AB-AV1 GUI v2.0.0 Release Notes

Breaking Changes

Change	Impact
Config file renamed	`av1_converter_config.json` → `ab_av1_gui_config.json` — existing settings will reset
History format changed	`conversion_history.json` uses a new schema with additional fields — existing data will reset
Python 3.11+ required	Was 3.8+; needed for `tomllib` (stdlib TOML parser) to read version from `pyproject.toml`
License changed	MIT → GPL-3.0

Analysis Tab — Browse folders and preview conversion estimates before committing. Four-level workflow: Discover (instant tree) → Basic Scan (ffprobe metadata) → Analyze (CRF search) → Convert (full encoding)
Convert Tab (Queue-based) — Replaces the single-folder model with a queue supporting individual files and folders, drag-to-reorder, and per-item operation selection
Statistics Tab — Canvas-based bar, pie, and line charts visualizing conversion history (space saved, time spent, file counts)
History Tab — Browse all processed files (converted, analyzed, skipped) with filtering and date sorting

Vendor Manager — Bundled ab-av1.exe removed; both ab-av1 and FFmpeg can now be downloaded directly from the Settings tab (FFmpeg previously required manual PATH installation). Displays installed versions and checks GitHub for updates.
Hardware-accelerated decoding — Optional NVIDIA CUVID and Intel QSV decoder support (toggle in Settings)
Duplicate file detection — Metadata-based detection (filename, size, duration) identifies identical files accessed via different paths (mapped drives, UNC paths) and shares history data across them
Queue operations — Choose between "Analyze" (CRF search only, no output file) and "Convert" (full encoding) per item
Queue preview dialog — Shows estimated time, size savings, and lists files that will be skipped (with reasons) before adding to queue
Version checking — App version displayed in Settings with GitHub update notifications
Cross-platform launcher — Added convert.sh for Linux/macOS alongside existing convert.bat

Time estimation — Now uses resolution-bucket grouping (4K/1440p/1080p/720p/SD × codec) instead of file-size similarity matching, providing more accurate predictions
Size estimation — Accounts for audio streams (copied unchanged, not re-encoded), improving accuracy for files with significant audio content
Estimation caching — Session-level percentile caching eliminates redundant O(n log n) sorting during batch operations
Progress display — Simplified "File X/Y" format instead of nested item/file counters
Audio metadata — All audio streams now stored with full codec, bitrate, and channel information
Queue management — "Clear Completed" button; controls enable/disable based on queue state; prevents reordering during active conversion

Code reorganization — Major refactoring extracted ~2,000+ lines from monolithic files into focused modules: history_index.py, folder_analysis.py, vendor_manager.py, video_metadata.py, hardware_accel.py, logging_setup.py, platform_utils.py, constants.py, charts.py, and others
Centralized defaults — ConfigDict TypedDict with CONFIG_DEFAULTS replaces scattered .get() calls
New documentation — TIME_ESTIMATION.md, HISTORY_FORMAT.md, AB_AV1_PARSING.md, ADR-001 (duplicate detection design)

Application renamed from "AV1 Video Converter" to "AB-AV1 GUI"