* [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec
@ 2022-06-29 10:12 Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 1/6] fflcms2: move to libavcodec Niklas Haas
` (6 more replies)
0 siblings, 7 replies; 13+ messages in thread
From: Niklas Haas @ 2022-06-29 10:12 UTC (permalink / raw)
To: ffmpeg-devel
Changes in v2:
- remove unnecessary import
- fixed assignment-instead-of-equality
- fixed #ifdef -> #if
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* [FFmpeg-devel] [PATCH v2 1/6] fflcms2: move to libavcodec
2022-06-29 10:12 [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
@ 2022-06-29 10:12 ` Niklas Haas
2022-07-06 14:18 ` Andreas Rheinhardt
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 2/6] avcodec/codec_internal: add cap for ICC profile support Niklas Haas
` (5 subsequent siblings)
6 siblings, 1 reply; 13+ messages in thread
From: Niklas Haas @ 2022-06-29 10:12 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Niklas Haas
From: Niklas Haas <git@haasn.dev>
We will need this helper inside libavcodec in the future, so move it
there, leaving behind an #include to the raw source file in its old
location in libvfilter. This approach is inspired by the handling of
vulkan.c, and avoids us needing to expose any of it publicly (or
semi-publicly) in e.g. libavutil, thus avoiding any ABI headaches.
It's debatable whether the actual code belongs in libavcodec or
libavfilter, but I decided to put it into libavcodec because it
conceptually deals with encoding and decoding ICC profiles, and will be
used to decode embedded ICC profiles in image files.
Signed-off-by: Niklas Haas <git@haasn.dev>
---
libavcodec/Makefile | 1 +
libavcodec/fflcms2.c | 311 ++++++++++++++++++++++++++++++++++++++++++
libavcodec/fflcms2.h | 87 ++++++++++++
libavfilter/fflcms2.c | 294 +--------------------------------------
libavfilter/fflcms2.h | 65 +--------
5 files changed, 401 insertions(+), 357 deletions(-)
create mode 100644 libavcodec/fflcms2.c
create mode 100644 libavcodec/fflcms2.h
diff --git a/libavcodec/Makefile b/libavcodec/Makefile
index 050934101c..5c4f62c631 100644
--- a/libavcodec/Makefile
+++ b/libavcodec/Makefile
@@ -1235,6 +1235,7 @@ SKIPHEADERS-$(CONFIG_AMF) += amfenc.h
SKIPHEADERS-$(CONFIG_D3D11VA) += d3d11va.h dxva2_internal.h
SKIPHEADERS-$(CONFIG_DXVA2) += dxva2.h dxva2_internal.h
SKIPHEADERS-$(CONFIG_JNI) += ffjni.h
+SKIPHEADERS-$(CONFIG_LCMS2) += fflcms2.h
SKIPHEADERS-$(CONFIG_LIBJXL) += libjxl.h
SKIPHEADERS-$(CONFIG_LIBVPX) += libvpx.h
SKIPHEADERS-$(CONFIG_LIBWEBP_ENCODER) += libwebpenc_common.h
diff --git a/libavcodec/fflcms2.c b/libavcodec/fflcms2.c
new file mode 100644
index 0000000000..fd370fb310
--- /dev/null
+++ b/libavcodec/fflcms2.c
@@ -0,0 +1,311 @@
+/*
+ * Copyright (c) 2022 Niklas Haas
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "libavutil/color_utils.h"
+#include "libavutil/csp.h"
+
+#include "fflcms2.h"
+
+static void log_cb(cmsContext ctx, cmsUInt32Number error, const char *str)
+{
+ FFIccContext *s = cmsGetContextUserData(ctx);
+ av_log(s->avctx, AV_LOG_ERROR, "lcms2: [%"PRIu32"] %s\n", error, str);
+}
+
+int ff_icc_context_init(FFIccContext *s, void *avctx)
+{
+ memset(s, 0, sizeof(*s));
+ s->avctx = avctx;
+ s->ctx = cmsCreateContext(NULL, s);
+ if (!s->ctx)
+ return AVERROR(ENOMEM);
+
+ cmsSetLogErrorHandlerTHR(s->ctx, log_cb);
+ return 0;
+}
+
+void ff_icc_context_uninit(FFIccContext *s)
+{
+ for (int i = 0; i < FF_ARRAY_ELEMS(s->curves); i++)
+ cmsFreeToneCurve(s->curves[i]);
+ cmsDeleteContext(s->ctx);
+ memset(s, 0, sizeof(*s));
+}
+
+static int get_curve(FFIccContext *s, enum AVColorTransferCharacteristic trc,
+ cmsToneCurve **out_curve)
+{
+ if (trc >= AVCOL_TRC_NB)
+ return AVERROR_INVALIDDATA;
+
+ if (s->curves[trc])
+ goto done;
+
+ switch (trc) {
+ case AVCOL_TRC_LINEAR:
+ s->curves[trc] = cmsBuildGamma(s->ctx, 1.0);
+ break;
+ case AVCOL_TRC_GAMMA22:
+ s->curves[trc] = cmsBuildGamma(s->ctx, 2.2);
+ break;
+ case AVCOL_TRC_GAMMA28:
+ s->curves[trc] = cmsBuildGamma(s->ctx, 2.8);
+ break;
+ case AVCOL_TRC_BT709:
+ case AVCOL_TRC_SMPTE170M:
+ case AVCOL_TRC_BT2020_10:
+ case AVCOL_TRC_BT2020_12:
+ s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 4, (double[5]) {
+ /* γ = */ 1/0.45,
+ /* a = */ 1/1.099296826809442,
+ /* b = */ 1 - 1/1.099296826809442,
+ /* c = */ 1/4.5,
+ /* d = */ 4.5 * 0.018053968510807,
+ });
+ break;
+ case AVCOL_TRC_SMPTE240M:
+ s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 4, (double[5]) {
+ /* γ = */ 1/0.45,
+ /* a = */ 1/1.1115,
+ /* b = */ 1 - 1/1.1115,
+ /* c = */ 1/4.0,
+ /* d = */ 4.0 * 0.0228,
+ });
+ break;
+ case AVCOL_TRC_LOG:
+ s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 8, (double[5]) {
+ /* a = */ 1.0,
+ /* b = */ 10.0,
+ /* c = */ 2.0,
+ /* d = */ -1.0,
+ /* e = */ 0.0
+ });
+ break;
+ case AVCOL_TRC_LOG_SQRT:
+ s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 8, (double[5]) {
+ /* a = */ 1.0,
+ /* b = */ 10.0,
+ /* c = */ 2.5,
+ /* d = */ -1.0,
+ /* e = */ 0.0
+ });
+ break;
+ case AVCOL_TRC_IEC61966_2_1:
+ s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 4, (double[5]) {
+ /* γ = */ 2.4,
+ /* a = */ 1/1.055,
+ /* b = */ 1 - 1/1.055,
+ /* c = */ 1/12.92,
+ /* d = */ 12.92 * 0.0031308,
+ });
+ break;
+ case AVCOL_TRC_SMPTE428:
+ s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 2, (double[3]) {
+ /* γ = */ 2.6,
+ /* a = */ pow(52.37/48.0, 1/2.6),
+ /* b = */ 0.0
+ });
+ break;
+
+ /* Can't be represented using the existing parametric tone curves.
