fix(realesrgan): fixed incorrect type for ncnn::Mat elemsize

Signed-off-by: k4yt3x <i@k4yt3x.com>
docs(video2x): updated command line help messages
2026-02-10 14:54:46 +08:00 · 2024-10-26 00:00:00 +00:00 · 2024-10-26 00:00:00 +00:00 · 2024-10-26 00:00:00 +00:00 · 2024-10-26 00:00:00 +00:00 · 2024-10-25 00:00:00 +00:00
43 changed files with 11179 additions and 339 deletions
--- a/.github/FUNDING.yml
+++ b/.github/FUNDING.yml
@@ -1,12 +1,2 @@
-# These are supported funding model platforms
-
 github: k4yt3x
 patreon: k4yt3x
-open_collective: # Replace with a single Open Collective username
-ko_fi: # Replace with a single Ko-fi username
-tidelift: # Replace with a single Tidelift platform-name/package-name e.g., npm/babel
-community_bridge: # Replace with a single Community Bridge project-name e.g., cloud-foundry
-liberapay: # Replace with a single Liberapay username
-issuehunt: # Replace with a single IssueHunt username
-otechie: # Replace with a single Otechie username
-custom: # Replace with up to 4 custom sponsorship URLs e.g., ['link1', 'link2']
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -30,7 +30,8 @@ jobs:
            libswscale-dev \
            libvulkan-dev \
            glslang-tools \
-            libomp-dev
+            libomp-dev \
+            libopencv-dev
      - name: Build Video2X
        run: |
          mkdir -p /tmp/build /tmp/install
@@ -63,6 +64,7 @@ jobs:
        run: |
          $ffmpegVersion = "7.1"
          $ncnnVersion = "20240820"
+          $opencvVersion = "4.10.0"

          git submodule update --init --recursive

@@ -73,9 +75,16 @@ jobs:
          curl -Lo ncnn-shared.zip "https://github.com/Tencent/ncnn/releases/download/$ncnnVersion/ncnn-$ncnnVersion-windows-vs2022-shared.zip"
          Expand-Archive -Path ncnn-shared.zip -DestinationPath third_party
          Rename-Item -Path "third_party/ncnn-$ncnnVersion-windows-vs2022-shared" -NewName ncnn-shared
+
+          curl -Lo opencv-shared.exe "https://github.com/opencv/opencv/releases/download/$opencvVersion/opencv-$opencvVersion-windows.exe"
+          7z x opencv-shared.exe -y
+          Move-Item -Path opencv -Destination third_party\opencv-shared
      - name: Build Video2X
+        shell: pwsh
        run: |
-          cmake -S . -B build -DUSE_SYSTEM_NCNN=OFF -DUSE_SYSTEM_SPDLOG=OFF -DCMAKE_BUILD_TYPE=Debug -DCMAKE_INSTALL_PREFIX=build/video2x_install
+          cmake -S . -B build `
+            -DUSE_SYSTEM_NCNN=OFF -DUSE_SYSTEM_SPDLOG=OFF -DUSE_SYSTEM_OPENCV=OFF `
+            -DCMAKE_BUILD_TYPE=Debug -DCMAKE_INSTALL_PREFIX=build/video2x_install
          cmake --build build --config Debug --parallel --target install
      - name: Upload artifacts
        uses: actions/upload-artifact@v4
--- a/.gitmodules
+++ b/.gitmodules
@@ -7,3 +7,6 @@
 [submodule "third_party/spdlog"]
 	path = third_party/spdlog
 	url = https://github.com/gabime/spdlog.git
+[submodule "third_party/opencv"]
+	path = third_party/opencv
+	url = https://github.com/opencv/opencv.git
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -1,5 +1,5 @@
 cmake_minimum_required(VERSION 3.10)
-project(video2x VERSION 6.0.0 LANGUAGES CXX C)
+project(video2x VERSION 6.0.0 LANGUAGES C CXX)

 # Set the C standard
 set(CMAKE_C_STANDARD 11)
@@ -28,17 +28,25 @@ if(CMAKE_BUILD_TYPE STREQUAL "Release")
    endif()
 endif()

+# Set global compile options for all targets
+if (CMAKE_CXX_COMPILER_ID STREQUAL "MSVC")
+    add_compile_options(/W4 /permissive-)
+elseif (CMAKE_CXX_COMPILER_ID STREQUAL "GNU" OR CMAKE_CXX_COMPILER_ID MATCHES "Clang")
+    add_compile_options(-Wall -Wextra -Wpedantic -Wconversion -Wshadow)
+endif()
+
 # Build options
 option(BUILD_SHARED_LIBS "Build libvideo2x as a shared library" ON)
 option(BUILD_VIDEO2X_CLI "Build the video2x executable" ON)
 option(USE_SYSTEM_SPDLOG "Use system spdlog library" ON)
+option(USE_SYSTEM_OPENCV "Use system OpenCV library" ON)
 option(USE_SYSTEM_NCNN "Use system ncnn library" ON)

 # Generate the version header file
 configure_file(
-  "${CMAKE_CURRENT_SOURCE_DIR}/include/libvideo2x/version.h.in"
-  "${CMAKE_CURRENT_BINARY_DIR}/libvideo2x/version.h"
-  @ONLY
+    "${CMAKE_CURRENT_SOURCE_DIR}/include/libvideo2x/version.h.in"
+    "${CMAKE_CURRENT_BINARY_DIR}/libvideo2x/version.h"
+    @ONLY
 )

 # Find the required packages
@@ -48,6 +56,7 @@ set(ALL_LIBRARIES)
 # spdlog
 if (USE_SYSTEM_SPDLOG)
    find_package(spdlog REQUIRED)
+    list(APPEND ALL_INCLUDE_DIRS ${spdlog_INCLUDE_DIRS})
    list(APPEND ALL_LIBRARIES spdlog::spdlog)
 else()
    add_subdirectory(third_party/spdlog)
@@ -59,6 +68,7 @@ if(WIN32)
    # Define base paths for FFmpeg and ncnn
    set(FFMPEG_BASE_PATH ${PROJECT_SOURCE_DIR}/third_party/ffmpeg-shared)
    set(NCNN_BASE_PATH ${PROJECT_SOURCE_DIR}/third_party/ncnn-shared/x64)
+    set(OPENCV_BASE_PATH ${PROJECT_SOURCE_DIR}/third_party/opencv-shared)

    # FFmpeg
    list(APPEND ALL_LIBRARIES
@@ -73,7 +83,9 @@ if(WIN32)

    # ncnn
    # TODO: Figure out why this file is not being copied to the install directory
-    set(SPIRV_BUILD_PATH ${CMAKE_BINARY_DIR}/realesrgan-prefix/src/realesrgan-build/ncnn/glslang/SPIRV)
+    set(SPIRV_BUILD_PATH
+        ${CMAKE_BINARY_DIR}/realesrgan-prefix/src/realesrgan-build/ncnn/glslang/SPIRV
+    )
    if (CMAKE_BUILD_TYPE STREQUAL "Release")
        set(SPIRV_LIB ${SPIRV_BUILD_PATH}/Release/SPIRV.lib)
    else()
@@ -85,8 +97,12 @@ if(WIN32)
        ${SPIRV_LIB}
    )
    list(APPEND ALL_INCLUDE_DIRS ${NCNN_BASE_PATH}/include/ncnn)
+
+    # OpenCV
+    list(APPEND ALL_LIBRARIES ${OPENCV_BASE_PATH}/build/x64/vc16/lib/opencv_world4100.lib)
+    list(APPEND ALL_INCLUDE_DIRS ${OPENCV_BASE_PATH}/build/include)
 else()
-    # Find the required packages using pkg-config
+    # FFmpeg
    find_package(PkgConfig REQUIRED)
    set(REQUIRED_PKGS
        libavcodec
@@ -103,11 +119,55 @@ else()
        list(APPEND ALL_INCLUDE_DIRS ${${PKG}_INCLUDE_DIRS})
        list(APPEND ALL_LIBRARIES ${${PKG}_LIBRARIES})
    endforeach()
-endif()

-# Remove duplicate entries
-list(REMOVE_DUPLICATES ALL_INCLUDE_DIRS)
-list(REMOVE_DUPLICATES ALL_LIBRARIES)
+    # OpenCV
+    if (USE_SYSTEM_OPENCV)
+        find_package(OpenCV REQUIRED)
+        list(APPEND ALL_INCLUDE_DIRS ${OpenCV_INCLUDE_DIRS}/opencv2)
+        list(APPEND ALL_LIBRARIES opencv_core opencv_videoio)
+    else()
+        option(BUILD_opencv_calib3d "" OFF)
+        option(BUILD_opencv_core "" ON)
+        option(BUILD_opencv_dnn "" OFF)
+        option(BUILD_opencv_features2d "" OFF)
+        option(BUILD_opencv_flann "" OFF)
+        option(BUILD_opencv_gapi "" OFF)
+        option(BUILD_opencv_highgui "" OFF)
+        option(BUILD_opencv_imgcodecs "" OFF)
+        option(BUILD_opencv_imgproc "" OFF)
+        option(BUILD_opencv_java "" OFF)
+        option(BUILD_opencv_js "" OFF)
+        option(BUILD_opencv_ml "" OFF)
+        option(BUILD_opencv_objc "" OFF)
+        option(BUILD_opencv_objdetect "" OFF)
+        option(BUILD_opencv_photo "" OFF)
+        option(BUILD_opencv_python "" OFF)
+        option(BUILD_opencv_python2 "" OFF)
+        option(BUILD_opencv_python3 "" OFF)
+        option(BUILD_opencv_stitching "" OFF)
+        option(BUILD_opencv_ts "" OFF)
+        option(BUILD_opencv_video "" OFF)
+        option(BUILD_opencv_videoio "" ON)
+        option(BUILD_opencv_world "" OFF)
+
+        option(WITH_PNG "" OFF)
+        option(WITH_JPEG "" OFF)
+        option(WITH_TIFF "" OFF)
+        option(WITH_WEBP "" OFF)
+        option(WITH_OPENJPEG "" OFF)
+        option(WITH_JASPER "" OFF)
+        option(WITH_OPENEXR "" OFF)
+        option(WITH_IMGCODEC "" OFF)
+
+        add_subdirectory(third_party/opencv)
+        list(APPEND ALL_INCLUDE_DIRS
+            ${PROJECT_SOURCE_DIR}/third_party/opencv/include
+            ${PROJECT_SOURCE_DIR}/third_party/opencv/modules/core/include
+            ${PROJECT_SOURCE_DIR}/third_party/opencv/modules/videoio/include
+        )
+        list(APPEND ALL_LIBRARIES opencv_core opencv_videoio)
+    endif() # USE_SYSTEM_OPENCV
+endif() # WIN32

 # Find ncnn package
 if(USE_SYSTEM_NCNN)
@@ -224,6 +284,10 @@ ExternalProject_Add(
    INSTALL_COMMAND ${CMAKE_COMMAND} --build . --target install --config ${CMAKE_BUILD_TYPE}
 )

