Chapter 12: Properties of The Fourier Transform

12.A Introduction.

The power of the Fourier transform derives principally from the many
theorems describing the properties of the transformation operation which
provide insight into the nature of physical systems. Most of these theorems have
been derived within the context of communications engineering to answer
questions framed like "if a time signal is manipulated in such-and-such a way,
what happens to its Fourier spectrum?" As a result, a way of thinking about the
transformation operation has developed in which a Fourier transform pair
y(t) ↔ Y( f ) is like the two sides of a coin, with the original time or space signal
on one side and its frequency spectrum on the other. The two halves of a Fourier
transform pair are thus complementary views of the same signal and so it makes
sense that if some operation is performed on one half of the pair, then some
equivalent operation is necessarily performed on the other half.

Many of the concepts underlying the theorems and properties described

below were introduced in Chapter 6 in the context of Fourier series. For the most
part, these theorems can be extended into the domain of the Fourier transform
simply by examining the limit as the length of the observation interval for the
signal grows without bound. Consequently, it will be sufficient here simply to
list the results. Rigorous proofs of these theorems may be found in most
standard textbooks (e.g. Bracewell).

12.B Theorems

Linearity

Scaling a function scales it's transform pair. Adding two functions corresponds
to adding the two frequency spectra.

If h(x) ↔ H( f ) then ah(x) ↔ aH( f ) [12.1]

h(x) ↔ H( f )
If then h(x) + g(x) ↔ H ( f ) + G( f ) [12.2]
g(x) ↔ G( f )

Scaling

Multiplication of the scale of the time/space reference frame changes by the

factor s inversely scales the frequency axis of the spectrum of the function.

If h(x) ↔ H( f ) then h(x /s) ↔ s H( f ⋅ s) [12.3]

and s h(x ⋅ s) ↔ H( f /s) [12.4]

Chapter 12: Properties of The Fourier Transform Page 109

A simple, but useful, implication of this theorem is that if h(x) ↔ H( f ) then

h(−x) ↔ H(− f ) . In words, flipping the time function about the origin
corresponds to flipping its spectrum about the origin.

Notice that this theorem differs from the corresponding theorem for discrete
spectra (Fig. 6.3) in that the ordinate scales inversely with the abscissa. This is
because the Fourier transform produces a spectral density function rather than a
spectral amplitude function, and therefore is sensitive to the scale of the frequency
axis.

Time/Space Shift

Displacement in time or space induces a phase shift proportional to frequency

and to the amount of displacement. This occurs because a given displacement
represents more cycles of phase shift for a high-frequency signal than for a low-
frequency signal.

h(x) ↔ H( f ) h(x − x 0 ) ↔ e− i2 fx 0
If then H( f ) [12.5]

Frequency Shift

Displacement in frequency multiplies the time/space function by a unit phasor

which has angle proportional to time/space and to the amount of displacement.

h(x) ↔ H( f ) h(x)ei2 ↔ H( f − f0 )
xf0
If then [12.6]

Modulation

Multiplication of a time/space function by a cosine wave splits the frequency

spectrum of the function. Half of the spectrum shifts left and half shifts right.
This is simply a variant of the shift theorem which makes use of Euler's
relationship cos(x) = (e ix + e −ix )/ 2

h(x)ei2 ↔ H( f − f0 )
xf0

if h(x) ↔ H( f ) then
h(x)e −i 2 ↔ H( f + f0 )
xf0

and therefore by the linearity theorem it follows that

H( f − f0 ) + H( f + f 0 )
h(x)cos(2 xf0 ) ↔
2

H( f − f 0) + H( f + f0 )
h(x)sin(2 xf0 ) ↔ [12.7]
2i

The modulation theorem is the basis of transmission of amplitude-modulated

radio broadcasts. When a low frequency audio signal is multiplied by a radio-
frequency carrier wave, the spectrum of the audio message is shifted to the radio
portion of the electromagnetic spectrum for transmission by an antenna.
Chapter 12: Properties of The Fourier Transform Page 110

Similarly, a method for recovery of the audio signal called super-heterodyne

demodulation involves multiplying the received signal by a sinusoid with the
same frequency as the carrier, thereby demodulating the audio component.

Differentiation

Differentiation of a function induces a 90° phase shift in the spectrum and scales
the magnitude of the spectrum in proportion to frequency. Repeated
differentiation leads to the general result:

d n h( x)
If h(x) ↔ H( f ) then n ↔ (i2 f )n H( f ) [12.8]
dx

This theorem explains why differentiation of a signal has the reputation for being
a noisy operation. Even if the signal is band-limited, noise will introduce high
frequency signals which are greatly amplified by differentiation.