+ * FIXME: use cmsBuildTabulatedToneCurveFloat instead */
+ case AVCOL_TRC_IEC61966_2_4:
+ case AVCOL_TRC_BT1361_ECG:
+ case AVCOL_TRC_SMPTE2084:
+ case AVCOL_TRC_ARIB_STD_B67:
+ return AVERROR_PATCHWELCOME;
+
+ default:
+ return AVERROR_INVALIDDATA;
+ }
+
+ if (!s->curves[trc])
+ return AVERROR(ENOMEM);
+
+done:
+ *out_curve = s->curves[trc];
+ return 0;
+}
+
+int ff_icc_profile_generate(FFIccContext *s,
+ enum AVColorPrimaries color_prim,
+ enum AVColorTransferCharacteristic color_trc,
+ cmsHPROFILE *out_profile)
+{
+ cmsToneCurve *tonecurve;
+ const AVColorPrimariesDesc *prim;
+ int ret;
+
+ if (!(prim = av_csp_primaries_desc_from_id(color_prim)))
+ return AVERROR_INVALIDDATA;
+ if ((ret = get_curve(s, color_trc, &tonecurve)) < 0)
+ return ret;
+
+ *out_profile = cmsCreateRGBProfileTHR(s->ctx,
+ &(cmsCIExyY) { av_q2d(prim->wp.x), av_q2d(prim->wp.y), 1.0 },
+ &(cmsCIExyYTRIPLE) {
+ .Red = { av_q2d(prim->prim.r.x), av_q2d(prim->prim.r.y), 1.0 },
+ .Green = { av_q2d(prim->prim.g.x), av_q2d(prim->prim.g.y), 1.0 },
+ .Blue = { av_q2d(prim->prim.b.x), av_q2d(prim->prim.b.y), 1.0 },
+ },
+ (cmsToneCurve *[3]) { tonecurve, tonecurve, tonecurve }
+ );
+
+ return *out_profile == NULL ? AVERROR(ENOMEM) : 0;
+}
+
+int ff_icc_profile_attach(FFIccContext *s, cmsHPROFILE profile, AVFrame *frame)
+{
+ cmsUInt32Number size;
+ AVBufferRef *buf;
+
+ if (!cmsSaveProfileToMem(profile, NULL, &size))
+ return AVERROR_EXTERNAL;
+
+ buf = av_buffer_alloc(size);
+ if (!buf)
+ return AVERROR(ENOMEM);
+
+ if (!cmsSaveProfileToMem(profile, buf->data, &size) || size != buf->size) {
+ av_buffer_unref(&buf);
+ return AVERROR_EXTERNAL;
+ }
+
+ if (!av_frame_new_side_data_from_buf(frame, AV_FRAME_DATA_ICC_PROFILE, buf)) {
+ av_buffer_unref(&buf);
+ return AVERROR(ENOMEM);
+ }
+
+ return 0;
+}
+
+static av_always_inline void XYZ_xy(cmsCIEXYZ XYZ, AVCIExy *xy)
+{
+ double k = 1.0 / (XYZ.X + XYZ.Y + XYZ.Z);
+ xy->x = av_d2q(k * XYZ.X, 100000);
+ xy->y = av_d2q(k * XYZ.Y, 100000);
+}
+
+int ff_icc_profile_read_primaries(FFIccContext *s, cmsHPROFILE profile,
+ AVColorPrimariesDesc *out_primaries)
+{
+ static const uint8_t testprimaries[4][3] = {
+ { 0xFF, 0, 0 }, /* red */
+ { 0, 0xFF, 0 }, /* green */
+ { 0, 0, 0xFF }, /* blue */
+ { 0xFF, 0xFF, 0xFF }, /* white */
+ };
+
+ AVWhitepointCoefficients *wp = &out_primaries->wp;
+ AVPrimaryCoefficients *prim = &out_primaries->prim;
+ cmsFloat64Number prev_adapt;
+ cmsHPROFILE xyz;
+ cmsHTRANSFORM tf;
+ cmsCIEXYZ dst[4];
+
+ xyz = cmsCreateXYZProfileTHR(s->ctx);
+ if (!xyz)
+ return AVERROR(ENOMEM);
+
+ /* We need to use an unadapted observer to get the raw values */
+ prev_adapt = cmsSetAdaptationStateTHR(s->ctx, 0.0);
+ tf = cmsCreateTransformTHR(s->ctx, profile, TYPE_RGB_8, xyz, TYPE_XYZ_DBL,
+ INTENT_ABSOLUTE_COLORIMETRIC,
+ /* Note: These flags mostly don't do anything
+ * anyway, but specify them regardless */
+ cmsFLAGS_NOCACHE |
+ cmsFLAGS_NOOPTIMIZE |
+ cmsFLAGS_LOWRESPRECALC |
+ cmsFLAGS_GRIDPOINTS(2));
+ cmsSetAdaptationStateTHR(s->ctx, prev_adapt);
+ cmsCloseProfile(xyz);
+ if (!tf) {
+ av_log(s->avctx, AV_LOG_ERROR, "Invalid ICC profile (e.g. CMYK)\n");
+ return AVERROR_INVALIDDATA;
+ }
+
+ cmsDoTransform(tf, testprimaries, dst, 4);
+ cmsDeleteTransform(tf);
+ XYZ_xy(dst[0], &prim->r);
+ XYZ_xy(dst[1], &prim->g);
+ XYZ_xy(dst[2], &prim->b);
+ XYZ_xy(dst[3], wp);
+ return 0;
+}
+
+int ff_icc_profile_detect_transfer(FFIccContext *s, cmsHPROFILE profile,
+ enum AVColorTransferCharacteristic *out_trc)
+{
+ /* 8-bit linear grayscale ramp */
+ static const uint8_t testramp[16][3] = {
+ { 1, 1, 1}, /* avoid exact zero due to log100 etc. */
+ { 17, 17, 17},
+ { 34, 34, 34},
+ { 51, 51, 51},
+ { 68, 68, 68},
+ { 85, 85, 85},
+ { 02, 02, 02},
+ {119, 119, 119},
+ {136, 136, 136},
+ {153, 153, 153},
+ {170, 170, 170},
+ {187, 187, 187},
+ {204, 204, 204},
+ {221, 221, 221},
+ {238, 238, 238},
+ {255, 255, 255},
+ };
+
+ double dst[FF_ARRAY_ELEMS(testramp)];
+
+ for (enum AVColorTransferCharacteristic trc = 0; trc < AVCOL_TRC_NB; trc++) {
+ cmsToneCurve *tonecurve;
+ cmsHPROFILE ref;
+ cmsHTRANSFORM tf;
+ double delta = 0.0;
+ if (get_curve(s, trc, &tonecurve) < 0)
+ continue;
+
+ ref = cmsCreateGrayProfileTHR(s->ctx, cmsD50_xyY(), tonecurve);
+ if (!ref)
+ return AVERROR(ENOMEM);
+
+ tf = cmsCreateTransformTHR(s->ctx, profile, TYPE_RGB_8, ref, TYPE_GRAY_DBL,
+ INTENT_RELATIVE_COLORIMETRIC,
+ cmsFLAGS_NOCACHE | cmsFLAGS_NOOPTIMIZE);
+ cmsCloseProfile(ref);
+ if (!tf) {
+ av_log(s->avctx, AV_LOG_ERROR, "Invalid ICC profile (e.g. CMYK)\n");
+ return AVERROR_INVALIDDATA;
+ }
+
+ cmsDoTransform(tf, testramp, dst, FF_ARRAY_ELEMS(dst));
+ cmsDeleteTransform(tf);
+
+ for (int i = 0; i < FF_ARRAY_ELEMS(dst); i++)
+ delta += fabs(testramp[i][0] / 255.0 - dst[i]);
+ if (delta < 0.01) {
+ *out_trc = trc;
+ return 0;
+ }
+ }
+
+ *out_trc = AVCOL_TRC_UNSPECIFIED;
+ return 0;
+}
diff --git a/libavcodec/fflcms2.h b/libavcodec/fflcms2.h
new file mode 100644
index 0000000000..af63c9a13c
--- /dev/null
+++ b/libavcodec/fflcms2.h
@@ -0,0 +1,87 @@
+/*
+ * Copyright (c) 2022 Niklas Haas
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/**
+ * @file
+ * Various functions for dealing with ICC profiles
+ */
+
+#ifndef AVCODEC_FFLCMS2_H
+#define AVCODEC_FFLCMS2_H
+
+#include "libavutil/csp.h"
+#include "libavutil/frame.h"
+#include "libavutil/pixfmt.h"
+
+#include <lcms2.h>
+
+typedef struct FFIccContext {
+ void *avctx;
+ cmsContext ctx;
+ cmsToneCurve *curves[AVCOL_TRC_NB]; /* tone curve cache */
+} FFIccContext;
+
+/**
+ * Initializes an FFIccContext. This must be done prior to using it.
+ *
+ * Returns 0 on success, or a negative error code.
+ */
+int ff_icc_context_init(FFIccContext *s, void *avctx);
+void ff_icc_context_uninit(FFIccContext *s);
+
+/**
+ * Generate an ICC profile for a given combination of color primaries and
+ * transfer function. Both values must be set to valid entries (not
+ * "undefined") for this function to work.
+ *
+ * Returns 0 on success, or a negative error code.
+ */
+int ff_icc_profile_generate(FFIccContext *s,
+ enum AVColorPrimaries color_prim,
+ enum AVColorTransferCharacteristic color_trc,
+ cmsHPROFILE *out_profile);
+
+/**
+ * Attach an ICC profile to a frame. Helper wrapper around cmsSaveProfileToMem
+ * and av_frame_new_side_data_from_buf.
+ *
+ * Returns 0 on success, or a negative error code.
+ */
+int ff_icc_profile_attach(FFIccContext *s, cmsHPROFILE profile, AVFrame *frame);
+
+/**
+ * Read the color primaries and white point coefficients encoded by an ICC
+ * profile, and return the raw values in `out_primaries`.
+ *
+ * Returns 0 on success, or a negative error code.
+ */
+int ff_icc_profile_read_primaries(FFIccContext *s, cmsHPROFILE profile,
+ AVColorPrimariesDesc *out_primaries);
+
+/**
+ * Attempt detecting the transfer characteristic that best approximates the
+ * transfer function encoded by an ICC profile. Sets `out_trc` to
+ * AVCOL_TRC_UNSPECIFIED if no clear match can be identified.
+ *
+ * Returns 0 on success (including no match), or a negative error code.
