See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/312494874

THE TERRELL-PENROSE ROTATION WHEN PHOTOGRAPHING A SPHERE AT

REST WITH A MOVING CAMERA

Article  in  Journal of Physics Conference Series · January 2017

DOI: 10.1088/1742-6596/1251/1/012035

CITATIONS READS
0 1,735

2 authors, including:

Vadim Matveev

24 PUBLICATIONS   20 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Vadim Matveev on 18 January 2017.

The user has requested enhancement of the downloaded file.

 THE TERRELL-PENROSE EFFECT WHEN PHOTOGRAPHING A
SPHERE AT REST WITH A MOVING CAMERA

V.N. Matveev, O.V. Matvejev

The paper deals with the Terrell-Penrose effect, the component parts of which are
independence of the form of the sphere, as recorded by a camera, from the speed of
motion of the sphere and the turn of the moving sphere. The present work shows that the
independence of the form of the moving sphere from the speed (the concealed nature of
the Lorentz contraction of the moving sphere) as well as its turn are optical illusions. In
examining the Terrell-Penrose effect in this paper, preference is given to a reference
frame where the sphere is at rest, whereas the camera (camera obscura) is in motion.
With this choice of the reference frame the mentioned optical illusions may be accounted
for by aberration of light reflected from the quiescent sphere, which has formed the
image on the photoreceptor of the moving camera. In the case of semitransparency of the
sphere and the presence of markers at the points of the sphere, which lie on the line of its
motion, the Lorentz contraction of the moving sphere would also be revealed by way of
photography.

Introduction

The Terrell-Penrose effect in question [1-2] relates to relativistic effects, which,

paradoxical as they may seem outwardly, attract the attention of those dealing with
special relativity. The component parts of this effect are the turn of the fast-moving
sphere and independence of its form from the speed of motion being recorded by a
photoreceptor. This behavior of the fast-moving sphere is often perceived as something
strange and contradicting the fact of the Lorentz contraction of moving bodies.
Simple as the Terrell-Penrose effect may seem at first glance, it proves to be
tangled under traditional examination. Teachers and methodologists fear that
acquaintance with this effect may cause difficulty in understanding the essence of the
Lorentz contraction. However, there is no contradiction of the effect of the integrity of
the visual form of a sphere to the Lorentz contraction, having in mind that the integrity of
the photographically recorded form of a sphere is theoretically described by Terrell and
Penrose having taken the Lorentz contraction of the moving sphere into account. The
imperfection of the description of the effect given by Terrell and Penrose lies in the fact
that it was carried out in a reference frame with the quiescent camera and the moving
sphere. This choice of a reference frame results from a tradition that examination of
relativistic effects is to be carried out in reference frames wherein observers or measuring
instruments are at rest and the objects of observation are in motion. Aberration of light
[3] is perhaps the only phenomenon, the explanation of which allows a moving observer
or a moving instrument. At the same time, as regards description of the physical side of
relativistic phenomena and effects, it is often expedient to use a reference frame wherein
the observed object is at rest and the observation facility is in motion. In the present
work, preference is given just to the reference frame wherein the sphere is at rest and the
camera (camera obscura) is in motion. Such choice of a reference frame makes the
Terrell-Penrose effect simpler and more visual.

1
 1. Photography of the rod and the sphere at rest with the camera obscura at
rest

Assume that between points A and B of X-axis on the plane XZ within the
Cartesian coordinate system K lies the flat rod (narrow thin plate) 2r long and s wide. The
ends of the rod are located symmetrically to the origin of coordinates O of the system of
coordinates K, the coordinates x of points A and B are respectively equal to -r and r.

Fig. 1. Photography of the rod AB from point F with a camera obscura at

rest. The ‘photograph’ to the right shows a geometrically undistorted image of the
rod at the center of the frame.

