[FFmpeg-devel] [PATCH v3 1/7] swscale/range_convert: saturate output instead of limiting input
Ramiro Polla
ramiro.polla at gmail.com
Sat Nov 30 17:23:36 EET 2024
For bit depths <= 14, the result is saturated to 15 bits.
For bit depths > 14, the result is saturated to 19 bits.
x86_64:
chrRangeFromJpeg8_1920_c: 5827.4 5804.5 ( 1.00x)
chrRangeFromJpeg16_1920_c: 5793.2 5792.8 ( 1.00x)
chrRangeToJpeg8_1920_c: 11726.2 9388.6 ( 1.25x)
chrRangeToJpeg16_1920_c: 10610.8 5796.5 ( 1.83x)
lumRangeFromJpeg8_1920_c: 4165.7 4147.9 ( 1.00x)
lumRangeFromJpeg16_1920_c: 4530.0 4529.0 ( 1.00x)
lumRangeToJpeg8_1920_c: 6044.8 5694.1 ( 1.06x)
lumRangeToJpeg16_1920_c: 5343.6 5334.2 ( 1.00x)
aarch64 A55:
chrRangeFromJpeg8_1920_c: 28839.3 28833.8 ( 1.00x)
chrRangeFromJpeg16_1920_c: 28843.8 28842.8 ( 1.00x)
chrRangeToJpeg8_1920_c: 44196.1 23070.6 ( 1.92x)
chrRangeToJpeg16_1920_c: 36526.7 17313.8 ( 2.11x)
lumRangeFromJpeg8_1920_c: 15384.3 15388.1 ( 1.00x)
lumRangeFromJpeg16_1920_c: 15390.1 15388.0 ( 1.00x)
lumRangeToJpeg8_1920_c: 23066.7 19226.2 ( 1.20x)
lumRangeToJpeg16_1920_c: 19224.6 19225.5 ( 1.00x)
aarch64 A76:
chrRangeFromJpeg8_1920_c: 6316.2 6317.8 ( 1.00x)
chrRangeFromJpeg16_1920_c: 6321.9 6322.9 ( 1.00x)
chrRangeToJpeg8_1920_c: 11389.3 9287.1 ( 1.23x)
chrRangeToJpeg16_1920_c: 9514.4 6104.9 ( 1.56x)
lumRangeFromJpeg8_1920_c: 4376.0 4359.1 ( 1.00x)
lumRangeFromJpeg16_1920_c: 4437.9 4358.8 ( 1.02x)
lumRangeToJpeg8_1920_c: 6667.0 5957.2 ( 1.12x)
lumRangeToJpeg16_1920_c: 6062.5 6072.5 ( 1.00x)
NOTE: all simd optimizations for range_convert have been disabled
except for x86, which already had the same behaviour.
they will be re-enabled when they are fixed for each architecture.
---
libswscale/aarch64/swscale.c | 5 +++++
libswscale/loongarch/swscale_init_loongarch.c | 5 +++++
libswscale/riscv/swscale.c | 5 +++++
libswscale/swscale.c | 21 ++++++++++++-------
libswscale/x86/range_convert.asm | 3 ---
5 files changed, 29 insertions(+), 10 deletions(-)
diff --git a/libswscale/aarch64/swscale.c b/libswscale/aarch64/swscale.c
index 5173359e09..1fce77df26 100644
--- a/libswscale/aarch64/swscale.c
+++ b/libswscale/aarch64/swscale.c
@@ -225,6 +225,10 @@ void ff_chrRangeToJpeg_neon(int16_t *dstU, int16_t *dstV, int width);
av_cold void ff_sws_init_range_convert_aarch64(SwsInternal *c)
{
+ /* This code is currently disabled because of changes in the base
+ * implementation of these functions. This code should be enabled
+ * again once those changes are ported to this architecture. */
+#if 0
int cpu_flags = av_get_cpu_flags();
if (have_neon(cpu_flags)) {
@@ -238,6 +242,7 @@ av_cold void ff_sws_init_range_convert_aarch64(SwsInternal *c)
}
}
}
+#endif
}
av_cold void ff_sws_init_swscale_aarch64(SwsInternal *c)
diff --git a/libswscale/loongarch/swscale_init_loongarch.c b/libswscale/loongarch/swscale_init_loongarch.c
index 5204a8b66d..f0b8a7db12 100644
--- a/libswscale/loongarch/swscale_init_loongarch.c
+++ b/libswscale/loongarch/swscale_init_loongarch.c
@@ -26,6 +26,10 @@
av_cold void ff_sws_init_range_convert_loongarch(SwsInternal *c)
{
+ /* This code is currently disabled because of changes in the base
+ * implementation of these functions. This code should be enabled
+ * again once those changes are ported to this architecture. */
+#if 0
int cpu_flags = av_get_cpu_flags();
if (have_lsx(cpu_flags)) {
@@ -52,6 +56,7 @@ av_cold void ff_sws_init_range_convert_loongarch(SwsInternal *c)