+ */
+int ff_icc_profile_detect_transfer(FFIccContext *s, cmsHPROFILE profile,
+ enum AVColorTransferCharacteristic *out_trc);
+
+#endif /* AVCODEC_FFLCMS2_H */
diff --git a/libavfilter/fflcms2.c b/libavfilter/fflcms2.c
index fd370fb310..822462de87 100644
--- a/libavfilter/fflcms2.c
+++ b/libavfilter/fflcms2.c
@@ -1,5 +1,4 @@
/*
- * Copyright (c) 2022 Niklas Haas
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
@@ -17,295 +16,4 @@
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
-#include "libavutil/color_utils.h"
-#include "libavutil/csp.h"
-
-#include "fflcms2.h"
-
-static void log_cb(cmsContext ctx, cmsUInt32Number error, const char *str)
-{
- FFIccContext *s = cmsGetContextUserData(ctx);
- av_log(s->avctx, AV_LOG_ERROR, "lcms2: [%"PRIu32"] %s\n", error, str);
-}
-
-int ff_icc_context_init(FFIccContext *s, void *avctx)
-{
- memset(s, 0, sizeof(*s));
- s->avctx = avctx;
- s->ctx = cmsCreateContext(NULL, s);
- if (!s->ctx)
- return AVERROR(ENOMEM);
-
- cmsSetLogErrorHandlerTHR(s->ctx, log_cb);
- return 0;
-}
-
-void ff_icc_context_uninit(FFIccContext *s)
-{
- for (int i = 0; i < FF_ARRAY_ELEMS(s->curves); i++)
- cmsFreeToneCurve(s->curves[i]);
- cmsDeleteContext(s->ctx);
- memset(s, 0, sizeof(*s));
-}
-
-static int get_curve(FFIccContext *s, enum AVColorTransferCharacteristic trc,
- cmsToneCurve **out_curve)
-{
- if (trc >= AVCOL_TRC_NB)
- return AVERROR_INVALIDDATA;
-
- if (s->curves[trc])
- goto done;
-
- switch (trc) {
- case AVCOL_TRC_LINEAR:
- s->curves[trc] = cmsBuildGamma(s->ctx, 1.0);
- break;
- case AVCOL_TRC_GAMMA22:
- s->curves[trc] = cmsBuildGamma(s->ctx, 2.2);
- break;
- case AVCOL_TRC_GAMMA28:
- s->curves[trc] = cmsBuildGamma(s->ctx, 2.8);
- break;
- case AVCOL_TRC_BT709:
- case AVCOL_TRC_SMPTE170M:
- case AVCOL_TRC_BT2020_10:
- case AVCOL_TRC_BT2020_12:
- s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 4, (double[5]) {
- /* γ = */ 1/0.45,
- /* a = */ 1/1.099296826809442,
- /* b = */ 1 - 1/1.099296826809442,
- /* c = */ 1/4.5,
- /* d = */ 4.5 * 0.018053968510807,
- });
- break;
- case AVCOL_TRC_SMPTE240M:
- s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 4, (double[5]) {
- /* γ = */ 1/0.45,
- /* a = */ 1/1.1115,
- /* b = */ 1 - 1/1.1115,
- /* c = */ 1/4.0,
- /* d = */ 4.0 * 0.0228,
- });
- break;
- case AVCOL_TRC_LOG:
- s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 8, (double[5]) {
- /* a = */ 1.0,
- /* b = */ 10.0,
- /* c = */ 2.0,
- /* d = */ -1.0,
- /* e = */ 0.0
- });
- break;
- case AVCOL_TRC_LOG_SQRT:
- s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 8, (double[5]) {
- /* a = */ 1.0,
- /* b = */ 10.0,
- /* c = */ 2.5,
- /* d = */ -1.0,
- /* e = */ 0.0
- });
- break;
- case AVCOL_TRC_IEC61966_2_1:
- s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 4, (double[5]) {
- /* γ = */ 2.4,
- /* a = */ 1/1.055,
- /* b = */ 1 - 1/1.055,
- /* c = */ 1/12.92,
- /* d = */ 12.92 * 0.0031308,
- });
- break;
- case AVCOL_TRC_SMPTE428:
- s->curves[trc] = cmsBuildParametricToneCurve(s->ctx, 2, (double[3]) {
- /* γ = */ 2.6,
- /* a = */ pow(52.37/48.0, 1/2.6),
- /* b = */ 0.0
- });
- break;
-
- /* Can't be represented using the existing parametric tone curves.
- * FIXME: use cmsBuildTabulatedToneCurveFloat instead */
- case AVCOL_TRC_IEC61966_2_4:
- case AVCOL_TRC_BT1361_ECG:
- case AVCOL_TRC_SMPTE2084:
- case AVCOL_TRC_ARIB_STD_B67:
- return AVERROR_PATCHWELCOME;
-
- default:
- return AVERROR_INVALIDDATA;
- }
-
- if (!s->curves[trc])
- return AVERROR(ENOMEM);
-
-done:
- *out_curve = s->curves[trc];
- return 0;
-}
-
-int ff_icc_profile_generate(FFIccContext *s,
- enum AVColorPrimaries color_prim,
- enum AVColorTransferCharacteristic color_trc,
- cmsHPROFILE *out_profile)
-{
- cmsToneCurve *tonecurve;
- const AVColorPrimariesDesc *prim;
- int ret;
-
- if (!(prim = av_csp_primaries_desc_from_id(color_prim)))
- return AVERROR_INVALIDDATA;
- if ((ret = get_curve(s, color_trc, &tonecurve)) < 0)
- return ret;
-
- *out_profile = cmsCreateRGBProfileTHR(s->ctx,
- &(cmsCIExyY) { av_q2d(prim->wp.x), av_q2d(prim->wp.y), 1.0 },
- &(cmsCIExyYTRIPLE) {
- .Red = { av_q2d(prim->prim.r.x), av_q2d(prim->prim.r.y), 1.0 },
- .Green = { av_q2d(prim->prim.g.x), av_q2d(prim->prim.g.y), 1.0 },
- .Blue = { av_q2d(prim->prim.b.x), av_q2d(prim->prim.b.y), 1.0 },
- },
- (cmsToneCurve *[3]) { tonecurve, tonecurve, tonecurve }
- );
-
- return *out_profile == NULL ? AVERROR(ENOMEM) : 0;
-}
-
-int ff_icc_profile_attach(FFIccContext *s, cmsHPROFILE profile, AVFrame *frame)
-{
- cmsUInt32Number size;
- AVBufferRef *buf;
-
- if (!cmsSaveProfileToMem(profile, NULL, &size))
- return AVERROR_EXTERNAL;
-
- buf = av_buffer_alloc(size);
- if (!buf)
- return AVERROR(ENOMEM);
-
- if (!cmsSaveProfileToMem(profile, buf->data, &size) || size != buf->size) {
- av_buffer_unref(&buf);
- return AVERROR_EXTERNAL;
- }
-
- if (!av_frame_new_side_data_from_buf(frame, AV_FRAME_DATA_ICC_PROFILE, buf)) {
- av_buffer_unref(&buf);
- return AVERROR(ENOMEM);
- }
-
- return 0;
-}
-
-static av_always_inline void XYZ_xy(cmsCIEXYZ XYZ, AVCIExy *xy)
-{
- double k = 1.0 / (XYZ.X + XYZ.Y + XYZ.Z);
- xy->x = av_d2q(k * XYZ.X, 100000);
- xy->y = av_d2q(k * XYZ.Y, 100000);
-}
-
-int ff_icc_profile_read_primaries(FFIccContext *s, cmsHPROFILE profile,
- AVColorPrimariesDesc *out_primaries)
-{
- static const uint8_t testprimaries[4][3] = {
- { 0xFF, 0, 0 }, /* red */
- { 0, 0xFF, 0 }, /* green */
- { 0, 0, 0xFF }, /* blue */
- { 0xFF, 0xFF, 0xFF }, /* white */
- };
-
- AVWhitepointCoefficients *wp = &out_primaries->wp;
- AVPrimaryCoefficients *prim = &out_primaries->prim;
- cmsFloat64Number prev_adapt;
- cmsHPROFILE xyz;
- cmsHTRANSFORM tf;
- cmsCIEXYZ dst[4];
-
- xyz = cmsCreateXYZProfileTHR(s->ctx);
- if (!xyz)
- return AVERROR(ENOMEM);
-
- /* We need to use an unadapted observer to get the raw values */
- prev_adapt = cmsSetAdaptationStateTHR(s->ctx, 0.0);
- tf = cmsCreateTransformTHR(s->ctx, profile, TYPE_RGB_8, xyz, TYPE_XYZ_DBL,
- INTENT_ABSOLUTE_COLORIMETRIC,
- /* Note: These flags mostly don't do anything
- * anyway, but specify them regardless */
- cmsFLAGS_NOCACHE |
- cmsFLAGS_NOOPTIMIZE |
- cmsFLAGS_LOWRESPRECALC |
- cmsFLAGS_GRIDPOINTS(2));
- cmsSetAdaptationStateTHR(s->ctx, prev_adapt);
- cmsCloseProfile(xyz);
- if (!tf) {
- av_log(s->avctx, AV_LOG_ERROR, "Invalid ICC profile (e.g. CMYK)\n");
- return AVERROR_INVALIDDATA;
- }
-
- cmsDoTransform(tf, testprimaries, dst, 4);
- cmsDeleteTransform(tf);