The camera obscura is located on the Y-axis at point F with the coordinate y = f.
The principal axis of the camera obscura (the straight line perpendicular to the front wall
of the camera obscura and passing through the center of the hole) is directed to the origin
of coordinates O (Fig. 1). The distance from the front wall of the camera to the film is
equal to h. The distance lOF between the camera and the center of the rod, equal to f, by
far exceeds the length 2r of the rod. The distance lAF from point A of the rod to point F
and the distance lBF from point B of the rod to point F are identical and equal r 2  f 2 .
The hole of the camera obscura is closed with a shutter. At a certain moment of time t the
shutter of the camera obscura is momentarily triggered, and the beams emitted by the
ends of the rod at points A and B at the moment of time t  r 2  f 2 c , where c is the
speed of light in the vacuum, pass through the hole of the camera obscura. After some
time the image of the rod is captured on the film. The photograph of the rod is
geometrically undistorted in the sense that in the photograph the length 2rk and the width

2
sk of the image of the rod are respectively proportional to the length and width of the rod
itself. Here the subscript k in notations rk and sk attests to the fact that we are talking
respectively about the length and width of the image of the rod, which is obtained on the
film in the scale k. Scale k in the case examined is equal to the ratio h/f.

Fig. 2. Photography of the rod AB from point G with the camera obscura at
rest. The ‘photograph’ to the right shows a geometrically undistorted image of the
rod shifted from the center of the frame.

Let us assume now that the principal optical axis of the camera obscura is directed
perpendicularly to the plane XZ, the film is parallel to this plane, but the photography of
the rod is performed at the moment of time t not from point F but from point G with the
coordinates x = g, y= f and z = 0 (Fig. 2). Despite the fact that the light from the rod to the
camera obscura travels in the coordinate system K at an angle of φ to the Y-axis and to the
principal optical axis, covering a greater distance than in the preceding case, the copying
scale k remains equal to h/f, and the length 2rk and the width sk of the rod in the
photograph remain the same as in the case shown in Fig. 1. This occurs because of the
equality of the ratio lO 'G lOG , where l O 'G and lOG respectively are the length of sections
OʹG and OG, to the ratio h f . One of conditions of the absence of distortion of the
image of the plane object obtained with the camera obscura is the parallelism of the plane
of the object to the plane of the film and the resulting similarity of triangles formed by
the segments of the lines connecting the points of a plane object and the corresponding
points of the image with each other and with the hole of the camera obscura. In the cases
examined such similar triangles are the triangles AFB and AʹFBʹ shown in Fig. 1 and the
triangles AGB and AʹGBʹ in Fig. 2.
In the case of the inclination of the film to the plane of the rod and to the X-axis,
the image gets distorted. The distortion of the image results from the absence of
similarity of the mentioned triangles.
By the moment of the passage of light through the hole of the camera obscura, the
light in Fig. 2 travels a different distance from points A and B. It becomes clear from the

3
right triangles AGE and BGE in Fig. 2 that from the distant point A the light passes the
distance ( g  r ) 2  f 2 , and from the near point B the light passes the distance
( g  r ) 2  f 2 . The difference Δl of the travelled distances can be represented by the
formula

l  ( g  r ) 2  f 2  ( g  r ) 2  f 2 . (1)

Multiplying and dividing the right side of the formula (1) by the conjugate
expression, we obtain

4rg
l  . (2)
(g  r)2  f 2  (g  r)2  f 2

Considering that the length of the rod is negligible in comparison with the
distance lOG from the origin of coordinates to the hole of the camera obscura and taking
values ( g  r ) 2  f 2 and ( g  r ) 2  f 2 as equal lOG, the formula (2) can be written
down in the form

2rg
l 
lOG

or, since g lOG  sin  ,

l  2r sin  . (3)

As the triangles AGB and AʹGBʹ in Fig. 2 are similar, it follows that the difference
lk of distances travelled by light from the hole of the camera obscura to the points Aʹ
and Bʹ of the image of the rod is related to the value Δl by the ratio l k  kl , which,
taking into account (3) and the equality rk  kr , is equivalent to the equality

l k  2rk sin  (4)

Using (4), the difference t k of the times of the propagation of light from the
hole of the camera obscura to the points Aʹ and Bʹ is possible to represent by the formula

2rk
t k  sin  . (5)
c

4
 Fig. 3. Photography of the rod AB and the sphere from point F with the
camera obscura at rest. The ‘photograph’ to the right shows a geometrically
undistorted image of the rod and the sphere in the center of the frame.

