Ricker, Ormsby; Klander, Bntterwo
A Choice of avelets
by Harold Ryan, Hi-Res Geoconsulting
Commercially available software used
to generate synthetic seismograms gives the
geophysicist a choice of up to four types of
standardized wavelets as well as the option
of a user-defined wavelet. This review arti-
cle will summarize the characteristics of
these wavelets, illustrate the shape of the
wavelets and their frequency spectrum and
present the parameters needed to define
them as well as their mathematical formu-
las.(In· all the following formulas, "t" is
time in seconds, "f" is frequency in hertz
and "1t" is the irrational number pi).
Illustrations of the wavelets and their fre-
quency spectrum were generated using the
scientific numeric computation software
MATLAB and the wavelets were chosen to
have as similar a bandwidth as possible for
ease of comparison.
Ricker wavelets (fig 1) are zero-phase
wavelets with a central peak and two
smaller side lobes. A Ricker wavelet can be
uniquely specified with only a single para-
meter,"f", it's peak frequency as seen on
the wavelet's frequency spectrum (fig 2).
The mathematical formula for a Ricker
wavelet is given by:
In some texts you will see the Ricker
wavelet's breadth, that is the time interval
between the centre of each of the two side
lobes, quoted as the reciprocal of the Ricker
wavelet's peak frequency. This is only an
approximation and indeed will give a
wavelet breadth 28% wider than the true
wavelet breadth. The correct formula for
the breadth of a Ricker wavelet is:
Ricker wavelet breadth = 'i{/1t/f
= 0.7797/f
Because of the simple inverse relation-
ship between the peak frequency and
breadth of a Ricker wavelet, the same
Ricker wavelet could be just as uniquely
described as a "31 ms Ricker wavelet" or as
a "25 Hz Ricker wavelet".
Ormsby wavelets are also zero-phase
wavelets although Ormsby, an aerospace
engineer, actually defined a filter so such
wavelets should really be called Ormsby-
filtered wavelets. The trapezoidal shape of
an Ormsby filter can be seen in the fre-
quency spectrum shown in figure 4.
Applying this filter to a unit impulse func-
8 CSEG Recorder September, 1994
tion results in the creation of the Ormsby
wavelet shown in figure 3. An Ormsby
wavelet will have numerous side lobes
unlike the simpler Ricker wavelet which
will always only have two side lobes. The
Ormsby wavelet will also become more
leggy the steeper the slope of the sides of
the trapezoidal filter. Four frequencies are
needed to specify the shape of an Ormsby
filter and which are also used to identify an
Ormsby wavelet (i.e. the 5-10-40-45 Hz
Ormsby wavelet of fig. 3). These frequen-
cies are "fl", the low-cut frequency; "f2",
the low-pass frequency; "f3", the high-pass
frequency and" f4 ", the high-cut frequency
which are all used in the following formula
to generate an Ormsby wavelet.
Ormsby(t) =
(1tf4? sinc 2(1tf4t) - (1tf3? sinc 2(1tf3t)
(1tf4-1tf3) (1tf4-1tf3)
(1th)2 sinc 2(1tf2t) - (1tfI? sinc 2(1tf1t)
-
(1tf2-1tfl) (1th-1tfl)
Like Ricker and Ormsby wavelets, a
Klauder wavelet (fig 5) is symmetrical
about a vertical line through its central peak
at time zero. A Klauder wavelet represents
the autocorrelation of a linearly swept fre-
quency-modulated sinusoidal signal used in
Vibroseis. It is defined by its terminal low
frequency, "fl"; its terminal high frequency,
"fl"; and the duration of the input signal
"T", often 6 or 7 seconds. The real part of
the following formula will generate a
Klauder wavelet.
Klauder(t) = wavelet and its frequency spectrum.
real [sin(1tkt(T-t))/(1tkt) exp (21tifot)]
where k = (f2 - fl)/T (rate of change of
frequency with time)
fo = (f2 + fl)/2 (midfrequency of
bandwidth)
i = ~ 1 (an imaginary number)
The frequency spectrum of a Klauder
wavelet (fig 6) clearly shows the substantial
similarities between a Klauder and an
Ormsby wavelet.
The most distinctive characteristics of
Butterworth wavelets is that they are mini-
mum-phase and physically realisable. A
Butterworth wavelet will start at time zero
while Ricker, Ormsby and Klauder
wavelets all have their peaks at time zero.
Butterworth defined a minimum-phase
-
filter with maximal flatness in the passband
so that applying a Butterworth filter to a
unit impulse function will generate a
wavelet such as in figure 7. Four parame-
ters are needed to specify a Butterworth fil-
ter, beginning with "fl" and "h", the lower
and upper frequencies of the bandpass.
Then the lower and upper cutoff rates, that
is the slope of the amplitude response out-
side the bandpass, are needed. They are
expressed as either "n", the order of the
Butterworth filter (a positive integer) or in
terms of "decibels/octave" which is "6n"
(i.e. the cutoff rate can be described as
either "third order" or as "18
decibels/octave"). As the cutoff rate of a
Butterworth filter increases, the wavelet
itself becomes more leggy. In an article by
Curtis(l975) there is an appendix entitled
"Generation of a Wavelet from Butterworth
Filter Spectral Characteristics" which pro-
vides a thorough mathematical treatment of
Butterworth filters. MATLAB has the
advantage of containing a function for
Butterworth filter design in its signal pro-
cessing toolbox so that Butterworth
wavelets and their frequency spectrum (fig.
8) can be generated quite easily.
Every time a synthetic seismogram is
generated, the geophysicist must first
choose the type of wavelet that will be used
and then the specific parameters needed to
define the wavelet. Making these choices
becomes much easier with a thorough
understanding of the differences and simi-
larities between Ricker, Ormsby, Klauder
and Butterworth wavelets and how varying
their parameters changes the shape of the
Bibliography
Curtis, J.W. 1975. A multi-trace syn-
thetic seismogram generator. Bulletin of the
Australian Society of Exploration
Geophysicists, v.6(4), p. 91-99.
Geyer, R.L. 1969. The vibroseis sys-
tem of seismic mapping. Journal of the
Canadian Society of Exploration
Geophysicists, v.6(l), p. 39-57.
Hosken, J.W. 1988. Ricker wavelets in
their various guises. First Break, v.6(l), p.
24-33.
Sengbush, R.L. 1983. Analog and digi-
tal filters. In: Seismic Exploration Methods.
IHRDC, Boston, p. 244-251.
 , Ft;;. 1 : 25 HZ R'lO<tR WAvo..cT
1
nc. 2: 25 HZ RIO<!:R WAVQ.[T

