Files
py-media-grader/croppa/main.py
PhatPhuckDave 5baa2572ea Add Cv2BufferedCap class for efficient video frame handling in VideoEditor
This commit introduces the Cv2BufferedCap class, which provides a buffered wrapper around cv2.VideoCapture. It implements frame caching with LRU eviction to optimize frame retrieval and reduce latency during video playback. The VideoEditor class has been updated to utilize this new class, enhancing performance and simplifying frame management. Unused frame cache methods have been removed to streamline the codebase.
2025-09-16 13:35:55 +02:00

2862 lines
116 KiB
Python

import os
import sys
import cv2
import argparse
import numpy as np
from pathlib import Path
from typing import List
import time
import re
import json
import threading
import queue
import subprocess
import ctypes
class Cv2BufferedCap:
"""Buffered wrapper around cv2.VideoCapture that handles frame loading, seeking, and caching correctly"""
def __init__(self, video_path, backend=None):
self.video_path = video_path
self.cap = cv2.VideoCapture(str(video_path), backend)
if not self.cap.isOpened():
raise ValueError(f"Could not open video: {video_path}")
# Video properties
self.total_frames = int(self.cap.get(cv2.CAP_PROP_FRAME_COUNT))
self.fps = self.cap.get(cv2.CAP_PROP_FPS)
self.frame_width = int(self.cap.get(cv2.CAP_PROP_FRAME_WIDTH))
self.frame_height = int(self.cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
# Frame cache
self.frame_cache = {}
self.cache_access_order = []
self.MAX_CACHE_FRAMES = 3000
# Current position tracking
self.current_frame = 0
def _manage_cache(self):
"""Manage cache size using LRU eviction"""
while len(self.frame_cache) > self.MAX_CACHE_FRAMES:
oldest_frame = self.cache_access_order.pop(0)
if oldest_frame in self.frame_cache:
del self.frame_cache[oldest_frame]
def _add_to_cache(self, frame_number, frame):
"""Add frame to cache"""
self.frame_cache[frame_number] = frame.copy()
if frame_number in self.cache_access_order:
self.cache_access_order.remove(frame_number)
self.cache_access_order.append(frame_number)
self._manage_cache()
def _get_from_cache(self, frame_number):
"""Get frame from cache and update LRU"""
if frame_number in self.frame_cache:
if frame_number in self.cache_access_order:
self.cache_access_order.remove(frame_number)
self.cache_access_order.append(frame_number)
return self.frame_cache[frame_number].copy()
return None
def get_frame(self, frame_number):
"""Get frame at specific index - always accurate"""
# Clamp frame number to valid range
frame_number = max(0, min(frame_number, self.total_frames - 1))
# Check cache first
cached_frame = self._get_from_cache(frame_number)
if cached_frame is not None:
self.current_frame = frame_number
return cached_frame
# Not in cache, seek to frame and read
self.cap.set(cv2.CAP_PROP_POS_FRAMES, frame_number)
ret, frame = self.cap.read()
if ret:
self._add_to_cache(frame_number, frame)
self.current_frame = frame_number
return frame
else:
raise ValueError(f"Failed to read frame {frame_number}")
def advance_frame(self, frames=1):
"""Advance by specified number of frames"""
new_frame = self.current_frame + frames
return self.get_frame(new_frame)
def release(self):
"""Release the video capture"""
if self.cap:
self.cap.release()
def isOpened(self):
"""Check if capture is opened"""
return self.cap and self.cap.isOpened()
def get_active_window_title():
"""Get the title of the currently active window"""
try:
# Get handle to foreground window
hwnd = ctypes.windll.user32.GetForegroundWindow()
# Get window title length
length = ctypes.windll.user32.GetWindowTextLengthW(hwnd)
# Create buffer and get window title
buffer = ctypes.create_unicode_buffer(length + 1)
ctypes.windll.user32.GetWindowTextW(hwnd, buffer, length + 1)
return buffer.value
except:
return ""
class ProjectView:
"""Project view that displays videos in current directory with progress bars"""
# Project view configuration
THUMBNAIL_SIZE = (200, 150) # Width, Height
THUMBNAIL_MARGIN = 20
PROGRESS_BAR_HEIGHT = 8
TEXT_HEIGHT = 30
# Colors
BG_COLOR = (40, 40, 40)
THUMBNAIL_BG_COLOR = (60, 60, 60)
PROGRESS_BG_COLOR = (80, 80, 80)
PROGRESS_FILL_COLOR = (0, 120, 255)
TEXT_COLOR = (255, 255, 255)
SELECTED_COLOR = (255, 165, 0)
def __init__(self, directory: Path, video_editor):
self.directory = directory
self.video_editor = video_editor
self.video_files = []
self.thumbnails = {}
self.progress_data = {}
self.selected_index = 0
self.scroll_offset = 0
self.items_per_row = 2 # Default to 2 items per row
self.window_width = 1200
self.window_height = 800
self._load_video_files()
self._load_progress_data()
def _calculate_thumbnail_size(self, window_width: int) -> tuple:
"""Calculate thumbnail size based on items per row and window width"""
available_width = window_width - self.THUMBNAIL_MARGIN
item_width = (available_width - (self.items_per_row - 1) * self.THUMBNAIL_MARGIN) // self.items_per_row
thumbnail_width = max(50, item_width) # Minimum 50px width
thumbnail_height = int(thumbnail_width * self.THUMBNAIL_SIZE[1] / self.THUMBNAIL_SIZE[0]) # Maintain aspect ratio
return (thumbnail_width, thumbnail_height)
def _load_video_files(self):
"""Load all video files from directory"""
self.video_files = []
for file_path in self.directory.iterdir():
if (file_path.is_file() and
file_path.suffix.lower() in self.video_editor.VIDEO_EXTENSIONS):
self.video_files.append(file_path)
self.video_files.sort(key=lambda x: x.name)
def _load_progress_data(self):
"""Load progress data from JSON state files"""
self.progress_data = {}
for video_path in self.video_files:
state_file = video_path.with_suffix('.json')
if state_file.exists():
try:
with open(state_file, 'r') as f:
state = json.load(f)
current_frame = state.get('current_frame', 0)
# Get total frames from video
cap = cv2.VideoCapture(str(video_path))
if cap.isOpened():
total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
cap.release()
if total_frames > 0:
progress = current_frame / (total_frames - 1)
self.progress_data[video_path] = {
'current_frame': current_frame,
'total_frames': total_frames,
'progress': progress
}
except Exception as e:
print(f"Error loading progress for {video_path.name}: {e}")
def refresh_progress_data(self):
"""Refresh progress data from JSON files (call when editor state changes)"""
self._load_progress_data()
def get_progress_for_video(self, video_path: Path) -> float:
"""Get progress (0.0 to 1.0) for a video"""
if video_path in self.progress_data:
return self.progress_data[video_path]['progress']
return 0.0
def get_thumbnail_for_video(self, video_path: Path, size: tuple = None) -> np.ndarray:
"""Get thumbnail for a video, generating it if needed"""
if size is None:
size = self.THUMBNAIL_SIZE
# Cache the original thumbnail by video path only (not size)
if video_path in self.thumbnails:
original_thumbnail = self.thumbnails[video_path]
# Resize the cached thumbnail to the requested size
return cv2.resize(original_thumbnail, size)
# Generate original thumbnail on demand (only once per video)
try:
cap = cv2.VideoCapture(str(video_path))
if cap.isOpened():
total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
if total_frames > 0:
middle_frame = total_frames // 2
cap.set(cv2.CAP_PROP_POS_FRAMES, middle_frame)
ret, frame = cap.read()
if ret:
# Store original thumbnail at original size
original_thumbnail = cv2.resize(frame, self.THUMBNAIL_SIZE)
self.thumbnails[video_path] = original_thumbnail
cap.release()
# Return resized version
return cv2.resize(original_thumbnail, size)
cap.release()
except Exception as e:
print(f"Error generating thumbnail for {video_path.name}: {e}")
# Return a placeholder if thumbnail generation failed
placeholder = np.full((size[1], size[0], 3),
self.THUMBNAIL_BG_COLOR, dtype=np.uint8)
return placeholder
def draw(self) -> np.ndarray:
"""Draw the project view"""
# Get actual window size dynamically
try:
# Try to get the actual window size from OpenCV
window_rect = cv2.getWindowImageRect("Project View")
if window_rect[2] > 0 and window_rect[3] > 0: # width and height > 0
actual_width = window_rect[2]
actual_height = window_rect[3]
else:
# Fallback to default size
actual_width = self.window_width
actual_height = self.window_height
except:
# Fallback to default size
actual_width = self.window_width
actual_height = self.window_height
canvas = np.full((actual_height, actual_width, 3), self.BG_COLOR, dtype=np.uint8)
if not self.video_files:
# No videos message
text = "No videos found in directory"
font = cv2.FONT_HERSHEY_SIMPLEX
text_size = cv2.getTextSize(text, font, 1.0, 2)[0]
text_x = (actual_width - text_size[0]) // 2
text_y = (actual_height - text_size[1]) // 2
cv2.putText(canvas, text, (text_x, text_y), font, 1.0, self.TEXT_COLOR, 2)
return canvas
# Calculate layout - use fixed items_per_row and calculate thumbnail size to fit
items_per_row = min(self.items_per_row, len(self.video_files)) # Don't exceed number of videos
# Calculate thumbnail size to fit the desired number of items per row
thumbnail_width, thumbnail_height = self._calculate_thumbnail_size(actual_width)
# Calculate item height dynamically based on thumbnail size
item_height = thumbnail_height + self.PROGRESS_BAR_HEIGHT + self.TEXT_HEIGHT + self.THUMBNAIL_MARGIN
item_width = (actual_width - (items_per_row + 1) * self.THUMBNAIL_MARGIN) // items_per_row
# Draw videos in grid
for i, video_path in enumerate(self.video_files):
row = i // items_per_row
col = i % items_per_row
# Skip if scrolled out of view
if row < self.scroll_offset:
continue
if row > self.scroll_offset + (actual_height // item_height):
break
# Calculate position
x = self.THUMBNAIL_MARGIN + col * (item_width + self.THUMBNAIL_MARGIN)
y = self.THUMBNAIL_MARGIN + (row - self.scroll_offset) * item_height
# Draw thumbnail background
cv2.rectangle(canvas,
(x, y),
(x + thumbnail_width, y + thumbnail_height),
self.THUMBNAIL_BG_COLOR, -1)
# Draw selection highlight
if i == self.selected_index:
cv2.rectangle(canvas,
(x - 2, y - 2),
(x + thumbnail_width + 2, y + thumbnail_height + 2),
self.SELECTED_COLOR, 3)
# Draw thumbnail
thumbnail = self.get_thumbnail_for_video(video_path, (thumbnail_width, thumbnail_height))
# Thumbnail is already the correct size, no need to resize
resized_thumbnail = thumbnail
# Ensure thumbnail doesn't exceed canvas bounds
end_y = min(y + thumbnail_height, actual_height)
end_x = min(x + thumbnail_width, actual_width)
thumb_height = end_y - y
thumb_width = end_x - x
if thumb_height > 0 and thumb_width > 0:
# Resize thumbnail to fit within bounds if necessary
if thumb_height != thumbnail_height or thumb_width != thumbnail_width:
resized_thumbnail = cv2.resize(thumbnail, (thumb_width, thumb_height))
canvas[y:end_y, x:end_x] = resized_thumbnail
# Draw progress bar
progress_y = y + thumbnail_height + 5
progress_width = thumbnail_width
progress = self.get_progress_for_video(video_path)
# Progress background
cv2.rectangle(canvas,
(x, progress_y),
(x + progress_width, progress_y + self.PROGRESS_BAR_HEIGHT),
self.PROGRESS_BG_COLOR, -1)
# Progress fill
if progress > 0:
fill_width = int(progress_width * progress)
cv2.rectangle(canvas,
(x, progress_y),
(x + fill_width, progress_y + self.PROGRESS_BAR_HEIGHT),
self.PROGRESS_FILL_COLOR, -1)
# Draw filename
filename = video_path.name
# Truncate if too long
if len(filename) > 25:
filename = filename[:22] + "..."