+# Remove duplicate entries
+list(REMOVE_DUPLICATES ALL_INCLUDE_DIRS)
+list(REMOVE_DUPLICATES ALL_LIBRARIES)
+
 # Add all source files for libvideo2x
 file(GLOB LIBVIDEO2X_SOURCES src/*.cpp)

@@ -250,7 +314,6 @@ target_include_directories(libvideo2x PRIVATE

 # Compile options for the shared library
 target_compile_options(libvideo2x PRIVATE
-    -Wall
    -fPIC
    $<$<CONFIG:Release>:-Ofast>
    $<$<CONFIG:Debug>:-g -DDEBUG>
@@ -287,10 +350,7 @@ if (BUILD_VIDEO2X_CLI)
    )

    # Compile options for the executable
-    target_compile_options(video2x PRIVATE
-        -Wall
-        $<$<CONFIG:Debug>:-g -DDEBUG>
-    )
+    target_compile_options(video2x PRIVATE $<$<CONFIG:Debug>:-g -DDEBUG>)

    # Link the executable with the shared library
    target_link_libraries(video2x PRIVATE ${ALL_LIBRARIES} libvideo2x)
@@ -309,6 +369,10 @@ else()
    set(MODEL_DESTINATION_DEFAULT "share/video2x")
 endif()

+# Explicitly set the output paths since OpenCV changes these variables
+set(EXECUTABLE_OUTPUT_PATH ${CMAKE_BINARY_DIR})
+set(LIBRARY_OUTPUT_PATH ${CMAKE_BINARY_DIR})
+
 # Set the installation directories
 set(INSTALL_BIN_DESTINATION ${BIN_DESTINATION_DEFAULT} CACHE STRING "")
 set(INSTALL_INCLUDE_DESTINATION ${INCLUDE_DESTINATION_DEFAULT} CACHE STRING "")
@@ -350,14 +414,36 @@ if(WIN32)
                    GROUP_READ GROUP_EXECUTE
                    WORLD_READ WORLD_EXECUTE
    )
-    install(FILES ${NCNN_BASE_PATH}/bin/ncnn.dll DESTINATION ${INSTALL_BIN_DESTINATION}
+    install(FILES ${FFMPEG_BASE_PATH}/bin/swscale-8.dll
+                  ${FFMPEG_BASE_PATH}/bin/avcodec-61.dll
+                  ${FFMPEG_BASE_PATH}/bin/avdevice-61.dll
+                  ${FFMPEG_BASE_PATH}/bin/avfilter-10.dll
+                  ${FFMPEG_BASE_PATH}/bin/avformat-61.dll
+                  ${FFMPEG_BASE_PATH}/bin/avutil-59.dll
+                  ${FFMPEG_BASE_PATH}/bin/postproc-58.dll
+                  ${FFMPEG_BASE_PATH}/bin/swresample-5.dll
+        DESTINATION ${INSTALL_BIN_DESTINATION}
+        PERMISSIONS OWNER_READ OWNER_WRITE OWNER_EXECUTE
+                    GROUP_READ GROUP_EXECUTE
+                    WORLD_READ WORLD_EXECUTE
+    )
+    install(FILES ${OPENCV_BASE_PATH}/build/x64/vc16/bin/opencv_world4100.dll
+                  ${OPENCV_BASE_PATH}/build/x64/vc16/bin/opencv_videoio_msmf4100_64.dll
+        DESTINATION ${INSTALL_BIN_DESTINATION}
+        PERMISSIONS OWNER_READ OWNER_WRITE OWNER_EXECUTE
+                    GROUP_READ GROUP_EXECUTE
+                    WORLD_READ WORLD_EXECUTE
+    )
+    install(FILES ${NCNN_BASE_PATH}/bin/ncnn.dll
+        DESTINATION ${INSTALL_BIN_DESTINATION}
        PERMISSIONS OWNER_READ OWNER_WRITE OWNER_EXECUTE
                    GROUP_READ GROUP_EXECUTE
                    WORLD_READ WORLD_EXECUTE
    )
 else()
    # Install Unix-specific dependencies
-    install(FILES ${REALESRGAN_LIB} DESTINATION ${INSTALL_LIB_DESTINATION}
+    install(FILES ${REALESRGAN_LIB}
+        DESTINATION ${INSTALL_LIB_DESTINATION}
        PERMISSIONS OWNER_READ OWNER_WRITE OWNER_EXECUTE
                    GROUP_READ GROUP_EXECUTE
                    WORLD_READ WORLD_EXECUTE
--- a/12
+++ b/12
@@ -1,14 +1,15 @@
 # Name: Video2X Dockerfile
 # Creator: K4YT3X
 # Date Created: February 3, 2022
-# Last Modified: October 7, 2024
+# Last Modified: October 21, 2024

 # stage 1: build the python components into wheels
 FROM docker.io/archlinux:latest AS builder

 # Install dependencies and create a non-root user
 RUN pacman -Syy --noconfirm \
-        base-devel ffmpeg ncnn git cmake make clang pkgconf vulkan-headers openmp sudo \
+        base-devel git cmake make clang pkgconf sudo \
+        ffmpeg ncnn vulkan-headers openmp spdlog opencv \
        nvidia-utils vulkan-radeon vulkan-intel vulkan-swrast \
    && useradd -m builder \
    && echo 'builder ALL=(ALL) NOPASSWD: ALL' > /etc/sudoers.d/builder
@@ -31,11 +32,12 @@ LABEL maintainer="K4YT3X <i@k4yt3x.com>" \

 ENV VK_ICD_FILENAMES=/usr/share/vulkan/icd.d/nvidia_icd.json\
 :/usr/share/vulkan/icd.d/radeon_icd.x86_64.json\
-:/usr/share/vulkan/icd.d/intel_icd.x86_64.json
+:/usr/share/vulkan/icd.d/intel_icd.x86_64.json\
+:/usr/share/vulkan/icd.d/lvp_icd.x86_64.json

 COPY --from=builder /tmp/video2x.pkg.tar.zst /video2x.pkg.tar.zst
-RUN pacman -Sy --noconfirm ffmpeg ncnn \
-        nvidia-utils vulkan-radeon vulkan-intel vulkan-swrast \
+RUN pacman -Sy --noconfirm nvidia-utils vulkan-radeon vulkan-intel vulkan-swrast \
+        ffmpeg ncnn spdlog opencv \
    && pacman -U --noconfirm /video2x.pkg.tar.zst \
    && rm -rf /video2x.pkg.tar.zst /var/cache/pacman/pkg/*

--- a/47
+++ b/47
@@ -1,9 +1,15 @@
-.PHONY: build static debug windows test-realesrgan test-libplacebo leakcheck clean
+.PHONY: build static debug clean \
+	test-realesrgan test-libplacebo \
+	memcheck-realesrgan memcheck-libplacebo \
+	heaptrack-realesrgan heaptrack-libplacebo

 BINDIR=build
 CC=clang
 CXX=clang++

+TEST_VIDEO=data/standard-test.mp4
+TEST_OUTPUT=data/output.mp4
+
 build:
 	cmake -S . -B $(BINDIR) \
 		-DCMAKE_EXPORT_COMPILE_COMMANDS=ON \
@@ -45,23 +51,27 @@ debian:
 		libvulkan-dev \
 		glslang-tools \
 		libomp-dev \
-		libspdlog-dev
+		libspdlog-dev \
+		libopencv-dev
 	cmake -B /tmp/build -S . -DUSE_SYSTEM_NCNN=OFF \
-		-DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ \
+		-DCMAKE_C_COMPILER=$(CC) -DCMAKE_CXX_COMPILER=$(CXX) \
 		-DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/tmp/install \
 		-DINSTALL_BIN_DESTINATION=. -DINSTALL_INCLUDE_DESTINATION=include \
 		-DINSTALL_LIB_DESTINATION=. -DINSTALL_MODEL_DESTINATION=.
 	cmake --build /tmp/build --config Release --target install --parallel

+clean:
+	rm -rf $(BINDIR)
+
 test-realesrgan:
-	LD_LIBRARY_PATH=$(BINDIR) $(BINDIR)/video2x -i data/standard-test.mp4 -o data/output.mp4 \
-		-f realesrgan -r 4 --model realesr-animevideov3
+	LD_LIBRARY_PATH=$(BINDIR) $(BINDIR)/video2x -i $(TEST_VIDEO) -o $(TEST_OUTPUT) \
+		-f realesrgan -r 4 -m realesr-animevideov3

 test-libplacebo:
-	LD_LIBRARY_PATH=$(BINDIR) $(BINDIR)/video2x -i data/standard-test.mp4 -o data/output.mp4 \
+	LD_LIBRARY_PATH=$(BINDIR) $(BINDIR)/video2x -i $(TEST_VIDEO) -o $(TEST_OUTPUT) \
 		-f libplacebo -w 1920 -h 1080 -s anime4k-mode-a

-leakcheck-realesrgan:
+memcheck-realesrgan:
 	LD_LIBRARY_PATH=$(BINDIR) valgrind \
 		--tool=memcheck \
 		--leak-check=full \
@@ -70,11 +80,11 @@ leakcheck-realesrgan:
 		--show-reachable=yes \
 		--verbose --log-file="valgrind.log" \
 		$(BINDIR)/video2x \
-		-i data/standard-test.mp4 -o data/output.mp4 \
-		-f realesrgan -r 2 --model realesr-animevideov3 \
+		-i $(TEST_VIDEO) -o $(TEST_OUTPUT) \
+		-f realesrgan -r 2 -m realesr-animevideov3 \
 		-p veryfast -b 1000000 -q 30

-leakcheck-libplacebo:
+memcheck-libplacebo:
 	LD_LIBRARY_PATH=$(BINDIR) valgrind \
 		--tool=memcheck \
 		--leak-check=full \
@@ -83,9 +93,20 @@ leakcheck-libplacebo:
 		--show-reachable=yes \
 		--verbose --log-file="valgrind.log" \
 		$(BINDIR)/video2x \
-		-i data/standard-test.mp4 -o data/output.mp4 \
+		-i $(TEST_VIDEO) -o $(TEST_OUTPUT) \
 		-f libplacebo -w 1920 -h 1080 -s anime4k-mode-a \
 		-p veryfast -b 1000000 -q 30

