JPS5816677B2 - Coding method for color television signals - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for successfully encoding color tenvision signals.

As is well known, in a system that samples a color television signal, quantizes it, encodes it, and returns it to the original signal, interference between the sampling frequency fs and the color subcarrier frequency fsc occurs in the reproduced signal. A beat signal is generated.

This beat signal is determined by the greatest common divisor of the sampling frequency fs and the color subcarrier frequency fsc, and corresponds to the beat frequency within the video band.

The band of the reproduced signal is the normal video band (0~4.5MH
z), and the relationship between fs and fsc is fs=m
sc (m is a positive integer), the beat frequency f
When selecting sc, the beat frequency is Ot fsc/2
+fsc.

Therefore, if fs=m×fsc from the relationship between the color subcarrier frequency, horizontal synchronization frequency, and vertical synchronization frequency, then i/2 s fs ”=(n−1) appears on the screen.

(From the perspective of visual interference improvement, an offset relationship smaller than 1/4 is not desirable.

) On the other hand, the color distortion caused by quantizing the subcarrier signal is averaged within two cycles because the point returns to the same location every second cycle from fs - mXfsc as described later, and the appearance is It decreases as the quantization width decreases.

However, it is preferable to synchronize the phase of the sampling pulse with the horizontal synchronization pulse and always align it on a straight line perpendicular to the scanning line, since it is preferable for the realization of a digital filter and its digital processing. Since the polarity at which color distortion occurs alternates for each line, in combination with the visual averaging effect, high-quality color reproduction is possible.

To make the bead signal associated with quantization less noticeable, increase the number of quantization levels and reduce one quantization level to reduce the high frequency level of the subcarrier, or make the sampling frequency very high. There are other means such as

For example, if the sampling frequency is extremely increased to infinity, no beat signal will be generated.

However, according to the sampling theorem, if encoding is performed at a sampling frequency that is more than twice the highest frequency of the signal to be transmitted, the sampling frequency will increase to a point where it is impossible to reproduce the original signal. Doing so would increase the required bandwidth more than necessary.

Note that the value of n is, for example, if the video band is 4.5 MHz in the NTSC system, the required band f is as described above.
s is 9 MHz, and as is well known, the subcarrier frequency fsc makes the integer n approximately 3 or more due to fsc.

Similarly, for systems other than NTSC, each n
will be determined.

Increasing the number of quantization levels also means increasing the number of parallel quanta per pixel, which, together with the sampling frequency, significantly increases the cost of the equipment when transmitting the signal.

The above is also a serious problem for encoders that encode signals.

In other words, particularly reducing the quantization level means dramatically increasing the precision of the encoder, and increasing the sampling frequency means that ultra-high-speed elements are required.
This will lead to a drastic increase in prices.

Therefore, in order to make the beat signal accompanying quantization less noticeable, it is preferable to select the relationship between <fs and fsc as described above, and to improve interference by offset.

The distortion of the color signal due to quantization mentioned earlier occurs because the beat frequency is concentrated in the color signal because the sampling frequency is an integer multiple of the subcarrier signal. Although it is reduced by increasing the quantization frequency, it ultimately remains within one quantization level, and there is a limit to its reduction.

Figure 1 shows that when the carrier color signal is quantized with 3Xfsc,
This is not enough to explain how distortion occurs.

In the figure, the sine wave shown as a solid line represents the subcarrier signal in the input signal, and the white circles above it are sampling points.

In the figure, the quantization level is shown to be roughly equal to the subcarrier level in order to make it easier to understand the process of generating distortion due to sampling.

At sampling phases 1.2 and 4.5, no quantization error occurs because the sampling point is above the output level, or at sampling phases 3 and 4.5, no quantization error occurs.
6, the sampling point is above the decision level, so the quantization value changes rapidly depending on the output level, resulting in a phase error (Δθ).

At the same time, a change in amplitude also occurs.