text_y = progress_y + self.PROGRESS_BAR_HEIGHT + 20
cv2.putText(canvas, filename, (x, text_y),
cv2.FONT_HERSHEY_SIMPLEX, 0.6, self.TEXT_COLOR, 2)
# Draw progress percentage
if video_path in self.progress_data:
progress_text = f"{progress * 100:.0f}%"
text_size = cv2.getTextSize(progress_text, cv2.FONT_HERSHEY_SIMPLEX, 0.4, 1)[0]
progress_text_x = x + progress_width - text_size[0]
cv2.putText(canvas, progress_text, (progress_text_x, text_y),
cv2.FONT_HERSHEY_SIMPLEX, 0.4, self.TEXT_COLOR, 1)
# Draw instructions
instructions = [
"Project View - Videos in current directory",
"WASD: Navigate | E: Open video | Q: Fewer items per row | Y: More items per row | q: Quit | ESC: Back to editor",
f"Showing {len(self.video_files)} videos | {items_per_row} per row | Thumbnail: {thumbnail_width}x{thumbnail_height}"
]
for i, instruction in enumerate(instructions):
y_pos = actual_height - 60 + i * 20
cv2.putText(canvas, instruction, (10, y_pos),
cv2.FONT_HERSHEY_SIMPLEX, 0.5, self.TEXT_COLOR, 1)
return canvas
def handle_key(self, key: int) -> str:
"""Handle keyboard input, returns action taken"""
if key == 27: # ESC
return "back_to_editor"
elif key == ord('q'): # lowercase q - Quit
return "quit"
elif key == ord('e') or key == ord('E'): # E - Open video
if self.video_files and 0 <= self.selected_index < len(self.video_files):
return f"open_video:{self.video_files[self.selected_index]}"
elif key == ord('w') or key == ord('W'): # W - Up
current_items_per_row = min(self.items_per_row, len(self.video_files))
if self.selected_index >= current_items_per_row:
self.selected_index -= current_items_per_row
else:
self.selected_index = 0
self._update_scroll()
elif key == ord('s') or key == ord('S'): # S - Down
current_items_per_row = min(self.items_per_row, len(self.video_files))
if self.selected_index + current_items_per_row < len(self.video_files):
self.selected_index += current_items_per_row
else:
self.selected_index = len(self.video_files) - 1
self._update_scroll()
elif key == ord('a') or key == ord('A'): # A - Left
if self.selected_index > 0:
self.selected_index -= 1
self._update_scroll()
elif key == ord('d') or key == ord('D'): # D - Right
if self.selected_index < len(self.video_files) - 1:
self.selected_index += 1
self._update_scroll()
elif key == ord('Q'): # uppercase Q - Fewer items per row (larger thumbnails)
if self.items_per_row > 1:
self.items_per_row -= 1
print(f"Items per row: {self.items_per_row}")
elif key == ord('y') or key == ord('Y'): # Y - More items per row (smaller thumbnails)
self.items_per_row += 1
print(f"Items per row: {self.items_per_row}")
return "none"
def _update_scroll(self):
"""Update scroll offset based on selected item"""
if not self.video_files:
return
# Use fixed items per row
items_per_row = min(self.items_per_row, len(self.video_files))
# Get window dimensions for calculations
try:
window_rect = cv2.getWindowImageRect("Project View")
if window_rect[2] > 0 and window_rect[3] > 0:
window_width = window_rect[2]
window_height = window_rect[3]
else:
window_width = self.window_width
window_height = self.window_height
except:
window_width = self.window_width
window_height = self.window_height
# Calculate thumbnail size and item height dynamically
thumbnail_width, thumbnail_height = self._calculate_thumbnail_size(window_width)
item_height = thumbnail_height + self.PROGRESS_BAR_HEIGHT + self.TEXT_HEIGHT + self.THUMBNAIL_MARGIN
selected_row = self.selected_index // items_per_row
visible_rows = max(1, window_height // item_height)
# Calculate how many rows we can actually show
total_rows = (len(self.video_files) + items_per_row - 1) // items_per_row
# If we can show all rows, no scrolling needed
if total_rows <= visible_rows:
self.scroll_offset = 0
return
# Update scroll to keep selected item visible
if selected_row < self.scroll_offset:
self.scroll_offset = selected_row
elif selected_row >= self.scroll_offset + visible_rows:
self.scroll_offset = selected_row - visible_rows + 1
# Ensure scroll offset doesn't go negative or beyond available content
self.scroll_offset = max(0, min(self.scroll_offset, total_rows - visible_rows))
class VideoEditor:
# Configuration constants
BASE_FRAME_DELAY_MS = 16 # ~60 FPS
SPEED_INCREMENT = 0.2
MIN_PLAYBACK_SPEED = 0.1
MAX_PLAYBACK_SPEED = 10.0
# Seek multiplier configuration
SEEK_MULTIPLIER_INCREMENT = 2.0
MIN_SEEK_MULTIPLIER = 1.0
MAX_SEEK_MULTIPLIER = 100.0
# Auto-repeat seeking configuration
AUTO_REPEAT_DISPLAY_RATE = 1.0
# Timeline configuration
TIMELINE_HEIGHT = 60
TIMELINE_MARGIN = 20
TIMELINE_BAR_HEIGHT = 12
TIMELINE_HANDLE_SIZE = 12
TIMELINE_COLOR_BG = (80, 80, 80)
TIMELINE_COLOR_PROGRESS = (0, 120, 255)
TIMELINE_COLOR_HANDLE = (255, 255, 255)
TIMELINE_COLOR_BORDER = (200, 200, 200)
TIMELINE_COLOR_CUT_POINT = (255, 0, 0)
# Progress bar configuration
PROGRESS_BAR_HEIGHT = 30
PROGRESS_BAR_MARGIN_PERCENT = 5 # 5% margin on each side
PROGRESS_BAR_TOP_MARGIN = 20 # Fixed top margin
PROGRESS_BAR_FADE_DURATION = 3.0 # seconds to fade out after completion
PROGRESS_BAR_COLOR_BG = (50, 50, 50)
PROGRESS_BAR_COLOR_FILL = (0, 255, 0) # Green when complete
PROGRESS_BAR_COLOR_PROGRESS = (0, 120, 255) # Blue during progress
PROGRESS_BAR_COLOR_BORDER = (200, 200, 200)
# Zoom and crop settings
MIN_ZOOM = 0.1
MAX_ZOOM = 10.0
ZOOM_INCREMENT = 0.1
# Supported video extensions
VIDEO_EXTENSIONS = {".mp4", ".avi", ".mov", ".mkv", ".wmv", ".flv", ".webm", ".m4v"}
# Supported image extensions
IMAGE_EXTENSIONS = {".jpg", ".jpeg", ".png", ".bmp", ".tiff", ".tif", ".webp", ".jp2", ".pbm", ".pgm", ".ppm", ".sr", ".ras"}
# Crop adjustment settings
CROP_SIZE_STEP = 15 # pixels to expand/contract crop
def __init__(self, path: str):
self.path = Path(path)
# Video file management
self.video_files = []
self.current_video_index = 0
# Media type tracking
self.is_image_mode = False # True if current file is an image
# Determine if path is file or directory
if self.path.is_file():
self.video_files = [self.path]
elif self.path.is_dir():
# Load all media files from directory
self.video_files = self._get_media_files_from_directory(self.path)
if not self.video_files:
raise ValueError(f"No media files found in directory: {path}")
else:
raise ValueError(f"Path does not exist: {path}")
# Mouse and keyboard interaction
self.mouse_dragging = False
self.timeline_rect = None
self.window_width = 1200
self.window_height = 800
# Auto-repeat seeking state
self.auto_repeat_active = False
self.auto_repeat_direction = 0
self.auto_repeat_shift = False
self.auto_repeat_ctrl = False
self.last_display_update = 0
# Crop settings
self.crop_rect = None # (x, y, width, height)
self.crop_selecting = False
self.crop_start_point = None
self.crop_preview_rect = None
self.crop_history = [] # For undo
# Zoom settings
self.zoom_factor = 1.0
self.zoom_center = None # (x, y) center point for zoom
# Rotation settings
self.rotation_angle = 0 # 0, 90, 180, 270 degrees
# Brightness and contrast settings
self.brightness = 0 # -100 to 100
self.contrast = 1.0 # 0.1 to 3.0
# Marker looping state
self.looping_between_markers = False
# Display offset for panning when zoomed
self.display_offset = [0, 0]
# Fullscreen state
self.is_fullscreen = False
# Progress bar state
self.progress_bar_visible = False
self.progress_bar_progress = 0.0 # 0.0 to 1.0
self.progress_bar_complete = False
self.progress_bar_complete_time = None
self.progress_bar_text = ""
self.progress_bar_fps = 0.0 # Current rendering FPS
# Feedback message state
self.feedback_message = ""
self.feedback_message_time = None
self.feedback_message_duration = 0.5 # seconds to show message
# Crop adjustment settings
self.crop_size_step = self.CROP_SIZE_STEP
# Render thread management
self.render_thread = None
self.render_cancelled = False
self.render_progress_queue = queue.Queue()
self.ffmpeg_process = None # Track FFmpeg process for cancellation
# Display optimization - track when redraw is needed
self.display_needs_update = True
self.last_display_state = None
# Cached transformations for performance
self.cached_transformed_frame = None
self.cached_frame_number = None
self.cached_transform_hash = None
# Project view mode
self.project_view_mode = False
self.project_view = None
# Initialize with first video
self._load_video(self.video_files[0])
# Load saved state after all attributes are initialized
self.load_state()
def _get_state_file_path(self) -> Path:
"""Get the state file path for the current media file"""
if not hasattr(self, 'video_path') or not self.video_path:
print("DEBUG: No video_path available for state file")
return None
state_path = self.video_path.with_suffix('.json')
print(f"DEBUG: State file path would be: {state_path}")
return state_path
def save_state(self):
"""Save current editor state to JSON file"""
state_file = self._get_state_file_path()
if not state_file:
print("No state file path available")
return False
try:
state = {
'timestamp': time.time(),
'current_frame': getattr(self, 'current_frame', 0),
'crop_rect': self.crop_rect,
'zoom_factor': self.zoom_factor,
'zoom_center': self.zoom_center,
'rotation_angle': self.rotation_angle,
'brightness': self.brightness,
'contrast': self.contrast,
'cut_start_frame': self.cut_start_frame,
'cut_end_frame': self.cut_end_frame,
'looping_between_markers': self.looping_between_markers,
'display_offset': self.display_offset,
'playback_speed': getattr(self, 'playback_speed', 1.0),
'seek_multiplier': getattr(self, 'seek_multiplier', 1.0),
'is_playing': getattr(self, 'is_playing', False)
}
with open(state_file, 'w') as f:
json.dump(state, f, indent=2)
print(f"State saved to {state_file}")
# Refresh project view progress data if project view is active
if self.project_view_mode and self.project_view:
self.project_view.refresh_progress_data()
return True
except Exception as e:
print(f"Error saving state: {e}")
return False
def load_state(self) -> bool:
"""Load editor state from JSON file"""
state_file = self._get_state_file_path()
if not state_file:
print("No state file path available")
return False
if not state_file.exists():
print(f"State file does not exist: {state_file}")
return False
print(f"Loading state from: {state_file}")
try:
with open(state_file, 'r') as f:
state = json.load(f)
print(f"State file contents: {state}")
# Restore state values
if 'current_frame' in state:
self.current_frame = state['current_frame']
print(f"Loaded current_frame: {self.current_frame}")
if 'crop_rect' in state and state['crop_rect'] is not None:
self.crop_rect = tuple(state['crop_rect'])
print(f"Loaded crop_rect: {self.crop_rect}")
if 'zoom_factor' in state:
self.zoom_factor = state['zoom_factor']
print(f"Loaded zoom_factor: {self.zoom_factor}")
if 'zoom_center' in state and state['zoom_center'] is not None:
self.zoom_center = tuple(state['zoom_center'])
print(f"Loaded zoom_center: {self.zoom_center}")
if 'rotation_angle' in state:
self.rotation_angle = state['rotation_angle']
print(f"Loaded rotation_angle: {self.rotation_angle}")
if 'brightness' in state:
self.brightness = state['brightness']
print(f"Loaded brightness: {self.brightness}")
if 'contrast' in state:
self.contrast = state['contrast']
print(f"Loaded contrast: {self.contrast}")
if 'cut_start_frame' in state:
self.cut_start_frame = state['cut_start_frame']
print(f"Loaded cut_start_frame: {self.cut_start_frame}")
if 'cut_end_frame' in state:
