

B&W 801 vs. CBT36 Ground‐Plane 
10/31/2011, v8.1 36
Measurements