JPH0779258B2 - Quantizer - Google Patents

Description

Detailed Description of the Invention

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a quantizer for converting a digital signal having a long word length into a digital signal having a short word length which is sampled at high speed. 2. Description of the Related Art In recent years, due to improvements in digital signal processing technology, signals that have been analog processed conventionally have been digitalized. Along with this, higher performance and lower cost of digital-to-analog converters have become more important. Quantizers are often used for these purposes. The block diagram is shown in FIG. 4 and is described (for example, Japanese Patent Application No.
-230114 publication). The adder 3, the local quantizer 101, the subtractor 2, and the delay unit 4 constitute a main loop 100 which is a single integral type noise shaping quantizer having a primary shaping order. Further, the adder 7, the local quantizer 6, the subtractor 8, and the feedback circuit 9 serve as a sub-loop 106 which serves as a double integral type noise shaping quantizer having a quadratic shaping order.
Is configured. The subtractor 2 gives the input / output difference of the local quantizer 101 and the feedback circuit 9 a delayed output to the double integral type noise shaping quantizer. Also,
The subtractor 8 supplies the feedback circuit 9 with the input / output difference of the local quantizer 6. Here, the input X is a 16-bit digital signal, and the local quantizers 101 and 6 perform the quantization as shown in Tables 1 and 2. The output is standardized by 11264. As you can see from this table, 33792 corresponds to 0 dB. Further, the feedback circuit 9 has a transfer function represented by the equation (1), and is specifically configured as shown in FIG. Here, 41 and 42 are delay devices, 43 is a subtracter, and 44 is a multiplier, which doubles the input value. H (z) = − 2z ⁻¹ + z ⁻² (1) Here, the quantization error generated by the local quantizer 101 is Vq1, and the quantization error generated by the local quantizer 6 is Vq1.
Assuming that q2, the relationship between the input X and the output Q1 of the main loop 100 and the relationship between the input X'and the output Q2 of the sub-loop 106 are represented by the expressions (2) and (3). Q1 = X + (1-z -1) · Vq1 ... (2) Q2 = X '+ (1-z -1) 2 · Vq2 ... (3) where the output of the adder 2 is the local quantizer 1 Since it is a difference between input and output, X '=-Vq1 (4) Therefore, after the output Q2 of the subloop 106 is differentiated by the differentiator 10 according to the shaping order of the local quantizer 101 of the main loop 100, the adder 12 Thus, when added to the output Q1 of the main loop 100, the term of Vq1 shown in the equation (2) is canceled out, and the relation between the input and output X and Y as a whole is as shown in the equation (5). Y = X + (1-z ^-1 ) ³ · Vq2 (5) Here, the value that Q1 can take, that is, the gradation is -3, -2,
..., there are seven values of +3 (seven values), and the possible values of Q2 are three values of -1, 0, and +1, so the possible values of Y are-.
11 values of 5, -4, ..., +5. That is, the input signal is compressed into 11 values (3.46 bits). Also,
Equation (5) shows that the quantization error in the low frequency band is driven to the high frequency band. Therefore, by configuring as shown in FIG. 4, the digital output that does not impair the dynamic range of the input digital signal is output. The bit number of the signal can be compressed, and a dynamic range of about 103 dB can be obtained by operating this circuit with 32 times oversampling. DISCLOSURE OF THE INVENTION Problems to be Solved by the Invention However, in the above-described configuration, when a value exceeding 33792 is input as the DC input, the quantization error Vq1 in the main loop 100 monotonically increases, and the main loop 100 oscillates. I will end up. Further, even if it is not a direct current input, for example, the sub-loop 106 that temporarily inputs a signal exceeding 33792 may oscillate. Further, oscillation can be prevented by increasing the gradation of the output of the local quantizer 101 in the main loop 100, but this causes a problem that the gradation of the quantizer as a whole increases. The present invention solves the above problem, and an object of the present invention is to provide a quantizer that does not oscillate even when a signal exceeding 0 dB is input. In order to solve the above problems, a quantizer according to the present invention includes a first local quantizer for quantizing an input signal, and a first local quantizer for noise shaping of a given input. No. 1 noise shaping type quantizer, a second noise shaping type quantizer which inputs the quantization error generated by the first local quantizer, and performs noise shaping of this quantization error, and 2 differentiating the noise shaping type quantizer output according to the shaping order of the first noise shaping type quantizer, and the differentiator output with respect to the output of the first local quantizer, Means for adding so that the quantization error generated by the first local quantizer is canceled, and the result of this addition is taken out as an output, wherein the state of the differentiator is Based on