self.cut_end_frame = state['cut_end_frame']
print(f"Loaded cut_end_frame: {self.cut_end_frame}")
if 'looping_between_markers' in state:
self.looping_between_markers = state['looping_between_markers']
print(f"Loaded looping_between_markers: {self.looping_between_markers}")
if 'display_offset' in state:
self.display_offset = state['display_offset']
print(f"Loaded display_offset: {self.display_offset}")
if 'playback_speed' in state:
self.playback_speed = state['playback_speed']
print(f"Loaded playback_speed: {self.playback_speed}")
if 'seek_multiplier' in state:
self.seek_multiplier = state['seek_multiplier']
print(f"Loaded seek_multiplier: {self.seek_multiplier}")
if 'is_playing' in state:
self.is_playing = state['is_playing']
print(f"Loaded is_playing: {self.is_playing}")
# Validate cut markers against current video length
if self.cut_start_frame is not None and self.cut_start_frame >= self.total_frames:
print(f"DEBUG: cut_start_frame {self.cut_start_frame} is beyond video length {self.total_frames}, clearing")
self.cut_start_frame = None
if self.cut_end_frame is not None and self.cut_end_frame >= self.total_frames:
print(f"DEBUG: cut_end_frame {self.cut_end_frame} is beyond video length {self.total_frames}, clearing")
self.cut_end_frame = None
# Calculate and show marker positions on timeline
if self.cut_start_frame is not None and self.cut_end_frame is not None:
start_progress = self.cut_start_frame / max(1, self.total_frames - 1)
end_progress = self.cut_end_frame / max(1, self.total_frames - 1)
print(f"Markers will be drawn at: Start {start_progress:.4f} ({self.cut_start_frame}/{self.total_frames}), End {end_progress:.4f} ({self.cut_end_frame}/{self.total_frames})")
# Validate and clamp values
self.current_frame = max(0, min(self.current_frame, getattr(self, 'total_frames', 1) - 1))
self.zoom_factor = max(self.MIN_ZOOM, min(self.MAX_ZOOM, self.zoom_factor))
self.brightness = max(-100, min(100, self.brightness))
self.contrast = max(0.1, min(3.0, self.contrast))
self.playback_speed = max(self.MIN_PLAYBACK_SPEED, min(self.MAX_PLAYBACK_SPEED, self.playback_speed))
self.seek_multiplier = max(self.MIN_SEEK_MULTIPLIER, min(self.MAX_SEEK_MULTIPLIER, self.seek_multiplier))
# Apply loaded settings
self.clear_transformation_cache()
self.load_current_frame()
print("Successfully loaded and applied all settings from state file")
return True
except Exception as e:
print(f"Error loading state: {e}")
return False
def _is_video_file(self, file_path: Path) -> bool:
"""Check if file is a supported video format"""
return file_path.suffix.lower() in self.VIDEO_EXTENSIONS
def _is_image_file(self, file_path: Path) -> bool:
"""Check if file is a supported image format"""
return file_path.suffix.lower() in self.IMAGE_EXTENSIONS
def _is_media_file(self, file_path: Path) -> bool:
"""Check if file is a supported media format (video or image)"""
return self._is_video_file(file_path) or self._is_image_file(file_path)
def _get_next_screenshot_filename(self, video_path: Path) -> str:
"""Generate the next available screenshot filename: video_frame_00001.jpg, video_frame_00002.jpg, etc."""
directory = video_path.parent
base_name = video_path.stem
# Pattern to match existing screenshot files: video_frame_00001.jpg, video_frame_00002.jpg, etc.
pattern = re.compile(rf"^{re.escape(base_name)}_frame_(\d{{5}})\.(jpg|jpeg|png)$")
existing_numbers = set()
for file_path in directory.iterdir():
if file_path.is_file():
match = pattern.match(file_path.name)
if match:
existing_numbers.add(int(match.group(1)))
# Find the next available number starting from 1
next_number = 1
while next_number in existing_numbers:
next_number += 1
return f"{base_name}_frame_{next_number:05d}.jpg"
def save_current_frame(self):
"""Save the current frame as a screenshot"""
if self.current_display_frame is None:
print("No frame to save")
return False
# Generate the next available screenshot filename
screenshot_name = self._get_next_screenshot_filename(self.video_path)
screenshot_path = self.video_path.parent / screenshot_name
# Apply current transformations (crop, zoom, rotation, brightness/contrast) to the frame
processed_frame = self.apply_crop_zoom_and_rotation(self.current_display_frame.copy())
if processed_frame is not None:
# Save the processed frame
success = cv2.imwrite(str(screenshot_path), processed_frame)
if success:
print(f"Screenshot saved: {screenshot_name}")
self.show_feedback_message(f"Screenshot saved: {screenshot_name}")
return True
else:
print(f"Error: Could not save screenshot to {screenshot_path}")
self.show_feedback_message("Error: Could not save screenshot")
return False
else:
print("Error: Could not process frame for screenshot")
self.show_feedback_message("Error: Could not process frame")
return False
def _get_media_files_from_directory(self, directory: Path) -> List[Path]:
"""Get all media files (video and image) from a directory, sorted by name"""
media_files = set()
for file_path in directory.iterdir():
if (
file_path.is_file()
and self._is_media_file(file_path)
):
media_files.add(file_path)
# Pattern to match edited files: basename_edited_001.ext, basename_edited_002.ext, etc.
edited_pattern = re.compile(r"^(.+)_edited_\d{3}$")
edited_base_names = set()
for file_path in media_files:
match = edited_pattern.match(file_path.stem)
if match:
edited_base_names.add(match.group(1))
non_edited_media = set()
for file_path in media_files:
# Skip if this is an edited file
if edited_pattern.match(file_path.stem):
continue
# Skip if there's already an edited version of this file
if file_path.stem in edited_base_names:
continue
non_edited_media.add(file_path)
return sorted(non_edited_media)
def _load_video(self, media_path: Path):
"""Load a media file (video or image) and initialize properties"""
if hasattr(self, "cap") and self.cap:
self.cap.release()
self.video_path = media_path
self.is_image_mode = self._is_image_file(media_path)
if self.is_image_mode:
# Load static image
self.static_image = cv2.imread(str(media_path))
if self.static_image is None:
raise ValueError(f"Could not load image file: {media_path}")
# Set up image properties to mimic video interface
self.frame_height, self.frame_width = self.static_image.shape[:2]
self.total_frames = 1
self.fps = 30 # Dummy FPS for image mode
self.cap = None
print(f"Loaded image: {self.video_path.name}")
print(f" Resolution: {self.frame_width}x{self.frame_height}")
else:
# Try different backends for better performance
# Order of preference: FFmpeg (best for video files), DirectShow (cameras), any available
backends_to_try = []
if hasattr(cv2, 'CAP_FFMPEG'): # FFmpeg - best for video files
backends_to_try.append(cv2.CAP_FFMPEG)
if hasattr(cv2, 'CAP_DSHOW'): # DirectShow - usually for cameras
backends_to_try.append(cv2.CAP_DSHOW)
backends_to_try.append(cv2.CAP_ANY) # Fallback
self.cap = None
for backend in backends_to_try:
try:
self.cap = Cv2BufferedCap(self.video_path, backend)
if self.cap.isOpened():
break
except Exception:
continue
if not self.cap or not self.cap.isOpened():
raise ValueError(f"Could not open video file: {media_path}")
# Video properties from buffered cap
self.total_frames = self.cap.total_frames
self.fps = self.cap.fps
self.frame_width = self.cap.frame_width
self.frame_height = self.cap.frame_height
# Get codec information for debugging
fourcc = int(self.cap.cap.get(cv2.CAP_PROP_FOURCC))
codec = "".join([chr((fourcc >> 8 * i) & 0xFF) for i in range(4)])
# Get backend information
backend_name = "FFmpeg" if hasattr(cv2, 'CAP_FFMPEG') and backend == cv2.CAP_FFMPEG else "Other"
print(f"Loaded video: {self.video_path.name} ({self.current_video_index + 1}/{len(self.video_files)})")
print(f" Codec: {codec} | Backend: {backend_name} | Resolution: {self.frame_width}x{self.frame_height}")
print(f" FPS: {self.fps:.2f} | Frames: {self.total_frames} | Duration: {self.total_frames/self.fps:.1f}s")
# Performance warning for known problematic cases
if codec in ['H264', 'H.264', 'AVC1', 'avc1'] and self.total_frames > 10000:
print(" Warning: Large H.264 video detected - seeking may be slow")
if self.frame_width * self.frame_height > 1920 * 1080:
print(" Warning: High resolution video - decoding may be slow")
if self.fps > 60:
print(" Warning: High framerate video - may impact playback smoothness")
# Set default values for video-specific properties
self.current_frame = 0
self.is_playing = False if self.is_image_mode else False # Images start paused
self.playback_speed = 1.0
self.seek_multiplier = 1.0
self.cut_start_frame = None
self.cut_end_frame = None
# Always reset these regardless of state
self.current_display_frame = None
def switch_to_video(self, index: int):
"""Switch to a specific video by index"""
if 0 <= index < len(self.video_files):
self.current_video_index = index
self._load_video(self.video_files[index])
self.load_current_frame()
def next_video(self):
"""Switch to the next video"""
self.save_state() # Save current video state before switching
next_index = (self.current_video_index + 1) % len(self.video_files)
self.switch_to_video(next_index)
def previous_video(self):
"""Switch to the previous video"""
self.save_state() # Save current video state before switching
prev_index = (self.current_video_index - 1) % len(self.video_files)
self.switch_to_video(prev_index)
def load_current_frame(self) -> bool:
"""Load the current frame into display cache"""
if self.is_image_mode:
# For images, just copy the static image
self.current_display_frame = self.static_image.copy()
return True
else:
# Use buffered cap to get frame
try:
self.current_display_frame = self.cap.get_frame(self.current_frame)
return True
except Exception as e:
print(f"Failed to load frame {self.current_frame}: {e}")
return False
def calculate_frame_delay(self) -> int:
"""Calculate frame delay in milliseconds based on playback speed"""
delay_ms = int(self.BASE_FRAME_DELAY_MS / self.playback_speed)
return max(1, delay_ms)
def seek_video(self, frames_delta: int):
"""Seek video by specified number of frames"""
target_frame = max(
0, min(self.current_frame + frames_delta, self.total_frames - 1)
)
self.current_frame = target_frame
self.load_current_frame()
self.display_needs_update = True
def seek_video_with_modifier(
self, direction: int, shift_pressed: bool, ctrl_pressed: bool
):
"""Seek video with different frame counts based on modifiers and seek multiplier"""
if ctrl_pressed:
base_frames = 60 # Ctrl: 60 frames
elif shift_pressed:
base_frames = 10 # Shift: 10 frames
else:
base_frames = 1 # Default: 1 frame
# Apply seek multiplier to the base frame count
frames = direction * int(base_frames * self.seek_multiplier)
self.seek_video(frames)
def start_auto_repeat_seek(self, direction: int, shift_pressed: bool, ctrl_pressed: bool):
"""Start auto-repeat seeking"""
if self.is_image_mode:
return
self.auto_repeat_active = True
self.auto_repeat_direction = direction
self.auto_repeat_shift = shift_pressed
self.auto_repeat_ctrl = ctrl_pressed
# Initialize last_display_update to prevent immediate auto-repeat
self.last_display_update = time.time()
self.seek_video_with_modifier(direction, shift_pressed, ctrl_pressed)
def stop_auto_repeat_seek(self):
"""Stop auto-repeat seeking"""
self.auto_repeat_active = False
self.auto_repeat_direction = 0
self.auto_repeat_shift = False
self.auto_repeat_ctrl = False
def update_auto_repeat_seek(self):