-clean:
-	rm -rf $(BINDIR)
+heaptrack-realesrgan:
+	LD_LIBRARY_PATH=$(BINDIR) HEAPTRACK_ENABLE_DEBUGINFOD=1 heaptrack \
+		$(BINDIR)/video2x \
+		-i $(TEST_VIDEO) -o $(TEST_OUTPUT) \
+		-f realesrgan -r 4 -m realesr-animevideov3 \
+		-p veryfast -b 1000000 -q 30
+
+heaptrack-libplacebo:
+	LD_LIBRARY_PATH=$(BINDIR) HEAPTRACK_ENABLE_DEBUGINFOD=1 heaptrack \
+		$(BINDIR)/video2x \
+		-i $(TEST_VIDEO) -o $(TEST_OUTPUT) \
+		-f libplacebo -w 1920 -h 1080 -s anime4k-mode-a \
+		-p veryfast -b 1000000 -q 30
--- a/6
+++ b/6
@@ -1,12 +1,12 @@
 pkgname=video2x
-pkgver=r840.ecbc512
+pkgver=r843.e09f348
 pkgrel=1
 pkgdesc="A machine learning-based lossless video super resolution framework"
 arch=('x86_64')
 url="https://github.com/k4yt3x/video2x"
 license=('AGPL3')
-depends=('ffmpeg' 'ncnn' 'vulkan-driver')
-makedepends=('git' 'cmake' 'make' 'clang' 'pkgconf' 'vulkan-headers' 'openmp' 'spdlog')
+depends=('ffmpeg' 'ncnn' 'vulkan-driver' 'opencv' 'spdlog')
+makedepends=('git' 'cmake' 'make' 'clang' 'pkgconf' 'vulkan-headers' 'openmp')

 pkgver() {
    printf "r%s.%s" "$(git rev-list --count HEAD)" "$(git rev-parse --short HEAD)"
--- a/README.md
+++ b/README.md
@@ -13,9 +13,9 @@

 ## 🌟 Version 6.0.0 Preview

-**[Direct download link for Windows (Installer)](https://github.com/k4yt3x/video2x/releases/download/6.0.0-beta.2/video2x-qt6-windows-amd64-installer.exe)**
+**[Direct download link for Windows (Installer)](https://github.com/k4yt3x/video2x/releases/download/6.0.0-beta.4/video2x-qt6-windows-amd64-installer.exe)**

-![6.0.0-beta-screenshot](https://github.com/user-attachments/assets/bde4e4e2-2f97-412f-8e34-848f384be720)
+![6.0.0-beta-screenshot](https://github.com/user-attachments/assets/bd9e0b82-67ee-4b62-a830-a694db6f5de3)

 Version 6.0.0 is a complete rewrite of this project in C/C++. It:

--- a/include/libvideo2x/decoder.h
+++ b/include/libvideo2x/decoder.h
@@ -9,10 +9,10 @@ extern "C" {
 int init_decoder(
    AVHWDeviceType hw_type,
    AVBufferRef *hw_ctx,
-    const char *input_filename,
+    const char *in_fname,
    AVFormatContext **fmt_ctx,
    AVCodecContext **dec_ctx,
-    int *video_stream_index
+    int *vstream_idx
 );

 #endif  // DECODER_H
--- a/include/libvideo2x/encoder.h
+++ b/include/libvideo2x/encoder.h
@@ -11,21 +11,21 @@ extern "C" {

 int init_encoder(
    AVBufferRef *hw_ctx,
-    const char *output_filename,
+    const char *out_fname,
    AVFormatContext *ifmt_ctx,
    AVFormatContext **ofmt_ctx,
    AVCodecContext **enc_ctx,
    AVCodecContext *dec_ctx,
    EncoderConfig *encoder_config,
-    int video_stream_index,
-    int **stream_mapping
+    int vstream_idx,
+    int **stream_map
 );

 int encode_and_write_frame(
    AVFrame *frame,
    AVCodecContext *enc_ctx,
    AVFormatContext *ofmt_ctx,
-    int video_stream_index
+    int vstream_idx
 );

 int flush_encoder(AVCodecContext *enc_ctx, AVFormatContext *ofmt_ctx);
--- a/include/libvideo2x/filter.h
+++ b/include/libvideo2x/filter.h
@@ -12,10 +12,10 @@ extern "C" {
 // Abstract base class for filters
 class Filter {
   public:
-    virtual ~Filter() {}
+    virtual ~Filter() = default;
    virtual int init(AVCodecContext *dec_ctx, AVCodecContext *enc_ctx, AVBufferRef *hw_ctx) = 0;
-    virtual int process_frame(AVFrame *input_frame, AVFrame **output_frame) = 0;
-    virtual int flush(std::vector<AVFrame *> &processed_frames) = 0;
+    virtual int process_frame(AVFrame *in_frame, AVFrame **out_frame) = 0;
+    virtual int flush(std::vector<AVFrame *> &flushed_frames) { return 0; }
 };

 #endif  // FILTER_H
--- a/include/libvideo2x/libplacebo.h
+++ b/include/libvideo2x/libplacebo.h
@@ -14,8 +14,8 @@ int init_libplacebo(
    AVFilterContext **buffersrc_ctx,
    AVFilterContext **buffersink_ctx,
    AVCodecContext *dec_ctx,
-    int output_width,
-    int output_height,
+    int out_width,
+    int out_height,
    const std::filesystem::path &shader_path
 );

--- a/include/libvideo2x/libplacebo_filter.h
+++ b/include/libvideo2x/libplacebo_filter.h
@@ -17,26 +17,27 @@ class LibplaceboFilter : public Filter {
    AVFilterGraph *filter_graph;
    AVFilterContext *buffersrc_ctx;
    AVFilterContext *buffersink_ctx;
-    int output_width;
-    int output_height;
+    int out_width;
+    int out_height;
    const std::filesystem::path shader_path;
-    AVRational output_time_base;
+    AVRational in_time_base;
+    AVRational out_time_base;

   public:
    // Constructor
    LibplaceboFilter(int width, int height, const std::filesystem::path &shader_path);

    // Destructor
-    virtual ~LibplaceboFilter();
+    virtual ~LibplaceboFilter() override;

    // Initializes the filter with decoder and encoder contexts
    int init(AVCodecContext *dec_ctx, AVCodecContext *enc_ctx, AVBufferRef *hw_ctx) override;

    // Processes an input frame and returns the processed frame
-    int process_frame(AVFrame *input_frame, AVFrame **output_frame) override;
+    int process_frame(AVFrame *in_frame, AVFrame **out_frame) override;

    // Flushes any remaining frames
-    int flush(std::vector<AVFrame *> &processed_frames) override;
+    int flush(std::vector<AVFrame *> &flushed_frames) override;
 };

 #endif  // LIBPLACEBO_FILTER_H
--- a/include/libvideo2x/libvideo2x.h
+++ b/include/libvideo2x/libvideo2x.h
@@ -41,8 +41,8 @@ enum Libvideo2xLogLevel {

 // Configuration for Libplacebo filter
 struct LibplaceboConfig {
-    int output_width;
-    int output_height;
+    int out_width;
+    int out_height;
    const char *shader_path;
 };

@@ -65,8 +65,8 @@ struct FilterConfig {

 // Encoder configuration
 struct EncoderConfig {
-    int output_width;
-    int output_height;
+    int out_width;
+    int out_height;
    bool copy_streams;
    enum AVCodecID codec;
    enum AVPixelFormat pix_fmt;
@@ -87,8 +87,8 @@ struct VideoProcessingContext {

 // C-compatible process_video function
 LIBVIDEO2X_API int process_video(
-    const char *input_filename,
-    const char *output_filename,
+    const char *in_fname,
+    const char *out_fname,
    enum Libvideo2xLogLevel log_level,
    bool benchmark,
    enum AVHWDeviceType hw_device_type,
--- a/include/libvideo2x/realesrgan_filter.h
+++ b/include/libvideo2x/realesrgan_filter.h
@@ -20,9 +20,9 @@ class RealesrganFilter : public Filter {
    const char *model;
    const std::filesystem::path custom_model_param_path;
    const std::filesystem::path custom_model_bin_path;
-    AVRational input_time_base;
-    AVRational output_time_base;
-    AVPixelFormat output_pix_fmt;
+    AVRational in_time_base;
+    AVRational out_time_base;
+    AVPixelFormat out_pix_fmt;

   public:
    // Constructor
@@ -31,21 +31,18 @@ class RealesrganFilter : public Filter {
        bool tta_mode = false,
        int scaling_factor = 4,
        const char *model = "realesr-animevideov3",
-        const std::filesystem::path custom_model_bin_pathmodel_param_path = std::filesystem::path(),
-        const std::filesystem::path custom_model_bin_pathmodel_bin_path = std::filesystem::path()
+        const std::filesystem::path custom_model_param_path = std::filesystem::path(),
+        const std::filesystem::path custom_model_bin_path = std::filesystem::path()
    );

    // Destructor
-    virtual ~RealesrganFilter();
+    virtual ~RealesrganFilter() override;

    // Initializes the filter with decoder and encoder contexts
    int init(AVCodecContext *dec_ctx, AVCodecContext *enc_ctx, AVBufferRef *hw_ctx) override;

    // Processes an input frame and returns the processed frame
-    int process_frame(AVFrame *input_frame, AVFrame **output_frame) override;
-
-    // Flushes any remaining frames (if necessary)
-    int flush(std::vector<AVFrame *> &processed_frames) override;
+    int process_frame(AVFrame *in_frame, AVFrame **out_frame) override;
 };