- XYZ_xy(dst[0], &prim->r);
- XYZ_xy(dst[1], &prim->g);
- XYZ_xy(dst[2], &prim->b);
- XYZ_xy(dst[3], wp);
- return 0;
-}
-
-int ff_icc_profile_detect_transfer(FFIccContext *s, cmsHPROFILE profile,
- enum AVColorTransferCharacteristic *out_trc)
-{
- /* 8-bit linear grayscale ramp */
- static const uint8_t testramp[16][3] = {
- { 1, 1, 1}, /* avoid exact zero due to log100 etc. */
- { 17, 17, 17},
- { 34, 34, 34},
- { 51, 51, 51},
- { 68, 68, 68},
- { 85, 85, 85},
- { 02, 02, 02},
- {119, 119, 119},
- {136, 136, 136},
- {153, 153, 153},
- {170, 170, 170},
- {187, 187, 187},
- {204, 204, 204},
- {221, 221, 221},
- {238, 238, 238},
- {255, 255, 255},
- };
-
- double dst[FF_ARRAY_ELEMS(testramp)];
-
- for (enum AVColorTransferCharacteristic trc = 0; trc < AVCOL_TRC_NB; trc++) {
- cmsToneCurve *tonecurve;
- cmsHPROFILE ref;
- cmsHTRANSFORM tf;
- double delta = 0.0;
- if (get_curve(s, trc, &tonecurve) < 0)
- continue;
-
- ref = cmsCreateGrayProfileTHR(s->ctx, cmsD50_xyY(), tonecurve);
- if (!ref)
- return AVERROR(ENOMEM);
-
- tf = cmsCreateTransformTHR(s->ctx, profile, TYPE_RGB_8, ref, TYPE_GRAY_DBL,
- INTENT_RELATIVE_COLORIMETRIC,
- cmsFLAGS_NOCACHE | cmsFLAGS_NOOPTIMIZE);
- cmsCloseProfile(ref);
- if (!tf) {
- av_log(s->avctx, AV_LOG_ERROR, "Invalid ICC profile (e.g. CMYK)\n");
- return AVERROR_INVALIDDATA;
- }
-
- cmsDoTransform(tf, testramp, dst, FF_ARRAY_ELEMS(dst));
- cmsDeleteTransform(tf);
-
- for (int i = 0; i < FF_ARRAY_ELEMS(dst); i++)
- delta += fabs(testramp[i][0] / 255.0 - dst[i]);
- if (delta < 0.01) {
- *out_trc = trc;
- return 0;
- }
- }
-
- *out_trc = AVCOL_TRC_UNSPECIFIED;
- return 0;
-}
+#include "libavcodec/fflcms2.c"
diff --git a/libavfilter/fflcms2.h b/libavfilter/fflcms2.h
index 0d238c679f..1ac29e357b 100644
--- a/libavfilter/fflcms2.h
+++ b/libavfilter/fflcms2.h
@@ -1,5 +1,4 @@
/*
- * Copyright (c) 2022 Niklas Haas
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
@@ -17,71 +16,9 @@
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
-/**
- * @file
- * Various functions for dealing with ICC profiles
- */
-
#ifndef AVFILTER_FFLCMS2_H
#define AVFILTER_FFLCMS2_H
-#include "libavutil/csp.h"
-#include "libavutil/frame.h"
-#include "libavutil/pixfmt.h"
-
-#include <lcms2.h>
-
-typedef struct FFIccContext {
- void *avctx;
- cmsContext ctx;
- cmsToneCurve *curves[AVCOL_TRC_NB]; /* tone curve cache */
-} FFIccContext;
-
-/**
- * Initializes an FFIccContext. This must be done prior to using it.
- *
- * Returns 0 on success, or a negative error code.
- */
-int ff_icc_context_init(FFIccContext *s, void *avctx);
-void ff_icc_context_uninit(FFIccContext *s);
-
-/**
- * Generate an ICC profile for a given combination of color primaries and
- * transfer function. Both values must be set to valid entries (not
- * "undefined") for this function to work.
- *
- * Returns 0 on success, or a negative error code.
- */
-int ff_icc_profile_generate(FFIccContext *s,
- enum AVColorPrimaries color_prim,
- enum AVColorTransferCharacteristic color_trc,
- cmsHPROFILE *out_profile);
-
-/**
- * Attach an ICC profile to a frame. Helper wrapper around cmsSaveProfileToMem
- * and av_frame_new_side_data_from_buf.
- *
- * Returns 0 on success, or a negative error code.
- */
-int ff_icc_profile_attach(FFIccContext *s, cmsHPROFILE profile, AVFrame *frame);
-
-/**
- * Read the color primaries and white point coefficients encoded by an ICC
- * profile, and return the raw values in `out_primaries`.
- *
- * Returns 0 on success, or a negative error code.
- */
-int ff_icc_profile_read_primaries(FFIccContext *s, cmsHPROFILE profile,
- AVColorPrimariesDesc *out_primaries);
-
-/**
- * Attempt detecting the transfer characteristic that best approximates the
- * transfer function encoded by an ICC profile. Sets `out_trc` to
- * AVCOL_TRC_UNSPECIFIED if no clear match can be identified.
- *
- * Returns 0 on success (including no match), or a negative error code.
- */
-int ff_icc_profile_detect_transfer(FFIccContext *s, cmsHPROFILE profile,
- enum AVColorTransferCharacteristic *out_trc);
+#include "libavcodec/fflcms2.h"
#endif /* AVFILTER_FFLCMS2_H */
--
2.36.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* [FFmpeg-devel] [PATCH v2 2/6] avcodec/codec_internal: add cap for ICC profile support
2022-06-29 10:12 [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 1/6] fflcms2: move to libavcodec Niklas Haas
@ 2022-06-29 10:12 ` Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 3/6] avcodec: add API for automatic handling of icc profiles Niklas Haas
` (4 subsequent siblings)
6 siblings, 0 replies; 13+ messages in thread
From: Niklas Haas @ 2022-06-29 10:12 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Niklas Haas
From: Niklas Haas <git@haasn.dev>
Codecs that can read/write ICC profiles deserve a special capability so
the common logic in encode.c/decode.c can decide whether or not there
needs to be any special handling for ICC profiles. The motivation here
is to be able to use it to decide whether or not an ICC profile needs to
be generated in the encode path, but it might as well get added to
decoders as well for purely informative reasons.
It's not entirely clear to me whether the "thp" and "smvjpeg" variants
of "mjpeg" should have this capability set or not, given that the code
technically supports it but I somehow doubt these files may contain
them. In either case, this cap is purely informative for decoders so it
doesn't matter too much either way.
It's also not entirely clear whether the "amv" encoder should signal ICC
profile support, but again erring on the side of caution, we probably
*shouldn't* be generating (and encoding!) ICC profiles for this type of
media file.
Signed-off-by: Niklas Haas <git@haasn.dev>
---
libavcodec/codec_internal.h | 4 ++++
libavcodec/libjxldec.c | 3 ++-
libavcodec/libjxlenc.c | 3 ++-
libavcodec/mjpegdec.c | 3 ++-
libavcodec/mjpegenc.c | 3 ++-
libavcodec/pngdec.c | 5 +++--
libavcodec/pngenc.c | 4 ++--
libavcodec/tiff.c | 3 ++-
libavcodec/webp.c | 2 +-
9 files changed, 20 insertions(+), 10 deletions(-)
diff --git a/libavcodec/codec_internal.h b/libavcodec/codec_internal.h
index 5df286ce52..b4e3813353 100644
--- a/libavcodec/codec_internal.h
+++ b/libavcodec/codec_internal.h
@@ -73,6 +73,10 @@
* internal logic derive them from AVCodecInternal.last_pkt_props.
*/
#define FF_CODEC_CAP_SETS_FRAME_PROPS (1 << 8)
+/**
+ * Codec supports embedded ICC profiles (AV_FRAME_DATA_ICC_PROFILE).
