From eae89cea4be5b6abb147cdeabdbadd844984b182 Mon Sep 17 00:00:00 2001
From: k4yt3x <i@k4yt3x.com>
Date: Sat, 18 Jan 2025 00:00:00 +0000
Subject: [PATCH] feat(libvideo2x): allow processing videos without PTS
 information

Signed-off-by: k4yt3x <i@k4yt3x.com>
---
 CHANGELOG.md                 | 1 +
 include/libvideo2x/encoder.h | 2 +-
 src/encoder.cpp              | 5 +----
 src/libvideo2x.cpp           | 8 +++++++-
 4 files changed, 10 insertions(+), 6 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 3d3f5ec..f4a2092 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -10,6 +10,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Added
 
 - Multi-versioning to critical functions to enhance performance in generic architecture builds.
+- Support for processing videos without PTS information (#1278).
 - The feature to copy input streams' metadata to the output streams (#1282).
 
 ### Changed
diff --git a/include/libvideo2x/encoder.h b/include/libvideo2x/encoder.h
index 5e98ed8..fba93a5 100644
--- a/include/libvideo2x/encoder.h
+++ b/include/libvideo2x/encoder.h
@@ -65,7 +65,7 @@ class Encoder {
         int in_vstream_idx
     );
 
-    int write_frame(AVFrame* frame, int64_t frame_idx);
+    int write_frame(AVFrame* frame);
     int flush();
 
     AVCodecContext* get_encoder_context() const;
diff --git a/src/encoder.cpp b/src/encoder.cpp
index 1ddce69..7954e9c 100644
--- a/src/encoder.cpp
+++ b/src/encoder.cpp
@@ -269,16 +269,13 @@ int Encoder::init(
 }
 
 [[gnu::target_clones("arch=x86-64-v4", "arch=x86-64-v3", "default")]]
-int Encoder::write_frame(AVFrame* frame, int64_t frame_idx) {
+int Encoder::write_frame(AVFrame* frame) {
     AVFrame* converted_frame = nullptr;
     int ret;
 
     // Let the encoder decide the frame type
     frame->pict_type = AV_PICTURE_TYPE_NONE;
 
-    // Calculate this frame's presentation timestamp (PTS)
-    frame->pts = av_rescale_q(frame_idx, av_inv_q(enc_ctx_->framerate), enc_ctx_->time_base);
-
     // Convert the frame to the encoder's pixel format if needed
     if (frame->format != enc_ctx_->pix_fmt) {
         converted_frame = conversions::convert_avframe_pix_fmt(frame, enc_ctx_->pix_fmt);
diff --git a/src/libvideo2x.cpp b/src/libvideo2x.cpp
index 751e9c3..fbcc185 100644
--- a/src/libvideo2x.cpp
+++ b/src/libvideo2x.cpp
@@ -1,4 +1,5 @@
 #include "libvideo2x.h"
+#include <libavcodec/avcodec.h>
 
 extern "C" {
 #include <libavutil/avutil.h>
@@ -152,6 +153,7 @@ int VideoProcessor::process_frames(
     AVCodecContext* dec_ctx = decoder.get_codec_context();
     int in_vstream_idx = decoder.get_video_stream_index();
     AVFormatContext* ofmt_ctx = encoder.get_format_context();
+    AVCodecContext* enc_ctx = encoder.get_encoder_context();
     int* stream_map = encoder.get_stream_map();
 
     // Reference to the previous frame does not require allocation
@@ -235,6 +237,10 @@ int VideoProcessor::process_frames(
                     return ret;
                 }
 
+                // Calculate this frame's presentation timestamp (PTS)
+                frame->pts =
+                    av_rescale_q(frame_idx_, av_inv_q(enc_ctx->framerate), enc_ctx->time_base);
+
                 // Process the frame based on the selected processing mode
                 AVFrame* proc_frame = nullptr;
                 switch (processor->get_processing_mode()) {
@@ -308,7 +314,7 @@ int VideoProcessor::write_frame(AVFrame* frame, encoder::Encoder& encoder) {
     int ret = 0;
 
     if (!benchmark_) {
-        ret = encoder.write_frame(frame, frame_idx_);
+        ret = encoder.write_frame(frame);
         if (ret < 0) {
             av_strerror(ret, errbuf, sizeof(errbuf));
             logger()->critical("Error encoding/writing frame: {}", errbuf);