Fractional quantization step sizes for high bit rates
A video codec uses fractional increments of quantization step size at high bit rates to permit a more continuous variation of quality and/or bit rate as the quantization scale changes. For high bit rate scenarios, the bit stream syntax includes an additional syntax element to specify fractional step increments (e.g., half step) of the normal quantizer scale step sizes.
1. In a computer system that implements a video decoder, a method comprising:
receiving encoded data in a bitstream for at least part of a video sequence; and
decoding the encoded data, including, for each of multiple frames of the video sequence:
decoding a quantizer index for the frame, wherein the quantizer index indicates a picture quantizer scale for the frame;
comparing the quantizer index to a comparison value;
responsive to the quantizer index being less than or equal to the comparison value, decoding a half-step quantization parameter field, the half-step quantization parameter field indicating a half-step increment for the picture quantizer scale for the frame, and selectively adjusting the picture quantizer scale for the frame according to the half-step increment, wherein syntax of the bitstream allows half-step picture quantization step sizes as specified by the quantizer index and the half-step quantization parameter field when the quantizer index is less than or equal to the comparison value but allows only integer-step picture quantization step sizes as specified by the quantizer index when the quantizer index is greater than the comparison value; and
inverse quantizing at least some quantized transform coefficients for the frame.
2. The method of claim 1 , wherein the quantizer index is a 5-bit field.
3. The method of claim 1 , wherein the quantizer index is signaled in a frame header.
4. The method of claim 1 , wherein the half-step quantization parameter field is a 1-bit field.
5. The method of claim 1 , wherein the half-step quantization parameter field is signaled in a frame header.
6. The method of claim 1 , wherein the selectively adjusting the picture quantizer scale for the frame includes:
if the half-step quantization parameter field is 1, adding ½ to the picture quantizer scale for the frame; and
otherwise, the half-step quantization parameter field being 0, not adjusting the picture quantizer scale for the frame.
7. The method of claim 1 , wherein the inverse quantizing includes performing operations according to dequant_coeff=quant_coeff*(2*quant_scale+halfstep), where dequant_coeff is a dequantized transform coefficient, quant_coeff is a quantized transform coefficient, quant_scale is the picture quantizer scale for the frame, and halfstep is the half-step quantization parameter field.
8. The method of claim 1 , wherein the inverse quantizing includes performing operations according to dequant_coeff=quant_coeff*(2* quant_scale+halfstep)+sign(quant_coeff)*quant_scale, where dequant_coeff is a dequantized transform coefficient, quant_coeff is a quantized transform coefficient, quant_scale is the picture quantizer scale for the frame, halfstep is the half-step quantization parameter field, and sign(x) is a function defined as +1 for x≥0, or −1 for x<0.
9. One or more computer-readable memory or storage devices having stored thereon computer-executable instructions for causing one or more processing units, when programmed thereby, to perform operations comprising:
receiving encoded data in a bitstream for at least part of a video sequence; and
decoding the encoded data, including, for each of multiple frames of the video sequence:
decoding a quantizer index for the frame, wherein the quantizer index indicates a picture quantizer scale for the frame;
comparing the quantizer index to a comparison value;
responsive to the quantizer index being less than or equal to the comparison value, decoding a half-step quantization parameter field, the half-step quantization parameter field indicating a half-step increment for the picture quantizer scale for the frame, and selectively adjusting the picture quantizer scale for the frame according to the half-step increment, wherein syntax of the bitstream allows half-step picture quantization step sizes as specified by the quantizer index and the half-step quantization parameter field when the quantizer index is less than or equal to the comparison value but allows only integer-step picture quantization step sizes as specified by the quantizer index when the quantizer index is greater than the comparison value; and
inverse quantizing at least some quantized transform coefficients for the frame.
10. The one or more computer-readable memory or storage devices of claim 9 , wherein the quantizer index is a 5-bit field, and wherein the quantizer index is signaled in a frame header.
11. The one or more computer-readable memory or storage devices of claim 9 , wherein the half-step quantization parameter field is a 1-bit field, and wherein the half-step quantization parameter field is signaled in a frame header.
12. The one or more computer-readable memory or storage devices of claim 9 , wherein the selectively adjusting the picture quantizer scale for the frame includes:
if the half-step quantization parameter field is 1, adding ½ to the picture quantizer scale for the frame; and
otherwise, the half-step quantization parameter field being 0, not adjusting the picture quantizer scale for the frame.
13. The one or more computer-readable memory or storage devices of claim 9 , wherein the inverse quantizing includes performing operations according to dequant_coeff=quant_coeff*(2*quant_scale+halfstep), where dequant_coeff is a dequantized transform coefficient, quant_coeff is a quantized transform coefficient, quant_scale is the picture quantizer scale for the frame, and halfstep is the half-step quantization parameter field.
14. The one or more computer-readable memory or storage devices of claim 9 , wherein the inverse quantizing includes performing operations according to dequant_coeff=quant_coeff*(2*quant_scale+halfstep)+sign(quant_coeff)*quant_scale, where dequant_coeff is a dequantized transform coefficient, quant_coeff is a quantized transform coefficient, quant_scale is the picture quantizer scale for the frame, halfstep is the half-step quantization parameter field, and sign(x) is a function defined as +1 for x≥0, or −1 for x<0.
15. A computer system comprising memory and one or more processing units, wherein the computer system implements a video decoder configured to perform operations comprising:
receiving encoded data in a bitstream for at least part of a video sequence; and
decoding the encoded data, including, for each of multiple frames of the video sequence:
decoding a quantizer index for the frame, wherein the quantizer index indicates a picture quantizer scale for the frame;
comparing the quantizer index to a comparison value;
responsive to the quantizer index being less than or equal to the comparison value, decoding a half-step quantization parameter field, the half-step quantization parameter field indicating a half-step increment for the picture quantizer scale for the frame, and selectively adjusting the picture quantizer scale for the frame according to the half-step increment, wherein syntax of the bitstream allows half-step picture quantization step sizes as specified by the quantizer index and the half-step quantization parameter field when the quantizer index is less than or equal to the comparison value but allows only integer-step picture quantization step sizes as specified by the quantizer index when the quantizer index is greater than the comparison value; and
inverse quantizing at least some quantized transform coefficients for the frame.
16. The computer system of claim 15 , wherein the quantizer index is a 5-bit field, and wherein the quantizer index is signaled in a frame header.
17. The computer system of claim 15 , wherein the half-step quantization parameter field is a 1-bit field, and wherein the half-step quantization parameter field is signaled in a frame header.
18. The computer system of claim 15 , wherein the selectively adjusting the picture quantizer scale for the frame includes:
if the half-step quantization parameter field is 1, adding ½ to the picture quantizer scale for the frame; and
otherwise, the half-step quantization parameter field being 0, not adjusting the picture quantizer scale for the frame.
19. The computer system of claim 15 , wherein the inverse quantizing includes performing operations according to dequant_coeff=quant_coeff*(2*quant_scale+halfstep), where dequant_coeff is a dequantized transform coefficient, quant_coeff is a quantized transform coefficient, quant_scale is the picture quantizer scale for the frame, and halfstep is the half-step quantization parameter field.
20. The computer system of claim 15 , wherein the inverse quantizing includes performing operations according to dequant_coeff=quant_coeff*(2*quant_scale+halfstep)+sign(quant_coeff)*quant_scale, where dequant_coeff is a dequantized transform coefficient, quant_coeff is a quantized transform coefficient, quant_scale is the picture quantizer scale for the frame, halfstep is the half-step quantization parameter field, and sign(x) is a function defined as +1 for x≥0, or −1 for x<0.