+ */
+#define FF_CODEC_CAP_ICC_PROFILES (1 << 9)
/**
* FFCodec.codec_tags termination value
diff --git a/libavcodec/libjxldec.c b/libavcodec/libjxldec.c
index d516d3b0ac..7c78627037 100644
--- a/libavcodec/libjxldec.c
+++ b/libavcodec/libjxldec.c
@@ -455,6 +455,7 @@ const FFCodec ff_libjxl_decoder = {
FF_CODEC_DECODE_CB(libjxl_decode_frame),
.close = libjxl_decode_close,
.p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_OTHER_THREADS,
- .caps_internal = FF_CODEC_CAP_AUTO_THREADS | FF_CODEC_CAP_INIT_CLEANUP,
+ .caps_internal = FF_CODEC_CAP_AUTO_THREADS | FF_CODEC_CAP_INIT_CLEANUP |
+ FF_CODEC_CAP_ICC_PROFILES,
.p.wrapper_name = "libjxl",
};
diff --git a/libavcodec/libjxlenc.c b/libavcodec/libjxlenc.c
index 6a948cc3ae..3a533db8a1 100644
--- a/libavcodec/libjxlenc.c
+++ b/libavcodec/libjxlenc.c
@@ -463,7 +463,8 @@ const FFCodec ff_libjxl_encoder = {
FF_CODEC_ENCODE_CB(libjxl_encode_frame),
.close = libjxl_encode_close,
.p.capabilities = AV_CODEC_CAP_OTHER_THREADS,
- .caps_internal = FF_CODEC_CAP_AUTO_THREADS | FF_CODEC_CAP_INIT_CLEANUP,
+ .caps_internal = FF_CODEC_CAP_AUTO_THREADS | FF_CODEC_CAP_INIT_CLEANUP |
+ FF_CODEC_CAP_ICC_PROFILES,
.p.pix_fmts = (const enum AVPixelFormat[]) {
AV_PIX_FMT_RGB24, AV_PIX_FMT_RGBA,
AV_PIX_FMT_RGB48, AV_PIX_FMT_RGBA64,
diff --git a/libavcodec/mjpegdec.c b/libavcodec/mjpegdec.c
index 32874a5a19..8e5878e9ee 100644
--- a/libavcodec/mjpegdec.c
+++ b/libavcodec/mjpegdec.c
@@ -3027,7 +3027,8 @@ const FFCodec ff_mjpeg_decoder = {
.p.priv_class = &mjpegdec_class,
.p.profiles = NULL_IF_CONFIG_SMALL(ff_mjpeg_profiles),
.caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_INIT_CLEANUP |
- FF_CODEC_CAP_SKIP_FRAME_FILL_PARAM | FF_CODEC_CAP_SETS_PKT_DTS,
+ FF_CODEC_CAP_SKIP_FRAME_FILL_PARAM | FF_CODEC_CAP_SETS_PKT_DTS |
+ FF_CODEC_CAP_ICC_PROFILES,
.hw_configs = (const AVCodecHWConfigInternal *const []) {
#if CONFIG_MJPEG_NVDEC_HWACCEL
HWACCEL_NVDEC(mjpeg),
diff --git a/libavcodec/mjpegenc.c b/libavcodec/mjpegenc.c
index 27217441a3..8787528b3e 100644
--- a/libavcodec/mjpegenc.c
+++ b/libavcodec/mjpegenc.c
@@ -652,7 +652,8 @@ const FFCodec ff_mjpeg_encoder = {
FF_CODEC_ENCODE_CB(ff_mpv_encode_picture),
.close = mjpeg_encode_close,
.p.capabilities = AV_CODEC_CAP_SLICE_THREADS | AV_CODEC_CAP_FRAME_THREADS,
- .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_INIT_CLEANUP,
+ .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_INIT_CLEANUP |
+ FF_CODEC_CAP_ICC_PROFILES,
.p.pix_fmts = (const enum AVPixelFormat[]) {
AV_PIX_FMT_YUVJ420P, AV_PIX_FMT_YUVJ422P, AV_PIX_FMT_YUVJ444P,
AV_PIX_FMT_YUV420P, AV_PIX_FMT_YUV422P, AV_PIX_FMT_YUV444P,
diff --git a/libavcodec/pngdec.c b/libavcodec/pngdec.c
index 6b44af59f2..5cd8a176fd 100644
--- a/libavcodec/pngdec.c
+++ b/libavcodec/pngdec.c
@@ -1727,7 +1727,7 @@ const FFCodec ff_apng_decoder = {
.update_thread_context = ONLY_IF_THREADS_ENABLED(update_thread_context),
.p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_FRAME_THREADS /*| AV_CODEC_CAP_DRAW_HORIZ_BAND*/,
.caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_INIT_CLEANUP |
- FF_CODEC_CAP_ALLOCATE_PROGRESS,
+ FF_CODEC_CAP_ALLOCATE_PROGRESS | FF_CODEC_CAP_ICC_PROFILES,
};
#endif
@@ -1744,6 +1744,7 @@ const FFCodec ff_png_decoder = {
.update_thread_context = ONLY_IF_THREADS_ENABLED(update_thread_context),
.p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_FRAME_THREADS /*| AV_CODEC_CAP_DRAW_HORIZ_BAND*/,
.caps_internal = FF_CODEC_CAP_SKIP_FRAME_FILL_PARAM | FF_CODEC_CAP_INIT_THREADSAFE |
- FF_CODEC_CAP_ALLOCATE_PROGRESS | FF_CODEC_CAP_INIT_CLEANUP,
+ FF_CODEC_CAP_ALLOCATE_PROGRESS | FF_CODEC_CAP_INIT_CLEANUP |
+ FF_CODEC_CAP_ICC_PROFILES,
};
#endif
diff --git a/libavcodec/pngenc.c b/libavcodec/pngenc.c
index d79b4e3895..b23ff84787 100644
--- a/libavcodec/pngenc.c
+++ b/libavcodec/pngenc.c
@@ -1210,7 +1210,7 @@ const FFCodec ff_png_encoder = {
AV_PIX_FMT_MONOBLACK, AV_PIX_FMT_NONE
},
.p.priv_class = &pngenc_class,
- .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE,
+ .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_ICC_PROFILES,
};
const FFCodec ff_apng_encoder = {
@@ -1232,5 +1232,5 @@ const FFCodec ff_apng_encoder = {
AV_PIX_FMT_NONE
},
.p.priv_class = &pngenc_class,
- .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE,
+ .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_ICC_PROFILES,
};
diff --git a/libavcodec/tiff.c b/libavcodec/tiff.c
index e4a5d3c537..c369c12f71 100644
--- a/libavcodec/tiff.c
+++ b/libavcodec/tiff.c
@@ -2187,6 +2187,7 @@ const FFCodec ff_tiff_decoder = {
.close = tiff_end,
FF_CODEC_DECODE_CB(decode_frame),
.p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_FRAME_THREADS,
- .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_INIT_CLEANUP,
+ .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_INIT_CLEANUP |
+ FF_CODEC_CAP_ICC_PROFILES,
.p.priv_class = &tiff_decoder_class,
};
diff --git a/libavcodec/webp.c b/libavcodec/webp.c
index 1b5e943a6e..2b1242615d 100644
--- a/libavcodec/webp.c
+++ b/libavcodec/webp.c
@@ -1565,5 +1565,5 @@ const FFCodec ff_webp_decoder = {
FF_CODEC_DECODE_CB(webp_decode_frame),
.close = webp_decode_close,
.p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_FRAME_THREADS,
- .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE,
+ .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_ICC_PROFILES,
};
--
2.36.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* [FFmpeg-devel] [PATCH v2 3/6] avcodec: add API for automatic handling of icc profiles
2022-06-29 10:12 [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 1/6] fflcms2: move to libavcodec Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 2/6] avcodec/codec_internal: add cap for ICC profile support Niklas Haas
@ 2022-06-29 10:12 ` Niklas Haas
2022-07-06 13:34 ` Anton Khirnov
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 4/6] avcodec: add common fflcms2 boilerplate Niklas Haas
` (3 subsequent siblings)
6 siblings, 1 reply; 13+ messages in thread
From: Niklas Haas @ 2022-06-29 10:12 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Niklas Haas
From: Niklas Haas <git@haasn.dev>
This functionally already exists, but as pointed out in #9672 and #9673,
requiring users to manually include filters is clumsy, error-prone and
hard to use together with tools like ffplay.
To streamline ICC profile support, add a new AVCodecContext option to
globally enable reading and writing ICC profiles, automatically, for all
appropriate media types.
Note that this commit only includes the new API. The implementation is
split off to separate commits for readability.
Signed-off-by: Niklas Haas <git@haasn.dev>
---
doc/APIchanges | 4 ++++
libavcodec/avcodec.h | 10 ++++++++++
libavcodec/version.h | 2 +-
3 files changed, 15 insertions(+), 1 deletion(-)
diff --git a/doc/APIchanges b/doc/APIchanges
index 20b944933a..b8acf86352 100644
--- a/doc/APIchanges
+++ b/doc/APIchanges
@@ -14,6 +14,10 @@ libavutil: 2021-04-27
API changes, most recent first:
+2022-06-28 - xxxxxxxxx - lavc 59.35.100 - avcodec.h
+ Add the icc_profiles option to AVCodecContext, to enable automatic reading
+ and writing of embedded ICC profiles in image files.
+
2022-06-12 - xxxxxxxxxx - lavf 59.25.100 - avio.h
Add avio_vprintf(), similar to avio_printf() but allow to use it
from within a function taking a variable argument list as input.
diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index 4dae23d06e..8ee9be22d0 100644
--- a/libavcodec/avcodec.h
+++ b/libavcodec/avcodec.h
@@ -979,6 +979,16 @@ typedef struct AVCodecContext {
*/
enum AVChromaLocation chroma_sample_location;
+ /**
+ * Whether to decode/encode ICC profiles. If set, libavcodec will
+ * generate/parse ICC profiles as appropriate for the type of file. No
+ * effect on codecs which cannot contain embedded ICC profiles, or when
+ * compiled without support for lcms2.
