Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~0'~42~4 RCA 69,895
1 The present invention relates to apparatus for
gamma correction of signals generated by a television film
camera, and in particular to negative gamma correction of
negative film used with a telecine camera.
In a television system, it is necessary to
process the video signals before transmission to compensate
for certain nonlinearities in the respective transmission
and receiving systems to ensure that the viewer sees a
picture which is a true reproduction of the televised scene.
Among the nonlinearities of the system for which compensation
must be made are the gamma characteristics of the television
receiver picture tube and the television camera pickup tube.
Gamma is defined as a numerical indication of the
degree of contrast in a television image. Kinescopes used
lS in television receivers generally have a nonlinear
characteristic such that the black portions of a video
signal are compressed and the white portions of a video
signal are stretched. The black to white range, or gray
scale, of a monochrome television signal or the luminance
portion of a color television signal is represented by
amplitude variations of the video signals. Therefore, a
video signal varying linearly in amplitude applied to a
nonlinear kinescope in a television receiver would result
in a picture the contrast range of which would be reduaed
undesirably according to the nonlinear transfer
characteristic of the kinescGpe. Accordingly, it is
desirable to gamma correct the video signal prior to
transmission in such a manner that the signal reproduced
in a television receiver has the desired contrast range.
Generally, gamma correction is accomplished by
~ 2 -
1~9'~214 RCA 69,895
1 passlng the vldeo signals derived from the television
camera through a nonlinear circuit having a predetermined
exponential relationship between input and output to
precorrect the signal for the subsequent nonlinear transfer
characteristic of the kinescope in the television receiver.
While the exponent may be any selected number, it is
generally accepted that the nonlinear circuit should provide
an output equal to its input raised to reciprocal of the
kinescope gamma, typically to the one-half power. The
nonlinear circuit is usually located in a video signal
processing amplifier coupled between the camera pickup
tube and the color encoder. When the image source for the
television camera is a positive film such as generally used
for television programming, the positive film has a gamma
characteristic which may be expressed in video signal
form as Vx = K-B~, where K i8 a constant and B is the
scene brightness. .For a typical positive color film,
gamma will be approximately +2. This gamma value of +2
.for a positive film when combined with the signal
processing gamma of +.5, previously described, results in
an overall gamma of approximately +1. This combined signal
gamma value of +l is further combined at the television
receiver kinescope whose gamma is typically +2. The
overall combined gamma from film to signal processing to
kinescope yields a gamma value of +1 which is desired in
order to accurately reproduce the film images.
Heretofore, it has been customary practice in
television programming to utilize positive color film in
television film cameras because such positive film is
color balanced; that is, equal amounts of red, green and
- 3 -
1094214
~CA 69,895
1 blue colors yield white, thereby permitting direct viewing
of the positive film by the camera system. However, the
making of positive film from the original negative film
requires at least one extra processing step which takes
time and results in a degradation of resolution as well as
color saturation of the resulting positive film relative
to the negative. It would be advantageous to use negative
film directly. However, negative film is not color
balanced; therefore, a masking operation is utilized for
balancing the video output signal from negative film to
correspond to the balanced video signals from a positive
film. Apparatus suitable for this masking operation is
described in detail in United States Patent No. 4,009,4~9,
issued February 22, 1977, Harold G. Seer, Jr.,
entitled, "Negative Color ~ask Correction".
Negative masking operations will provide for the
development of balanced video signals when utilizing
negative film; however, such masking operations do not
correct for the negative gamma characteristic of negative
film. As was previously described in conjunction with the
transfer characteristic of television systems, known
positive gamma circuits readily accommodate a positive
film, whereas negative gamma image signal sources cannot
be readily accommodated without additional circuitry.
Prior art efforts involved inverting the video signal
and providing negative curve shaping by means of networks
utilizing semiconductors and resistive components in an
attempt to develop a video signal with a positive gamma
from negative film sources. These typical prior art
-- 4
.
1~34Z14
RCA 69,895
1 attempts were only marginally successful in regard to
satisfactory curve shaping of the inverted video signal
as well as tracking repeatability between the three
color channels.
A gamma circuit for developing a video output
signal having a positive gamma coefficient from an input
signal having a negative gamma coefficient, comprises an
input terminal for receiving a first signal (Vx) having
a negative gamma coefficient. First combining means
coupled to the first signal and to a first reference (VB)
for providing a first additive combination (Vx + VB) f
the first signal and the first reference signal. Divider
means responsive to the first additive combination signal
and a second reference source (V~) for developing an
output signal equal to the second reference source divided
by the first additive combination signal. Second combining
means coupled to the divider output signal and a third
reference source (P) for additively combining the divider
output signal and the third reference source.