Integration

Integration of a function induces a -90° phase shift in the spectrum and scales the
magnitude of the spectrum inversely with frequency.
x
If h(x) ↔ H( f ) then ∫ −∞
h(u) du ↔ H( f )/(i2 f ) + constant [12.9]

From this theorem we see that integration is analagous to a low-pass filter which
blurs the signal.

Transform of a transform

We normally think of using the inverse Fourier transform to move from the
frequency spectrum back to the time/space function. However, if instead the
spectrum is subjected to the forward Fourier transform, the result is a time/space
function which has been flipped about the y-axis. This gives some appreciation
for why the kernels of the two transforms are complex conjugates of each other:
the change in sign in the reverse transform flips the function about the y-axis a
second time so that the result matches the original function.

h(t) 
 → H( f ) H( f ) 
F → h(−t)
F
If then [12.10]

One practical implication of this theorem is a 2-for-1 bonus: every transform pair
brings with it a second transform pair at no extra cost.

If h(t) ↔ H( f) then H(t) ↔ h(− f ) [12.11]

For example, rect(t) ↔ sinc( f ) implies sinc(t) ↔ rect(− f ).

This theorem highlights the fact that the Fourier transform operation is
fundamentally a mathematical relation that can be completely divorced from the
physical notions of time and frequency. It is simply a method for transforming a
Chapter 12: Properties of The Fourier Transform Page 111

function of one variable into a function of another variable. So, for example, in
probability theory the Fourier transform is used to convert a probability density
function into a moment-generating function, neither of which bear the slightest
resemblance to the time or frequency domains.

Central ordinate

By analogy with the mean Fourier coefficient a0 , the central ordinate value H(0)
(analog of a0 in discrete spectra) represents the total area under the function h(x).
∞ ∞

∫−∞ h(u) e du = ∫
−i 0
If h(x) ↔ H( f ) then H(0) = h(u) du
−∞

∞
h(0) = ∫−∞ H(u) ei 0du
∞
For the inverse transform, = ∫−∞ H(u) du [12.12]
∞ ∞
= ∫−∞ Re[H(u)] du + i ∫−∞ Im[ H(u)] du
Note that for a real-valued function h(t) the imaginary portion of the spectru will
have odd symmetry, so the area under the real part of the spectrum is all that
needs to be computed to find h(0).

For example, in optics the line-spread function (LSF) and the optical transfer
function (OTF) are Fourier transform pairs. Therefore, according to the central-
ordinat theorem, the central point of the LSF is equal to the area under the OTF.
In two dimensions, the transform relationship exists between the point-spread
function (PSF) and the OTF. In such 2D cases, the integral must be taken over an
area, in which case the result is interpreted as the volume under the 2D surface.

Equivalent width

A corollary the central ordinate theorm is

∞
∫
∞
h(u) du = H(0)
If
−∞
then
∫ −∞
h(u) du
=
H (0)
[12.13]
∞

∫−∞ H(u) du
∞
∫
h(0) = H (u) du h(0)
−∞

The ratio on the left side of this last expression is called the "equivalent width" of
the given function h because it represents the width of a rectangle with the same
central ordinate and the same area as h. Likewise, the ratio on the right is the
inverse of the equivalent width of H. Thus we conclude that the equivalent
width of a function in one domain is the inverse of the equivalent width in the
other domain as illustrated in Fig. 12.1. For example, as a pulse in the time
domain gets shorter, its frequency spectrum gets longer. This theorem quantifies
that relationship for one particular measure of width.
Chapter 12: Properties of The Fourier Transform Page 112

Fig. 12.1 Equivalent Width Theorem

Space/time Domain Frequency Domain

w 1
w

Convolution

The convolution operation (denoted by an asterisk) is a way of combining two

functions to produce a new function. By definition,
∞
p = h∗g means p(x) = ∫−∞ g(u)h(x − u) du [12.14]

Convolution will be described in detail in section 12C. Here it is sufficient to

state the convolution theorem:

h(x) ↔ H( f ) h(x) ∗ g(x) ↔ H( f )⋅ G( f )

If then [12.15]
g(x) ↔ G( f ) h(x) ⋅ g(x) ↔ H( f ) ∗ G( f )