 #endif
--- a/models/libplacebo/anime4k-a+a.glsl
+++ b/models/libplacebo/anime4k-a+a.glsl
--- a/models/libplacebo/anime4k-a.glsl
+++ b/models/libplacebo/anime4k-a.glsl
--- a/models/libplacebo/anime4k-b+b.glsl
+++ b/models/libplacebo/anime4k-b+b.glsl
--- a/models/libplacebo/anime4k-b.glsl
+++ b/models/libplacebo/anime4k-b.glsl
--- a/models/libplacebo/anime4k-c+a.glsl
+++ b/models/libplacebo/anime4k-c+a.glsl
--- a/models/libplacebo/anime4k-c.glsl
+++ b/models/libplacebo/anime4k-c.glsl
--- a/models/realesrgan/realesr-animevideov3-x2.bin
+++ b/models/realesrgan/realesr-animevideov3-x2.bin
--- a/models/realesrgan/realesr-animevideov3-x2.param
+++ b/models/realesrgan/realesr-animevideov3-x2.param
--- a/models/realesrgan/realesr-animevideov3-x3.bin
+++ b/models/realesrgan/realesr-animevideov3-x3.bin
--- a/models/realesrgan/realesr-animevideov3-x3.param
+++ b/models/realesrgan/realesr-animevideov3-x3.param
--- a/models/realesrgan/realesr-animevideov3-x4.bin
+++ b/models/realesrgan/realesr-animevideov3-x4.bin
--- a/models/realesrgan/realesr-animevideov3-x4.param
+++ b/models/realesrgan/realesr-animevideov3-x4.param
--- a/models/realesrgan/realesrgan-plus-anime-x4.bin
+++ b/models/realesrgan/realesrgan-plus-anime-x4.bin
--- a/models/realesrgan/realesrgan-plus-anime-x4.param
+++ b/models/realesrgan/realesrgan-plus-anime-x4.param
--- a/models/realesrgan/realesrgan-plus-x4.bin
+++ b/models/realesrgan/realesrgan-plus-x4.bin
--- a/models/realesrgan/realesrgan-plus-x4.param
+++ b/models/realesrgan/realesrgan-plus-x4.param
--- a/scripts/download_anime4k_glsl.py
+++ b/scripts/download_anime4k_glsl.py
@@ -1,46 +0,0 @@
-#!/usr/bin/python
-# -*- coding: utf-8 -*-
-import shutil
-from pathlib import Path
-
-import requests
-
-ANIME4K_COMMIT = "master"
-GITHUB_GLSL_ROOT = (
-    f"https://raw.githubusercontent.com/bloc97/Anime4K/{ANIME4K_COMMIT}/glsl"
-)
-SHADERS_DIR = Path(__file__).parent.parent / "data"
-
-
-def download_and_combine_files():
-
-    modes = {
-        "ModeA": [
-            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Clamp_Highlights.glsl",
-            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Restore_CNN_VL.glsl",
-            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_VL.glsl",
-            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x2.glsl",
-            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x4.glsl",
-            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_M.glsl",
-        ]
-    }
-
-    for mode in modes:
-        file_contents = ""
-        for file in modes[mode]:
-            response = requests.get(file, timeout=5)
-            response.raise_for_status()
-            file_contents += response.text + "\n"
-
-        with (SHADERS_DIR / Path(f"Anime4K_{mode}.glsl")).open("w") as output_file:
-            output_file.write(file_contents)
-
-
-if __name__ == "__main__":
-    # clear shaders directory
-    if SHADERS_DIR.exists():
-        shutil.rmtree(SHADERS_DIR)
-    SHADERS_DIR.mkdir(exist_ok=True)
-
-    # download and combine shaders
-    download_and_combine_files()
--- a/scripts/download_merge_anime4k_glsl.py
+++ b/scripts/download_merge_anime4k_glsl.py
@@ -0,0 +1,86 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+import shutil
+from pathlib import Path
+
+import requests
+
+ANIME4K_COMMIT = "master"
+GITHUB_GLSL_ROOT = (
+    f"https://raw.githubusercontent.com/bloc97/Anime4K/{ANIME4K_COMMIT}/glsl"
+)
+SHADERS_DIR = Path(__file__).parent.parent / "models" / "libplacebo"
+
+
+def download_and_combine_files():
+    modes = {
+        "a": [
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Clamp_Highlights.glsl",
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Restore_CNN_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x2.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x4.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_M.glsl",
+        ],
+        "b": [
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Clamp_Highlights.glsl",
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Restore_CNN_Soft_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x2.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x4.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_M.glsl",
+        ],
+        "c": [
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Clamp_Highlights.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale+Denoise/Anime4K_Upscale_Denoise_CNN_x2_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x2.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x4.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_M.glsl",
+        ],
+        "a+a": [
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Clamp_Highlights.glsl",
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Restore_CNN_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Restore_CNN_M.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x2.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x4.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_M.glsl",
+        ],
+        "b+b": [
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Clamp_Highlights.glsl",
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Restore_CNN_Soft_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x2.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x4.glsl",
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Restore_CNN_Soft_M.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_M.glsl",
+        ],
+        "c+a": [
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Clamp_Highlights.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale+Denoise/Anime4K_Upscale_Denoise_CNN_x2_VL.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x2.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_AutoDownscalePre_x4.glsl",
+            f"{GITHUB_GLSL_ROOT}/Restore/Anime4K_Restore_CNN_M.glsl",
+            f"{GITHUB_GLSL_ROOT}/Upscale/Anime4K_Upscale_CNN_x2_M.glsl",
+        ],
+    }
+
+    for mode in modes:
+        file_contents = ""
+        for file in modes[mode]:
+            response = requests.get(file, timeout=5)
+            response.raise_for_status()
+            file_contents += response.text + "\n"
+
+        with (SHADERS_DIR / Path(f"anime4k-{mode}.glsl")).open("w") as output_file:
+            output_file.write(file_contents)
+
+
+if __name__ == "__main__":
+    # clear shaders directory
+    if SHADERS_DIR.exists():
+        shutil.rmtree(SHADERS_DIR)
+    SHADERS_DIR.mkdir(exist_ok=True)
+
+    # download and combine shaders
+    download_and_combine_files()
--- a/src/conversions.cpp
+++ b/src/conversions.cpp
@@ -1,5 +1,6 @@
 #include "conversions.h"

+#include <cstddef>
 #include <cstdio>

 #include <spdlog/spdlog.h>
@@ -79,14 +80,16 @@ ncnn::Mat avframe_to_ncnn_mat(AVFrame *frame) {
    // Allocate a new ncnn::Mat and copy the data
    int width = converted_frame->width;
    int height = converted_frame->height;
-    ncnn::Mat ncnn_image = ncnn::Mat(width, height, (size_t)3, 3);  // BGR has 3 channels
+    ncnn::Mat ncnn_image = ncnn::Mat(width, height, static_cast<size_t>(3), 3);

    // Manually copy the pixel data from AVFrame to the new ncnn::Mat
    const uint8_t *src_data = converted_frame->data[0];
    for (int y = 0; y < height; y++) {
        uint8_t *dst_row = ncnn_image.row<uint8_t>(y);
        const uint8_t *src_row = src_data + y * converted_frame->linesize[0];
-        memcpy(dst_row, src_row, width * 3);  // Copy 3 channels (BGR) per pixel
+
+        // Copy 3 channels (BGR) per pixel
+        memcpy(dst_row, src_row, static_cast<size_t>(width) * 3);
    }

    // If we allocated a converted frame, free it
@@ -143,7 +146,9 @@ AVFrame *ncnn_mat_to_avframe(const ncnn::Mat &mat, AVPixelFormat pix_fmt) {
    for (int y = 0; y < mat.h; y++) {
        uint8_t *dst_row = bgr_frame->data[0] + y * bgr_frame->linesize[0];
        const uint8_t *src_row = mat.row<const uint8_t>(y);
-        memcpy(dst_row, src_row, mat.w * 3);  // Copy 3 channels (BGR) per pixel
+
+        // Copy 3 channels (BGR) per pixel
+        memcpy(dst_row, src_row, static_cast<size_t>(mat.w) * 3);
    }

    // Step 3: Convert the BGR frame to the desired pixel format
--- a/src/decoder.cpp
+++ b/src/decoder.cpp
@@ -9,7 +9,7 @@
 static enum AVPixelFormat hw_pix_fmt = AV_PIX_FMT_NONE;

 // Callback function to choose the hardware-accelerated pixel format
-static enum AVPixelFormat get_hw_format(AVCodecContext *ctx, const enum AVPixelFormat *pix_fmts) {
+static enum AVPixelFormat get_hw_format(AVCodecContext *_, const enum AVPixelFormat *pix_fmts) {
    for (const enum AVPixelFormat *p = pix_fmts; *p != AV_PIX_FMT_NONE; p++) {
        if (*p == hw_pix_fmt) {
            return *p;
@@ -22,17 +22,17 @@ static enum AVPixelFormat get_hw_format(AVCodecContext *ctx, const enum AVPixelF
 int init_decoder(
    AVHWDeviceType hw_type,
    AVBufferRef *hw_ctx,
-    const char *input_filename,
+    const char *in_fname,
    AVFormatContext **fmt_ctx,
    AVCodecContext **dec_ctx,
-    int *video_stream_index
+    int *vstream_idx
 ) {
    AVFormatContext *ifmt_ctx = NULL;
    AVCodecContext *codec_ctx = NULL;
    int ret;

-    if ((ret = avformat_open_input(&ifmt_ctx, input_filename, NULL, NULL)) < 0) {
-        spdlog::error("Could not open input file '{}'", input_filename);
+    if ((ret = avformat_open_input(&ifmt_ctx, in_fname, NULL, NULL)) < 0) {
+        spdlog::error("Could not open input file '{}'", in_fname);
        return ret;
    }

@@ -55,7 +55,8 @@ int init_decoder(
    const AVCodec *decoder = avcodec_find_decoder(video_stream->codecpar->codec_id);
    if (!decoder) {
        spdlog::error(
-            "Failed to find decoder for codec ID {}", (int)video_stream->codecpar->codec_id
+            "Failed to find decoder for codec ID {}",
+            static_cast<int>(video_stream->codecpar->codec_id)
        );
        return AVERROR_DECODER_NOT_FOUND;
    }
@@ -109,7 +110,7 @@ int init_decoder(

    *fmt_ctx = ifmt_ctx;
    *dec_ctx = codec_ctx;
-    *video_stream_index = stream_index;
+    *vstream_idx = stream_index;

    return 0;
 }
--- a/src/encoder.cpp
+++ b/src/encoder.cpp
@@ -19,21 +19,21 @@ static enum AVPixelFormat get_encoder_default_pix_fmt(const AVCodec *encoder) {

 int init_encoder(
    AVBufferRef *hw_ctx,
-    const char *output_filename,
+    const char *out_fname,
    AVFormatContext *ifmt_ctx,
    AVFormatContext **ofmt_ctx,
    AVCodecContext **enc_ctx,
    AVCodecContext *dec_ctx,
    EncoderConfig *encoder_config,
-    int video_stream_index,
-    int **stream_mapping
+    int vstream_idx,
+    int **stream_map
 ) {
    AVFormatContext *fmt_ctx = NULL;
    AVCodecContext *codec_ctx = NULL;
    int stream_index = 0;
    int ret;

-    avformat_alloc_output_context2(&fmt_ctx, NULL, NULL, output_filename);
+    avformat_alloc_output_context2(&fmt_ctx, NULL, NULL, out_fname);
    if (!fmt_ctx) {
        spdlog::error("Could not create output context");
        return AVERROR_UNKNOWN;
@@ -67,8 +67,8 @@ int init_encoder(
    }

    // Set encoding parameters
-    codec_ctx->height = encoder_config->output_height;
-    codec_ctx->width = encoder_config->output_width;
+    codec_ctx->height = encoder_config->out_height;
+    codec_ctx->width = encoder_config->out_width;
    codec_ctx->sample_aspect_ratio = dec_ctx->sample_aspect_ratio;
    codec_ctx->bit_rate = encoder_config->bit_rate;

@@ -93,7 +93,7 @@ int init_encoder(

    // Set the CRF and preset for any codecs that support it
    char crf_str[16];
-    snprintf(crf_str, sizeof(crf_str), "%.f", encoder_config->crf);
+    snprintf(crf_str, sizeof(crf_str), "%.f", static_cast<double>(encoder_config->crf));
    av_opt_set(codec_ctx->priv_data, "crf", crf_str, 0);
    av_opt_set(codec_ctx->priv_data, "preset", encoder_config->preset, 0);

@@ -116,57 +116,58 @@ int init_encoder(

    if (encoder_config->copy_streams) {
        // Allocate the stream map
-        *stream_mapping = (int *)av_malloc_array(ifmt_ctx->nb_streams, sizeof(**stream_mapping));
-        if (!*stream_mapping) {
+        *stream_map =
+            reinterpret_cast<int *>(av_malloc_array(ifmt_ctx->nb_streams, sizeof(**stream_map)));
+        if (!*stream_map) {
            spdlog::error("Could not allocate stream mapping");
            return AVERROR(ENOMEM);
        }