"""Update auto-repeat seeking"""
if not self.auto_repeat_active or self.is_image_mode:
return
current_time = time.time()
if current_time - self.last_display_update >= self.AUTO_REPEAT_DISPLAY_RATE:
self.seek_video_with_modifier(
self.auto_repeat_direction,
self.auto_repeat_shift,
self.auto_repeat_ctrl
)
self.last_display_update = current_time
def seek_to_frame(self, frame_number: int):
"""Seek to specific frame"""
self.current_frame = max(0, min(frame_number, self.total_frames - 1))
self.load_current_frame()
def advance_frame(self) -> bool:
"""Advance to next frame - handles playback speed and marker looping"""
if not self.is_playing:
return True
# Calculate how many frames to advance based on speed
frames_to_advance = max(1, int(self.playback_speed))
new_frame = self.current_frame + frames_to_advance
# Handle marker looping bounds
if self.looping_between_markers and self.cut_start_frame is not None and self.cut_end_frame is not None:
if new_frame >= self.cut_end_frame:
# Loop back to start marker
new_frame = self.cut_start_frame
elif new_frame >= self.total_frames:
# Loop to beginning
new_frame = 0
# Update current frame and load it
self.current_frame = new_frame
return self.load_current_frame()
def apply_crop_zoom_and_rotation(self, frame):
"""Apply current crop, zoom, rotation, and brightness/contrast settings to frame"""
if frame is None:
return None
# Create a hash of the transformation parameters for caching
transform_hash = hash((
self.crop_rect,
self.zoom_factor,
self.rotation_angle,
self.brightness,
self.contrast,
tuple(self.display_offset)
))
# Check if we can use cached transformation during auto-repeat seeking
if (self.auto_repeat_active and
self.cached_transformed_frame is not None and
self.cached_frame_number == self.current_frame and
self.cached_transform_hash == transform_hash):
return self.cached_transformed_frame.copy()
# Work in-place when possible to avoid unnecessary copying
processed_frame = frame
# Apply brightness/contrast first (to original frame for best quality)
processed_frame = self.apply_brightness_contrast(processed_frame)
# Apply crop
if self.crop_rect:
x, y, w, h = self.crop_rect
x, y, w, h = int(x), int(y), int(w), int(h)
# Ensure crop is within frame bounds
x = max(0, min(x, processed_frame.shape[1] - 1))
y = max(0, min(y, processed_frame.shape[0] - 1))
w = min(w, processed_frame.shape[1] - x)
h = min(h, processed_frame.shape[0] - y)
if w > 0 and h > 0:
processed_frame = processed_frame[y : y + h, x : x + w]
# Apply rotation
if self.rotation_angle != 0:
processed_frame = self.apply_rotation(processed_frame)
# Apply zoom
if self.zoom_factor != 1.0:
height, width = processed_frame.shape[:2]
new_width = int(width * self.zoom_factor)
new_height = int(height * self.zoom_factor)
processed_frame = cv2.resize(
processed_frame, (new_width, new_height), interpolation=cv2.INTER_LINEAR
)
# Handle zoom center and display offset
if new_width > self.window_width or new_height > self.window_height:
# Calculate crop from zoomed image to fit window
start_x = max(0, self.display_offset[0])
start_y = max(0, self.display_offset[1])
end_x = min(new_width, start_x + self.window_width)
end_y = min(new_height, start_y + self.window_height)
processed_frame = processed_frame[start_y:end_y, start_x:end_x]
# Cache the result for auto-repeat seeking performance
if self.auto_repeat_active:
self.cached_transformed_frame = processed_frame.copy()
self.cached_frame_number = self.current_frame
self.cached_transform_hash = transform_hash
return processed_frame
def clear_transformation_cache(self):
"""Clear the cached transformation to force recalculation"""
self.cached_transformed_frame = None
self.cached_frame_number = None
self.cached_transform_hash = None
def apply_rotation(self, frame):
"""Apply rotation to frame"""
if self.rotation_angle == 0:
return frame
elif self.rotation_angle == 90:
return cv2.rotate(frame, cv2.ROTATE_90_CLOCKWISE)
elif self.rotation_angle == 180:
return cv2.rotate(frame, cv2.ROTATE_180)
elif self.rotation_angle == 270:
return cv2.rotate(frame, cv2.ROTATE_90_COUNTERCLOCKWISE)
return frame
def rotate_clockwise(self):
"""Rotate video 90 degrees clockwise"""
self.rotation_angle = (self.rotation_angle + 90) % 360
self.clear_transformation_cache()
def apply_brightness_contrast(self, frame):
"""Apply brightness and contrast adjustments to frame"""
if self.brightness == 0 and self.contrast == 1.0:
return frame
# Convert brightness from -100/100 range to -255/255 range
brightness_value = self.brightness * 2.55
# Apply brightness and contrast: new_pixel = contrast * old_pixel + brightness
adjusted = cv2.convertScaleAbs(
frame, alpha=self.contrast, beta=brightness_value
)
return adjusted
def adjust_brightness(self, delta: int):
"""Adjust brightness by delta (-100 to 100)"""
self.brightness = max(-100, min(100, self.brightness + delta))
self.clear_transformation_cache()
self.display_needs_update = True
def adjust_contrast(self, delta: float):
"""Adjust contrast by delta (0.1 to 3.0)"""
self.contrast = max(0.1, min(3.0, self.contrast + delta))
self.clear_transformation_cache()
self.display_needs_update = True
def show_progress_bar(self, text: str = "Processing..."):
"""Show progress bar with given text"""
self.progress_bar_visible = True
self.progress_bar_progress = 0.0
self.progress_bar_complete = False
self.progress_bar_complete_time = None
self.progress_bar_text = text
self.display_needs_update = True
def update_progress_bar(self, progress: float, text: str = None, fps: float = None):
"""Update progress bar progress (0.0 to 1.0) and optionally text and FPS"""
if self.progress_bar_visible:
self.progress_bar_progress = max(0.0, min(1.0, progress))
if text is not None:
self.progress_bar_text = text
if fps is not None:
self.progress_bar_fps = fps
# Mark as complete when reaching 100%
if self.progress_bar_progress >= 1.0 and not self.progress_bar_complete:
self.progress_bar_complete = True
self.progress_bar_complete_time = time.time()
def hide_progress_bar(self):
"""Hide progress bar"""
self.progress_bar_visible = False
self.progress_bar_complete = False
self.progress_bar_complete_time = None
self.progress_bar_fps = 0.0
def show_feedback_message(self, message: str):
"""Show a feedback message on screen for a few seconds"""
self.feedback_message = message
self.feedback_message_time = time.time()
self.display_needs_update = True
def toggle_fullscreen(self):
"""Toggle between windowed and fullscreen mode"""
window_title = "Image Editor" if self.is_image_mode else "Video Editor"
if self.is_fullscreen:
# Switch to windowed mode
self.is_fullscreen = False
cv2.setWindowProperty(window_title, cv2.WND_PROP_FULLSCREEN, cv2.WINDOW_NORMAL)
cv2.resizeWindow(window_title, 1200, 800)
print("Switched to windowed mode")
else:
# Switch to fullscreen mode
self.is_fullscreen = True
cv2.setWindowProperty(window_title, cv2.WND_PROP_FULLSCREEN, cv2.WINDOW_FULLSCREEN)
print("Switched to fullscreen mode")
self.display_needs_update = True
def toggle_project_view(self):
"""Toggle between editor and project view mode"""
if self.project_view_mode:
# Switch back to editor mode
self.project_view_mode = False
if self.project_view:
cv2.destroyWindow("Project View")
self.project_view = None
print("Switched to editor mode")
else:
# Switch to project view mode
self.project_view_mode = True
# Create project view for the current directory
if self.path.is_dir():
project_dir = self.path
else:
project_dir = self.path.parent
self.project_view = ProjectView(project_dir, self)
# Create separate window for project view
cv2.namedWindow("Project View", cv2.WINDOW_AUTOSIZE)
print("Switched to project view mode")
self.display_needs_update = True
def open_video_from_project_view(self, video_path: Path):
"""Open a video from project view in editor mode"""
print(f"Attempting to open video: {video_path}")
print(f"Video path exists: {video_path.exists()}")
# Save current state before switching
self.save_state()
# Find the video in our video_files list
try:
video_index = self.video_files.index(video_path)
self.current_video_index = video_index
self._load_video(video_path)
self.load_current_frame()
# Load the saved state for this video (same logic as normal video loading)
self.load_state()
print(f"Opened video: {video_path.name}")
except ValueError:
print(f"Video not found in current session: {video_path.name}")
# If video not in current session, reload the directory
self.path = video_path.parent
self.video_files = self._get_media_files_from_directory(self.path)
if video_path in self.video_files:
video_index = self.video_files.index(video_path)
self.current_video_index = video_index
self._load_video(video_path)
self.load_current_frame()
# Load the saved state for this video (same logic as normal video loading)
self.load_state()
print(f"Opened video: {video_path.name}")
else:
print(f"Could not find video: {video_path.name}")
return
# Keep project view open but switch focus to video editor
# Don't destroy the project view window - just let the user switch between them
def draw_feedback_message(self, frame):
"""Draw feedback message on frame if visible"""
if not self.feedback_message or not self.feedback_message_time:
return
# Check if message should still be shown
elapsed = time.time() - self.feedback_message_time
if elapsed > self.feedback_message_duration:
self.feedback_message = ""
self.feedback_message_time = None
return
height, width = frame.shape[:2]
# Calculate message position (center of frame)
font = cv2.FONT_HERSHEY_SIMPLEX
font_scale = 1.0
thickness = 2
# Get text size
text_size = cv2.getTextSize(self.feedback_message, font, font_scale, thickness)[0]
text_x = (width - text_size[0]) // 2
text_y = (height + text_size[1]) // 2
# Draw background rectangle
padding = 10
rect_x1 = text_x - padding
rect_y1 = text_y - text_size[1] - padding
rect_x2 = text_x + text_size[0] + padding
rect_y2 = text_y + padding
# Semi-transparent background
overlay = frame.copy()
cv2.rectangle(overlay, (rect_x1, rect_y1), (rect_x2, rect_y2), (0, 0, 0), -1)
alpha = 0.7
cv2.addWeighted(overlay, alpha, frame, 1 - alpha, 0, frame)
# Draw text with shadow
cv2.putText(frame, self.feedback_message, (text_x + 2, text_y + 2), font, font_scale, (0, 0, 0), thickness + 1)
cv2.putText(frame, self.feedback_message, (text_x, text_y), font, font_scale, (255, 255, 255), thickness)
def draw_progress_bar(self, frame):
"""Draw progress bar on frame if visible - positioned at top with full width"""
if not self.progress_bar_visible:
return
# Check if we should fade out
if self.progress_bar_complete and self.progress_bar_complete_time:
elapsed = time.time() - self.progress_bar_complete_time
if elapsed > self.PROGRESS_BAR_FADE_DURATION:
self.hide_progress_bar()
return
# Calculate fade alpha (1.0 at start, 0.0 at end)
fade_alpha = max(0.0, 1.0 - (elapsed / self.PROGRESS_BAR_FADE_DURATION))
else:
fade_alpha = 1.0
height, width = frame.shape[:2]
# Calculate progress bar position (top of frame with 5% margins)
margin_width = int(width * self.PROGRESS_BAR_MARGIN_PERCENT / 100)
bar_width = width - (2 * margin_width)
bar_x = margin_width
bar_y = self.PROGRESS_BAR_TOP_MARGIN
# Apply fade alpha to colors
bg_color = tuple(int(c * fade_alpha) for c in self.PROGRESS_BAR_COLOR_BG)
border_color = tuple(
int(c * fade_alpha) for c in self.PROGRESS_BAR_COLOR_BORDER
