Note: Descriptions are shown in the official language in which they were submitted.
_ield of the Invention
The present invention relates to methods and
apparatus for processing video signals and particularly to a
method and apparatus for processing a stored single frame NTSC
colour signal for viewing on a NTSC TV monitor.
Background of_the Invention
As colour television was introduced in order to
maintain compatibility with black-and-white requirements~
chrominance information was superimposed on the black-and white
intensity signal in the form of a phase and amplitude modulated
sub-carrier signal. The intensity (luminance) signal, having a
bandwidth of 4 Megahertz but having most of its power in the
lower portion of the spectrum, did not suffer appreciably from
the 1.5 Mhz wide chrominance signal centred around a frequency
of 3.579545 Mhz, the so-called colour sub-carrier. However, to
minimize the interference caused by the colour sub-carrier with
the basic luminance information signal, it was necessary for
the phase of the colour sub-carrier 'to appear inverted with
respect to the luminance signal from frame to frame. Thus the
interference, in the form of a ripple superimposed on the
intensity pattern of picture lines, is averaged out due to the
persistance of human vision.
Such an apparent phase inversion was simply
achieved by choosing a colour sub-carrier frequency which is a
half odd-multiple of -the frame frequency. Given a frame frequency
oi 29.9770026 per second (close enough to the original monochrome
frame rate of 30 frames/second) the colour sub-carrier was chosen
to be exactly 119437.5 times that, which yields the sub-carrier
.,
~8~99~
frequency of 3.579545 Mhz. A more rigorous description of ~hat
actually happens is that due to this half odd-multiple relation-
ship the peaks of energy in the colour sub-carrier fall between
the peaks in the luminance signal spectrum, and so are of
minimum visibility when viewed on a monochrome receiver. This
and a rPlated succinct discussion of the NTSC colour format is
given in part of a paper by John O. Limb et al titled Digital
Coding of Colour Video Signals - A Review and published November
1977 in IEE (Institute of Electrical and Electronics Engineers)
Transactions on Communications, Vol. Com-25, No. 11, at pages
1353 to 1355.
To summarize, this apparent phase shift of the
colour sub-carrier between successive frames is a result of the
half odd-multiple relationship between sub-carrier and frame
frequencies.
A problem, however, arises when a single video
frame is stored for viewing as a sti]l picture. Whether a frame
is stored in a VTR (Video Tape Recorcler) with "Stopped frame"
capability or in a single frame digital image store, in
repetiti~e retrieval of the stored frame the same information is
repeated over and over. Consequently, the reference colour sub-
carrier burst always appears in the same fixed phase relationship
with respect to all other elements of the composite signal,
horizontal synchronization pulse and all. What this actually
means is that an abrupt phase change of 180 occurs in the
chrominance signal r including the reference sub-carrier burst, at
the beginning of each frame. Since this burst is used as
reference for l~cking the TV monitor's internal sub-carrier
-- 2
:
309~
oscil~ator, the oscillator must adjust its output phase by 180
to maintain phase synchronization with the incoming signal in
order to decode the chrominance signal correctly. This however
typically takes a few milliseconds because such oscillators are
designed to have a very tight phase lock (high inertia) since
they must maintain their phase without correction for the
duration of a full line after the short period of the reference
burst at the beginning of every line. The long period required
for the reference oscillator to lock in on the new phase would
cause a good portion of the picture (frame) to be unacceptably
distorted in colour.
Prior Art of the Invention
In an article published August 1976 in the SMPTE
(Society of Motion Picture and Television Engineers) Journal,
Vol. 85, No. 8, entitled, "The Electronic Still Store: A Digital
System for the Storage and Display of Still Pictures", the
authors W.G. Connoly and J. Diermann describe briefly that -
luminance is separated from chrominance by means of digital
filtering - chroma phase is inverted to reproduce the four-field
NTSC colour signal from the two fields stored on the disc, and
the chrominance and luminance signals are recombined. The Digital
signal is then transformed back to the analog domain and,
finally, sync and burst are reinserted via a processing
amplifier".
An Electronic Still Store utilizing the above
quoted technique was manufactured and sold by Ampex Corporation,
of Redwood City, California, U.S.A.
It will be appreciated by those skilled in the
-- 3 --
art that the process of separating the luminance and
chrominance signals, be it performed by analog or digital
techniques, in order to generate the 4-field (i.e. 2-frame)
signal necessary for the display of a colour still picture, is
somewhat complex and still exhibits problems due to residual
chrominance after filtering.
Summary of- the Invention
An object of the present invention is to simplify
the process of generating two frames suitable for colour TV
display from the single stored frame, whether the frame is
stored in analog or in digital form.
A feature of the present invention is that it
utilizes a simple analoy delay and a few logic components to
achieve this object.
A disadvantage of the present invention, however,
is that the resultant 2-frame signal is not strictly in
adherence with the NTSC standard format. The discrepancy,
though, from the standard format is so minor that no noticeable
degradation in the picture has been observed.
The solution of the present invention is to simply
delay every alternate frame by a period equal to one-half the
period of colour sub-carrier frequency, a delay of a few
nanoseconds. The 1/2 period delay, very small compared to the
duration of a picture line, causes the colour reference burst to
reach the TV monitor (or receiver) with the phase in approximate
coincidence with the phase of the internal reference oscillator
of the monitor. Thus the tightly locked oscillator remains in
proper phase-lock and the displayed colour thereby undistorted.