        // Map the video stream
-        (*stream_mapping)[video_stream_index] = stream_index++;
+        (*stream_map)[vstream_idx] = stream_index++;

        // Loop through each stream in the input file
-        for (int i = 0; i < ifmt_ctx->nb_streams; i++) {
+        for (int i = 0; i < static_cast<int>(ifmt_ctx->nb_streams); i++) {
            AVStream *in_stream = ifmt_ctx->streams[i];
            AVCodecParameters *in_codecpar = in_stream->codecpar;

-            if (i == video_stream_index) {
+            if (i == vstream_idx) {
                // Video stream is already handled
                continue;
            }

            if (in_codecpar->codec_type != AVMEDIA_TYPE_AUDIO &&
                in_codecpar->codec_type != AVMEDIA_TYPE_SUBTITLE) {
-                (*stream_mapping)[i] = -1;
+                (*stream_map)[i] = -1;
                continue;
            }

            // Create corresponding output stream
-            AVStream *out_stream = avformat_new_stream(fmt_ctx, NULL);
-            if (!out_stream) {
+            AVStream *out_copied_stream = avformat_new_stream(fmt_ctx, NULL);
+            if (!out_copied_stream) {
                spdlog::error("Failed allocating output stream");
                return AVERROR_UNKNOWN;
            }

-            ret = avcodec_parameters_copy(out_stream->codecpar, in_codecpar);
+            ret = avcodec_parameters_copy(out_copied_stream->codecpar, in_codecpar);
            if (ret < 0) {
                spdlog::error("Failed to copy codec parameters");
                return ret;
            }
-            out_stream->codecpar->codec_tag = 0;
+            out_copied_stream->codecpar->codec_tag = 0;

            // Copy time base
-            out_stream->time_base = in_stream->time_base;
+            out_copied_stream->time_base = in_stream->time_base;

-            (*stream_mapping)[i] = stream_index++;
+            (*stream_map)[i] = stream_index++;
        }
    }

    // Open the output file
    if (!(fmt_ctx->oformat->flags & AVFMT_NOFILE)) {
-        ret = avio_open(&fmt_ctx->pb, output_filename, AVIO_FLAG_WRITE);
+        ret = avio_open(&fmt_ctx->pb, out_fname, AVIO_FLAG_WRITE);
        if (ret < 0) {
-            spdlog::error("Could not open output file '{}'", output_filename);
+            spdlog::error("Could not open output file '{}'", out_fname);
            return ret;
        }
    }
@@ -181,20 +182,20 @@ int encode_and_write_frame(
    AVFrame *frame,
    AVCodecContext *enc_ctx,
    AVFormatContext *ofmt_ctx,
-    int video_stream_index
+    int vstream_idx
 ) {
+    AVFrame *converted_frame = nullptr;
    int ret;

    // Convert the frame to the encoder's pixel format if needed
    if (frame->format != enc_ctx->pix_fmt) {
-        AVFrame *converted_frame = convert_avframe_pix_fmt(frame, enc_ctx->pix_fmt);
+        converted_frame = convert_avframe_pix_fmt(frame, enc_ctx->pix_fmt);
        if (!converted_frame) {
            spdlog::error("Error converting frame to encoder's pixel format");
            return AVERROR_EXTERNAL;
        }

        converted_frame->pts = frame->pts;
-        frame = converted_frame;
    }

    AVPacket *enc_pkt = av_packet_alloc();
@@ -203,7 +204,12 @@ int encode_and_write_frame(
        return AVERROR(ENOMEM);
    }

-    ret = avcodec_send_frame(enc_ctx, frame);
+    if (converted_frame != nullptr) {
+        ret = avcodec_send_frame(enc_ctx, converted_frame);
+        av_frame_free(&converted_frame);
+    } else {
+        ret = avcodec_send_frame(enc_ctx, frame);
+    }
    if (ret < 0) {
        spdlog::error("Error sending frame to encoder");
        av_packet_free(&enc_pkt);
@@ -223,9 +229,9 @@ int encode_and_write_frame(

        // Rescale packet timestamps
        av_packet_rescale_ts(
-            enc_pkt, enc_ctx->time_base, ofmt_ctx->streams[video_stream_index]->time_base
+            enc_pkt, enc_ctx->time_base, ofmt_ctx->streams[vstream_idx]->time_base
        );
-        enc_pkt->stream_index = video_stream_index;
+        enc_pkt->stream_index = vstream_idx;

        // Write the packet
        ret = av_interleaved_write_frame(ofmt_ctx, enc_pkt);
--- a/src/fsutils.cpp
+++ b/src/fsutils.cpp
@@ -11,7 +11,7 @@
 #include <spdlog/spdlog.h>

 #if _WIN32
-std::filesystem::path get_executable_directory() {
+static std::filesystem::path get_executable_directory() {
    std::vector<wchar_t> filepath(MAX_PATH);

    // Get the executable path, expanding the buffer if necessary
@@ -36,7 +36,7 @@ std::filesystem::path get_executable_directory() {
    return execpath.parent_path();
 }
 #else   // _WIN32
-std::filesystem::path get_executable_directory() {
+static std::filesystem::path get_executable_directory() {
    std::error_code ec;
    std::filesystem::path filepath = std::filesystem::read_symlink("/proc/self/exe", ec);

--- a/src/libplacebo.cpp
+++ b/src/libplacebo.cpp
@@ -13,8 +13,8 @@ int init_libplacebo(
    AVFilterContext **buffersrc_ctx,
    AVFilterContext **buffersink_ctx,
    AVCodecContext *dec_ctx,
-    int output_width,
-    int output_height,
+    int out_width,
+    int out_height,
    const std::filesystem::path &shader_path
 ) {
    char args[512];
@@ -77,8 +77,8 @@ int init_libplacebo(
        filter_args,
        sizeof(filter_args),
        "w=%d:h=%d:upscaler=ewa_lanczos:custom_shader_path=%s",
-        output_width,
-        output_height,
+        out_width,
+        out_height,
        shader_path_string.c_str()
    );

--- a/src/libplacebo_filter.cpp
+++ b/src/libplacebo_filter.cpp
@@ -7,12 +7,16 @@
 #include "fsutils.h"
 #include "libplacebo.h"

-LibplaceboFilter::LibplaceboFilter(int width, int height, const std::filesystem::path &shader_path)
+LibplaceboFilter::LibplaceboFilter(
+    int out_width,
+    int out_height,
+    const std::filesystem::path &shader_path
+)
    : filter_graph(nullptr),
      buffersrc_ctx(nullptr),
      buffersink_ctx(nullptr),
-      output_width(width),
-      output_height(height),
+      out_width(out_width),
+      out_height(out_height),
      shader_path(std::move(shader_path)) {}

 LibplaceboFilter::~LibplaceboFilter() {
@@ -38,8 +42,9 @@ int LibplaceboFilter::init(AVCodecContext *dec_ctx, AVCodecContext *enc_ctx, AVB
        shader_full_path = shader_path;
    } else {
        // Construct the fallback path using std::filesystem
-        shader_full_path =
-            find_resource_file(std::filesystem::path("models") / (shader_path.string() + ".glsl"));
+        shader_full_path = find_resource_file(
+            std::filesystem::path("models") / "libplacebo" / (shader_path.string() + ".glsl")
+        );
    }

    // Check if the shader file exists
@@ -49,52 +54,61 @@ int LibplaceboFilter::init(AVCodecContext *dec_ctx, AVCodecContext *enc_ctx, AVB
    }

    // Save the output time base
-    output_time_base = enc_ctx->time_base;
+    in_time_base = dec_ctx->time_base;
+    out_time_base = enc_ctx->time_base;

-    return init_libplacebo(
+    // Initialize the libplacebo filter
+    int ret = init_libplacebo(
        hw_ctx,
        &filter_graph,
        &buffersrc_ctx,
        &buffersink_ctx,
        dec_ctx,
-        output_width,
-        output_height,
+        out_width,
+        out_height,
        shader_full_path
    );
+
+    // Set these resources to nullptr since they are already freed by `avfilter_graph_free`
+    if (ret < 0) {
+        buffersrc_ctx = nullptr;
+        buffersink_ctx = nullptr;
+        filter_graph = nullptr;
+    }
+    return ret;
 }

-int LibplaceboFilter::process_frame(AVFrame *input_frame, AVFrame **output_frame) {
+int LibplaceboFilter::process_frame(AVFrame *in_frame, AVFrame **out_frame) {
    int ret;

    // Get the filtered frame
-    *output_frame = av_frame_alloc();
-    if (*output_frame == nullptr) {
+    *out_frame = av_frame_alloc();
+    if (*out_frame == nullptr) {
        spdlog::error("Failed to allocate output frame");
        return -1;
    }

    // Feed the frame to the filter graph
-    ret = av_buffersrc_add_frame(buffersrc_ctx, input_frame);
+    ret = av_buffersrc_add_frame(buffersrc_ctx, in_frame);
    if (ret < 0) {
        spdlog::error("Error while feeding the filter graph");
        return ret;
    }

-    ret = av_buffersink_get_frame(buffersink_ctx, *output_frame);
+    ret = av_buffersink_get_frame(buffersink_ctx, *out_frame);
    if (ret < 0) {
-        av_frame_free(output_frame);
+        av_frame_free(out_frame);
        return ret;
    }

    // Rescale PTS to encoder's time base
-    (*output_frame)->pts =
-        av_rescale_q((*output_frame)->pts, buffersink_ctx->inputs[0]->time_base, output_time_base);
+    (*out_frame)->pts = av_rescale_q((*out_frame)->pts, in_time_base, out_time_base);

    // Return the processed frame to the caller
    return 0;
 }

-int LibplaceboFilter::flush(std::vector<AVFrame *> &processed_frames) {
+int LibplaceboFilter::flush(std::vector<AVFrame *> &flushed_frames) {
    int ret = av_buffersrc_add_frame(buffersrc_ctx, nullptr);
    if (ret < 0) {
        spdlog::error("Error while flushing filter graph");
@@ -119,11 +133,10 @@ int LibplaceboFilter::flush(std::vector<AVFrame *> &processed_frames) {
        }

        // Rescale PTS to encoder's time base
-        filt_frame->pts =
-            av_rescale_q(filt_frame->pts, buffersink_ctx->inputs[0]->time_base, output_time_base);
+        filt_frame->pts = av_rescale_q(filt_frame->pts, in_time_base, out_time_base);

        // Add to processed frames
-        processed_frames.push_back(filt_frame);
+        flushed_frames.push_back(filt_frame);
    }

    return 0;
--- a/src/libvideo2x.cpp
+++ b/src/libvideo2x.cpp
@@ -3,10 +3,10 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-#include <cstdint>
 #include <thread>