0.8
0.9
0.5
0.6
0.7
0.4
o 0.6
~
i -"
6
0.2
] 0.5
0.4
o I
0.3
-0.2
0.2
-<J.4

-0.6
-200 -'50 -'00 -50 o 50 •.00
'50 200
O. ,

0
0
J
20 '0
\ 60 80 100 '20 "0
lime(ms) fr eClJOncy(tu)

, rlG. .3: S- 10-40-45 HZ ORMSSY' WAYnET , FIC. 4: 5-'0-40-45 HZ 0RII.S8Y WAVELET

O. 9
0.5
0.8
0.6
0.7
0.' 0.6

-
~
'!.
0

0.2
-]
>-
0.5
6 0
0.4
0.3
-0.2
0.2
-<J.' 0.1

-0.6 0
-200 -'50 -'00 -50 o 50 150 200 0 20 '0 60 80 100 '20 "0
lime(ms) freQ.Jency(IU)

, FIC. 5: '0-40 HZ <LAUDER WAVELET (T • 7 S~CCl'OS) , FIC. 6: 10-40 HZ KLAUOER WAVELET (T • 75£CO"CS)

0.9
0.5
0.8
0.6
0.7
0.' 0.6
,• -]"
-
'0

0.2 0.5
~
6 0
0.4

0.3
-0.2
0.2
-0.4 O. ,

-0.6 0
-200 -'50 -100 -50 50 '00 150 200 0 20 40 60 80 100 120 140
freQoency(tu)

Ff 7: '0-40 HZ BUTTERWORTH WAVELET(ClJTOFF RATE. '2/'2 DB/OCT) FIC. 5: '0-40 HZ BUTTERWORTH WAVELET(CUTOFF RATE - '2/'2 OS/OCT)

0.5 0.: "'"

0.5
0.6
0.7
0.'
0.6
-]"
0

-6
~
'!. 0.2 0.5

0 0.4

0.3
-0.2
0.2
-<J.4
o.,
-0.6 0
0 50 100 '50 200 250 300 350 400 0 20 40 60 80 140
'00 '20
lime(ms) freQJency(h')

September, 1994 CSEG Recorder 9