Let us assume now that in the coordinate system K besides the rod examined by
us there is a semi-transparent sphere with the center at the origin of coordinates O, with a
radius r, through which the rod is visible (Fig. 3).
Let the shutter of the camera obscura at the moment of time t be momentarily
triggered, and the image of the sphere and of the rod be captured on the film.
If the photography is performed from a remote point F, and the diameter dAB of
the sphere is equal to the length of the rod, then the geometrically undistorted image of
the rod as described above and a geometrically undistorted image of the sphere lying on
the plane XZ of a large circle will appear in the photograph (see the ‘photograph’ to the
right in Fig. 3). As the distance from the points of the intersection of the sphere and the
plane XZ to the hole of the camera obscura is identical and equals r 2  f 2 , the light
emitted from these points at the moment of time t  r 2  f 2 c simultaneously reaches
the hole of the camera obscura. The fact that the light beams passing through the hole of
the camera from other elements of the sphere are emitted not at the moment of time
t  r 2  f 2 c but at other moments of time does not affect the form of the image of the
sphere since ‘these other elements of the sphere’ form an internal part of the image of the
sphere. If at the point of intersection of the hemisphere facing the camera obscura and the
Y-axis there is a marker (in the figure it is the white point in the upper part of the sphere),
then in the photograph the marker will find itself in its center. If the surface of the sphere
at the points of its contact with the ends of the rod also has markers, then these markers
(white points) in the figure will coincide with the ends of the rod.

5
 But if the photographing is performed from point G (Fig. 4), then at the moment
of time t when the shutter of the camera obscura is triggered, the beams reach the hole of
the camera obscura, simultaneously emitted some time before by the points of the sphere,

Fig. 4. Photography of the rod AB and the sphere from point G with the
camera obscura at rest. The plane of the film is parallel to the plane XZ. The
‘photograph’ to the right shows the undistorted image of the rod and the stretched
image of the sphere that are shifted from the center of the frame.

which lie on the large circle formed by the surface of the sphere and by the diametrical
section located perpendicular to the line OG. Points C and H of the diameter dCH shown
in Fig. 4 also belong to these points. The time of propagation of light beams from these
points to the hole of the camera obscura equals l 2 OG  r 2 c . It is these beams that form
the image of the sphere.
The beams simultaneously emitted by points C and H, which have simultaneously
passed through the hole of the camera obscura from the indicated points because of the
inclination of the line OOʹ to the plane of the film, do not simultaneously fall on the film
at an angle of φ. This results from the absence of the similarity of the triangles CGH and
CʹGHʹ. Because of the inclination of the line OOʹ to the plane of the film, the longitudinal
diameter of the circle in the photograph turns out to be longer than the transverse
diameter 1/cos φ times, and the circle takes the form of the ellipsis elongated in the
direction of the X- axis. However, the image of the rod as in the above case (see Fig. 2)
will not be distorted because of the parallelism of the rod and the film and the similarity
of the triangles AGB and AʹGBʹ. The ‘photograph’ that conditionally appears on the right
side of Fig. 4 shows the distorted image of the circle in the form of the elongated ellipsis
and the undistorted rod found inside the ellipsis, the ends of which do not reach the points
of the ellipsis. It is clear that noncoincidence of the ellipse-shaped image of the sphere

6
with the ends of the rod does not prove the elongated dimension of the sphere itself along
the X-axis. Indeed, the size of the sphere along this axis is indicated with the markers on
the sphere (white points in the figure), which coincide with the ends of the rod.
If the hole of the camera obscura is turned to the sphere (the principal optical axis,
in this case, will pass through the origin of coordinates) or, preserving the position of the
principal optical axis, we place the film perpendicular to the line OG (Fig. 5) and make a
photograph, then the undistorted sphere will be captured in the photograph. In this case
the rod, because of its nonparallelism to the film, will prove to be distorted (shortened).

Fig. 5. Photography of the rod AB and the sphere from point G with the
camera obscura at rest. The plane of the film is tilted to the plane XZ. The
'photograph' to the right shows the distorted (contracted) image of the rod and the
undistorted image of the sphere.

It is clear that if on the surface of the sphere any details are visible, then in the
cases shown in Figs. 4 and 5 they will show that the sphere in the photograph is turned at
an angle φ. Namely, the above-mentioned marker on the hemisphere will be shifted in the
photograph from the center. The displacement of the marker cannot be considered as the
consequence of the turn of the sphere within the space of the reference frame K, in which
both the sphere and the camera obscura are at rest since this displacement is a simple
consequence of the choice of the surveying point.
Thus, taking a photo of the rod and the sphere from point G and changing the
orientation of the film, it is possible to obtain either the undistorted image of the rod with
the distortion of the sphere or the undistorted image of the sphere with the distortion of
the rod. This is caused by nonparallelism of the rod and the large cross section of the
sphere seen from point G. It is clear that such a disagreement of the distortions of the rod
and the sphere does not attest to the fact that the rod AB lying on the X-axis and the
diameter dAB of the sphere lying on the same axis behave differently when photographed.

