Extended quantizer with finer step size
A video bitstream including a current block in a current picture is received. A set of candidate quantization step sizes of a current block is determined from a plurality of sets of candidate quantization step sizes associated with quantization parameters (QPs). Each of the plurality of sets of candidate quantization step sizes includes a different number of the candidate quantization step sizes. One or more quantization step sizes are determined from the set of candidate quantization step sizes for the current block. Transform coefficients of the current block are determined based on the determined one or more quantization step sizes. The current block is reconstructed based on the determined transform coefficients of the current block.
1 . A method of video decoding, the method comprising:
receiving a video bitstream including a current block in a current picture;
determining a set of candidate quantization step sizes of the current block from a plurality of sets of candidate quantization step sizes associated with quantization parameters (QPs), each of the plurality of sets of candidate quantization step sizes including a different number of the candidate quantization step sizes that is N times 64, N being an integer greater than 1;
determining one or more quantization step sizes from the set of candidate quantization step sizes for the current block;
determining transform coefficients of the current block based on the determined one or more quantization step sizes; and
reconstructing the current block based on the determined transform coefficients of the current block.
2 . The method of claim 1 , further comprising:
based on a first quantization parameter of the quantization parameters being equal to a multiple of a second quantization parameter of the quantization parameters of the current picture,
determining that a first quantization step size of the set of candidate quantization step sizes associated with the first quantization parameter is equal to a second quantization step size of the set of candidate quantization step sizes associated with the second quantization parameter.
3 . The method of claim 1 , wherein a number of the candidate quantization step sizes of the set of candidate quantization step sizes for the current block is N times 64, N being an integer greater than 1.
4 . The method of claim 1 , wherein:
the determining the transform coefficients includes determining a first transform coefficient of the transform coefficients of the current block based on a product of a first quantized transform coefficient and a first value obtained from a two-dimensional array,
row indices in a row dimension of the two-dimensional array include zero and one, and
a number of column indices in a column dimension of the two-dimensional array is equal to a multiple of six.
5 . The method of claim 4 , wherein the first value is obtained based on an entry value of the two-dimensional array that is indicated by a row index of the row indices and a column index of the column indices, the row index being determined based on a flag, the column index being determined based on a modulo operation that is performed based on (i) a sum of a quantization parameter of the current block and one and (ii) the multiple of six, a value of the quantization parameter of the current block being within 0 and a multiple of 64 minus 1.
6 . The method of claim 1 , wherein the determining the set of candidate quantization step sizes further comprises:
determining a set of initial quantization step sizes; and
interpolating one or more quantization step sizes between at least two adjacent initial quantization step sizes of the set of initial quantization step sizes based on one of a linear interpolation and a non-linear model.
7 . The method of claim 1 , wherein the set of candidate quantization step sizes includes a first subset of candidate quantization step sizes for a luma component of the current block and a second subset of candidate quantization step sizes for a chroma component of the current block, a number of the first subset of candidate quantization step sizes being different from a number of the second subset of candidate quantization step sizes.
8 . The method of claim 1 , further comprising:
determining a first set of quantization parameters for a first one of the current picture, a tile of the current picture, a CTU of the current picture, and the current block of the current picture; and
determining a second set of quantization parameters for a second one of the current picture, a tile of the current picture, a CTU of the current picture, and the current block of the current picture, wherein:
a number of the first set of quantization parameters is different from a number of the second set of quantization parameters.
9 . The method of claim 1 , wherein:
candidate values of quantization parameters associated with a luma component of the current block are in a first range, and
candidate values of quantization parameter offsets between the luma component and a chroma component of the current block are in a second range that is different from the first range.
10 . The method of claim 1 , wherein:
the determining the set of candidate quantization step sizes further includes determining a first set of candidate quantization step sizes and a second set of candidate quantization step sizes for the current block, and
a first clipping range of candidate values of quantization parameter offsets associated with the first set of candidate quantization step sizes is different from a second clipping range of candidate values of quantization parameter offsets associated with the second set of candidate quantization step sizes.