)
if self.progress_bar_complete:
fill_color = tuple(
int(c * fade_alpha) for c in self.PROGRESS_BAR_COLOR_FILL
)
else:
fill_color = tuple(
int(c * fade_alpha) for c in self.PROGRESS_BAR_COLOR_PROGRESS
)
# Draw background
cv2.rectangle(
frame,
(bar_x, bar_y),
(bar_x + bar_width, bar_y + self.PROGRESS_BAR_HEIGHT),
bg_color,
-1,
)
# Draw progress fill
fill_width = int(bar_width * self.progress_bar_progress)
if fill_width > 0:
cv2.rectangle(
frame,
(bar_x, bar_y),
(bar_x + fill_width, bar_y + self.PROGRESS_BAR_HEIGHT),
fill_color,
-1,
)
# Draw border
cv2.rectangle(
frame,
(bar_x, bar_y),
(bar_x + bar_width, bar_y + self.PROGRESS_BAR_HEIGHT),
border_color,
2,
)
# Draw progress percentage on the left
percentage_text = f"{self.progress_bar_progress * 100:.1f}%"
text_color = tuple(int(255 * fade_alpha) for _ in range(3))
cv2.putText(
frame,
percentage_text,
(bar_x + 12, bar_y + 22),
cv2.FONT_HERSHEY_SIMPLEX,
0.5,
(0, 0, 0),
4,
)
cv2.putText(
frame,
percentage_text,
(bar_x + 10, bar_y + 20),
cv2.FONT_HERSHEY_SIMPLEX,
0.5,
text_color,
2,
)
# Draw FPS on the right if available
if self.progress_bar_fps > 0:
fps_text = f"{self.progress_bar_fps:.1f} FPS"
fps_text_size = cv2.getTextSize(fps_text, cv2.FONT_HERSHEY_SIMPLEX, 0.5, 1)[
0
]
fps_x = bar_x + bar_width - fps_text_size[0] - 10
cv2.putText(
frame,
fps_text,
(fps_x + 2, bar_y + 22),
cv2.FONT_HERSHEY_SIMPLEX,
0.5,
(0, 0, 0),
4,
)
cv2.putText(
frame,
fps_text,
(fps_x, bar_y + 20),
cv2.FONT_HERSHEY_SIMPLEX,
0.5,
text_color,
2,
)
# Draw main text in center
if self.progress_bar_text:
text_size = cv2.getTextSize(
self.progress_bar_text, cv2.FONT_HERSHEY_SIMPLEX, 0.5, 1
)[0]
text_x = bar_x + (bar_width - text_size[0]) // 2
text_y = bar_y + 20
# Draw text shadow for better visibility
cv2.putText(
frame,
self.progress_bar_text,
(text_x + 2, text_y + 2),
cv2.FONT_HERSHEY_SIMPLEX,
0.5,
(0, 0, 0),
4,
)
cv2.putText(
frame,
self.progress_bar_text,
(text_x, text_y),
cv2.FONT_HERSHEY_SIMPLEX,
0.5,
text_color,
2,
)
def draw_timeline(self, frame):
"""Draw timeline at the bottom of the frame"""
# Don't draw timeline for images
if self.is_image_mode:
return
height, width = frame.shape[:2]
# Timeline background area
timeline_y = height - self.TIMELINE_HEIGHT
cv2.rectangle(frame, (0, timeline_y), (width, height), (40, 40, 40), -1)
# Calculate timeline bar position
bar_y = timeline_y + (self.TIMELINE_HEIGHT - self.TIMELINE_BAR_HEIGHT) // 2
bar_x_start = self.TIMELINE_MARGIN
bar_x_end = width - self.TIMELINE_MARGIN
bar_width = bar_x_end - bar_x_start
self.timeline_rect = (bar_x_start, bar_y, bar_width, self.TIMELINE_BAR_HEIGHT)
# Draw timeline background
cv2.rectangle(
frame,
(bar_x_start, bar_y),
(bar_x_end, bar_y + self.TIMELINE_BAR_HEIGHT),
self.TIMELINE_COLOR_BG,
-1,
)
cv2.rectangle(
frame,
(bar_x_start, bar_y),
(bar_x_end, bar_y + self.TIMELINE_BAR_HEIGHT),
self.TIMELINE_COLOR_BORDER,
1,
)
# Draw progress
if self.total_frames > 0:
progress = self.current_frame / max(1, self.total_frames - 1)
progress_width = int(bar_width * progress)
if progress_width > 0:
cv2.rectangle(
frame,
(bar_x_start, bar_y),
(bar_x_start + progress_width, bar_y + self.TIMELINE_BAR_HEIGHT),
self.TIMELINE_COLOR_PROGRESS,
-1,
)
# Draw current position handle
handle_x = bar_x_start + progress_width
handle_y = bar_y + self.TIMELINE_BAR_HEIGHT // 2
cv2.circle(
frame,
(handle_x, handle_y),
self.TIMELINE_HANDLE_SIZE // 2,
self.TIMELINE_COLOR_HANDLE,
-1,
)
cv2.circle(
frame,
(handle_x, handle_y),
self.TIMELINE_HANDLE_SIZE // 2,
self.TIMELINE_COLOR_BORDER,
2,
)
# Draw cut points
if self.cut_start_frame is not None:
cut_start_progress = self.cut_start_frame / max(
1, self.total_frames - 1
)
cut_start_x = bar_x_start + int(bar_width * cut_start_progress)
cv2.line(
frame,
(cut_start_x, bar_y),
(cut_start_x, bar_y + self.TIMELINE_BAR_HEIGHT),
self.TIMELINE_COLOR_CUT_POINT,
3,
)
cv2.putText(
frame,
"1",
(cut_start_x - 5, bar_y - 5),
cv2.FONT_HERSHEY_SIMPLEX,
0.4,
self.TIMELINE_COLOR_CUT_POINT,
1,
)
if self.cut_end_frame is not None:
cut_end_progress = self.cut_end_frame / max(1, self.total_frames - 1)
cut_end_x = bar_x_start + int(bar_width * cut_end_progress)
cv2.line(
frame,
(cut_end_x, bar_y),
(cut_end_x, bar_y + self.TIMELINE_BAR_HEIGHT),
self.TIMELINE_COLOR_CUT_POINT,
3,
)
cv2.putText(
frame,
"2",
(cut_end_x - 5, bar_y - 5),
cv2.FONT_HERSHEY_SIMPLEX,
0.4,
self.TIMELINE_COLOR_CUT_POINT,
1,
)
def display_current_frame(self):
"""Display the current frame with all overlays"""
if self.current_display_frame is None:
return
# Check if display needs update (optimization)
current_state = (
self.current_frame,
self.crop_rect,
self.zoom_factor,
self.rotation_angle,
self.brightness,
self.contrast,
self.display_offset,
self.progress_bar_visible,
self.feedback_message
)
# Always update display when paused to ensure UI elements are visible
if not self.display_needs_update and current_state == self.last_display_state and self.is_playing:
return # Skip redraw if nothing changed and playing
self.last_display_state = current_state
self.display_needs_update = False
# Apply crop, zoom, and rotation transformations for preview
display_frame = self.apply_crop_zoom_and_rotation(
self.current_display_frame
)
if display_frame is None:
return
# Resize to fit window while maintaining aspect ratio
height, width = display_frame.shape[:2]
available_height = self.window_height - (0 if self.is_image_mode else self.TIMELINE_HEIGHT)
scale = min(self.window_width / width, available_height / height)
if scale < 1.0:
new_width = int(width * scale)
new_height = int(height * scale)
display_frame = cv2.resize(display_frame, (new_width, new_height))
# Create canvas with timeline space
canvas = np.zeros((self.window_height, self.window_width, 3), dtype=np.uint8)
# Center the frame on canvas
frame_height, frame_width = display_frame.shape[:2]
start_y = (available_height - frame_height) // 2
start_x = (self.window_width - frame_width) // 2
canvas[start_y : start_y + frame_height, start_x : start_x + frame_width] = (
display_frame
)
# Draw crop selection preview during Shift+Click+Drag
if self.crop_preview_rect:
x, y, w, h = self.crop_preview_rect
cv2.rectangle(
canvas, (int(x), int(y)), (int(x + w), int(y + h)), (0, 255, 0), 2
)
# Add info overlay
rotation_text = (
f" | Rotation: {self.rotation_angle}°" if self.rotation_angle != 0 else ""
)
brightness_text = (
f" | Brightness: {self.brightness}" if self.brightness != 0 else ""
)
contrast_text = (
f" | Contrast: {self.contrast:.1f}" if self.contrast != 1.0 else ""
)
seek_multiplier_text = (
f" | Seek: {self.seek_multiplier:.1f}x" if self.seek_multiplier != 1.0 else ""
)
if self.is_image_mode:
info_text = f"Image | Zoom: {self.zoom_factor:.1f}x{rotation_text}{brightness_text}{contrast_text}"
else:
info_text = f"Frame: {self.current_frame}/{self.total_frames} | Speed: {self.playback_speed:.1f}x | Zoom: {self.zoom_factor:.1f}x{seek_multiplier_text}{rotation_text}{brightness_text}{contrast_text} | {'Playing' if self.is_playing else 'Paused'}"
cv2.putText(
canvas,
info_text,
(10, 30),
cv2.FONT_HERSHEY_SIMPLEX,
0.7,
(255, 255, 255),
2,
)
cv2.putText(
canvas, info_text, (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (0, 0, 0), 1
)
# Add video navigation info
if len(self.video_files) > 1:
video_text = f"Video: {self.current_video_index + 1}/{len(self.video_files)} - {self.video_path.name}"
cv2.putText(
canvas,
video_text,
(10, 60),
cv2.FONT_HERSHEY_SIMPLEX,
0.6,
(255, 255, 255),
2,
)
cv2.putText(
canvas,
video_text,
(10, 60),
cv2.FONT_HERSHEY_SIMPLEX,
0.6,
(0, 0, 0),
1,
)
y_offset = 90
else:
y_offset = 60
# Add crop info
if self.crop_rect:
crop_text = f"Crop: {int(self.crop_rect[0])},{int(self.crop_rect[1])} {int(self.crop_rect[2])}x{int(self.crop_rect[3])}"
cv2.putText(
canvas,
crop_text,
(10, y_offset),
cv2.FONT_HERSHEY_SIMPLEX,
0.6,
(255, 255, 255),
2,
)
cv2.putText(
canvas,
crop_text,
(10, y_offset),
cv2.FONT_HERSHEY_SIMPLEX,
0.6,
(0, 0, 0),
1,
)
y_offset += 30
# Add cut info
if self.cut_start_frame is not None or self.cut_end_frame is not None:
cut_text = (
f"Cut: {self.cut_start_frame or '?'} - {self.cut_end_frame or '?'}"
)
cv2.putText(
canvas,
cut_text,
(10, y_offset),
cv2.FONT_HERSHEY_SIMPLEX,
0.6,
(255, 255, 255),
2,
)
cv2.putText(
canvas,
cut_text,
(10, y_offset),
cv2.FONT_HERSHEY_SIMPLEX,
0.6,
(0, 0, 0),
1,
)
# Draw timeline
self.draw_timeline(canvas)
# Draw progress bar (if visible)
self.draw_progress_bar(canvas)
# Draw feedback message (if visible)
self.draw_feedback_message(canvas)
window_title = "Image Editor" if self.is_image_mode else "Video Editor"
cv2.imshow(window_title, canvas)
def mouse_callback(self, event, x, y, flags, _):
"""Handle mouse events"""
# Handle timeline interaction (not for images)
if self.timeline_rect and not self.is_image_mode:
bar_x_start, bar_y, bar_width, bar_height = self.timeline_rect
bar_x_end = bar_x_start + bar_width
if bar_y <= y <= bar_y + bar_height + 10:
if event == cv2.EVENT_LBUTTONDOWN:
if bar_x_start <= x <= bar_x_end:
self.mouse_dragging = True
self.seek_to_timeline_position(x, bar_x_start, bar_width)
elif event == cv2.EVENT_MOUSEMOVE and self.mouse_dragging:
if bar_x_start <= x <= bar_x_end:
self.seek_to_timeline_position(x, bar_x_start, bar_width)
elif event == cv2.EVENT_LBUTTONUP:
self.mouse_dragging = False
return
# Handle crop selection (Shift + click and drag)
if flags & cv2.EVENT_FLAG_SHIFTKEY:
if event == cv2.EVENT_LBUTTONDOWN:
self.crop_selecting = True
self.crop_start_point = (x, y)
self.crop_preview_rect = None
elif event == cv2.EVENT_MOUSEMOVE and self.crop_selecting:
if self.crop_start_point:
start_x, start_y = self.crop_start_point
width = abs(x - start_x)
height = abs(y - start_y)
crop_x = min(start_x, x)
crop_y = min(start_y, y)
self.crop_preview_rect = (crop_x, crop_y, width, height)
elif event == cv2.EVENT_LBUTTONUP and self.crop_selecting:
if self.crop_start_point and self.crop_preview_rect:
# Convert screen coordinates to video coordinates
self.set_crop_from_screen_coords(self.crop_preview_rect)
self.crop_selecting = False
self.crop_start_point = None
self.crop_preview_rect = None
# Handle zoom center (Ctrl + click)
if flags & cv2.EVENT_FLAG_CTRLKEY and event == cv2.EVENT_LBUTTONDOWN:
self.zoom_center = (x, y)
# Handle scroll wheel for zoom (Ctrl + scroll)
if flags & cv2.EVENT_FLAG_CTRLKEY:
if event == cv2.EVENT_MOUSEWHEEL:
if flags > 0: # Scroll up
self.zoom_factor = min(
self.MAX_ZOOM, self.zoom_factor + self.ZOOM_INCREMENT
)
else: # Scroll down
self.zoom_factor = max(
self.MIN_ZOOM, self.zoom_factor - self.ZOOM_INCREMENT
)
self.clear_transformation_cache()
def set_crop_from_screen_coords(self, screen_rect):
"""Convert screen coordinates to video frame coordinates and set crop"""
x, y, w, h = screen_rect
if self.current_display_frame is None:
return
# Get the original frame dimensions
original_height, original_width = self.current_display_frame.shape[:2]
available_height = self.window_height - (0 if self.is_image_mode else self.TIMELINE_HEIGHT)
# Calculate how the original frame is displayed (after crop/zoom/rotation)
display_frame = self.apply_crop_zoom_and_rotation(
self.current_display_frame.copy()
)
if display_frame is None:
return
display_height, display_width = display_frame.shape[:2]
# Calculate scale for the display frame
scale = min(
self.window_width / display_width, available_height / display_height