In words, this theorem says that if two functions are multiplied in one domain,
then their Fourier transforms are convolved in the other domain. Unlike the
cross-correlation operation described next, convolution obeys the commutiative,
associative, and distributive laws of algebra. That is,

commutative law h∗g = g∗h

associative law f ∗ (g ∗ h) = ( f ∗ g) ∗h [12.16]
distributive law f ∗ ( g + h) = f ∗ g + f ∗ h

Derivative of a convolution

Combining the derivative theorem with the convolution theorm leads to the
conclusion

dh df dg
If h(x) = f (x) ∗ g(x) then = ∗g = f ∗ [12.17]
dx dx dx

In words, this theorem states that the derivative of a convolution is equal to the
convolution of either of the functions with the derivative of the other.

Cross-correlation

The cross-correlation operation (denoted by a pentagram) is a way of combining

two functions to produce a new function that is similar to convolution. By
definition,
Chapter 12: Properties of The Fourier Transform Page 113

∞
q = h★g means q(x) = ∫−∞ g(u − x)h(u) du [12.18]

The cross-correlation theorem is

h(x)★g(x) ↔ H( f )⋅ G(− f )
h(x) ↔ H( f )
If then h( −x) ⋅ g(x) ↔ H( f )★G( f ) [12.19]
g(x) ↔ G( f )
h(x) ⋅ g(x) ↔ H(− f )★G( f )

Combining eqns. [12.14] and [12.16] indicates the spectrum of the product of two
functions can be computed two ways, h ⋅ g ↔ H( f ) ∗G( f ) and
h ⋅ g ↔ H(− f )★G( f ). Since the spectrum of a function is unique, the implication
is that

H( f ) ∗G( f ) = H(− f )★G( f ) [12.20]

which shows the relationship between correlation and convolution.

Auto-correlation

The auto-correlation theorem is the special case of the cross-correlation theorem

when the two functions h and g are the same function. In this case, we use the
Hermitian symmetry of the Fourier transform to show that:

If h(x) ↔ H( f ) then h(x)★h(x) ↔ H( f )⋅ H ∗ ( f ) [12.21]

The quantity h★h is known as the autocorrelation function of h and the quantity
HH* is called the power spectral density function of h. This theorem says that the
autocorrelation function and the power spectral density function comprise a
Fourier transform pair.

Parseval/Rayleigh

Parseval's energy conservation theorem developed in the context of Fourier

series is often called Rayleigh's theorem in the context of Fourier transforms .
∞ ∞
h(x) ↔ H( f ) ∫ h(x) dx = ∫
2 2
If then H( f ) df [12.22]
−∞ −∞

The left hand integral is interpreted as the total amount of energy in the signal as
computed in the time domain, whereas the right hand integral is the total
amount of energy computed in the frequency domain. The modulus symbols
(|~|) serve as a reminder that the integrands are in general complex valued, in
which case it is the magnitude of these complex quantities which is being
integrated.

A more general formulation of Parseval's theorem is as follows:

Chapter 12: Properties of The Fourier Transform Page 114

h(x) ↔ H( f ) ∞ ∞

∫−∞ h(x)g (x) dx = ∫−∞ H ( f )G ( f ) df

∗ ∗
If then [12.23]
g(x) ↔ G( f )

which is the analog of eqn. [7.6] developed for discrete spectra.

In many physical interpretations, the product of functions h and g correspond to

instantaneous or local power (e.g., current times voltage, or force times velocity)
and so the theorem says that the total power can be obtained either by
integrating over the space/time domain or over the frequency domain.

12.C The convolution operation

Convolution is the mathematical operation which describes many physical

processes in which the response or output of a system is the result of
superposition of many individual responses. Consider, for example, the
response of a "low-pass" electronic filter to a unit pulse of voltage as illustrated in
Fig. 12.2. Such a response is called the "impulse" response of the filter. Low-pass
filters have the characteristic of "memory" for the input signal so that, although
the input exists only briefly, the output continues for a much longer period of
time. Consequently, if several impulses arrive at the filter at different times, then
a series of output waveforms will be generated by the filter. If the shape of these
output waveforms is exactly the same, regardless of when the input impuses
arrived, and are scaled in amplitude in proportion to the strength of the input
pulse, then the filter is said to be time invariant. Now suppose that the input
pulses arrive in rapid succession such that the the output waveforms will begin
to overlap. If the actual output waveform is the linear sum of all of the
individual impulse responses, then the filter is said to be a linear filter.