- 4 -
~8(~9~L
Thus, in accordance with the present invention,
there is provided apparatus for continuously displaying on an
NTSC television monitor a single video frame NTSC colour signal
stored in a video signal source, characterized by delay means
adapted to delay said NTSC colour signal by a predetermined time
interval, and video signal switching means controllable to
alternately switch to an output said NTSC colour signal output
by one of said video source and said delay means.
Further according to the present invention, a
signal processing me~hod for continuously displaying on an NTSC
television monitor a stored single frame NTSC colour signal,
characterized by the step of delayin~ every alternate retrieved
frame by a time interval substantial:Ly equivalent to one-half
the period of the colour subcarrier 'Erequency in said stored
single frame NTSC colour signal prio:r to application of said
every alternate retrieved frame to said NTSC television monitcr;
the other retrieved frames being applied to the monitor undelayed
but processed through a substantially identical signal path and
not including said delaying step.
Normally, such predetermined time interval would
be equal to one-half the period of the colour sub-carrier
frequency of the stored single frame. But clearly odd multiples
of the one-half period are equivalent to one-half period for
purposes of phase locking and would result in the same effect,
although the NTSC format, and thus the displayed picture, would
become progressively degraded as the odd multiple grows.
In the foregoing, the words television and
monitor are to be considered synonymous. The important point is
-- 5 --
~L8/D~9~L
that the "set" be capable of accepting a base-band video signal
directly, otherwise it is necessary to modulate the base-band
signal on a suitable carrier to be applied to the TV set in the
same manner as, say, a cable signal.
Brief Description of the Drawings
A preferred embodiment of the present invention
will now be described in conjunction with the accompanying
drawings in which:
Figure 1 is an illustration of the phase relation-
ship between two lines one from a delayed and one from a non-
delayed retrieved frame; and
Figure 2 is a schematic diagram of an alternate
frame shift circuit suitable for implementing the present
invention.
Description of the Preferred Embodiment
In Figure 1 of the drawings, the top graph
depicts the beginning of a line in a non-delayed retrieved frame.
The synchronizing colour sub-carrier reference burst (about 8 to
10 cycles thereof) is superimposed on the "back porch" of each
horizontal blanking pulse. The phase of the local oscillator in
the TV monitor is locked to this reference burst at the line
beginning so that its output may serve as a phase reference for
the actual chrominance information. The latter is superimposed
on the luminance signal in the form of the amplitude and phase
modulated colour sub-carrier~ The bottom graph in Figure 1
depicts a line from a delayed retrieved frame, and as may be seen,
the sole difference is in the brief l/2-period delay on the
whole line from the point in time when the line should have
-- 6 --
':
94
normally occurred if the frame containing it were not delayed.
Of course, every line in that frame is also delayed by the same
1/2-period.
Figure 2 of the drawings shows a simple block
diagram for the alternate frame shift circuit. A video source
10 outputs a repeated two-field video frame in analog form, which
signal is applied in parallel to a delay line 20 and to input A
of switching video amplifier 30. The output of the delay line 20
is applied input B of the switching video amplifier 30. The
output of the amplifier 30 labelled "Video-Output" delivers the
desired end signal to the TV monitor or receiver (standard NTSC
receiver). The video source 10 also delivers a vertical drive
pulse and a frame identifier pulse, which are both applied to a
control logic circuit 40. The output of the control logic 40
is applied to the switching control input 5 of the switching
video amplifier 30.
In the present embodiment the video source 10
comprises a frame store 11 and a vidlso timebase generator 12.
The frame store 11 may be digital or analog, but would ultimately
have an analog output if digital by decoding the signal in a video
decoder. The timebase generator 12 could be a device such as
manufactured by Leitch part No. CTG 200/120N. The control logic
circuit 40 simply comprises two flip-flops 41 and 42 inter-
connected as shown. It operates in the following manner.
Flip-flops 41 and 42 form a 2 bit cascaded counter clocked by
the vertical drive pulse from the generator 12 through inverter
43. Vertical drive occurs once each field time during the
vertical retrace period; as a result, the flip-flop 41 toggles
-- 7
8~9~
at the field rate and the flip-flop 42 at the frame rate. The
presence of the frame identifier pulse at the R (reset) inputs of
the flip-flops 41 and 42 ensures synchronization between the
counter "state" and the two frame cycle of the video signal.
~onsequently, the Q output of the flip-flop 42 specifies the
frame number ("O" = odd, "1" - even) of the continuous video
signal. With this Q output fed into the switching input S of
the amplifier 30, operation is such that the undelayed video
signal (input A) appears at the output during odd frame periods
and the delayed signal (input s) appears at the output
during the even frame periods. The time delay introduced by
the delay line 20 must be equal to one half the period of the
colour sub-carrier frequency, or 140 nanoseconds. ~ delay
line to provide the delay of 140 na~oseconds at a bandwidth of
ca. 5.5 Megahertz is readily available from several video
equipment manufacturers (i.e. Barnett and Longmore, Coventry,
U.K.).
It is, of course, also possible to delay every
alternate digitally encoded video frame prior to decoding by
the same amount of delay with the same result, provided that
the two signal paths (delayed and undelayed) are identical
except for the delay.
.