)
if scale < 1.0:
final_display_width = int(display_width * scale)
final_display_height = int(display_height * scale)
else:
final_display_width = display_width
final_display_height = display_height
scale = 1.0
start_x = (self.window_width - final_display_width) // 2
start_y = (available_height - final_display_height) // 2
# Convert screen coordinates to display frame coordinates
display_x = (x - start_x) / scale
display_y = (y - start_y) / scale
display_w = w / scale
display_h = h / scale
# Clamp to display frame bounds
display_x = max(0, min(display_x, display_width))
display_y = max(0, min(display_y, display_height))
display_w = min(display_w, display_width - display_x)
display_h = min(display_h, display_height - display_y)
# Now we need to convert from the display frame coordinates back to original frame coordinates
# The display frame is the result of: original -> crop -> rotation -> zoom
# Step 1: Reverse zoom
if self.zoom_factor != 1.0:
display_x = display_x / self.zoom_factor
display_y = display_y / self.zoom_factor
display_w = display_w / self.zoom_factor
display_h = display_h / self.zoom_factor
# Step 2: Reverse rotation
if self.rotation_angle != 0:
# Get the dimensions of the frame after crop but before rotation
if self.crop_rect:
crop_w, crop_h = int(self.crop_rect[2]), int(self.crop_rect[3])
else:
crop_w, crop_h = original_width, original_height
# Apply inverse rotation to coordinates
# The key insight: we need to use the dimensions of the ROTATED frame for the coordinate transformation
# because the coordinates we have are in the rotated coordinate system
if self.rotation_angle == 90:
# 90° clockwise rotation: (x,y) -> (y, rotated_width-x-w)
# The rotated frame has dimensions: height x width (swapped)
rotated_w, rotated_h = crop_h, crop_w
new_x = display_y
new_y = rotated_w - display_x - display_w
new_w = display_h
new_h = display_w
elif self.rotation_angle == 180:
# 180° rotation: (x,y) -> (width-x-w, height-y-h)
new_x = crop_w - display_x - display_w
new_y = crop_h - display_y - display_h
new_w = display_w
new_h = display_h
elif self.rotation_angle == 270:
# 270° clockwise rotation: (x,y) -> (rotated_height-y-h, x)
# The rotated frame has dimensions: height x width (swapped)
rotated_w, rotated_h = crop_h, crop_w
new_x = rotated_h - display_y - display_h
new_y = display_x
new_w = display_h
new_h = display_w
else:
new_x, new_y, new_w, new_h = display_x, display_y, display_w, display_h
display_x, display_y, display_w, display_h = new_x, new_y, new_w, new_h
# Step 3: Convert from cropped frame coordinates to original frame coordinates
original_x = display_x
original_y = display_y
original_w = display_w
original_h = display_h
# Add the crop offset to get back to original frame coordinates
if self.crop_rect:
crop_x, crop_y, crop_w, crop_h = self.crop_rect
original_x += crop_x
original_y += crop_y
# Clamp to original frame bounds
original_x = max(0, min(original_x, original_width))
original_y = max(0, min(original_y, original_height))
original_w = min(original_w, original_width - original_x)
original_h = min(original_h, original_height - original_y)
if original_w > 10 and original_h > 10: # Minimum size check
# Save current crop for undo
if self.crop_rect:
self.crop_history.append(self.crop_rect)
self.crop_rect = (original_x, original_y, original_w, original_h)
self.clear_transformation_cache()
self.save_state() # Save state when crop is set
def seek_to_timeline_position(self, mouse_x, bar_x_start, bar_width):
"""Seek to position based on mouse click on timeline"""
relative_x = mouse_x - bar_x_start
position_ratio = max(0, min(1, relative_x / bar_width))
target_frame = int(position_ratio * (self.total_frames - 1))
self.seek_to_frame(target_frame)
def undo_crop(self):
"""Undo the last crop operation"""
if self.crop_history:
self.crop_rect = self.crop_history.pop()
else:
self.crop_rect = None
self.clear_transformation_cache()
self.save_state() # Save state when crop is undone
def toggle_marker_looping(self):
"""Toggle looping between cut markers"""
# Check if both markers are set
if self.cut_start_frame is None or self.cut_end_frame is None:
print("Both markers must be set to enable looping. Use '1' and '2' to set markers.")
return False
if self.cut_start_frame >= self.cut_end_frame:
print("Invalid marker range - start frame must be before end frame")
return False
self.looping_between_markers = not self.looping_between_markers
if self.looping_between_markers:
print(f"Marker looping ENABLED: frames {self.cut_start_frame} - {self.cut_end_frame}")
# Jump to start marker when enabling
self.seek_to_frame(self.cut_start_frame)
else:
print("Marker looping DISABLED")
self.save_state() # Save state when looping is toggled
return True
def adjust_crop_size(self, direction: str, expand: bool, amount: int = None):
"""
Adjust crop size in given direction
direction: 'up', 'down', 'left', 'right'
expand: True to expand, False to contract
amount: pixels to adjust by (uses self.crop_size_step if None)
"""
if amount is None:
amount = self.crop_size_step
if not self.crop_rect:
# If no crop exists, create a default one in the center
center_x = self.frame_width // 2
center_y = self.frame_height // 2
default_size = min(self.frame_width, self.frame_height) // 4
self.crop_rect = (
center_x - default_size // 2,
center_y - default_size // 2,
default_size,
default_size
)
return
x, y, w, h = self.crop_rect
if direction == 'up':
if expand:
# Expand upward - decrease y, increase height
new_y = max(0, y - amount)
new_h = h + (y - new_y)
self.crop_rect = (x, new_y, w, new_h)
else:
# Contract from bottom - decrease height
new_h = max(10, h - amount) # Minimum size of 10 pixels
self.crop_rect = (x, y, w, new_h)
elif direction == 'down':
if expand:
# Expand downward - increase height
new_h = min(self.frame_height - y, h + amount)
self.crop_rect = (x, y, w, new_h)
else:
# Contract from top - increase y, decrease height
amount = min(amount, h - 10) # Don't make it smaller than 10 pixels
new_y = y + amount
new_h = h - amount
self.crop_rect = (x, new_y, w, new_h)
elif direction == 'left':
if expand:
# Expand leftward - decrease x, increase width
new_x = max(0, x - amount)
new_w = w + (x - new_x)
self.crop_rect = (new_x, y, new_w, h)
else:
# Contract from right - decrease width
new_w = max(10, w - amount) # Minimum size of 10 pixels
self.crop_rect = (x, y, new_w, h)
elif direction == 'right':
if expand:
# Expand rightward - increase width
new_w = min(self.frame_width - x, w + amount)
self.crop_rect = (x, y, new_w, h)
else:
# Contract from left - increase x, decrease width
amount = min(amount, w - 10) # Don't make it smaller than 10 pixels
new_x = x + amount
new_w = w - amount
self.crop_rect = (new_x, y, new_w, h)
self.clear_transformation_cache()
self.save_state() # Save state when crop is adjusted
def render_video(self, output_path: str):
"""Render video or save image with current edits applied"""
if self.is_image_mode:
return self._render_image(output_path)
else:
return self._render_video_threaded(output_path)
def _render_video_threaded(self, output_path: str):
"""Start video rendering in a separate thread"""
# Check if already rendering
if self.render_thread and self.render_thread.is_alive():
print("Render already in progress! Use 'X' to cancel first.")
return False
# Reset render state
self.render_cancelled = False
# Start render thread
self.render_thread = threading.Thread(
target=self._render_video_worker,
args=(output_path,),
daemon=True
)
self.render_thread.start()
print(f"Started rendering to {output_path} in background thread...")
print("You can continue editing while rendering. Press 'X' to cancel.")
return True
def _render_video_worker(self, output_path: str):
"""Worker method that runs in the render thread"""
render_cap = None
try:
if not output_path.endswith(".mp4"):
output_path += ".mp4"
start_time = time.time()
# Send progress update to main thread
self.render_progress_queue.put(("init", "Initializing render...", 0.0, 0.0))
# No need to create VideoCapture since we use FFmpeg directly
# Determine frame range
start_frame = self.cut_start_frame if self.cut_start_frame is not None else 0
end_frame = (
self.cut_end_frame
if self.cut_end_frame is not None
else self.total_frames - 1
)
if start_frame >= end_frame:
self.render_progress_queue.put(("error", "Invalid cut range!", 1.0, 0.0))
return False
# Send progress update
self.render_progress_queue.put(("progress", "Calculating output dimensions...", 0.05, 0.0))
# Calculate output dimensions (accounting for rotation)
if self.crop_rect:
crop_width = int(self.crop_rect[2])
crop_height = int(self.crop_rect[3])
else:
crop_width = self.frame_width
crop_height = self.frame_height
# Swap dimensions if rotation is 90 or 270 degrees
if self.rotation_angle == 90 or self.rotation_angle == 270:
output_width = int(crop_height * self.zoom_factor)
output_height = int(crop_width * self.zoom_factor)
else:
output_width = int(crop_width * self.zoom_factor)
output_height = int(crop_height * self.zoom_factor)
# Ensure dimensions are divisible by 2 for H.264 encoding
output_width = output_width - (output_width % 2)
output_height = output_height - (output_height % 2)
# Send progress update
self.render_progress_queue.put(("progress", "Setting up FFmpeg encoder...", 0.1, 0.0))
# Debug output dimensions
print(f"Output dimensions: {output_width}x{output_height}")
print(f"Zoom factor: {self.zoom_factor}")
print(f"Crop dimensions: {crop_width}x{crop_height}")
# Skip all the OpenCV codec bullshit and go straight to FFmpeg
print("Using FFmpeg for encoding with OpenCV transformations...")
return self._render_with_ffmpeg_pipe(output_path, start_frame, end_frame, output_width, output_height)
except Exception as e:
error_msg = str(e)
# Handle specific FFmpeg threading errors
if "async_lock" in error_msg or "pthread_frame" in error_msg:
error_msg = "FFmpeg threading error - try restarting the application"
elif "Assertion" in error_msg:
error_msg = "Video codec error - the video file may be corrupted or incompatible"
self.render_progress_queue.put(("error", f"Render error: {error_msg}", 1.0, 0.0))
print(f"Render error: {error_msg}")
return False
finally:
# No cleanup needed since we don't create VideoCapture
pass
def update_render_progress(self):
"""Process progress updates from the render thread"""
try:
while True:
# Non-blocking get from queue
update_type, text, progress, fps = self.render_progress_queue.get_nowait()
if update_type == "init":
self.show_progress_bar(text)
elif update_type == "progress":
self.update_progress_bar(progress, text, fps)
elif update_type == "complete":
self.update_progress_bar(progress, text, fps)
# Handle file overwrite if this was an overwrite operation
if hasattr(self, 'overwrite_temp_path') and self.overwrite_temp_path:
self._handle_overwrite_completion()
elif update_type == "error":
self.update_progress_bar(progress, text, fps)
# Also show error as feedback message for better visibility
self.show_feedback_message(f"ERROR: {text}")
elif update_type == "cancelled":
self.hide_progress_bar()
self.show_feedback_message("Render cancelled")
except queue.Empty:
# No more updates in queue
pass
def _handle_overwrite_completion(self):
"""Handle file replacement after successful render"""
try:
print("Replacing original file...")