+ * - encoding: Set by user
+ * - decoding: Set by user
+ */
+ int icc_profiles;
+
/**
* Number of slices.
* Indicates number of picture subdivisions. Used for parallelized
diff --git a/libavcodec/version.h b/libavcodec/version.h
index 0ef6c991f3..1008fead27 100644
--- a/libavcodec/version.h
+++ b/libavcodec/version.h
@@ -29,7 +29,7 @@
#include "version_major.h"
-#define LIBAVCODEC_VERSION_MINOR 34
+#define LIBAVCODEC_VERSION_MINOR 35
#define LIBAVCODEC_VERSION_MICRO 100
#define LIBAVCODEC_VERSION_INT AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \
--
2.36.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* [FFmpeg-devel] [PATCH v2 4/6] avcodec: add common fflcms2 boilerplate
2022-06-29 10:12 [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
` (2 preceding siblings ...)
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 3/6] avcodec: add API for automatic handling of icc profiles Niklas Haas
@ 2022-06-29 10:12 ` Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 5/6] avcodec/decode: parse ICC profiles Niklas Haas
` (2 subsequent siblings)
6 siblings, 0 replies; 13+ messages in thread
From: Niklas Haas @ 2022-06-29 10:12 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Niklas Haas
From: Niklas Haas <git@haasn.dev>
Handling this in general code makes more sense than handling it in
individual codec files, because it would be a lot of unnecessary code
duplication for the plenty of formats that support exporting ICC
profiles (jpg, png, tiff, webp, jxl, ...).
encode.c and decode.c will be in charge of initializing this state as
needed, so we merely need to make sure to uninit it afterwards from the
common destructor path.
Signed-off-by: Niklas Haas <git@haasn.dev>
---
configure | 2 +-
libavcodec/Makefile | 1 +
libavcodec/avcodec.c | 4 ++++
libavcodec/decode.c | 4 ++++
libavcodec/internal.h | 8 ++++++++
5 files changed, 18 insertions(+), 1 deletion(-)
diff --git a/configure b/configure
index 0de9b2abcb..d3f151e246 100755
--- a/configure
+++ b/configure
@@ -3807,7 +3807,7 @@ swresample_suggest="libm libsoxr stdatomic"
swscale_deps="avutil"
swscale_suggest="libm stdatomic"
-avcodec_extralibs="pthreads_extralibs iconv_extralibs dxva2_extralibs"
+avcodec_extralibs="pthreads_extralibs iconv_extralibs dxva2_extralibs lcms2_extralibs"
avfilter_extralibs="pthreads_extralibs"
avutil_extralibs="d3d11va_extralibs nanosleep_extralibs pthreads_extralibs vaapi_drm_extralibs vaapi_x11_extralibs vdpau_x11_extralibs"
diff --git a/libavcodec/Makefile b/libavcodec/Makefile
index 5c4f62c631..d2f969ba52 100644
--- a/libavcodec/Makefile
+++ b/libavcodec/Makefile
@@ -114,6 +114,7 @@ OBJS-$(CONFIG_INTRAX8) += intrax8.o intrax8dsp.o msmpeg4data.o
OBJS-$(CONFIG_IVIDSP) += ivi_dsp.o
OBJS-$(CONFIG_JNI) += ffjni.o jni.o
OBJS-$(CONFIG_JPEGTABLES) += jpegtables.o
+OBJS-$(CONFIG_LCMS2) += fflcms2.o
OBJS-$(CONFIG_LLAUDDSP) += lossless_audiodsp.o
OBJS-$(CONFIG_LLVIDDSP) += lossless_videodsp.o
OBJS-$(CONFIG_LLVIDENCDSP) += lossless_videoencdsp.o
diff --git a/libavcodec/avcodec.c b/libavcodec/avcodec.c
index 5f6e71a39e..3524214546 100644
--- a/libavcodec/avcodec.c
+++ b/libavcodec/avcodec.c
@@ -481,6 +481,10 @@ av_cold int avcodec_close(AVCodecContext *avctx)
av_channel_layout_uninit(&avci->initial_ch_layout);
+#if CONFIG_LCMS2
+ ff_icc_context_uninit(&avci->icc);
+#endif
+
av_freep(&avctx->internal);
}
diff --git a/libavcodec/decode.c b/libavcodec/decode.c
index 1893caa6a6..d1414c599b 100644
--- a/libavcodec/decode.c
+++ b/libavcodec/decode.c
@@ -49,6 +49,10 @@
#include "internal.h"
#include "thread.h"
+#if CONFIG_LCMS2
+# include "fflcms2.h"
+#endif
+
static int apply_param_change(AVCodecContext *avctx, const AVPacket *avpkt)
{
int ret;
diff --git a/libavcodec/internal.h b/libavcodec/internal.h
index 17e1de8127..e284816705 100644
--- a/libavcodec/internal.h
+++ b/libavcodec/internal.h
@@ -33,6 +33,10 @@
#include "avcodec.h"
#include "config.h"
+#if CONFIG_LCMS2
+# include "fflcms2.h"
+#endif
+
#define FF_SANE_NB_CHANNELS 512U
#if HAVE_SIMD_ALIGN_64
@@ -148,6 +152,10 @@ typedef struct AVCodecInternal {
uint64_t initial_channel_layout;
#endif
AVChannelLayout initial_ch_layout;
+
+#if CONFIG_LCMS2
+ FFIccContext icc; /* used to read and write embedded ICC profiles */
+#endif
} AVCodecInternal;
/**
--
2.36.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* [FFmpeg-devel] [PATCH v2 5/6] avcodec/decode: parse ICC profiles
2022-06-29 10:12 [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
` (3 preceding siblings ...)
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 4/6] avcodec: add common fflcms2 boilerplate Niklas Haas
@ 2022-06-29 10:12 ` Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 6/6] avcodec/encode:: generate " Niklas Haas
2022-07-02 21:05 ` [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
6 siblings, 0 replies; 13+ messages in thread
From: Niklas Haas @ 2022-06-29 10:12 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Niklas Haas
From: Niklas Haas <git@haasn.dev>
Implementation for the decode side of the ICC profile API, roughly
matching the behavior of the existing vf_iccdetect filter.
Closes: #9673
Signed-off-by: Niklas Haas <git@haasn.dev>
---
libavcodec/decode.c | 61 +++++++++++++++++++++++++++++++++++++++++----
1 file changed, 56 insertions(+), 5 deletions(-)
diff --git a/libavcodec/decode.c b/libavcodec/decode.c
index d1414c599b..e81a7fb4d5 100644
--- a/libavcodec/decode.c
+++ b/libavcodec/decode.c
@@ -49,10 +49,6 @@
#include "internal.h"
#include "thread.h"
-#if CONFIG_LCMS2
-# include "fflcms2.h"
-#endif
-
static int apply_param_change(AVCodecContext *avctx, const AVPacket *avpkt)
{
int ret;
@@ -508,6 +504,54 @@ FF_ENABLE_DEPRECATION_WARNINGS
return ret < 0 ? ret : 0;
}
+#if CONFIG_LCMS2
+static int detect_colorspace(AVCodecContext *avctx, AVFrame *frame)
+{
+ AVCodecInternal *avci = avctx->internal;
+ enum AVColorTransferCharacteristic trc;
+ AVColorPrimariesDesc coeffs;
+ enum AVColorPrimaries prim;
+ cmsHPROFILE profile;
+ AVFrameSideData *sd;
+ int ret;
+ if (!avctx->icc_profiles)
+ return 0;
+
+ sd = av_frame_get_side_data(frame, AV_FRAME_DATA_ICC_PROFILE);
+ if (!sd || !sd->size)
+ return 0;
+
+ if (!avci->icc.avctx) {
+ ret = ff_icc_context_init(&avci->icc, avctx);
+ if (ret < 0)
+ return ret;
+ }
+
+ profile = cmsOpenProfileFromMemTHR(avci->icc.ctx, sd->data, sd->size);
+ if (!profile)
+ return AVERROR_INVALIDDATA;
+
+ ret = ff_icc_profile_read_primaries(&avci->icc, profile, &coeffs);
+ if (!ret)
+ ret = ff_icc_profile_detect_transfer(&avci->icc, profile, &trc);
+ cmsCloseProfile(profile);
+ if (ret < 0)
+ return ret;
+
+ prim = av_csp_primaries_id_from_desc(&coeffs);
+ if (prim != AVCOL_PRI_UNSPECIFIED)
+ frame->color_primaries = prim;
+ if (trc != AVCOL_TRC_UNSPECIFIED)
+ frame->color_trc = trc;
+ return 0;
+}
+#else /* !CONFIG_LCMS2 */
+static int detect_colorspace(av_unused AVCodecContext *c, av_unused AVFrame *f)
+{
+ return 0;
+}
+#endif
+
static int decode_simple_receive_frame(AVCodecContext *avctx, AVFrame *frame)
{
int ret;
@@ -528,7 +572,7 @@ static int decode_receive_frame_internal(AVCodecContext *avctx, AVFrame *frame)
{
AVCodecInternal *avci = avctx->internal;
const FFCodec *const codec = ffcodec(avctx->codec);
- int ret;
+ int ret, ok;
av_assert0(!frame->buf[0]);
@@ -542,6 +586,13 @@ static int decode_receive_frame_internal(AVCodecContext *avctx, AVFrame *frame)
if (ret == AVERROR_EOF)
avci->draining_done = 1;
+ /* preserve ret */
+ ok = detect_colorspace(avctx, frame);
+ if (ok < 0) {
+ av_frame_unref(frame);
+ return ok;
+ }
+
if (!(codec->caps_internal & FF_CODEC_CAP_SETS_FRAME_PROPS) &&
IS_EMPTY(avci->last_pkt_props)) {
// May fail if the FIFO is empty.