11 . The method of claim 1 , wherein:
the determining the set of candidate quantization step sizes includes determining a first set of candidate quantization step sizes and a second set of candidate quantization step sizes for the current block, and
a first range of candidate offset values associated with the first set of candidate quantization step sizes that is used in a deblock filter is different from a second range of candidate offset values associated with the second set of candidate quantization step sizes that is used in the deblock filter.
12 . The method of claim 1 , wherein:
the determining the set of candidate quantization step sizes further includes determining a first set of candidate quantization step sizes associated with a first coding mode and a second set of candidate quantization step sizes associated with a second coding mode for the current block, and
a first range of quantization parameter offset values associated with the first set of candidate quantization step sizes is different from a second range of quantization parameter offset values associated with the second set of candidate quantization step sizes.
13 . The method of claim 7 , wherein:
the number of the first subset of candidate quantization step sizes is equal to a first value, the number of the second subset of candidate quantization step sizes is equal to a second value, a first QP bitdepth offset value associated with the luma component is QpBDOffsetY, and a second QP bitdepth offset value associated with the chroma component is QpBDOffsetC,
an input range of a piecewise linear model for a luma-to-chroma QP mapping table is [−QpBdOffsetY, first value−1], and
an output range of the piecewise linear model for the luma-to-chroma QP mapping table is [−QpBdOffsetC, second value−1].
14 . A method of video encoding, the method comprising:
determining a set of candidate quantization step sizes of a current block in a current picture from a plurality of sets of candidate quantization step sizes associated with quantization parameters (QPs), each of the plurality of sets of candidate quantization step sizes including a different number of the candidate quantization step sizes that is N times 64, N being an integer greater than 1;
determining one or more quantization step sizes from the set of candidate quantization step sizes for the current block;
determining transform coefficients of the current block based on the determined one or more quantization step sizes; and
encoding the current block in a bitstream based on the determined transform coefficients of the current block.
15 . The method of claim 14 , further comprising:
based on a first quantization parameter of the quantization parameters being equal to a multiple of a second quantization parameter of the quantization parameters of the current picture,
determining that a first quantization step size of the set of candidate quantization step sizes associated with the first quantization parameter is equal to a second quantization step size of the set of candidate quantization step sizes associated with the second quantization parameter.
16 . The method of claim 14 , wherein the set of candidate quantization step sizes includes a first subset of candidate quantization step sizes for a luma component of the current block and a second subset of candidate quantization step sizes for a chroma component of the current block, a number of the first subset of candidate quantization step sizes being different from a number of the second subset of candidate quantization step sizes.
17 . The method of claim 14 , further comprising:
determining a first set of quantization parameters for a first one of the current picture, a tile of the current picture, a CTU of the current picture, and the current block of the current picture; and
determining a second set of quantization parameters for a second one of the current picture, a tile of the current picture, a CTU of the current picture, and the current block of the current picture, wherein:
a number of the first set of quantization parameters is different from a number of the second set of quantization parameters.
18 . The method of claim 14 , wherein:
candidate values of quantization parameters associated with a luma component of the current block are in a first range, and
candidate values of quantization parameter offsets between the luma component and a chroma component of the current block are in a second range that is different from the first range.
19 . The method of claim 14 , wherein:
the determining the set of candidate quantization step sizes further includes determining a first set of candidate quantization step sizes and a second set of candidate quantization step sizes for the current block, and
a first clipping range of candidate values of quantization parameter offsets associated with the first set of candidate quantization step sizes is different from a second clipping range of candidate values of quantization parameter offsets associated with the second set of candidate quantization step sizes.
20 . A non-transitory computer-readable storage medium storing instructions which, when executed by a processor, cause the processor to perform a method of encoding a bitstream comprising:
determining a set of candidate quantization step sizes of a current block in a current picture from a plurality of sets of candidate quantization step sizes associated with quantization parameters (QPs), each of the plurality of sets of candidate quantization step sizes including a different number of the candidate quantization step sizes that is N times 64, N being an integer greater than 1;
determining one or more quantization step sizes from the set of candidate quantization step sizes for the current block;
determining transform coefficients of the current block based on the determined one or more quantization step sizes;
encoding the current block in the bitstream based on the determined transform coefficients of the current block; and
transmitting the encoded bitstream.