# Release current video capture before replacing the file
if hasattr(self, 'cap') and self.cap:
self.cap.release()
# Replace the original file with the temporary file
import shutil
print(f"DEBUG: Moving {self.overwrite_temp_path} to {self.overwrite_target_path}")
try:
shutil.move(self.overwrite_temp_path, self.overwrite_target_path)
print("DEBUG: File move successful")
except Exception as e:
print(f"DEBUG: File move failed: {e}")
# Try to clean up temp file
if os.path.exists(self.overwrite_temp_path):
os.remove(self.overwrite_temp_path)
raise
# Small delay to ensure file system operations are complete
time.sleep(0.1)
try:
self._load_video(self.video_path)
self.load_current_frame()
print("File reloaded successfully")
except Exception as e:
print(f"Warning: Could not reload file after overwrite: {e}")
print("The file was saved successfully, but you may need to restart the editor to continue editing it.")
except Exception as e:
print(f"Error during file overwrite: {e}")
finally:
# Clean up overwrite state
self.overwrite_temp_path = None
self.overwrite_target_path = None
def cancel_render(self):
"""Cancel the current render operation"""
if self.render_thread and self.render_thread.is_alive():
self.render_cancelled = True
print("Render cancellation requested...")
return True
return False
def is_rendering(self):
"""Check if a render operation is currently active"""
return self.render_thread and self.render_thread.is_alive()
def cleanup_render_thread(self):
"""Clean up render thread resources"""
if self.render_thread and self.render_thread.is_alive():
self.render_cancelled = True
# Terminate FFmpeg process if running
if self.ffmpeg_process:
try:
self.ffmpeg_process.terminate()
self.ffmpeg_process.wait(timeout=1.0)
except:
try:
self.ffmpeg_process.kill()
except:
pass
self.ffmpeg_process = None
# Wait a bit for the thread to finish gracefully
self.render_thread.join(timeout=2.0)
if self.render_thread.is_alive():
print("Warning: Render thread did not finish gracefully")
self.render_thread = None
self.render_cancelled = False
def _render_image(self, output_path: str):
"""Save image with current edits applied"""
# Get the appropriate file extension
original_ext = self.video_path.suffix.lower()
if not output_path.endswith(original_ext):
output_path += original_ext
print(f"Saving image to {output_path}...")
# Apply all transformations to the image
processed_image = self.apply_crop_zoom_and_rotation(self.static_image.copy())
if processed_image is not None:
# Save the image
success = cv2.imwrite(output_path, processed_image)
if success:
print(f"Image saved successfully to {output_path}")
return True
else:
print(f"Error: Could not save image to {output_path}")
return False
else:
print("Error: Could not process image")
return False
def _process_frame_for_render(self, frame, output_width: int, output_height: int):
"""Process a single frame for rendering (optimized for speed)"""
try:
# Apply crop (vectorized operation)
if self.crop_rect:
x, y, w, h = map(int, self.crop_rect)
# Clamp coordinates to frame bounds
h_frame, w_frame = frame.shape[:2]
x = max(0, min(x, w_frame - 1))
y = max(0, min(y, h_frame - 1))
w = min(w, w_frame - x)
h = min(h, h_frame - y)
if w > 0 and h > 0:
frame = frame[y : y + h, x : x + w]
else:
return None
# Apply brightness and contrast
frame = self.apply_brightness_contrast(frame)
# Apply rotation
if self.rotation_angle != 0:
frame = self.apply_rotation(frame)
# Apply zoom and resize directly to final output dimensions
if self.zoom_factor != 1.0:
height, width = frame.shape[:2]
# Calculate what the zoomed dimensions would be
zoomed_width = int(width * self.zoom_factor)
zoomed_height = int(height * self.zoom_factor)
# If zoomed dimensions match output, use them; otherwise resize directly to output
if zoomed_width == output_width and zoomed_height == output_height:
frame = cv2.resize(
frame, (zoomed_width, zoomed_height), interpolation=cv2.INTER_LINEAR
)
else:
# Resize directly to final output dimensions
frame = cv2.resize(
frame, (output_width, output_height), interpolation=cv2.INTER_LINEAR
)
else:
# No zoom, just resize to output dimensions if needed
if frame.shape[1] != output_width or frame.shape[0] != output_height:
frame = cv2.resize(
frame, (output_width, output_height), interpolation=cv2.INTER_LINEAR
)
return frame
except Exception as e:
print(f"Error processing frame: {e}")
return None
def _render_with_ffmpeg_pipe(self, output_path: str, start_frame: int, end_frame: int, output_width: int, output_height: int):
"""Render video with transformations"""
try:
# Test FFmpeg with a simple command first
try:
test_result = subprocess.run(['ffmpeg', '-version'], capture_output=True, text=True, timeout=10)
if test_result.returncode != 0:
print(f"FFmpeg test failed with return code {test_result.returncode}")
print(f"FFmpeg stderr: {test_result.stderr}")
error_msg = "FFmpeg is not working properly"
self.render_progress_queue.put(("error", error_msg, 1.0, 0.0))
return False
except (subprocess.CalledProcessError, FileNotFoundError, subprocess.TimeoutExpired) as e:
error_msg = f"FFmpeg not found or not working: {e}"
print(error_msg)
self.render_progress_queue.put(("error", error_msg, 1.0, 0.0))
return False
self.render_progress_queue.put(("progress", "Starting encoder...", 0.0, 0.0))
import tempfile
import os
temp_file = tempfile.NamedTemporaryFile(delete=False, suffix='.raw')
temp_file.close()
# Use a simpler, more Windows-compatible FFmpeg command
ffmpeg_cmd = [
'ffmpeg', '-y',
'-f', 'rawvideo',
'-s', f'{output_width}x{output_height}',
'-pix_fmt', 'bgr24',
'-r', str(self.fps),
'-i', temp_file.name,
'-c:v', 'libx264',
'-preset', 'fast',
'-crf', '18',
'-pix_fmt', 'yuv420p',
output_path
]
self.temp_file_name = temp_file.name
render_cap = cv2.VideoCapture(str(self.video_path))
render_cap.set(cv2.CAP_PROP_POS_FRAMES, start_frame)
total_frames = end_frame - start_frame + 1
frames_written = 0
start_time = time.time()
last_progress_update = 0
self.render_progress_queue.put(("progress", f"Processing {total_frames} frames...", 0.1, 0.0))
with open(self.temp_file_name, 'wb') as temp_file:
for i in range(total_frames):
if self.render_cancelled:
render_cap.release()
self.render_progress_queue.put(("cancelled", "Render cancelled", 0.0, 0.0))
return False
ret, frame = render_cap.read()
if not ret:
break
processed_frame = self._process_frame_for_render(frame, output_width, output_height)
if processed_frame is not None:
if i == 0:
print(f"Processed frame dimensions: {processed_frame.shape[1]}x{processed_frame.shape[0]}")
print(f"Expected dimensions: {output_width}x{output_height}")
temp_file.write(processed_frame.tobytes())
frames_written += 1
current_time = time.time()
progress = 0.1 + (0.8 * (i + 1) / total_frames)
if current_time - last_progress_update > 0.5:
elapsed = current_time - start_time
fps_rate = frames_written / elapsed if elapsed > 0 else 0
self.render_progress_queue.put(("progress", f"Processed {i+1}/{total_frames} frames", progress, fps_rate))
last_progress_update = current_time
render_cap.release()
self.render_progress_queue.put(("progress", "Encoding...", 0.9, 0.0))
# Use subprocess.run() with timeout for better Windows reliability
result = subprocess.run(
ffmpeg_cmd,
capture_output=True,
text=True,
timeout=300, # 5 minute timeout
creationflags=subprocess.CREATE_NO_WINDOW if hasattr(subprocess, 'CREATE_NO_WINDOW') else 0
)
return_code = result.returncode
stdout = result.stdout
stderr = result.stderr
# Debug output
print(f"FFmpeg return code: {return_code}")
if stdout:
print(f"FFmpeg stdout: {stdout}")
if stderr:
print(f"FFmpeg stderr: {stderr}")
if os.path.exists(self.temp_file_name):
try:
os.unlink(self.temp_file_name)
except OSError:
pass
if return_code == 0:
total_time = time.time() - start_time
avg_fps = frames_written / total_time if total_time > 0 else 0
self.render_progress_queue.put(("complete", f"Rendered {frames_written} frames", 1.0, avg_fps))
print(f"Successfully rendered {frames_written} frames (avg {avg_fps:.1f} FPS)")
return True
else:
error_details = stderr if stderr else "No error details available"
print(f"Encoding failed with return code {return_code}")
print(f"Error: {error_details}")
self.render_progress_queue.put(("error", f"Encoding failed: {error_details}", 1.0, 0.0))
return False
except Exception as e:
error_msg = str(e)
print(f"Rendering exception: {error_msg}")
print(f"Exception type: {type(e).__name__}")
if "Errno 22" in error_msg or "invalid argument" in error_msg.lower():
error_msg = "File system error - try using a different output path"
elif "BrokenPipeError" in error_msg:
error_msg = "Process terminated unexpectedly"
elif "FileNotFoundError" in error_msg or "ffmpeg" in error_msg.lower():
error_msg = "FFmpeg not found - please install FFmpeg and ensure it's in your PATH"
self.render_progress_queue.put(("error", f"Rendering failed: {error_msg}", 1.0, 0.0))
return False
def run(self):
"""Main editor loop"""
if self.is_image_mode:
print("Image Editor Controls:")
print(" E/Shift+E: Increase/Decrease brightness")
print(" R/Shift+R: Increase/Decrease contrast")
print(" -: Rotate clockwise 90°")
print()
print("Crop Controls:")
print(" Shift+Click+Drag: Select crop area")
print(" h/j/k/l: Contract crop (left/down/up/right)")
print(" H/J/K/L: Expand crop (left/down/up/right)")
print(" U: Undo crop")
print(" C: Clear crop")
print()
print("Other Controls:")
print(" Ctrl+Scroll: Zoom in/out")
print(" Shift+S: Save screenshot")
print(" f: Toggle fullscreen")
print(" p: Toggle project view")
if len(self.video_files) > 1:
print(" N: Next file")
print(" n: Previous file")
print(" Enter: Save image (overwrites if '_edited_' in name)")
print(" b: Save image as _edited_edited")
print(" Q/ESC: Quit")
print()
else:
print("Video Editor Controls:")
print(" Space: Play/Pause")
print(" A/D: Seek backward/forward (1 frame)")
print(" Shift+A/D: Seek backward/forward (10 frames)")
print(" Ctrl+A/D: Seek backward/forward (60 frames)")
print(" W/S: Increase/Decrease speed")
print(" Q/Y: Increase/Decrease seek multiplier")
print(" E/Shift+E: Increase/Decrease brightness")
print(" R/Shift+R: Increase/Decrease contrast")
print(" -: Rotate clockwise 90°")
print()
print("Crop Controls:")
print(" Shift+Click+Drag: Select crop area")
print(" h/j/k/l: Contract crop (left/down/up/right)")
print(" H/J/K/L: Expand crop (left/down/up/right)")
print(" U: Undo crop")
print(" C: Clear crop")
print()
print("Other Controls:")
print(" Ctrl+Scroll: Zoom in/out")
print(" Shift+S: Save screenshot")
print(" f: Toggle fullscreen")
print(" p: Toggle project view")
print(" 1: Set cut start point")
print(" 2: Set cut end point")