7
The behavior of the diameter dAB is in strict agreement with the behavior of the rod; both
the rod and the diameter dAB of the sphere produce images of identical length in the
photograph. It is not the rod AB and the diameter dAB that behave differently, but it is the
rod AB and the diameter dCH, which is another diameter of a large circle of the sphere
considered here. If we, for example, compress the rod and flatten the sections of the
sphere adjacent to the X-axis in the direction of the X-axis, then in the photograph the rod
and the diameter dAB of the sphere visible through the semi-transparent sphere will prove
shortened. However, if with the compression of the sphere the sections seen from point G
are not affected by deformation, then the contours of the sphere, which are formed in the
photograph by the beams from the points of the large circle lying on the plane
perpendicular to the line OG, will not lose their form.
Further on we will show that something similar occurs also with the photography
of the sphere by the camera obscura if they are in motion relative to each other.

2. Photography of the rod and the sphere that are at rest with the moving
camera obscura

Let us pass on to the examination of the photography of the rod and the sphere at
the above-mentioned points F and G with the camera moving parallel to the X-axis and
staying at the distance f from the X-axis (Figs. 6 and 7).
Assume now that the camera obscura is moving parallel to the Х-axis at a speed v,
hitting the surveying point and at this point momentarily triggering the shutter at the
moment of time t. Let us note that the information, which is at the given point passing
through the momentarily opened aperture of the camera obscura, does not depend on
whether the camera is in motion or not. The light information practically instantly passing
through the hole of the quiescent camera shown in Fig. 3 as well as the light information
practically instantly passing through the hole of the moving camera shown in Fig. 6 are
identical. The same can also be said about the cases shown in Figs. 4 and 7. With all that,
the images of the quiescent object photographed with the cameras that are at rest and in
motion, are different. The factor causing a difference in the images on the quiescent and
moving film is the behavior of the film at the stage after the light has passed the hole of
the camera and before the incidence of the light on the film. The photography results are
primarily affected by the Lorentz contraction of the film and its displacement during the
motion of the light beams from the hole of the camera obscura to the film.
In the case shown in Fig. 6, the light from the ends of the rod and from the large
circle of the sphere lying on the plane XY at the moment of time t hits point F and enters
the hole of the moving camera obscura. Instantly the light falls on the film, forming the
image. In the coordinate system K, the light beams simultaneously reflected from the
ends of the rod, simultaneously falling on the moving film. Because of the motion of the
film the

8
 Fig. 6. Photography of the rod AB and the sphere from point F with the
moving camera obscura. The ‘photograph’ to the right shows the distorted
(stretched) image of the rod and the sphere, shifted aside.

Fig. 7. Photography of the rod AB and the sphere from point G with the camera
obscura moving at a speed v. The ‘photograph’ to the right shows the shortened
image of the rod and the undistorted sphere in the center of the frame.

9
image on it over the time of the light passage of the distance from the hole of the camera
obscura to the film shifts, taking into account the inequality 2rk  h , by the distance
vh/ c . The observer who is holding the camera can account for this displacement if he
considers the issue within the framework of the coordinate system K and if he recognizes
that he is moving together with the camera obscura relative to the rod and the sphere at a
speed v by his own motion and aberration of light the sine of the angle φ of which, as is
known [3], makes v / c .
Because of the Lorentz contraction of the film moving in the coordinate system K,
the image of the rod and of the sphere obtained after visualization will prove to be
stretched lengthwise (in the direction of motion) 1 1  v 2 c 2 times (Fig. 6, the
‘photograph’ to the right). The length λʹk of the image of the rod captured on the film (the
proper length λʹk in the coordinate system Kʹ) will prove to be increased and related to the
value 2rk by the ratio

 'k  2rk 1 v2 c2 .