The upper limit value and the lower limit value of the output of the first local quantizer are controlled. Operation With the above configuration, the second noise shaping quantizer which receives the quantization error generated by the first local quantizer of the first noise shaping quantizer as an input and performs the noise shaping of this quantization error The output of is differentiated by the differentiator according to the shaping order of the first noise shaping quantizer, and the output of this differentiator is canceled so as to cancel the quantization error generated by the first local quantizer. First
When the output of the noise shaping type quantizer is added and the addition result is taken out as an output, the upper limit value of the output of the first local quantizer of the first shaping type quantizer,
Alternatively, since the lower limit value is controlled to expand or contract according to the state of the differentiator, there is no increase in the gradation of the output of the quantizer as a whole, and the sign of the quantization error generated by the main loop is randomized. It is possible to prevent oscillation of the main loop. Similarly, oscillation can be prevented in the sub loop. Embodiment An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the concept of a quantizer according to the present invention. In FIG. 1, reference numeral 20 denotes a main loop, which, like the main loop 100 shown in FIG. 4, performs single integral type noise shaping, and only the local quantizer 1 is different. Reference numeral 21 denotes a sub-loop, which performs the same double integral type noise shaping as the sub-loop 106 shown in FIG. 4, for example. 10 is a differentiator. The local quantizer 1 quantizes the input signal, for example, as shown in Table 3. However, 0 dB is 33792 as in the case of the conventional example. Since the output of the sub-loop 21 is any of ± 1.0 as in Table 2, the output of the differentiator 10 is -2 to +2. The local quantizer 1 is in the state of the differentiator 10, where the differentiator 10
The upper and lower limit values of the output are changed according to the output value of, for example, as shown in Table 4, when the control input by the differentiator 10 is +2, the upper limit value of the output is +3, and the output of the differentiator 10 When is -2, the lower limit value of the output is -3. Considering the output of the differentiator 10 here, there are only three values that can be input to the differentiator 10, namely -1, 0, and +1. Therefore, assuming that the output of the differentiator 10 is +2, the next output value is Is always 0 or less. Because, the output of the differentiator 10 is 2 means that its input has increased by +2. Since there are three kinds of inputs, -1, 0, +1, in order to output +2, it has to be input in the order of -1, +1. What value will be input next? In both cases, a value of 0 or less will be output. In other words, the local quantizer 1 is always +5 every two degrees,
Alternatively, it means that a value of -5 can be output. With this configuration, the quantization error Vq generated by the main loop can be increased even if an input of 0 dB or more is applied.
1 does not increase monotonically, can operate normally even with a larger input, and the output value of the entire quantizer is suppressed to 11 values from -5 to +5. Therefore, the gradation does not increase. FIG. 2 is a block diagram showing another concrete embodiment of the quantizer according to the present invention. In addition, in FIG.
Components having the same functions as those in FIG. 4 are designated by the same reference numerals, and detailed description thereof will be omitted. In the figure, a main loop 20 which is a single integral type noise shaping quantizer by an adder 3, a local quantizer 5, a subtractor 2 and a delay device 4
And the adder 7, the local quantizer 6, the subtractor 8 and the feedback circuit 9 constitute a sub-loop 21 which is a double integral type noise shaping quantizer. Also differentiator
Reference numeral 10 is composed of a delay unit 14 composed of a register and a subtracter 13, and in this embodiment, the local quantizer 5 is controlled by the output of the delay unit 14. As shown in FIG. 3, the local quantizer 5 is composed of a local quantizer 31 and a limiter 32 that limits the output of the local quantizer 31, and the local quantizer 31 quantizes as shown in Table 3. I do. Further, the limiter 32 is controlled by the output of the delay device 14, and limits the output value as shown below according to the output thereof. If delay 14 output = +1, output upper limit = 5, output lower limit = -33 If delay 14 output = 0, output upper limit = 4, output lower limit = -4 If delay 14 output = -1, output Upper limit value = 3, output lower limit value = −5 Here, consider a case where a value exceeding 0 dB as the input X, for example, 35000 is input as the DC value.