Fig. 12.2 Impulse Response of Linear Filter

Input pulse Impulse Response
h(t)
a
Linear b
c
Filter
1 2 3
Time

Fig. 12.3 illustrates a specific example of the superpostion of three responses

to three input pulses which arrive at times t=0, t=1, and t=2. The amplitudes of
these three input pulses are 4, 5, and 6 units, respectively. If we wish to calculate
the response at some specific time t0 , then there are three ways to go about the
computation. The most direct way is to draw the three impulse response
waveforms appropriately scaled vertically and displaced along the time axis and
then add up the ordinate values at time t0 . That is,

r(t 0 ) = 4 ⋅ h(t 0 − 0) + 5⋅ h(t0 − 1)+ 6 ⋅ h(t0 − 2) [12.24]

Chapter 12: Properties of The Fourier Transform Page 115

Fig. 12.3 Superposition of Impulse Responses

Input pulses Response
6 h(t)
5 6a
r(t=3)=6a+5b+4c h(t-1) h(t-2)
4
5b
Linear h(t-0)
4c
Filter
0 1 2 Time 0 1 2 Time

The other two methods for getting the same answer correspond to the idea of
convolution. In Fig. 12.4 the unit impulse response is drawn without
displacement along the x-axis. In the same figure we also draw the input pulses,
but notice that they are drawn in reverse sequence. We then shift the input train
of impulses along the x-axis by the amount t0 , which in this example is 3 units,
and overlay the result on top of the impulse response. Now the arithmetic is as
follows: using the x-location of each impulse in turn, locate the corresponding
point on the unit impulse response function, and scale the ordinate value of h(t)
by the height of the impulse. Repeat for each impuse in the input sequence and
add the results. The result will be exactly the same as given above in eqn. [12.24].

Fig. 12.4 Convolution Method #1

Reversed Input pulses r(t=3)=6h(1)+5h(2)+4h(3)
6 =6a+5b+4c
5 6
4 5
a 4
b
h(t) c
-2 -1 0 0 1 2 3
Time

An equivalent method for computing the strength of the response at the

particular instant in time, t=3, is illustrated in Fig. 12.5. This time it is the
impulse response which is reversed in time. For clarity we give this reversed
function a different name, h´, and let the dummy variable u stand for the
reversed time axis. This new function h´(u) is then translated to the right by 3
units of time and overlaid upon the input function plotted without reversal.

Fig. 12.5 Convolution Method #2

Reversed impulse r(t=3)=6h(1)+5h(2)+4h(3)
response 6
5 =6a+5b+4c
h'(u)=h(-t) 4 a
b
c h'(u-3)=h(3-t)
-3 -2 -1 0 0 1 2 3
u u
Chapter 12: Properties of The Fourier Transform Page 116

The arithmetic for evaluating the response in Fig. 12.5 is the same as in Fig.
12.4: multiply each ordinate value of the impulse response function by the
amplitude of the corresponding impulse and add the results. In fact, this is
nothing more than an inner product. To see this, we write the sequence of input
pulses as a stimulus vector s=(s0 ,s1 ,s2 ) = (4,5,6) and the strength of the impulse
response at the same points in time could be written as the vector h=(h 0 , h 1 , h 2 )=
(a,b,c). The operation of reversing the impulse response to plot it along the u-axis
would change the impulse response vector to h´=(h 2 , h1 , h0 )=(c,b,a). Accordingly,
the method described above for computing the response at time t0 is
3
r(t 0 ) = ∑ sk ⋅ hk′
k= 1 [12.25]
= s • h′

Although this result was illustrated by the particular example of t0 = 3, the same
method obviously applies for any point in time and so the subscript notation
may be dropped at this point without loss of meaning.

If we now generalize the above ideas so that the input signal is a continuous
function s(t), then the inner product of vectors in eqn. [12.25] becomes the inner
product between continuous functions.
∞
r(t) = ∫
−∞
s(u)h ′(u − t) du
∞
= ∫
−∞
s(u)h(t −u) du [12.26]
= s(t)∗ h(t )

Notice that the abscissa variable in Fig. 12.5 becomes a dummy variable of
integration u in eqn. 12.26 and so we recognize the result as the convolution of
the stimulus and impulse response. Therefore, we conclude that convolution
yields the superposition of responses to a collection of point stimuli. This is a major
result because any stimulus can be considered a collection of point stimuli.