print(" T: Toggle loop between markers")
if len(self.video_files) > 1:
print(" N: Next video")
print(" n: Previous video")
print(" Enter: Render video (overwrites if '_edited_' in name)")
print(" b: Render video")
print(" x: Cancel render")
print(" Q/ESC: Quit")
print()
window_title = "Image Editor" if self.is_image_mode else "Video Editor"
cv2.namedWindow(window_title, cv2.WINDOW_NORMAL)
cv2.resizeWindow(window_title, self.window_width, self.window_height)
cv2.setMouseCallback(window_title, self.mouse_callback)
self.load_current_frame()
while True:
# Update auto-repeat seeking if active
self.update_auto_repeat_seek()
# Update render progress from background thread
self.update_render_progress()
# Update display
self.display_current_frame()
# Handle project view window if it exists
if self.project_view_mode and self.project_view:
# Draw project view in its own window
project_canvas = self.project_view.draw()
cv2.imshow("Project View", project_canvas)
# Calculate appropriate delay based on playback state
if self.is_playing and not self.is_image_mode:
# Use calculated frame delay for proper playback speed
delay_ms = self.calculate_frame_delay()
else:
# Use minimal delay when not playing for responsive UI
delay_ms = 1
# Key capture with appropriate delay
key = cv2.waitKey(delay_ms) & 0xFF
# Route keys based on window focus
if key != 255: # Key was pressed
active_window = get_active_window_title()
if "Project View" in active_window:
# Project view window has focus - handle project view keys
if self.project_view_mode and self.project_view:
action = self.project_view.handle_key(key)
if action == "back_to_editor":
self.toggle_project_view()
elif action == "quit":
return # Exit the main loop
elif action.startswith("open_video:"):
video_path_str = action.split(":", 1)[1]
video_path = Path(video_path_str)
self.open_video_from_project_view(video_path)
continue # Skip main window key handling
elif "Video Editor" in active_window or "Image Editor" in active_window:
# Main window has focus - handle editor keys
pass # Continue to main window key handling below
else:
# Neither window has focus, ignore key
continue
# Handle auto-repeat - stop if no key is pressed
if key == 255 and self.auto_repeat_active: # 255 means no key pressed
self.stop_auto_repeat_seek()
if key == ord("q") or key == 27: # ESC
self.stop_auto_repeat_seek()
self.save_state()
break
elif key == ord("p"): # P - Toggle project view
self.toggle_project_view()
elif key == ord(" "):
# Don't allow play/pause for images
if not self.is_image_mode:
self.stop_auto_repeat_seek() # Stop seeking when toggling play/pause
self.is_playing = not self.is_playing
elif key == ord("a") or key == ord("A"):
# Seeking only for videos
if not self.is_image_mode:
# Check if it's uppercase A (Shift+A)
if key == ord("A"):
if not self.auto_repeat_active:
self.start_auto_repeat_seek(-1, True, False) # Shift+A: -10 frames
else:
if not self.auto_repeat_active:
self.start_auto_repeat_seek(-1, False, False) # A: -1 frame
elif key == ord("d") or key == ord("D"):
# Seeking only for videos
if not self.is_image_mode:
# Check if it's uppercase D (Shift+D)
if key == ord("D"):
if not self.auto_repeat_active:
self.start_auto_repeat_seek(1, True, False) # Shift+D: +10 frames
else:
if not self.auto_repeat_active:
self.start_auto_repeat_seek(1, False, False) # D: +1 frame
elif key == 1: # Ctrl+A
# Seeking only for videos
if not self.is_image_mode:
if not self.auto_repeat_active:
self.start_auto_repeat_seek(-1, False, True) # Ctrl+A: -60 frames
elif key == 4: # Ctrl+D
# Seeking only for videos
if not self.is_image_mode:
if not self.auto_repeat_active:
self.start_auto_repeat_seek(1, False, True) # Ctrl+D: +60 frames
elif key == ord("-") or key == ord("_"):
self.rotate_clockwise()
print(f"Rotated to {self.rotation_angle}°")
elif key == ord("f"):
self.toggle_fullscreen()
elif key == ord("s"): # Shift+S - Save screenshot
self.save_current_frame()
elif key == ord("W"):
# Speed control only for videos
if not self.is_image_mode:
self.playback_speed = min(
self.MAX_PLAYBACK_SPEED, self.playback_speed + self.SPEED_INCREMENT
)
elif key == ord("S"):
# Speed control only for videos
if not self.is_image_mode:
self.playback_speed = max(
self.MIN_PLAYBACK_SPEED, self.playback_speed - self.SPEED_INCREMENT
)
elif key == ord("Q"):
# Seek multiplier control only for videos
if not self.is_image_mode:
self.seek_multiplier = min(
self.MAX_SEEK_MULTIPLIER, self.seek_multiplier + self.SEEK_MULTIPLIER_INCREMENT
)
print(f"Seek multiplier: {self.seek_multiplier:.1f}x")
elif key == ord("Y"):
# Seek multiplier control only for videos
if not self.is_image_mode:
self.seek_multiplier = max(
self.MIN_SEEK_MULTIPLIER, self.seek_multiplier - self.SEEK_MULTIPLIER_INCREMENT
)
print(f"Seek multiplier: {self.seek_multiplier:.1f}x")
elif key == ord("e") or key == ord("E"):
# Brightness adjustment: E (increase), Shift+E (decrease)
if key == ord("E"):
self.adjust_brightness(-5)
print(f"Brightness: {self.brightness}")
else:
self.adjust_brightness(5)
print(f"Brightness: {self.brightness}")
elif key == ord("r") or key == ord("R"):
# Contrast adjustment: R (increase), Shift+R (decrease)
if key == ord("R"):
self.adjust_contrast(-0.1)
print(f"Contrast: {self.contrast:.1f}")
else:
self.adjust_contrast(0.1)
print(f"Contrast: {self.contrast:.1f}")
elif key == ord("u"):
self.undo_crop()
elif key == ord("c"):
if self.crop_rect:
self.crop_history.append(self.crop_rect)
self.crop_rect = None
self.zoom_factor = 1.0
self.clear_transformation_cache()
self.save_state() # Save state when crop is cleared
elif key == ord("1"):
# Cut markers only for videos
if not self.is_image_mode:
self.cut_start_frame = self.current_frame
print(f"Set cut start at frame {self.current_frame}")
self.save_state() # Save state when cut start is set
elif key == ord("2"):
# Cut markers only for videos
if not self.is_image_mode:
self.cut_end_frame = self.current_frame
print(f"Set cut end at frame {self.current_frame}")
self.save_state() # Save state when cut end is set
elif key == ord("N"):
if len(self.video_files) > 1:
self.previous_video()
elif key == ord("n"):
if len(self.video_files) > 1:
self.next_video()
elif key == ord("b"):
directory = self.video_path.parent
base_name = self.video_path.stem
extension = self.video_path.suffix
# Remove any existing _edited_ suffix to get clean base name
clean_base = base_name.replace("_edited", "")
# Find next available number
counter = 1
while True:
new_name = f"{clean_base}_edited_{counter:05d}{extension}"
output_path = directory / new_name
if not output_path.exists():
break
counter += 1
success = self.render_video(str(output_path))
elif key == 13: # Enter
# Only overwrite if file already contains "_edited_" in name
print(f"DEBUG: Checking if '{self.video_path.stem}' contains '_edited_'")
if "_edited_" in self.video_path.stem:
print("DEBUG: File contains '_edited_', proceeding with overwrite")
print(f"DEBUG: Original file path: {self.video_path}")
print(f"DEBUG: Original file exists: {self.video_path.exists()}")
output_path = str(self.video_path)
# If we're overwriting the same file, use a temporary file first
import tempfile
temp_dir = self.video_path.parent
temp_fd, temp_path = tempfile.mkstemp(suffix=self.video_path.suffix, dir=temp_dir)
os.close(temp_fd) # Close the file descriptor, we just need the path
print(f"DEBUG: Created temp file: {temp_path}")
print("Rendering to temporary file first...")
success = self.render_video(temp_path)
# Store the temp path so we can replace the file when render completes
self.overwrite_temp_path = temp_path
self.overwrite_target_path = str(self.video_path)
else:
print(f"DEBUG: File '{self.video_path.stem}' does not contain '_edited_'")
print("Enter key only overwrites files with '_edited_' in the name. Use 'n' to create new files.")
elif key == ord("t"):
# Marker looping only for videos
if not self.is_image_mode:
self.toggle_marker_looping()
elif key == ord("x"):
# Cancel render if active
if self.is_rendering():
self.cancel_render()
print("Render cancellation requested")
else:
print("No render operation to cancel")
# Individual direction controls using shift combinations we can detect
elif key == ord("J"): # Shift+i - expand up
self.adjust_crop_size('up', False)
print(f"Expanded crop upward by {self.crop_size_step}px")
elif key == ord("K"): # Shift+k - expand down
self.adjust_crop_size('down', False)
print(f"Expanded crop downward by {self.crop_size_step}px")
elif key == ord("L"): # Shift+j - expand left
self.adjust_crop_size('left', False)
print(f"Expanded crop leftward by {self.crop_size_step}px")
elif key == ord("H"): # Shift+l - expand right
self.adjust_crop_size('right', False)
print(f"Expanded crop rightward by {self.crop_size_step}px")
# Contract in specific directions
elif key == ord("k"): # i - contract from bottom (reduce height from bottom)
self.adjust_crop_size('up', True)
print(f"Contracted crop from bottom by {self.crop_size_step}px")
elif key == ord("j"): # k - contract from top (reduce height from top)
self.adjust_crop_size('down', True)
print(f"Contracted crop from top by {self.crop_size_step}px")
elif key == ord("h"): # j - contract from right (reduce width from right)
self.adjust_crop_size('left', True)
print(f"Contracted crop from right by {self.crop_size_step}px")
elif key == ord("l"): # l - contract from left (reduce width from left)
self.adjust_crop_size('right', True)
print(f"Contracted crop from left by {self.crop_size_step}px")
# Auto advance frame when playing (videos only)
if self.is_playing and not self.is_image_mode:
self.advance_frame()
self.save_state()
self.cleanup_render_thread()
if hasattr(self, 'cap') and self.cap:
self.cap.release()
cv2.destroyAllWindows()
def main():
parser = argparse.ArgumentParser(
description="Fast Media Editor - Crop, Zoom, and Edit videos and images"
)
parser.add_argument(
"media", help="Path to media file or directory containing videos/images"
)
try:
args = parser.parse_args()
except SystemExit:
# If launched from context menu without arguments, this might fail
input("Argument parsing failed. Press Enter to exit...")
return
if not os.path.exists(args.media):
error_msg = f"Error: {args.media} does not exist"
print(error_msg)
input("Press Enter to exit...") # Keep window open in context menu
sys.exit(1)
try:
editor = VideoEditor(args.media)
editor.run()
except Exception as e:
error_msg = f"Error initializing media editor: {e}"
print(error_msg)
import traceback
traceback.print_exc() # Full error trace for debugging
input("Press Enter to exit...") # Keep window open in context menu
sys.exit(1)
if __name__ == "__main__":
main()