The observer who recognizes his motion in the coordinate system K can also
explain this stretched length by his own motion and by the Lorentz contraction of the film
moving together with him. But if the observer moving together with the camera obscura
considers the issue within the inertial reference system Kʹ, in which he and the camera
obscura are at rest while the rod and the sphere are in motion, then he will explain the
stretched length of the sphere and of the rod in the photograph and the shift of the image
by an inclination of the beam of light coming into the hole of the camera obscura and by
non-simultaneity of the emission of light, which has simultaneously passed through the
hole of the camera at the moment of triggering the shutter, by the ends A and B of the
moving rod. Indeed, from his ‘point of view,’ the light from the moving rod and the
sphere came to him at an angle   arcsin(v / c) .
If the photography of the sphere and the rod is performed with the camera moving
in the coordinate system K not from point F but from point G (Fig. 7), then the beams
that have arrived from the ends of the rod and that at the moment of triggering the shutter
have passed through the hole of the camera obscura because of their inclination to the
plane of the film do not simultaneously (in the reference frame K) fall on this plane. If the
film were not in motion, then the length of its image would be equal 2rk because of the
mutual parallelism of the film and the rod. However, the camera obscura and the film
move at a speed v. The motion of the film has a dual effect on the image. On the one
hand, because of the Lorentz contraction, the image on the film becomes 1 1  v 2 c 2
times stretched, in the result of which the length of the image of the rod must become
equal to 2rk 1  v 2 c 2 , as in the case shown in Fig. 6. On the other hand, over the time
Δtk between the incidence of the beam coming from point A, and the incidence of the
beam coming from point B, the displacement of the film equals vt k . Considering the
formula (5), this displacement vt k of the film can be represented by the formula

10
 2rk v
vt k  sin  . (6)
c

The displacement vt k shortens the length 2rk 1  v 2 c 2 of the image of the
2rk v
rod on the film by the value sin  , and the length λʹk of the image of the rod
c 1 v2 c2
AB falling on the moving film, captured in the reference frame K (the proper length λʹk in
the coordinate system Kʹ) becomes equal to

2rk v
 'k  (1  sin ) . (7)
1 v c 2 c2

Since the angle sin  equals v/c , then (7) is transformed to

2rk
 'k  (1  v 2 c 2 )
1 v c 2 2

or

 'k  2rk 1  v 2 c 2 . (8)

The diameter dAB of the sphere coincides with the rod AB, therefore its length
dʹk,AB in the photograph is expressed by the formula

d ' k , AB  d k , AB 1  v 2 c 2 ,

analogous to formula (8).

As regards the photograph of the sphere, at the moment of triggering the shutter
of the camera obscura, the beams come to the hole of the camera obscura. These beams
are simultaneously emitted by the points of the sphere lying on a large circle formed by
the surface of the sphere and the central section located perpendicular to the line OG. If
the camera obscura were at rest, then the image of the sphere falling on the plane in
which the film is moving would acquire, as was noted above and shown in Fig. 4, the
elliptic form stretched 1/cosφ times along the X-axis. The length dʹk,CH of the image of the
section CH, which is the diameter dCH, would make 2rk/cosφ. However, the camera
obscura moves at a speed v, and over the time Δtk between the incidence of the beam
passing along the side GHʹ of the triangle GHʹCʹ and the incidence of the beam passing
along the longer side GCʹ of this triangle the film gets displaced by vt k . Considering the
lengthening of the image 1/cosα times, it is easy to understand that the displacement vt k
represented by the formula (6) in the reference frame K shortens the length 2rk cos of

11
 2rk v
the image of the ellipsis falling on the moving film by the value sin , i.e. the
c cos
length λk of the image falling on the moving film becomes equal

2rk v
k  (1   sin  ) . (9)
cos c

If sin  equals v/c, then (9) becomes

2rk
k  (1  sin 2  )
cos

and, since 1  sin 2   cos2  , becomes

k  2rk cos .

Considering the Lorentz contraction of the film, on the condition that

1 1  v 2 c 2  1 1  sin 2   1 cos , the length λʹk of the image of the ellipsis on the
film becomes equal 2rk , i.e. on the condition that sin   v / c the image of the sphere
proves to be undistorted. However, the absence of distortion relates not to the
longitudinal size dAB of the sphere but to its size expressed by the diameter dCH. The
longitudinal size of the sphere is reproduced by the size between the markers of the
sphere, which coincides with the ends of the rod, i.e. the sphere shortens in the direction
of its motion 1 1  v 2 c 2 times.