If the development of this result had centered on Fig. 12.4 instead of Fig. 12.5,
the last equation would have been:
∞
r(t) = ∫
−∞
h(u) s′(u − t0 ) du
∞
= ∫
−∞
h(u)s(t −u) du [12.27]
= h(t) ∗ s(t)

Since we observed that the same result is achieved regardless of whether it is the
stimulus or the impulse response that is reversed and shifted, this demonstrates
that the order of the functions is immaterial for convolution. That is, s*h = h *s.
which is the commutative law stated earlier.
Chapter 12: Properties of The Fourier Transform Page 117

In summary, for a linear, shift-invariant system, the response to an arbitrary

input is equal to the convolution of the input with the impulse response of the
system. This result is the foundation of engineering analysis of linear systems. Because
the impulse response of a linear system can be used (by way of the convolution
theorem) to predict the response to any input, it is a complete description of the
system's characteristics. An equivalent description is the Fourier transform of the
impulse response, which is called the transfer function of the system. According
to the convolution theorem, the prescribed convolution of input with impulse
response is equivalent to multiplication of the input spectrum with the transfer
function of the system. Since multiplication is an easier operation to perform
than is convolution, much analysis may be done in the frequency domain and
only the final result transformed back into the time/space domain for
interpretation.

12.D Delta functions

Although the transition from Fourier series to the Fourier transform is a major
advance, it is also a retreat since not all functions are eligible for Fourier analysis.
In particular, the sinusoidal functions which were the very basis of Fourier series
are excluded by the preceding development of the Fourier transform operation.
This is because one condition for the existence of the Fourier transform for any
particular function is that the function be "absolutely integrable", that is, the
integral of the absolute value over the range -∞ to +∞ must be finite, and a true
sinusoid lasting for all time does not satisfy this requirement. The same is true
for constant signals. On the other hand, any physical signal that an
experimentalist encounters will have started at some definite time and will
inevitably finish at some time. Thus, empirical signals will always have Fourier
transforms, but our mathematical models of these signals may not. Since the
function sin(x) is a very important element of mathematical models, we must
show some ingenuity and find a way to bring them into the domain of Fourier
analysis. That is the purpose of delta functions.

Recall that the transition from Fourier series to transforms was accompanied
by a change in viewpoint: the spectrum is now a display of amplitude density.
As a result, the emphasis shifted from the ordinate values of a spectrum to the
area under the spectrum within some bandwidth. This is why a pure sinusoid
has a perfectly good Fourier series representation, but fails to make the transition
to a Fourier transform: we would need to divide by the bandwidth of the signal,
which is zero for a pure sinusoid. On the other hand, if what is important
anyway is the area under the transform curve, which corresponds to the total
amplitude in the signal, then we have a useful work-around. The idea is to
invent a function which looks like a narrow pulse (so the bandwidth is small)
that is zero almost everywhere but which has unit area. Obviously as the width
approaches zero, the height of this pulse will have to become infinitely large to
maintain its area. However, we should not let this little conundrum worry us
since the only time we will be using this new function is inside an integral, in
which case only the area of the pulse is relevant. This new function is called a
unit delta function and it is defined by the two conditions:
Chapter 12: Properties of The Fourier Transform Page 118

(t) = 0 for t ≠ 0
∞ [12.28]
∫ −∞
(t) du = 1

To indicate a pulse at any other time, a, we write (x-a).

One important consequence of this definition is that the integral of an

arbitrary function times a delta function equals one point on the original
function. This is called the sifting property of delta functions and occurs because
the delta function is zero everywhere except when the argument is zero. That is,

∞ a− a+ ∞
∫−∞
g(u) (u − a) du = ∫−∞
g(u) (u − a) du + g(a)∫
a −
(u − a) du + ∫ + g(u) (u − a) du
a [12.29]
= 0 + g(a) + 0 = g(a)

Applying this result to the convolution integral we see that convolution of any
function with a delta function located at x=a reproduces the function at x=a
∞
g(t)∗ (t − a) = ∫ −∞
g(u) (t − a − u) du
+
[12.30]
= g(t − a)∫ (t − a − u) du = g(t − a)
−

Consider now the Fourier transform of a delta function. By the sifting

property of the delta function, if y(x) = (x) then
∞
Y( f ) = ∫ (x)e − i2 xf
dx
−∞ [12.31]
=e −i 2 0 f
=1