[When delayer 14 output = −1] The local quantizer 5 is supposed to output 4 originally, but outputs 3 by the limiter 32. Therefore, Vq1 gradually decreases. Since −Vq1 is input to the sub loop, the sum of the output β of the feedback circuit 9 and the output β gradually increases,
The output of the local quantizer 6 becomes zero. Therefore, the output of the delay unit 14 = 0, the upper limit value of the output of the local quantizer 5 = 4 is obtained, and the main loop operates normally without oscillating.

[When Delayer 14 Output = 0 or 1] The local quantizer 5 outputs +4 which should be output originally, and then outputs +3 several times. It is obvious that the output of the local quantizer 5 does not have any limitation and thus operates normally. In addition, the ratio of the number of outputs that outputs +4 and +3 is
As shown in the value 3.107, which is a standardized value of 35000 with 11264, 0.10
7: (1-0.107) = 1: 8.35. That is, about 1 in 9
This means that +4 should be output once, in other words, the value of the delay device 14 should be 0 or more once every 9 times. With this configuration, the maximum value of the amplitude of the input signal can be expanded without increasing the output gradation of the quantizer as a whole. In the case of this embodiment, the input signal level can be increased to approximately +2 dB as compared with the conventional example. In the above embodiment, the local quantizer 1 that outputs 11 values from -5 to +5 is used, but the present invention is not limited to this, and 9 values from -4 to +4, or
It goes without saying that the value may be 11 or more. Also, the main loop does not have to be a single-integration type noise shaping circuit, and the point is that the upper and lower limits of the output depend on the state of the differentiator in which the local quantizer in this loop is connected to the subloop. Anything that is controlled is good. EFFECTS OF THE INVENTION As described above, according to the present invention, there is provided a first local quantizer for quantizing an input signal, and a first noise shaping quantizer for performing noise shaping on a given input. A second noise shaping type quantizer for inputting the quantization error generated by the first local quantizer and noise shaping the quantization error, and an output of the second noise shaping type quantizer A differentiator that differentiates according to the shaping order of a first noise shaping quantizer, and the first local quantizer outputs the differentiator output with respect to the output of the first local quantizer. A quantizer configured to add so as to cancel the generated quantization error, and to take out the addition result as an output, based on the state of the differentiator, the first local quantizer output By controlling the upper and lower limit values of, even if a signal that exceeds 0 dB is input, circuit oscillation is prevented without increasing the gradation of the value output by the quantizer, and It has an excellent effect that a quantizer which operates stably can be provided.

[Brief description of drawings]

FIG. 1 is a block diagram showing the concept of the quantizer according to the present invention, FIG. 2 is a block diagram showing another concrete embodiment of the quantizer according to the present invention, and FIG. 3 is a local portion of the quantizer. FIG. 4 is a block diagram showing a concrete example of the quantizer, FIG. 4 is a block diagram showing a conventional quantizer, and FIG. 5 is a block diagram showing a concrete example of a feedback circuit in the quantizer. 1,5,6 …… Local quantizer, 4 …… Delayer, 9 …… Feedback circuit, 10 …… Differentiator, 13 …… Subtractor, 14 …… Delayer, 20…
... Main loop (first noise shaping quantizer), 21 ... Sub loop (second noise shaping quantizer).

Continuation of the front page (56) Reference JP 62-265810 (JP, A) JP 3-289809 (JP, A) JP 61-84914 (JP, A) JP 1-274510 (JP , A) JP 3-289808 (JP, A) JP 4-30618 (JP, A) JP 4-30619 (JP, A) JP 4-56407 (JP, A) JP 3-289810 (JP, A) JP-A-3-289709 (JP, A)

Claims (2)

[Claims] 1. A first noise-shaping quantizer having a first local quantizer for quantizing an input signal, for noise-shaping a given input, and the first local quantizer. Noise shaping type quantizer for inputting the quantization error generated by the quantizer, and the second noise shaping type quantizer output for the first noise shaping type quantizer. A differentiator that differentiates according to the shaping order of the atomizer;
Means for adding the differentiator output to the output of the local quantizer so that the quantization error generated by the first local quantizer is canceled, and the addition result is taken out as an output. Based on the state of the differentiator, the upper limit value of the output of the first local quantizer,
A quantizer configured to control a lower bound. 2. A differentiator is composed of a delayer and a subtractor for subtracting a delayer output from a delayer input, and an upper limit value of an output value of the first local quantizer based on a value of the delayer,
The quantizer according to claim 1, wherein the quantizer is configured to change the lower limit value.