FIGURE l is a block diagram of a negative gamma
circuit embodying the present invention,
FIGURE 2 is a more detailed schematic of the
negative gamma circuit of FIGURE l; and
FIGURE 3 illustrates typical characteristic
curves of the circuits of FIGURES l and 2.
In FIGURE l, an input video signal Vx is
additively combined with a voltage source VB and the
additive combination of Vx + VB is coupled to one input
-- 5 --
1~4Z14
RCA 69!895
.
1 terminal of the differential input terminals of a four-
quadrant multiplier 100. The output terminal of multiplier
100 is coupled to one terminal of a voltage divider
comprising resistors 201, 202. A voltage source V~ is
coupled to the other terminal of resistors 201, 202. The
common junction of resistors 201, 202 is coupled to the
input terminal of an amplifier 300. The output signal VO
of amplifier 300 is coupled to a second input terminal of
multiplier 100 and also additively combined with a
voltage source P. As illustrated, multiplier 100, which
is connected in the feedback loop of operational amplifier
300, forms an electronic divider in a known manner; i.e.,
the output signal VO equals the signal V~ divided by
(VX VB~. .
The signal transfer characteristic of the circuit
of FIGURE 1 may be expressed as:
(1) KO = O X B) M 2 + V~R
where:
KA = amplifier gain
X~ = multiplier constant
VX = signal input
V~ = divider constant
2S VO = signal output
and:
VOp = VO + P
where:
P = D.C. Pedestal
-- 6 --
10~42~4
RCA 69,895
1 therefore:
(2) VO (Vx + VB)K~1R2KA ~ (Rl + R2)
and: .
Rl \
(3) Vo = ~Rl R2)
(VX + Vg)KM(Rl + R2) KA
and:
Vi~ ( 1 ) .
( ) OP/ R2
(VX + VB)KM ~Rl + RJ KA
If the gain of amplifier 300 approaches infinity
and voltage sources V and P are made zero, the circuit
B
of FIGURE l would form a perfect divider circuit.
Equation 4, which defines the transfer characteristic of
~ the circuit of FIGURE l under these conditions, becomes:
VffRl
(5) V
where Rl, R2 and K~ are constants, or:
(6) V~
and:
(7) VOp = KVff~Vx) l
-- 7 --
?42~4
RCA 69,895
1 wherein V~ can be adjusted to control the output level VO.
The contrast or density of a negative film
which determines the gamma characteristic may be expressed
mathematically as:
(8) DenSity = Light Transmission Of Film
and the light transmission of the film is:
10 (9)
C ~ BY
where C is a constant and s is the scene brightness which
may be further expressed as:
.
(10) i 1 ~ B( Y) : .
.:
Since the video signal from an image pickup tube is
directly proportional to the amount of light passed by
the film and imaged onto a photosensitive electrode of
20 ~ the pickup tube, the output signal from the image pickup
tube is:
,
( 11 ) ' ( )
2S or:
(12) V = K ~ B(-Y)
where K is a constant and B is scene brightness.
Therefore, if the video signal Vx is coupled to
-- 8 --
1094214 RCA 69,895
1 the input terminal Vx, the circuit of FIGURE l may be used
to provide negative gamma characteristics other than -l by
adjusting voltages VB, V~ and P.
In a typical application, the gamma of a negative
film is within the range of -.35 to -.65 from which it is
desired to develop a signal for processing which has the
proper positive gamma characteristic of +l. A correction
of gamma to +l for a negative film will most faithfully
reproduce the original scene content, whereas in a positive
film having a typical gamma of +2, the reproduction is
faithful only as to the film image. Attempts to alter
the positive gamma of positive film results in a loss
of color saturation and is therefore rarely attempted.
The ability to reproduce the original scene from negative
film provides an improved image quality.
The operation of the circuit of FIGURE l with
a typical negative film having a gamma of -.5 can be best
described by reference to equation (4) which represents
the transfer chiracteristic of the circuit of FIGURE l and
equation (12) which represents the light transmission of
the film in terms of brightness as follows:
( l 2)
(4) VOp / R2
(VX + Vg)KM ~Rl + R2) KA
Initial conditions for equation (4) are established by
aelection of values for Rl, R2, KA and Kr~ Satisfactory
operation has been achieved with the following values:
Rl + R2 .l
_ g _
1~)9421~
RCA 69,895
R2
= g
Rl + R2 .
K = 30
S KM
therefore:
(13) VOp V~(.l) 1 + P
(VX + VB) ( 9) ~ -
As previously stated, a properly shaped negative
gamma output VOp in terms of Vx may be achieved by
adjustment of the values for V~, VB, and P. Suitable
values of V~, VB, and P for a negative gamma film of -.5
may be derived from equation (13) as follows:
In equation (12):
V = K B Y
X
which relates video output Vx to scene brightness. For a
value VOp = 1 at 100% scene brightness,equation (12) ylelds
VOp = 1 = K ~ B = .14142, which also determines the
constant K as .14142.