}
}
#endif // #if HAVE_LASX
+#endif
}
av_cold void ff_sws_init_swscale_loongarch(SwsInternal *c)
diff --git a/libswscale/riscv/swscale.c b/libswscale/riscv/swscale.c
index 3bdaf20af4..49c492f153 100644
--- a/libswscale/riscv/swscale.c
+++ b/libswscale/riscv/swscale.c
@@ -28,6 +28,10 @@ void ff_range_chr_from_jpeg_16_rvv(int16_t *, int16_t *, int);
av_cold void ff_sws_init_range_convert_riscv(SwsInternal *c)
{
+ /* This code is currently disabled because of changes in the base
+ * implementation of these functions. This code should be enabled
+ * again once those changes are ported to this architecture. */
+#if 0
#if HAVE_RVV
int flags = av_get_cpu_flags();
@@ -47,6 +51,7 @@ av_cold void ff_sws_init_range_convert_riscv(SwsInternal *c)
c->chrConvertRange = convs[from].chr;
}
#endif
+#endif
}
#define RVV_INPUT(name) \
diff --git a/libswscale/swscale.c b/libswscale/swscale.c
index 3c4637c0a1..048c1765c8 100644
--- a/libswscale/swscale.c
+++ b/libswscale/swscale.c
@@ -160,8 +160,10 @@ static void chrRangeToJpeg_c(int16_t *dstU, int16_t *dstV, int width)
{
int i;
for (i = 0; i < width; i++) {
- dstU[i] = (FFMIN(dstU[i], 30775) * 4663 - 9289992) >> 12; // -264
- dstV[i] = (FFMIN(dstV[i], 30775) * 4663 - 9289992) >> 12; // -264
+ int U = (dstU[i] * 4663 - 9289992) >> 12; // -264
+ int V = (dstV[i] * 4663 - 9289992) >> 12; // -264
+ dstU[i] = FFMIN(U, (1 << 15) - 1);
+ dstV[i] = FFMIN(V, (1 << 15) - 1);
}
}
@@ -177,8 +179,10 @@ static void chrRangeFromJpeg_c(int16_t *dstU, int16_t *dstV, int width)
static void lumRangeToJpeg_c(int16_t *dst, int width)
{
int i;
- for (i = 0; i < width; i++)
- dst[i] = (FFMIN(dst[i], 30189) * 19077 - 39057361) >> 14;
+ for (i = 0; i < width; i++) {
+ int Y = (dst[i] * 19077 - 39057361) >> 14;
+ dst[i] = FFMIN(Y, (1 << 15) - 1);
+ }
}
static void lumRangeFromJpeg_c(int16_t *dst, int width)
@@ -194,8 +198,10 @@ static void chrRangeToJpeg16_c(int16_t *_dstU, int16_t *_dstV, int width)
int32_t *dstU = (int32_t *) _dstU;
int32_t *dstV = (int32_t *) _dstV;
for (i = 0; i < width; i++) {
- dstU[i] = ((int)(FFMIN(dstU[i], 30775 << 4) * 4663U - (9289992 << 4))) >> 12; // -264
- dstV[i] = ((int)(FFMIN(dstV[i], 30775 << 4) * 4663U - (9289992 << 4))) >> 12; // -264
+ int U = (dstU[i] * 4663U - (9289992 << 4)) >> 12; // -264
+ int V = (dstV[i] * 4663U - (9289992 << 4)) >> 12; // -264
+ dstU[i] = FFMIN(U, (1 << 19) - 1);
+ dstV[i] = FFMIN(V, (1 << 19) - 1);
}
}
@@ -215,7 +221,8 @@ static void lumRangeToJpeg16_c(int16_t *_dst, int width)
int i;
int32_t *dst = (int32_t *) _dst;
for (i = 0; i < width; i++) {
- dst[i] = ((int)(FFMIN(dst[i], 30189 << 4) * 4769U - (39057361 << 2))) >> 12;
+ int Y = ((int)(FFMIN(dst[i], 30189 << 4) * 4769U - (39057361 << 2))) >> 12;
+ dst[i] = FFMIN(Y, (1 << 19) - 1);
}
}
diff --git a/libswscale/x86/range_convert.asm b/libswscale/x86/range_convert.asm
index 97c7525448..ffda009c4e 100644
--- a/libswscale/x86/range_convert.asm
+++ b/libswscale/x86/range_convert.asm
@@ -40,9 +40,6 @@ lum_from_offset: times 4 dd 33561947
SECTION .text
-; NOTE: there is no need to clamp the input when converting to jpeg range
-; (like we do in the C code) because packssdw will saturate the output.
-
;-----------------------------------------------------------------------------
; lumConvertRange
;
--
2.39.5
More information about the ffmpeg-devel
mailing list