 #include <spdlog/spdlog.h>
+#include <opencv2/videoio.hpp>

 #include "decoder.h"
 #include "encoder.h"
@@ -24,12 +24,12 @@
 * @param[in] dec_ctx Decoder context
 * @param[in] enc_ctx Encoder context
 * @param[in] filter Filter instance
- * @param[in] video_stream_index Index of the video stream in the input format context
- * @param[in] stream_mapping Array mapping input stream indexes to output stream indexes
+ * @param[in] vstream_idx Index of the video stream in the input format context
+ * @param[in] stream_map Array mapping input stream indexes to output stream indexes
 * @param[in] benchmark Flag to enable benchmarking mode
 * @return int 0 on success, negative value on error
 */
-int process_frames(
+static int process_frames(
    EncoderConfig *encoder_config,
    VideoProcessingContext *proc_ctx,
    AVFormatContext *ifmt_ctx,
@@ -37,8 +37,8 @@ int process_frames(
    AVCodecContext *dec_ctx,
    AVCodecContext *enc_ctx,
    Filter *filter,
-    int video_stream_index,
-    int *stream_mapping,
+    int vstream_idx,
+    int *stream_map,
    bool benchmark = false
 ) {
    int ret;
@@ -46,17 +46,21 @@ int process_frames(
    std::vector<AVFrame *> flushed_frames;
    char errbuf[AV_ERROR_MAX_STRING_SIZE];

-    // Get the total number of frames in the video
-    AVStream *video_stream = ifmt_ctx->streams[video_stream_index];
-    proc_ctx->total_frames = video_stream->nb_frames;
+    // Get the total number of frames in the video with OpenCV
+    spdlog::debug("Unable to estimate total number of frames; reading with OpenCV");
+    cv::VideoCapture cap(ifmt_ctx->url);
+    if (!cap.isOpened()) {
+        spdlog::error("Failed to open video file with OpenCV");
+        return -1;
+    }
+    proc_ctx->total_frames = static_cast<int64_t>(cap.get(cv::CAP_PROP_FRAME_COUNT));
+    cap.release();

-    // If nb_frames is not set, calculate total frames using duration and frame rate
+    // Check if the total number of frames is still 0
    if (proc_ctx->total_frames == 0) {
-        int64_t duration = video_stream->duration;
-        AVRational frame_rate = video_stream->avg_frame_rate;
-        if (duration != AV_NOPTS_VALUE && frame_rate.num != 0 && frame_rate.den != 0) {
-            proc_ctx->total_frames = duration * frame_rate.num / frame_rate.den;
-        }
+        spdlog::warn("Unable to determine total number of frames");
+    } else {
+        spdlog::debug("{} frames to process", proc_ctx->total_frames);
    }

    // Get start time
@@ -75,10 +79,16 @@ int process_frames(
    while (!proc_ctx->abort) {
        ret = av_read_frame(ifmt_ctx, &packet);
        if (ret < 0) {
-            break;  // End of file or error
+            if (ret == AVERROR_EOF) {
+                spdlog::debug("Reached end of file");
+                break;
+            }
+            av_strerror(ret, errbuf, sizeof(errbuf));
+            spdlog::error("Error reading packet: {}", errbuf);
+            goto end;
        }

-        if (packet.stream_index == video_stream_index) {
+        if (packet.stream_index == vstream_idx) {
            // Send the packet to the decoder
            ret = avcodec_send_packet(dec_ctx, &packet);
            if (ret < 0) {
@@ -98,6 +108,7 @@ int process_frames(

                ret = avcodec_receive_frame(dec_ctx, frame);
                if (ret == AVERROR(EAGAIN) || ret == AVERROR_EOF) {
+                    spdlog::debug("Frame not ready");
                    break;
                } else if (ret < 0) {
                    av_strerror(ret, errbuf, sizeof(errbuf));
@@ -111,9 +122,8 @@ int process_frames(
                if (ret == 0 && processed_frame != nullptr) {
                    // Encode and write the processed frame
                    if (!benchmark) {
-                        ret = encode_and_write_frame(
-                            processed_frame, enc_ctx, ofmt_ctx, video_stream_index
-                        );
+                        ret =
+                            encode_and_write_frame(processed_frame, enc_ctx, ofmt_ctx, vstream_idx);
                        if (ret < 0) {
                            av_strerror(ret, errbuf, sizeof(errbuf));
                            spdlog::error("Error encoding/writing frame: {}", errbuf);
@@ -134,9 +144,9 @@ int process_frames(
                    "Processed frame {}/{}", proc_ctx->processed_frames, proc_ctx->total_frames
                );
            }
-        } else if (encoder_config->copy_streams && stream_mapping[packet.stream_index] >= 0) {
+        } else if (encoder_config->copy_streams && stream_map[packet.stream_index] >= 0) {
            AVStream *in_stream = ifmt_ctx->streams[packet.stream_index];
-            int out_stream_index = stream_mapping[packet.stream_index];
+            int out_stream_index = stream_map[packet.stream_index];
            AVStream *out_stream = ofmt_ctx->streams[out_stream_index];

            // Rescale packet timestamps
@@ -165,7 +175,7 @@ int process_frames(

    // Encode and write all flushed frames
    for (AVFrame *&flushed_frame : flushed_frames) {
-        ret = encode_and_write_frame(flushed_frame, enc_ctx, ofmt_ctx, video_stream_index);
+        ret = encode_and_write_frame(flushed_frame, enc_ctx, ofmt_ctx, vstream_idx);
        if (ret < 0) {
            av_strerror(ret, errbuf, sizeof(errbuf));
            spdlog::error("Error encoding/writing flushed frame: {}", errbuf);
@@ -197,46 +207,54 @@ end:
 }

 // Cleanup resources after processing the video
-void cleanup(
+static void cleanup(
    AVFormatContext *ifmt_ctx,
    AVFormatContext *ofmt_ctx,
    AVCodecContext *dec_ctx,
    AVCodecContext *enc_ctx,
    AVBufferRef *hw_ctx,
-    int *stream_mapping,
+    int *stream_map,
    Filter *filter
 ) {
    if (ifmt_ctx) {
        avformat_close_input(&ifmt_ctx);
+        ifmt_ctx = nullptr;
    }
    if (ofmt_ctx && !(ofmt_ctx->oformat->flags & AVFMT_NOFILE)) {
        avio_closep(&ofmt_ctx->pb);
+        ofmt_ctx->pb = nullptr;
    }
    if (ofmt_ctx) {
        avformat_free_context(ofmt_ctx);
+        ofmt_ctx = nullptr;
    }
    if (dec_ctx) {
        avcodec_free_context(&dec_ctx);
+        dec_ctx = nullptr;
    }
    if (enc_ctx) {
        avcodec_free_context(&enc_ctx);
+        enc_ctx = nullptr;
    }
    if (hw_ctx) {
        av_buffer_unref(&hw_ctx);
+        hw_ctx = nullptr;
    }
-    if (stream_mapping) {
-        av_free(stream_mapping);
+    if (stream_map) {
+        av_free(stream_map);
+        stream_map = nullptr;
    }
    if (filter) {
        delete filter;
+        filter = nullptr;
    }
 }

 /**
 * @brief Process a video file using the selected filter and encoder settings.
 *
- * @param[in] input_filename Path to the input video file
- * @param[in] output_filename Path to the output video file
+ * @param[in] in_fname Path to the input video file
+ * @param[in] out_fname Path to the output video file
 * @param[in] log_level Log level
 * @param[in] benchmark Flag to enable benchmarking mode
 * @param[in] hw_type Hardware device type
@@ -246,8 +264,8 @@ void cleanup(
 * @return int 0 on success, non-zero value on error
 */
 extern "C" int process_video(
-    const char *input_filename,
-    const char *output_filename,
+    const char *in_fname,
+    const char *out_fname,
    Libvideo2xLogLevel log_level,
    bool benchmark,
    AVHWDeviceType hw_type,
@@ -260,9 +278,9 @@ extern "C" int process_video(
    AVCodecContext *dec_ctx = nullptr;
    AVCodecContext *enc_ctx = nullptr;
    AVBufferRef *hw_ctx = nullptr;
-    int *stream_mapping = nullptr;
+    int *stream_map = nullptr;
    Filter *filter = nullptr;
-    int video_stream_index = -1;
+    int vstream_idx = -1;
    char errbuf[AV_ERROR_MAX_STRING_SIZE];
    int ret = 0;

@@ -313,11 +331,11 @@ extern "C" int process_video(
    }

    // Initialize input
-    ret = init_decoder(hw_type, hw_ctx, input_filename, &ifmt_ctx, &dec_ctx, &video_stream_index);
+    ret = init_decoder(hw_type, hw_ctx, in_fname, &ifmt_ctx, &dec_ctx, &vstream_idx);
    if (ret < 0) {
        av_strerror(ret, errbuf, sizeof(errbuf));
        spdlog::error("Failed to initialize decoder: {}", errbuf);
-        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
        return ret;
    }

@@ -325,33 +343,39 @@ extern "C" int process_video(
    int output_width = 0, output_height = 0;
    switch (filter_config->filter_type) {
        case FILTER_LIBPLACEBO:
-            output_width = filter_config->config.libplacebo.output_width;
-            output_height = filter_config->config.libplacebo.output_height;
+            output_width = filter_config->config.libplacebo.out_width;
+            output_height = filter_config->config.libplacebo.out_height;
            break;
        case FILTER_REALESRGAN:
            // Calculate the output dimensions based on the scaling factor
            output_width = dec_ctx->width * filter_config->config.realesrgan.scaling_factor;
            output_height = dec_ctx->height * filter_config->config.realesrgan.scaling_factor;
+            break;
+        default:
+            spdlog::error("Unknown filter type");
+            cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
+            return -1;
    }
+    spdlog::info("Output video dimensions: {}x{}", output_width, output_height);

    // Initialize output encoder
-    encoder_config->output_width = output_width;
-    encoder_config->output_height = output_height;
+    encoder_config->out_width = output_width;
+    encoder_config->out_height = output_height;
    ret = init_encoder(
        hw_ctx,
-        output_filename,
+        out_fname,
        ifmt_ctx,
        &ofmt_ctx,
        &enc_ctx,
        dec_ctx,
        encoder_config,
-        video_stream_index,
-        &stream_mapping
+        vstream_idx,
+        &stream_map
    );
    if (ret < 0) {
        av_strerror(ret, errbuf, sizeof(errbuf));
        spdlog::error("Failed to initialize encoder: {}", errbuf);
-        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
        return ret;
    }

@@ -360,7 +384,7 @@ extern "C" int process_video(
    if (ret < 0) {
        av_strerror(ret, errbuf, sizeof(errbuf));
        spdlog::error("Error occurred when opening output file: {}", errbuf);
-        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
        return ret;
    }

@@ -372,19 +396,19 @@ extern "C" int process_video(
            // Validate shader path
            if (!config.shader_path) {
                spdlog::error("Shader path must be provided for the libplacebo filter");
-                cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+                cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
                return -1;
            }

            // Validate output dimensions
-            if (config.output_width <= 0 || config.output_height <= 0) {
+            if (config.out_width <= 0 || config.out_height <= 0) {
                spdlog::error("Output dimensions must be provided for the libplacebo filter");
-                cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+                cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
                return -1;
            }

            filter = new LibplaceboFilter{
-                config.output_width, config.output_height, std::filesystem::path(config.shader_path)
+                config.out_width, config.out_height, std::filesystem::path(config.shader_path)
            };
            break;
        }
@@ -394,14 +418,14 @@ extern "C" int process_video(
            // Validate model name
            if (!config.model) {
                spdlog::error("Model name must be provided for the RealESRGAN filter");
-                cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+                cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
                return -1;
            }