Figure 1 bye shows the change in the output of the quantized signal, and the dotted line in Figure 1 a shows the relationship between the phase and amplitude of the waveform when this quantized signal is faced with an analog signal. It is.

This distortion occurs in a specific phase relationship between the subcarrier and the sampling pulse as long as a sampling pulse that is an integral multiple of the subcarrier is used, resulting in an error in the phase and amplitude of the transmitted color signal.

To eliminate this distortion, either increase the number of quantizations and make one quantization level below the detection limit, or break the specific phase relationship between the subcarrier and the sampling pulse (change the frequency of the sampling pulse to the subcarrier frequency). (not fixed to an integer multiple of ).

However, in the former case, as mentioned above, the cost of the equipment increases significantly, and in the latter case, there are problems such as beat interference, and it is difficult to determine which is the case.

In the generation of a beat signal, as mentioned above, the sampling pulse averages the distortion caused by this sampling within two periods of the subcarrier, since interference can be reduced by placing the beat fringes in an offset relationship. One solution is to do so.

Next, phase-locking the sampling pulse to the horizontal synchronization signal so that the top location of each scan line is sampled is
The rate of change in the chroma signal of the SC signal is small, and when there is no color change on adjacent scan lines, the chroma signal phase differs by 180 degrees on adjacent scan lines. In addition, since the sampling points on adjacent scanning lines are placed at symmetrical positions on the carrier color signal, if the sampling frequency is not an even multiple of the subcarrier, quantization is not possible. They work to mutually correct the distortion caused by the

FIG. 2 shows a state in which the quantization of the carrier color signal differs on adjacent scanning lines in the case of fs=3fsc.

In consideration of these points, the present invention obtains a sampling pulse that has a color television number, is phase synchronized with a horizontal synchronization signal, and has sampling points arranged on a straight line perpendicular to the scanning line, thereby generating a preview signal. This is what we are trying to encode.

FIG. 3 shows one embodiment of the invention, in which 101 is a signal separation section, 102 and 106 are phase comparison sections, 103 is a controlled oscillation section, 104, 105, and 108 are frequency step-down sections, and 107 is a waveform shaping section. , 109 is a code exchange unit.

Input signal 201 (NTSC) from terminal A is signal separated 1
01 and is supplied to the code conversion unit 109.

The signal separation unit 101 separates the input signal 201 into a burst signal 202 and a horizontal synchronizing pulse 203.

Controlled oscillator 103 outputs signal 206.

The signal 206 becomes the down-down section comparison signal 204.

The phase comparison signal 204 is phase-compared with the output signal 202 from the signal separation section 101 by the phase comparator 102.

In the phase comparator 102, the signals 204 and 202 are compared during the period in which the signal 202 exists (burst period),
The signal 205 is output as a comparison signal.

This comparison signal has a constant potential determined by the amount of error.

This signal 205 controls the control transmitter 103, and changes the excavation frequency and phase of the burst signal 202 to the lowering part 1.
05 and 108 are also included.

The down-converter 105 converts the signal 206 to 1/910×(
Output n.

This signal 207 is phase-compared with the output signal 203 of the signal separation section 101 in the phase comparator 106, and the phase difference is determined as the signal 208.
is output as

Signal 208 is applied to step-down section 105 to control the step-down number of step 105, resulting in the phase of signal 2070 being changed to signal 2030.
Match the phase.

The signal 207 is pulsed by the waveform shaping section 107 and is used as a pulse signal 211 to control the descending section of the pond.

The step-down section 108 converts the output 206 of the controlled oscillation section 103 into 1/
The pulse signal 209 whose phase is corrected for each horizontal scanning line by the pulse signal 211
occurs.

The pulse signal 209 thus obtained is a sampling pulse signal that has an input multiple, is phase synchronized with the horizontal synchronizing pulse, and has sampling points arranged on a straight line perpendicular to the scanning line.

This signal 209 drives the encoder 109, samples the input signal 201 from the terminal A, encodes it, and outputs it as a signal 210.

FIG. 4 shows the input and output signals of each part.

