Film grain parameters adaptation based on viewing environment
Methods, systems, and bitstream syntax are described for metadata signaling and film-grain parameter adaptation based on a viewing environment which may differ from a reference environment. Example adaptation models are provided for viewing parameters that include: ambient room illumination, viewing distance, and pixels per inch in a target display. Example systems include a single reference viewing environment model and a multi-reference viewing environment model supporting adaptation of film-grain model parameters via adaptation functions or interpolation.
1. A method to process film grain metadata, the method comprising:
receiving an input video bitstream and associated input film grain information;
parsing the input film grain information to generate input film grain parameters ( 301 ) for generating film noise for a target display;
accessing measured viewing parameters ( 312 ) for the target display, wherein the viewing parameters comprise an ambient light value, a pixels per inch value, and a viewing distance value;
accessing reference viewing parameters for a reference display;
adjusting ( 315 ) one or more of the input film grain parameters based on the measured viewing parameters and the reference viewing parameters to generate adjusted film grain parameters, wherein the adjusting comprises:
generating ratio values between corresponding parameters in the measured and reference viewing parameters;
adjusting film grain size and/or strength proportionally to a ratio of measured ambient luminance over reference ambient luminance;
adjusting film grain size and/or strength proportionally to a ratio of measured viewing distance over reference viewing distance; and
adjusting film grain size and/or strength proportionally to a ratio of measured pixels per inch over a reference pixels per inch;
generating output film noise based at least on the adjusted film grain parameters;
decoding the input video bitstream to generate decoded video pictures; and
blending ( 320 ) the output film noise with the decoded video pictures to generate output video pictures on the target display.
2. The method of claim 1 , wherein the reference viewing parameters for the reference display are generated ( 310 ) by parsing metadata associated with the input video bitstream.
3. The method of claim 1 , wherein the reference viewing parameters for the reference display are generated based on predetermined values or values recommended by standards BT. 2035 or SMPTE 2080.
4. The method of claim 1 , wherein an adjustment function of film grain noise comprises computing:
f
(
σ
,
L
,
p
,
d
)
=
σ
*
(
a
L
+
b
L
*
L
)
*
(
a
p
+
b
p
*
p
)
*
(
a
d
+
b
d
*
d
)
,
where σ denotes noise standard deviation as determined in the input film grain parameters, L, p, and d denote the ratio values between corresponding parameters in the measured and reference viewing parameters for ambient luminance, pixels per inch, and viewing distance, and a L , b L , a p , b p , a d , and b d denote film grain model adaptation constants.
5. A method to process film grain metadata, the method comprising:
receiving an input video bitstream and accessing two or more sets of associated input film grain information ( 331 ), each set corresponding to a distinct target viewing environment, and each set comprising film grain parameters for generating film noise for a target display;
accessing measured viewing parameters ( 312 ) for the target display;
based on the measured viewing parameters, interpolating ( 325 ) parameters from the two or more sets of input film grain information to generate output film grain parameters,
wherein when there are only two received sets of input film grain information, one corresponding to a lower bound of an environment viewing parameter and one corresponding to an upper bound of the environment viewing parameter, then computing an interpolation function for the P-th film-grain parameter comprises computing
f ( P )=δ* P ( P U )+(1−δ)* P ( P L ),
wherein
δ
=
P
m
-
P
L
P
U
-
P
L
,
where P m denotes a measured value of the environment viewing parameter between P L and P U , P L and P U denote the lower and upper bounds of the environment viewing parameter, and P(P L ) and P(P U ) denote corresponding film grain parameters in the two received sets of input film grain information,
generating output film noise based at least on the output film grain parameters;
decoding the input video bitstream to generate decoded video pictures; and
blending ( 320 ) the output film noise with the decoded video pictures to generate output video pictures on the target display.
6. The method of claim 5 , wherein generating the output film grain parameters comprises:
based on the measured viewing parameters, selecting from the two or more sets of input film grain information a first set and a second set of input film grain information with parameters closest to the measured viewing parameters; and
generating the output film grain parameters by applying an interpolation function based on corresponding parameters in the first set and the second set of the input film grain information.
7. The method of claim 6 , wherein the first set and the second set of the input film grain information are selected as the two sets from the two or more sets of input film grain information for which a distance measure between the measured viewing parameters and corresponding parameters in the two or more sets of input film grain information has the two smallest values of the distance measure.
8. The method of claim 7 , wherein computing the distance measure comprises:
D
(
i
)
=
(
∑
k
=
1
M
w
k
(
P
(
k
)
m
-
P
(
k
)
r
(
i
)
)
2
)
1
/
2
,
wherein w k , k=1, . . . , M, denotes normalized weighting factors in [0, 1] for M viewing environment parameters (M>0), P(k) m denotes a measured value for the k-th viewing parameter, and P(k) r(i) denotes a corresponding viewing parameter value in the i-th film grain information set.
9. The method of claim 8 , wherein computing an interpolation function for the P-th film-grain parameter comprises computing:
f
(
P
)
=
D
2
D
1
+
D
2
P
(
P
1
)
+
D
1
D
1
+
D
2
P
(
P
2
)
,
where D1 and D2 denote distortion values computed for the selected first set and the second set of the input film grain information, and P(P 1 ) and P(P 2 ) denote the values of the P-th film grain parameters defined in the first set and the second set of the input film grain information.
10. The method of claim 7 , wherein the M viewing environment parameters comprise one or more of
an ambient light value in the viewing environment;
a viewing distance from a target display in the viewing environment;
a maximum luminance value of the target display;
pixels per inch in the target display; and
a ratio of pixels per inch in the target display over the viewing distance from the target display.
11. A non-transitory computer-readable storage medium having stored thereon computer-executable instructions for executing with one or more processors a method in accordance with claim 1 .