--
2.36.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* [FFmpeg-devel] [PATCH v2 6/6] avcodec/encode:: generate ICC profiles
2022-06-29 10:12 [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
` (4 preceding siblings ...)
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 5/6] avcodec/decode: parse ICC profiles Niklas Haas
@ 2022-06-29 10:12 ` Niklas Haas
2022-07-06 13:35 ` Anton Khirnov
2022-07-02 21:05 ` [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
6 siblings, 1 reply; 13+ messages in thread
From: Niklas Haas @ 2022-06-29 10:12 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Niklas Haas
From: Niklas Haas <git@haasn.dev>
Only if requested, and only if the codec signals support for ICC
profiles. Implementation roughly matches the functionality of the
existing vf_iccgen filter, albeit with some reduced flexibility and no
caching.
Ideally, we'd also only do this on the first frame (e.g. mjpeg, apng),
but there's no meaningful way for us to distinguish between this case
and e.g. somebody using the image2 muxer, in which case we'd want to
attach ICC profiles to every frame in the stream.
Closes: #9672
Signed-off-by: Niklas Haas <git@haasn.dev>
---
libavcodec/encode.c | 50 +++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 50 insertions(+)
diff --git a/libavcodec/encode.c b/libavcodec/encode.c
index b68bf1e184..d1fb6f3a75 100644
--- a/libavcodec/encode.c
+++ b/libavcodec/encode.c
@@ -308,6 +308,50 @@ static int encode_receive_packet_internal(AVCodecContext *avctx, AVPacket *avpkt
return ret;
}
+#if CONFIG_LCMS2
+static int encode_generate_icc_profile(AVCodecContext *avctx, AVFrame *frame)
+{
+ enum AVColorTransferCharacteristic trc = frame->color_trc;
+ enum AVColorPrimaries prim = frame->color_primaries;
+ const FFCodec *const codec = ffcodec(avctx->codec);
+ AVCodecInternal *avci = avctx->internal;
+ cmsHPROFILE profile;
+ int ret;
+
+ if (!avctx->icc_profiles || !(codec->caps_internal & FF_CODEC_CAP_ICC_PROFILES))
+ return 0; /* don't generate ICC profiles if disabled or unsupported */
+
+ if (trc == AVCOL_TRC_UNSPECIFIED)
+ trc = avctx->color_trc;
+ if (prim == AVCOL_PRI_UNSPECIFIED)
+ prim = avctx->color_primaries;
+ if (trc == AVCOL_TRC_UNSPECIFIED || prim == AVCOL_PRI_UNSPECIFIED)
+ return 0; /* can't generate ICC profile with missing csp tags */
+
+ if (av_frame_get_side_data(frame, AV_FRAME_DATA_ICC_PROFILE))
+ return 0; /* don't overwrite existing ICC profile */
+
+ if (!avci->icc.avctx) {
+ ret = ff_icc_context_init(&avci->icc, avctx);
+ if (ret < 0)
+ return ret;
+ }
+
+ ret = ff_icc_profile_generate(&avci->icc, prim, trc, &profile);
+ if (ret < 0)
+ return ret;
+
+ ret = ff_icc_profile_attach(&avci->icc, profile, frame);
+ cmsCloseProfile(profile);
+ return ret;
+}
+#else /* !CONFIG_LCMS2 */
+static int encode_generate_icc_profile(av_unused AVCodecContext *c, av_unused AVFrame *f)
+{
+ return 0;
+}
+#endif
+
static int encode_send_frame_internal(AVCodecContext *avctx, const AVFrame *src)
{
AVCodecInternal *avci = avctx->internal;
@@ -352,6 +396,12 @@ static int encode_send_frame_internal(AVCodecContext *avctx, const AVFrame *src)
return ret;
}
+ if (avctx->codec->type == AVMEDIA_TYPE_VIDEO) {
+ ret = encode_generate_icc_profile(avctx, dst);
+ if (ret < 0)
+ return ret;
+ }
+
return 0;
}
--
2.36.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec
2022-06-29 10:12 [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
` (5 preceding siblings ...)
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 6/6] avcodec/encode:: generate " Niklas Haas
@ 2022-07-02 21:05 ` Niklas Haas
6 siblings, 0 replies; 13+ messages in thread
From: Niklas Haas @ 2022-07-02 21:05 UTC (permalink / raw)
To: ffmpeg-devel
On Wed, 29 Jun 2022 12:12:45 +0200 Niklas Haas <ffmpeg@haasn.xyz> wrote:
> Changes in v2:
> - remove unnecessary import
> - fixed assignment-instead-of-equality
> - fixed #ifdef -> #if
Bump. Any thoughts on this?
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [FFmpeg-devel] [PATCH v2 3/6] avcodec: add API for automatic handling of icc profiles
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 3/6] avcodec: add API for automatic handling of icc profiles Niklas Haas
@ 2022-07-06 13:34 ` Anton Khirnov
0 siblings, 0 replies; 13+ messages in thread
From: Anton Khirnov @ 2022-07-06 13:34 UTC (permalink / raw)
To: FFmpeg development discussions and patches; +Cc: Niklas Haas
Quoting Niklas Haas (2022-06-29 12:12:48)
> From: Niklas Haas <git@haasn.dev>
>
> This functionally already exists, but as pointed out in #9672 and #9673,
> requiring users to manually include filters is clumsy, error-prone and
> hard to use together with tools like ffplay.
>
> To streamline ICC profile support, add a new AVCodecContext option to
> globally enable reading and writing ICC profiles, automatically, for all
> appropriate media types.
>
> Note that this commit only includes the new API. The implementation is
> split off to separate commits for readability.
>
> Signed-off-by: Niklas Haas <git@haasn.dev>
> ---
> doc/APIchanges | 4 ++++
> libavcodec/avcodec.h | 10 ++++++++++
> libavcodec/version.h | 2 +-
> 3 files changed, 15 insertions(+), 1 deletion(-)
>
> diff --git a/doc/APIchanges b/doc/APIchanges
> index 20b944933a..b8acf86352 100644
> --- a/doc/APIchanges
> +++ b/doc/APIchanges
> @@ -14,6 +14,10 @@ libavutil: 2021-04-27
>
> API changes, most recent first:
>
> +2022-06-28 - xxxxxxxxx - lavc 59.35.100 - avcodec.h
> + Add the icc_profiles option to AVCodecContext, to enable automatic reading
> + and writing of embedded ICC profiles in image files.
> +
> 2022-06-12 - xxxxxxxxxx - lavf 59.25.100 - avio.h
> Add avio_vprintf(), similar to avio_printf() but allow to use it
> from within a function taking a variable argument list as input.
> diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
> index 4dae23d06e..8ee9be22d0 100644
> --- a/libavcodec/avcodec.h
> +++ b/libavcodec/avcodec.h
> @@ -979,6 +979,16 @@ typedef struct AVCodecContext {
> */
> enum AVChromaLocation chroma_sample_location;
>
> + /**
> + * Whether to decode/encode ICC profiles. If set, libavcodec will
> + * generate/parse ICC profiles as appropriate for the type of file. No
> + * effect on codecs which cannot contain embedded ICC profiles, or when
> + * compiled without support for lcms2.
> + * - encoding: Set by user
> + * - decoding: Set by user
> + */
> + int icc_profiles;
This is an ABI break. You could also just add a new AV_CODEC_FLAG
instead.
Also, why should this not be enabled by default?
--
Anton Khirnov
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [FFmpeg-devel] [PATCH v2 6/6] avcodec/encode:: generate ICC profiles
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 6/6] avcodec/encode:: generate " Niklas Haas
@ 2022-07-06 13:35 ` Anton Khirnov
2022-07-19 12:18 ` Niklas Haas
0 siblings, 1 reply; 13+ messages in thread
From: Anton Khirnov @ 2022-07-06 13:35 UTC (permalink / raw)
To: FFmpeg development discussions and patches; +Cc: Niklas Haas
Quoting Niklas Haas (2022-06-29 12:12:51)
> From: Niklas Haas <git@haasn.dev>
>
> Only if requested, and only if the codec signals support for ICC
> profiles. Implementation roughly matches the functionality of the
> existing vf_iccgen filter, albeit with some reduced flexibility and no
> caching.