In other words, a delta function at the origin has a flat Fourier spectrum, which
means that all frequencies are present to an equal degree. Likewise, the inverse
Fourier transform of a unit delta function at the origin in the frequency domain is
a constant (d.c.) value. These results are shown pictorially in Fig. 12.6 with the
delta function represented by a spike or arrow along with a number indicating
the area under the spike. Thus we see that the Fourier transform of a constant is a
delta function at the origin, 1 ↔ (0) . Applying the modulation theorem to this
result we may infer that the spectrum of a cosine or sine wave is a pair of delta
functions

( f − f0 ) + ( f + f0 ) ( f − f0 ) − ( f + f0 )
cos(2 f0 x) ↔ sin(2 f0 x) ↔ [12.32]
2 2i

and the spectrum of a pair of delta functions is a cosine or sine wave

(x − x0 ) + (x + x 0 ) (x + x0 ) − (x − x 0 )
↔ cos(2 fx0 ) ↔ sin(2 fx0 ) [12.33]
2 2i
Chapter 12: Properties of The Fourier Transform Page 119

Fig. 12.6 Fourier Transform of Delta Function

Space/time Domain Frequency Domain
1

0 0

0 0

Fig. 12.7 Fourier Transform of Coisine Function

Space/time Domain Frequency Domain

0 0

0 0

12.E Complex conjugate relations

If the complex conjugate is taken of a function, its spectrum is reflected about the
origin. This statement, and related results, are summarized in Table 12.1

Table 12.1 Complex conjugate relations of the Fourier transform

y(x) Y(f)

h* ( x) H* (− f )

h* (−x) H* ( f )

h(−x) H(− f)

2Re{h(x) } H( f ) + H * (− f )

2Im{h(x) } H( f ) − H * (− f )

h(x) + h* (−x) 2Re {H( f )}

h(x) − h* (−x) 2Im {H( f )}

Chapter 12: Properties of The Fourier Transform Page 120

12.F Symmetry relations

We saw in Ch. 5 that any function y(x) can be written as the sum of even and odd
components, y(x) = E(x) + O(x) , with E(x) and O(x) in general being complex-
valued. Applying this fact to the definition of the Fourier transform yields
∞ ∞
Y( f ) = ∫−∞
y(x) cos(2 xf ) df − i ∫ y(x) sin(2 xf ) df
−∞
∞ ∞
[12.34]
= ∫ E(x) cos(2 xf ) df −i ∫ O(x) sin(2 xf ) df
−∞ −∞

from which we may deduce the symmetry relations of Table 12.2 between the
function y(x) in the space/time domain and its Fourier spectrum, Y(f). A
graphical illustration of these relations may be found in Ch. 2 of Bracewell (1978).

Table 12.2 Symmetry relations of the Fourier transform

y(x) Y(f)

real and even real and even

real and odd imaginary and odd

imaginary and even imaginary and even

imaginary and odd real and odd

complex and even complex and even

complex and odd complex and odd

real and asymmetrical complex and hermitian

imaginary and asymmetrical complex and antihermitian

real even, imaginary odd real

real odd, imaginary even imaginary

even even

odd odd
Chapter 12: Properties of The Fourier Transform Page 121

12.G Variations on the convolution theorem

The convolution theorem h ∗ g ↔ H ⋅ G of eqn. [12.15] and correlation theorem

h★g ↔ H ⋅ G(−) of eqn. [12.19] take on different forms when we allow for
complex conjugation and sign reversal of the variables (see Table 12.3).

Table 12.3 Convolution and correlation relations of the Fourier transform

y(x) Y(f)

h∗g H⋅G

h ∗ g(−) H ⋅ G(−)

h(−)∗ g(−) H(−) ⋅ G(−)

h ∗ g* (−) H ⋅ G*

h ∗ g* H ⋅ G* (−)

h(−)∗ g * (−) H(−) ⋅ G*

h(−)∗ g * H(−) ⋅ G* (−)

h* (−) ∗ g* (−) H* ⋅ G*

h* (−) ∗ g* H* ⋅ G* (−)

h* ∗ g* H* (−) ⋅G * (−)

h∗h H2

h(−)∗ h(−) [H(−)]2

h* (−) ∗ h*(−) [H * ]2

h* ∗ h* [H * (−)]2

h★h H ⋅ H(−)

h(−)★h(−) H ⋅ H(−)

h* (−)★h* (−) H* ⋅ H* (−)

h*★h* H* ⋅ H* (−)