            // Validate scaling factor
            if (config.scaling_factor <= 0) {
                spdlog::error("Scaling factor must be provided for the RealESRGAN filter");
-                cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+                cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
                return -1;
            }

@@ -412,16 +436,15 @@ extern "C" int process_video(
        }
        default:
            spdlog::error("Unknown filter type");
-            cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+            cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
            return -1;
    }

    // Initialize the filter
    ret = filter->init(dec_ctx, enc_ctx, hw_ctx);
    if (ret < 0) {
-        av_strerror(ret, errbuf, sizeof(errbuf));
-        spdlog::error("Failed to initialize filter: {}", errbuf);
-        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+        spdlog::error("Failed to initialize filter");
+        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
        return ret;
    }

@@ -434,14 +457,14 @@ extern "C" int process_video(
        dec_ctx,
        enc_ctx,
        filter,
-        video_stream_index,
-        stream_mapping,
+        vstream_idx,
+        stream_map,
        benchmark
    );
    if (ret < 0) {
        av_strerror(ret, errbuf, sizeof(errbuf));
        spdlog::error("Error processing frames: {}", errbuf);
-        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+        cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);
        return ret;
    }

@@ -449,7 +472,7 @@ extern "C" int process_video(
    av_write_trailer(ofmt_ctx);

    // Cleanup before returning
-    cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_mapping, filter);
+    cleanup(ifmt_ctx, ofmt_ctx, dec_ctx, enc_ctx, hw_ctx, stream_map, filter);

    if (ret < 0 && ret != AVERROR_EOF) {
        av_strerror(ret, errbuf, sizeof(errbuf));
--- a/src/realesrgan_filter.cpp
+++ b/src/realesrgan_filter.cpp
@@ -32,16 +32,16 @@ RealesrganFilter::~RealesrganFilter() {
    }
 }

-int RealesrganFilter::init(AVCodecContext *dec_ctx, AVCodecContext *enc_ctx, AVBufferRef *hw_ctx) {
+int RealesrganFilter::init(AVCodecContext *dec_ctx, AVCodecContext *enc_ctx, AVBufferRef *_) {
    // Construct the model paths using std::filesystem
    std::filesystem::path model_param_path;
    std::filesystem::path model_bin_path;

    if (model) {
        // Find the model paths by model name if provided
-        model_param_path = std::filesystem::path("models") /
+        model_param_path = std::filesystem::path("models") / "realesrgan" /
                           (std::string(model) + "-x" + std::to_string(scaling_factor) + ".param");
-        model_bin_path = std::filesystem::path("models") /
+        model_bin_path = std::filesystem::path("models") / "realesrgan" /
                         (std::string(model) + "-x" + std::to_string(scaling_factor) + ".bin");
    } else if (!custom_model_param_path.empty() && !custom_model_bin_path.empty()) {
        // Use the custom model paths if provided
@@ -71,9 +71,9 @@ int RealesrganFilter::init(AVCodecContext *dec_ctx, AVCodecContext *enc_ctx, AVB
    realesrgan = new RealESRGAN(gpuid, tta_mode);

    // Store the time bases
-    input_time_base = dec_ctx->time_base;
-    output_time_base = enc_ctx->time_base;
-    output_pix_fmt = enc_ctx->pix_fmt;
+    in_time_base = dec_ctx->time_base;
+    out_time_base = enc_ctx->time_base;
+    out_pix_fmt = enc_ctx->pix_fmt;

    // Load the model
    if (realesrgan->load(model_param_full_path, model_bin_full_path) != 0) {
@@ -100,38 +100,33 @@ int RealesrganFilter::init(AVCodecContext *dec_ctx, AVCodecContext *enc_ctx, AVB
    return 0;
 }

-int RealesrganFilter::process_frame(AVFrame *input_frame, AVFrame **output_frame) {
+int RealesrganFilter::process_frame(AVFrame *in_frame, AVFrame **out_frame) {
    int ret;

    // Convert the input frame to RGB24
-    ncnn::Mat input_mat = avframe_to_ncnn_mat(input_frame);
-    if (input_mat.empty()) {
+    ncnn::Mat in_mat = avframe_to_ncnn_mat(in_frame);
+    if (in_mat.empty()) {
        spdlog::error("Failed to convert AVFrame to ncnn::Mat");
        return -1;
    }

    // Allocate space for ouptut ncnn::Mat
-    int output_width = input_mat.w * realesrgan->scale;
-    int output_height = input_mat.h * realesrgan->scale;
-    ncnn::Mat output_mat = ncnn::Mat(output_width, output_height, (size_t)3, 3);
+    int output_width = in_mat.w * realesrgan->scale;
+    int output_height = in_mat.h * realesrgan->scale;
+    ncnn::Mat out_mat = ncnn::Mat(output_width, output_height, static_cast<size_t>(3), 3);

-    ret = realesrgan->process(input_mat, output_mat);
+    ret = realesrgan->process(in_mat, out_mat);
    if (ret != 0) {
        spdlog::error("RealESRGAN processing failed");
        return ret;
    }

    // Convert ncnn::Mat to AVFrame
-    *output_frame = ncnn_mat_to_avframe(output_mat, output_pix_fmt);
+    *out_frame = ncnn_mat_to_avframe(out_mat, out_pix_fmt);

    // Rescale PTS to encoder's time base
-    (*output_frame)->pts = av_rescale_q(input_frame->pts, input_time_base, output_time_base);
+    (*out_frame)->pts = av_rescale_q(in_frame->pts, in_time_base, out_time_base);

    // Return the processed frame to the caller
    return ret;
 }
-
-int RealesrganFilter::flush(std::vector<AVFrame *> &processed_frames) {
-    // No special flushing needed for RealESRGAN
-    return 0;
-}
--- a/src/video2x.c
+++ b/src/video2x.c
@@ -22,23 +22,6 @@

 #include "getopt.h"

-// Set UNIX terminal input to non-blocking mode
-#ifndef _WIN32
-void set_nonblocking_input(bool enable) {
-    static struct termios oldt, newt;
-    if (enable) {
-        tcgetattr(STDIN_FILENO, &oldt);
-        newt = oldt;
-        newt.c_lflag &= ~(ICANON | ECHO);
-        tcsetattr(STDIN_FILENO, TCSANOW, &newt);
-        fcntl(STDIN_FILENO, F_SETFL, O_NONBLOCK);
-    } else {
-        tcsetattr(STDIN_FILENO, TCSANOW, &oldt);
-        fcntl(STDIN_FILENO, F_SETFL, 0);
-    }
-}
-#endif
-
 // Define command line options
 static struct option long_options[] = {
    {"loglevel", required_argument, NULL, 0},
@@ -73,13 +56,23 @@ static struct option long_options[] = {
    {0, 0, 0, 0}
 };

+// List of valid RealESRGAN models
+const char *valid_realesrgan_models[] = {
+    "realesrgan-plus",
+    "realesrgan-plus-anime",
+    "realesr-animevideov3",
+};
+
+// Indicate if a newline needs to be printed before the next output
+bool newline_required = false;
+
 // Structure to hold parsed arguments
 struct arguments {
    // General options
    const char *loglevel;
    bool noprogress;
-    const char *input_filename;
-    const char *output_filename;
+    const char *in_fname;
+    const char *out_fname;
    const char *filter_type;
    const char *hwaccel;
    bool nocopystreams;
@@ -94,8 +87,8 @@ struct arguments {

    // libplacebo options
    const char *shader_path;
-    int output_width;
-    int output_height;
+    int out_width;
+    int out_height;

    // RealESRGAN options
    int gpuid;
@@ -111,30 +104,51 @@ struct ProcessVideoThreadArguments {
    struct VideoProcessingContext *proc_ctx;
 };

-const char *valid_models[] = {
-    "realesrgan-plus",
-    "realesrgan-plus-anime",
-    "realesr-animevideov3",
-};
+// Set UNIX terminal input to non-blocking mode
+#ifndef _WIN32
+void set_nonblocking_input(bool enable) {
+    static struct termios oldt, newt;
+    if (enable) {
+        tcgetattr(STDIN_FILENO, &oldt);
+        newt = oldt;
+        newt.c_lflag &= ~(tcflag_t)(ICANON | ECHO);
+        tcsetattr(STDIN_FILENO, TCSANOW, &newt);
+        fcntl(STDIN_FILENO, F_SETFL, O_NONBLOCK);
+    } else {
+        tcsetattr(STDIN_FILENO, TCSANOW, &oldt);
+        fcntl(STDIN_FILENO, F_SETFL, 0);
+    }
+}
+#endif
+
+// Newline-safe log callback for FFmpeg
+void newline_safe_ffmpeg_log_callback(void *ptr, int level, const char *fmt, va_list vl) {
+    if (level <= av_log_get_level() && newline_required) {
+        putchar('\n');
+        newline_required = false;
+    }
+    av_log_default_callback(ptr, level, fmt, vl);
+}

 int is_valid_realesrgan_model(const char *model) {
    if (!model) {
        return 0;
    }
-    for (int i = 0; i < sizeof(valid_models) / sizeof(valid_models[0]); i++) {
-        if (strcmp(model, valid_models[i]) == 0) {
+    for (unsigned long i = 0;
+         i < sizeof(valid_realesrgan_models) / sizeof(valid_realesrgan_models[0]);
+         i++) {
+        if (strcmp(model, valid_realesrgan_models[i]) == 0) {
            return 1;
        }
    }
    return 0;
 }

-void print_help() {
+void print_help(void) {
    printf("Usage: video2x [OPTIONS]\n");
    printf("\nOptions:\n");
-    printf(
-        "  --loglevel		Set log level (trace, debug, info, warn, error, critical, none)\n"
-    );
+    printf("  --loglevel		Set log level \n");
+    printf("			(trace, debug, info, warn, error, critical, none)\n");
    printf("  --noprogress		Do not display the progress bar\n");
    printf("  -v, --version		Print program version\n");
    printf("  -?, --help		Display this help page\n");
@@ -154,7 +168,9 @@ void print_help() {
    printf("  -q, --crf		Constant Rate Factor (default: 20.0)\n");

    printf("\nlibplacebo Options:\n");
-    printf("  -s, --shader		Name or path to custom GLSL shader file\n");
+    printf("  -s, --shader		Name or path of the GLSL shader file to use \n");
+    printf("			(built-in: 'anime4k-a', 'anime4k-b', 'anime4k-c',\n");
+    printf("			'anime4k-a+a', 'anime4k-b+b', 'anime4k-c+a')\n");
    printf("  -w, --width		Output width\n");
    printf("  -h, --height		Output height\n");