Although 207 and 207 in FIG. 4 are the same signal, 207 is the signal when there is a phase error, and 207 is the signal when there is a phase error.
07 is when the phase error signal 208 does not exist.

Similarly, the signal 209° 209 is a partially expanded version of the signal 209, and the signals 209 and 209 show how the phases of the two signals are on adjacent scanning lines. .

The signal 210 shows how the signal 201 is encoded by the sampling pulse, and the code content is A-B-C-D.
It is indicated by...

Note that the output signal 206 of the controlled oscillation unit 103 only needs to be such that the sampling points on each scanning line are phase-locked with the horizontal synchronizing pulse and aligned on a straight line perpendicular to the horizontal synchronizing pulse. The influence will be felt on ゜108.

As mentioned above, the signal 210 thus obtained has a beat frequency of fsc and 1/2 fsc, and the beat disturbance is not noticeable.

Since the sampling points are in the same phase every two periods of the subcarrier, there is less distortion due to quantization.

Since the sampling points for each scanning line are aligned on the same line, the vertical correlation of the images is good and it is easy to form a comb filter.

The phase information of the sampling point (each sampling point has specific phase information because it is in a specific phase with the sampling pulse or the burst signal) is encoded with an integral multiple of the sampling pulse. Although it is somewhat inferior, it retains the characteristics of chroma signals and can easily realize separation of chroma signals (synchronization force daily calculation) using horizontal correlation as the first priority.

In addition to the signal 209, the down-down section 108 also outputs a sampling pulse 209 having the same repetition frequency and a phase difference of 180°.
-1 can be generated at the same time, and as shown in FIG. 3, this signal 209-1 is used to drive the encoder 109-1 of the same type of music as 109, and the input signal 201 from terminal A is
If the signal 210 is encoded, an output 210-1 sampled at a phase 180° different from the signal 210 can be obtained.

The signal 210 and the signal 210-1 are arranged as shown in FIG.
, B, C, C, . . . , the signal 212 can be equivalent to two input signals 201.

Furthermore, as shown in FIG.
The signal 213 obtained by selectively switching the signal 201 from the terminal A and encoding it as one sampling pulse 209-2 for each field) has the characteristics of the signal 210 described above, as well as for each field, Alternatively, since the phase of the sampling pulse with respect to the horizontal synchronizing pulse differs from the phase of the sampled signal by 180° for each frame, the sampling point is interpolated in the horizontal direction.

This appears to be an attempt to expand the band by improving the resolution without increasing the frequency of the sampling pulse.

As is clear from the above description, according to the present invention, the beat frequency becomes fsc or 1/2 fsc, making the beat disturbance less noticeable.

Furthermore, since the sampling points for each scanning line are lined up on the same line, the vertical correlation of the images is good and the formation of a comb-shaped filter is easy.

[Brief explanation of drawings]

FIG. 1 is a diagram showing a state in which the carrier color signal is quantized with three times the subcarrier, FIG. 2 is a diagram showing a state in which the quantization of the carrier color signal is different on adjacent scanning lines when fs=3fsc, and FIG. FIG. 3 is a block diagram of a device using the color array vision signal encoding method according to an embodiment of the present invention, FIG. 4 is a diagram showing input and output signals of each part, and FIG. 5 is a diagram of an embodiment of a song. The block diagram and FIG. 6 are signal waveform diagrams thereof. 103... Controlled oscillation section, 104, 105°10
8... Frequency down-down unit, 109... Code conversion unit.

Claims (1)

[Claims] Having a frequency of (integer), phase synchronized with the horizontal synchronization pulse,
A method for encoding a color 7 revision signal, characterized in that the color television signal is encoded using a sampling pulse in which sampling points are arranged on a straight line perpendicular to a scanning line. A first sampling pulse, which has a frequency of 1. A method of encoding a color television signal, characterized in that a color preview signal is entirely encoded using a sampling pulse obtained by selectively switching a sampling pulse for each field frequency or frame frequency.