Columns 1 and 2 of following Table A illustrates
calculated values of Vx for scene brightness B from 100%
to 2%, whereas Column 3 shows actual measured values
obtained from the circuit of FIGURE 1.
-- 10 --
~'34~l4
RCA 69,895
TABLE A
Column 1 Column 2 Column 3
B Vx = .1412 X B Vop measured
1.0 .14142 1.000
.7 .16903 .623
.5 .20000 .432
.35 .23905 307
.28 .26726 .251
.2 .31623 .188
.14 .37796 .140
.1 .44721 .106
.07 .53452 .0779
.05 .63246 .0574
.035 .75593 . .0402
.028 .84515 .0313
- .02 1.0000 .020
Selecting values of Vx for 100%, 14% and 2% scene
brightness from Table A yields:
VX = .14142
VX = ,37796
VX = 1 . 0000 ,
and the corresponding VOp signal desired will be:
VOp = 1
VOp = .14
VOp = .02
Sub~tituting the values of Vx corresponding to the desired
light transmission of the negative film at 100%, 14%, and -
2% in equation (13) results in three equations in which
V~, VB and P are unknowns. Solving the three equations
-- 11 --
10~4;Z~4
RCA 69,895
1 simultaneously results in values for V~, VB and P of:
Vff = +.4482
VB = .05627
P = -.03493
therefore:
(14) VOp4482(.1) 1 ~ 03493
(VX ~ .05627) ( 9) 30
which when solved for values of Vx, as shown in Columns 1
and 2 of Table A, results in an overall transfer
characteristic approximating gamma equal to +1. The values
of VOp versus brightness B are plotted and illustrated in
FIGURE 3 by curve I which represents an ideal power law.
Curve C of FIGURE 3 and previously mentioned Column 3
of Table A illustrate the actual transfer characteristic
obtained in practice from the circuit of FIGURE 1 utilizing
the selected initial condition values.
FIGURE 2 is a more detailed schematic circuit
diagram of the negative gamma circuit of FIGURE 1 in which
functional groups of components have been identified with
the same reference nùmerals used in FIGURE 1.
Multiplier 110 is a four-quadrant multiplier in
commercially available integrated circuit form, for
example, the Motorola MC-1595L. Resistors 101, 102, 103,
2S 104, 105, 106, 107, 108, 109, 113, 114, and 115 are used
to configure the multiplier 110 for a basic divide mode of
operation in conjunction with operational amplifier 300.
Transistors 301, 304, 309, 307, 308, 313 and the associated
resistors 302, 303, 305, 311, 312, 314, 316, 306 form such
an operational amplifier 300. Amplifier 300 further
- 12 -
10'34Z14
RCA 69,895
I lncludes diodes 310 and 315 for providing temperature
stability of amplifier 300.
An input signal Vx is coupled to a first input
circuit of multiplier 110 and a voltage -VB derived from
voltage divider 400 is advantageously coupled to the
signal inverting terminal of the same first input circuit,
thereby adding Vx to VB (Vx + VB). A second input
circuit of muItiplier 110 receives an input signal VO from
the output terminal of amplifier 300 while the corres-
ponding signal inverting terminal of the second multiplierinput circuit is coupled to a reference voltage divider
109 to balance the first and second input circuits of
multiplier 110. In this configuration, multiplier 110
provides a differential output signal at its output
terminals (2, 14) comprising the product of the two input
cirouits of multiplier 110 in the~form of -VO (Vx + VB)KM.
KM, the multiplier constant is determined by resistors
~ 103: and 104. The input terminals comprising the respective
: base electrodes of transistors 301 and 304 of differential
input amplifier 300~are coupled to output terminals 2 and
14 of muitiplier lIO as shown. A voltage V~ derived from
.
voltage divider 206 is coupled by resistors 205 and 115
(R2 and Rl), respectively, for controlling the output
level VO of amplifier~ 300. A voltage P is added in later
circuitry (not shown) to provide the D.C. pedestal yielding
on output VOp.
AB illustrated, the circuit of FIGURE 2 satisfies
the transfer characteristic of equations (4) and (13),
thereby providing the desired.negative gamma correction
by adjustment of V~, VB and P.
: - 13 -
~ .
10~4Z14 RCA 69,895
I As previously described in connection with a
negative film of gamma equal to -.5, similar computations
may be made for any value of gamma within the range of
-.35 to -l; however, in practice, such computation is
tedious; alternatively, changes in negative gamma may
be readily accomplished by inserting a negative image of
a standard gray scale, for example, EIA Logarithmic Gray
Scale, and adjusting V$, VB and P while observing a
television camera waveform monitor display.
- 14 -