@@ -164,7 +180,7 @@ void print_help() {
    printf("  -r, --scale		Scaling factor (2, 3, or 4)\n");

    printf("\nExamples Usage:\n");
-    printf("  video2x -i in.mp4 -o out.mp4 -f libplacebo -s anime4k-mode-a -w 3840 -h 2160\n");
+    printf("  video2x -i in.mp4 -o out.mp4 -f libplacebo -s anime4k-a+a -w 3840 -h 2160\n");
    printf("  video2x -i in.mp4 -o out.mp4 -f realesrgan -m realesr-animevideov3 -r 4\n");
 }

@@ -175,8 +191,8 @@ void parse_arguments(int argc, char **argv, struct arguments *arguments) {
    // Default argument values
    arguments->loglevel = "info";
    arguments->noprogress = false;
-    arguments->input_filename = NULL;
-    arguments->output_filename = NULL;
+    arguments->in_fname = NULL;
+    arguments->out_fname = NULL;
    arguments->filter_type = NULL;
    arguments->hwaccel = "none";
    arguments->nocopystreams = false;
@@ -191,8 +207,8 @@ void parse_arguments(int argc, char **argv, struct arguments *arguments) {

    // libplacebo options
    arguments->shader_path = NULL;
-    arguments->output_width = 0;
-    arguments->output_height = 0;
+    arguments->out_width = 0;
+    arguments->out_height = 0;

    // RealESRGAN options
    arguments->gpuid = 0;
@@ -204,10 +220,10 @@ void parse_arguments(int argc, char **argv, struct arguments *arguments) {
            )) != -1) {
        switch (c) {
            case 'i':
-                arguments->input_filename = optarg;
+                arguments->in_fname = optarg;
                break;
            case 'o':
-                arguments->output_filename = optarg;
+                arguments->out_fname = optarg;
                break;
            case 'f':
                arguments->filter_type = optarg;
@@ -232,7 +248,7 @@ void parse_arguments(int argc, char **argv, struct arguments *arguments) {
                }
                break;
            case 'q':
-                arguments->crf = atof(optarg);
+                arguments->crf = (float)atof(optarg);
                if (arguments->crf < 0.0 || arguments->crf > 51.0) {
                    fprintf(stderr, "Error: CRF must be between 0 and 51.\n");
                    exit(1);
@@ -242,15 +258,15 @@ void parse_arguments(int argc, char **argv, struct arguments *arguments) {
                arguments->shader_path = optarg;
                break;
            case 'w':
-                arguments->output_width = atoi(optarg);
-                if (arguments->output_width <= 0) {
+                arguments->out_width = atoi(optarg);
+                if (arguments->out_width <= 0) {
                    fprintf(stderr, "Error: Output width must be greater than 0.\n");
                    exit(1);
                }
                break;
            case 'h':
-                arguments->output_height = atoi(optarg);
-                if (arguments->output_height <= 0) {
+                arguments->out_height = atoi(optarg);
+                if (arguments->out_height <= 0) {
                    fprintf(stderr, "Error: Output height must be greater than 0.\n");
                    exit(1);
                }
@@ -301,12 +317,12 @@ void parse_arguments(int argc, char **argv, struct arguments *arguments) {
    }

    // Check for required arguments
-    if (!arguments->input_filename) {
+    if (!arguments->in_fname) {
        fprintf(stderr, "Error: Input file path is required.\n");
        exit(1);
    }

-    if (!arguments->output_filename && !arguments->benchmark) {
+    if (!arguments->out_fname && !arguments->benchmark) {
        fprintf(stderr, "Error: Output file path is required.\n");
        exit(1);
    }
@@ -317,8 +333,7 @@ void parse_arguments(int argc, char **argv, struct arguments *arguments) {
    }

    if (strcmp(arguments->filter_type, "libplacebo") == 0) {
-        if (!arguments->shader_path || arguments->output_width == 0 ||
-            arguments->output_height == 0) {
+        if (!arguments->shader_path || arguments->out_width == 0 || arguments->out_height == 0) {
            fprintf(
                stderr,
                "Error: For libplacebo, shader name/path (-s), width (-w), "
@@ -371,8 +386,8 @@ int process_video_thread(void *arg) {

    // Call the process_video function
    int result = process_video(
-        arguments->input_filename,
-        arguments->output_filename,
+        arguments->in_fname,
+        arguments->out_fname,
        log_level,
        arguments->benchmark,
        hw_device_type,
@@ -400,8 +415,8 @@ int main(int argc, char **argv) {
    struct FilterConfig filter_config;
    if (strcmp(arguments.filter_type, "libplacebo") == 0) {
        filter_config.filter_type = FILTER_LIBPLACEBO;
-        filter_config.config.libplacebo.output_width = arguments.output_width;
-        filter_config.config.libplacebo.output_height = arguments.output_height;
+        filter_config.config.libplacebo.out_width = arguments.out_width;
+        filter_config.config.libplacebo.out_height = arguments.out_height;
        filter_config.config.libplacebo.shader_path = arguments.shader_path;
    } else if (strcmp(arguments.filter_type, "realesrgan") == 0) {
        filter_config.filter_type = FILTER_REALESRGAN;
@@ -433,8 +448,8 @@ int main(int argc, char **argv) {

    // Setup encoder configuration
    struct EncoderConfig encoder_config = {
-        .output_width = 0,   // To be filled by libvideo2x
-        .output_height = 0,  // To be filled by libvideo2x
+        .out_width = 0,   // To be filled by libvideo2x
+        .out_height = 0,  // To be filled by libvideo2x
        .copy_streams = !arguments.nocopystreams,
        .codec = codec->id,
        .pix_fmt = pix_fmt,
@@ -472,6 +487,10 @@ int main(int argc, char **argv) {
        .proc_ctx = &proc_ctx
    };

+    // Register a newline-safe log callback for FFmpeg
+    // This will ensure that log messages are printed on a new line after the progress bar
+    av_log_set_callback(newline_safe_ffmpeg_log_callback);
+
    // Create a thread for video processing
    thrd_t processing_thread;
    if (thrd_create(&processing_thread, process_video_thread, &thread_args) != thrd_success) {
@@ -509,8 +528,9 @@ int main(int argc, char **argv) {
            }
        } else if (ch == 'q' || ch == 'Q') {
            // Abort processing
-            printf("Aborting processing...\n");
+            printf("\nAborting processing...\n");
            proc_ctx.abort = true;
+            newline_required = false;
            break;
        }

@@ -521,15 +541,16 @@ int main(int argc, char **argv) {
                proc_ctx.processed_frames,
                proc_ctx.total_frames,
                proc_ctx.total_frames > 0
-                    ? proc_ctx.processed_frames * 100.0 / proc_ctx.total_frames
+                    ? (double)proc_ctx.processed_frames * 100.0 / (double)proc_ctx.total_frames
                    : 0.0,
                time(NULL) - proc_ctx.start_time
            );
            fflush(stdout);
+            newline_required = true;
        }

        // Sleep for 50ms
-        thrd_sleep(&(struct timespec){.tv_sec = 0, .tv_nsec = 50000000}, NULL);
+        thrd_sleep(&(struct timespec){.tv_sec = 0, .tv_nsec = 100000000}, NULL);
    }

 // Restore terminal to blocking mode
@@ -542,35 +563,36 @@ int main(int argc, char **argv) {
    thrd_join(processing_thread, &process_result);

    // Print a newline if progress bar was displayed
-    if (!arguments.noprogress && process_result == 0) {
-        puts("");
+    if (newline_required) {
+        putchar('\n');
    }

+    // Print final message based on processing result
    if (proc_ctx.abort) {
        fprintf(stderr, "Video processing aborted\n");
        return 2;
-    }
-
-    if (process_result != 0) {
+    } else if (process_result != 0) {
        fprintf(stderr, "Video processing failed\n");
        return process_result;
+    } else {
+        printf("Video processing completed successfully\n");
    }

    // Calculate statistics
    time_t time_elapsed = time(NULL) - proc_ctx.start_time;
    float average_speed_fps =
-        (float)proc_ctx.processed_frames / (time_elapsed > 0 ? time_elapsed : 1);
+        (float)proc_ctx.processed_frames / (time_elapsed > 0 ? (float)time_elapsed : 1);

    // Print processing summary
    printf("====== Video2X %s summary ======\n", arguments.benchmark ? "Benchmark" : "Processing");
-    printf("Video file processed: %s\n", arguments.input_filename);
+    printf("Video file processed: %s\n", arguments.in_fname);
    printf("Total frames processed: %ld\n", proc_ctx.processed_frames);
    printf("Total time taken: %lds\n", time_elapsed);
    printf("Average processing speed: %.2f FPS\n", average_speed_fps);

    // Print additional information if not in benchmark mode
    if (!arguments.benchmark) {
-        printf("Output written to: %s\n", arguments.output_filename);
+        printf("Output written to: %s\n", arguments.out_fname);
    }

    return 0;
--- a/third_party/opencv
+++ b/third_party/opencv
Author	SHA1	Message	Date
k4yt3x	ebd5e8eb3d	fix(realesrgan): fixed incorrect type for ncnn::Mat elemsize Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-26 00:00:00 +00:00
k4yt3x	50d5cb16d1	docs(video2x): updated command line help messages Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-26 00:00:00 +00:00
k4yt3x	8eac1a7393	feat(libplacebo): added more modes for Anime4K v4 Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-26 00:00:00 +00:00
k4yt3x	f0f3166d92	chore(models): moved model files into subdirectories Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-26 00:00:00 +00:00
k4yt3x	482e82f9c4	style(*): reduced implicit casting and other styling issues Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-25 00:00:00 +00:00
k4yt3x	58ea9e4b35	feat(cmake): removed CMake policy CMP0146 for compatibility Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-22 00:00:00 +00:00
k4yt3x	e077849a2c	docs(readme): updated links for 6.0.0-beta.4 Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-22 00:00:00 +00:00
k4yt3x	0fb547589d	feat(*): updated Dockerfile and PKGBUILD for 6.0.0-beta.4 Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-22 00:00:00 +00:00
k4yt3x	1f18b8507d	feat(cmake): added installation rules for FFmpeg and OpenCV DLLs	2024-10-21 21:43:11 -04:00
k4yt3x	eb7d411f49	perf(libvideo2x): removed unnecessary read of nb_frames Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-22 00:00:00 +00:00
k4yt3x	640d9cd52b	fix(encoder): fixed memory leak in the encoder Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-21 00:00:00 +00:00
K4YT3X	bc168d11ab	feat(libvideo2x): use OpenCV to retrieve total frame count (#1194 )	2024-10-21 23:54:22 +00:00
Owen Quinlan	e09f348890	Fix missing spdlog in docker build [Fix for #1185 ] (#1186 ) * Fix missing spdlog in docker build * fix: added the spdlog dependency for the Dockerfile building stage --------- Co-authored-by: K4YT3X <i@k4yt3x.com>	2024-10-14 16:51:40 +00:00
k4yt3x	f3caf22765	docs(readme): updated links for 6.0.0-beta.3 Signed-off-by: k4yt3x <i@k4yt3x.com>	2024-10-14 00:00:00 +00:00