>
> Ideally, we'd also only do this on the first frame (e.g. mjpeg, apng),
> but there's no meaningful way for us to distinguish between this case
> and e.g. somebody using the image2 muxer, in which case we'd want to
> attach ICC profiles to every frame in the stream.
AV_CODEC_FLAG_GLOBAL_HEADER?
--
Anton Khirnov
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [FFmpeg-devel] [PATCH v2 1/6] fflcms2: move to libavcodec
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 1/6] fflcms2: move to libavcodec Niklas Haas
@ 2022-07-06 14:18 ` Andreas Rheinhardt
2022-07-12 12:09 ` Niklas Haas
0 siblings, 1 reply; 13+ messages in thread
From: Andreas Rheinhardt @ 2022-07-06 14:18 UTC (permalink / raw)
To: ffmpeg-devel
Niklas Haas:
> From: Niklas Haas <git@haasn.dev>
>
> We will need this helper inside libavcodec in the future, so move it
> there, leaving behind an #include to the raw source file in its old
> location in libvfilter. This approach is inspired by the handling of
> vulkan.c, and avoids us needing to expose any of it publicly (or
> semi-publicly) in e.g. libavutil, thus avoiding any ABI headaches.
This is only correct if you work under the assumption that when building
with static libraries, all the static libraries come from the same
commit. If you e.g. allow to build with an older libavfilter and a newer
libavcodec (like with shared libs), this assumption breaks down: if the
newer version of this file exports another function, you get linker
(ODR) errors because both versions might be pulled in.
Behaviour/signature changes by any function or modifications to any of
the structs would be similarly catastrophic.
A way out of this mess would be to version everything in the header like so:
#define FFLCMS2_VERSION 1
void AV_JOIN(ff_icc_context_uninit, FFLCMS2_VERSION)(FFIccContext *s);
(of course, there should be a dedicated macro for this to reduce typing.)
A patch that makes any of the "catastrophic" modifications described
above would need to bump the version. If one uses compatible versions,
the files would be deduplicated for static builds.
It would of course have the downside that these macros would be
everywhere where it is used, not only in fflcms2.[ch].
>
> It's debatable whether the actual code belongs in libavcodec or
> libavfilter, but I decided to put it into libavcodec because it
> conceptually deals with encoding and decoding ICC profiles, and will be
> used to decode embedded ICC profiles in image files.
>
> Signed-off-by: Niklas Haas <git@haasn.dev>
> ---
> libavcodec/Makefile | 1 +
> libavcodec/fflcms2.c | 311 ++++++++++++++++++++++++++++++++++++++++++
> libavcodec/fflcms2.h | 87 ++++++++++++
> libavfilter/fflcms2.c | 294 +--------------------------------------
> libavfilter/fflcms2.h | 65 +--------
> 5 files changed, 401 insertions(+), 357 deletions(-)
> create mode 100644 libavcodec/fflcms2.c
> create mode 100644 libavcodec/fflcms2.h
>
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [FFmpeg-devel] [PATCH v2 1/6] fflcms2: move to libavcodec
2022-07-06 14:18 ` Andreas Rheinhardt
@ 2022-07-12 12:09 ` Niklas Haas
0 siblings, 0 replies; 13+ messages in thread
From: Niklas Haas @ 2022-07-12 12:09 UTC (permalink / raw)
To: ffmpeg-devel
On Wed, 06 Jul 2022 16:18:21 +0200 Andreas Rheinhardt <andreas.rheinhardt@outlook.com> wrote:
> Niklas Haas:
> > From: Niklas Haas <git@haasn.dev>
> >
> > We will need this helper inside libavcodec in the future, so move it
> > there, leaving behind an #include to the raw source file in its old
> > location in libvfilter. This approach is inspired by the handling of
> > vulkan.c, and avoids us needing to expose any of it publicly (or
> > semi-publicly) in e.g. libavutil, thus avoiding any ABI headaches.
>
> This is only correct if you work under the assumption that when building
> with static libraries, all the static libraries come from the same
> commit. If you e.g. allow to build with an older libavfilter and a newer
> libavcodec (like with shared libs), this assumption breaks down: if the
> newer version of this file exports another function, you get linker
> (ODR) errors because both versions might be pulled in.
> Behaviour/signature changes by any function or modifications to any of
> the structs would be similarly catastrophic.
>
> A way out of this mess would be to version everything in the header like so:
> #define FFLCMS2_VERSION 1
>
> void AV_JOIN(ff_icc_context_uninit, FFLCMS2_VERSION)(FFIccContext *s);
>
> (of course, there should be a dedicated macro for this to reduce typing.)
>
> A patch that makes any of the "catastrophic" modifications described
> above would need to bump the version. If one uses compatible versions,
> the files would be deduplicated for static builds.
> It would of course have the downside that these macros would be
> everywhere where it is used, not only in fflcms2.[ch].
I suppose you could also version this API on a symbol-by-symbol basis,
with v1 being the current version (and then v2 getting a 2 suffix, and
so on).
In that case, we don't need to do anything currently, rather the first
patch to modify this API would have to worry about making sure the new
symbol doesn't conflict.
>
> >
> > It's debatable whether the actual code belongs in libavcodec or
> > libavfilter, but I decided to put it into libavcodec because it
> > conceptually deals with encoding and decoding ICC profiles, and will be
> > used to decode embedded ICC profiles in image files.
> >
> > Signed-off-by: Niklas Haas <git@haasn.dev>
> > ---
> > libavcodec/Makefile | 1 +
> > libavcodec/fflcms2.c | 311 ++++++++++++++++++++++++++++++++++++++++++
> > libavcodec/fflcms2.h | 87 ++++++++++++
> > libavfilter/fflcms2.c | 294 +--------------------------------------
> > libavfilter/fflcms2.h | 65 +--------
> > 5 files changed, 401 insertions(+), 357 deletions(-)
> > create mode 100644 libavcodec/fflcms2.c
> > create mode 100644 libavcodec/fflcms2.h
> >
>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [FFmpeg-devel] [PATCH v2 6/6] avcodec/encode:: generate ICC profiles
2022-07-06 13:35 ` Anton Khirnov
@ 2022-07-19 12:18 ` Niklas Haas
0 siblings, 0 replies; 13+ messages in thread
From: Niklas Haas @ 2022-07-19 12:18 UTC (permalink / raw)
To: FFmpeg development discussions and patches
On Wed, 06 Jul 2022 15:35:41 +0200 Anton Khirnov <anton@khirnov.net> wrote:
> Quoting Niklas Haas (2022-06-29 12:12:51)
> > From: Niklas Haas <git@haasn.dev>
> >
> > Only if requested, and only if the codec signals support for ICC
> > profiles. Implementation roughly matches the functionality of the
> > existing vf_iccgen filter, albeit with some reduced flexibility and no
> > caching.
> >
> > Ideally, we'd also only do this on the first frame (e.g. mjpeg, apng),
> > but there's no meaningful way for us to distinguish between this case
> > and e.g. somebody using the image2 muxer, in which case we'd want to
> > attach ICC profiles to every frame in the stream.
>
> AV_CODEC_FLAG_GLOBAL_HEADER?
This seems marginally unrelated - apng etc. don't have extradata, and
none of the relevant codecs set this flag either. At best we could
introduce a new flag, but to be honest I don't think it's a huge issue
because apng and mjpeg with ICC profiles are exceptionally rare, and the
worst case is merely wasting some file space.
>
> --
> Anton Khirnov
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 13+ messages in thread
end of thread, other threads:[~2022-07-19 12:18 UTC | newest]
Thread overview: 13+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-06-29 10:12 [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 1/6] fflcms2: move to libavcodec Niklas Haas
2022-07-06 14:18 ` Andreas Rheinhardt
2022-07-12 12:09 ` Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 2/6] avcodec/codec_internal: add cap for ICC profile support Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 3/6] avcodec: add API for automatic handling of icc profiles Niklas Haas
2022-07-06 13:34 ` Anton Khirnov
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 4/6] avcodec: add common fflcms2 boilerplate Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 5/6] avcodec/decode: parse ICC profiles Niklas Haas
2022-06-29 10:12 ` [FFmpeg-devel] [PATCH v2 6/6] avcodec/encode:: generate " Niklas Haas
2022-07-06 13:35 ` Anton Khirnov
2022-07-19 12:18 ` Niklas Haas
2022-07-02 21:05 ` [FFmpeg-devel] [PATCH v2 0/6] ICC profile support in avcodec Niklas Haas
Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
This inbox may be cloned and mirrored by anyone:
git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git
# If you have public-inbox 1.1+ installed, you may
# initialize and index your mirror using the following commands:
public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
ffmpegdev@gitmailbox.com
public-inbox-index ffmpegdev
Example config snippet for mirrors.
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git