3. Photography of the rod and the sphere with the camera obscura moving
along the circle

Imagine that the camera obscura is in the center of the circle with the radius R,
and the sphere with the radius r (r << R) is moving along the circle, turning during its
motion in such a way that a certain hemisphere finds itself constantly facing the center of
the circle. If, at the moment of time t the center of the sphere moving along the circle for
a short period of time finds itself on the principal optical axis of the camera obscura, and
at the moment of time t  r 2  R 2 c the shutter of the camera is momentarily triggered,
then the image of the ‘flattened’ sphere because of the Lorentz contraction will be
captured on the film of the camera in the direction of the motion of the sphere. The
turning of the sphere will not be revealed since at any moment of time one and the same
hemisphere faces the center of the circle and the beams come only from the points of the
large cross section of the sphere, which separates its part that is visible from the center of
the circle from the invisible part.

12
 Now, imagine that the camera obscura is moving along the circle, slowly turning
so that the principal optical axis of the camera obscura is constantly directed to the center
of the circle where the sphere is located.
If at a certain moment of time the shutter of the camera is triggered and the light
beams simultaneously emitted by the points of the circle of the large cross section of the
sphere enter the hole of the camera obscura, then the light falling on the film forms the
image of the sphere ‘stretched’ 1 1  v 2 c 2 times as in the case shown in Fig. 6.

4. Photography of the point source of monochromatic radiation with the

moving camera obscura

Imagine that in the center of the sphere examined by us above there is a point
source of monochromatic green light and the photography is performed using color film.
It is clear that, in the cases of the sphere and the camera obscura, which are at rest relative
to each other, shown in Fig.1-5, the image of the point will prove to be green.
In the case shown in Fig. 6, the color of the source shifts into the dark-blue
spectral region. Within the reference frame K, this shift can be accounted for by time
dilation of the moving camera. Because of time dilation the film moving across the green
beam reproduces the color of the beam of the green source as having turned blue.
In the case shown in Fig. 7, the light from the source within the coordinate system
K falls on the film at an angle φ. This results in the fact that in the coordinate system K
the frequency of emission falling on the film disregarding time dilation makes

f  f 0 [1  (v c) sin ]

or, as in the case considered here sin   v / c ,

f  f 0 (1  v 2 c 2 ) . (10)

Because of time dilation the frequency of light perceived in the coordinate system
turns out to be 1 1  v 2 c 2 times greater than is given by the formula (10), i.e.

f  f0 1 v2 c2 .

The explanation of a change in the color of the source, given within the
framework of the coordinate system Kʹ, is different, but it yields the same result. In the
case shown in Fig. 6, the light from the source moving relative to the camera obscura
falls in the coordinate system Kʹ on the film at an angle φ, which leads to the Doppler
shift of the color of the point on the image into the dark-blue spectral region. The
frequency f of the light falling on the film at an angle φ to the normal is connected with
the frequency f0 of the source light by the formula

13
 f0 1 v2 c2
f  . (11)
1  (v c) sin

The value sin  in the case considered here equals v / c , which is why the formula (11)
takes the form
f0
f  . (12)
1 v2 c2

In the case shown in Fig. 7 the light from the source falls in the coordinate
system K on the film vertically, but the color of the coming light proves to be shifted into
the red spectral region because of time dilation of the moving source, which produces the
transverse Doppler effect.
When photographing the point source at rest in the inertial coordinate system K
with the camera obscura moving around this source, the color of the source image will, in
any event, prove to be ‘having turned blue’. Within the framework of the inertial
coordinate system K, ‘having turned blue’ is caused by the absorption of green light by
the film subjected to time dilation. Within the framework of the instantly associated
inertial system Kʹ, this ‘having turned blue’ can be accounted for by the Doppler effect,
according to which the shift of frequency of the light wave occurs in accordance with the
formula (12).
When photographing the point source circling around the camera obscura, one can
only obtain the image of the ‘reddened’ light source. Within the framework of the inertial
coordinate system K with the camera obscura at rest and the moving light source, the
effect is accounted for by the Doppler time dilation of the moving source.

References

1. R. Penrose, "The Apparent Shape of a Relativistically Moving Sphere", Proc.

Camb. Phil. Soc., vol 55 Jul 1958.
2. J. Terrell, "Invisibility of the Lorentz Contraction", Phys. Rev. vol 116 no. 4 pgs
1041–1045 (1959).
3. R.P. Feynman, R.B. Leightonn and M. Sands, “The Feynman Lectures on
Physics”, Vol.1 (Addison-Wesley, Reading, 1963).

14

View publication stats