APNG Disassembler: Extract Frames and Metadata from Animated PNGs

APNG Disassembler: Extract Frames and Metadata from Animated PNGs

What it is

• A tool (script or application) that reads APNG files and splits them into individual PNG frames while extracting timing, frame-control flags, and other metadata.

Key features

• Frame extraction: exports each frame as a standalone PNG (preserving alpha).
• Metadata extraction: reads acTL, fcTL, fdAT chunks and outputs frame durations, sequence numbers, disposal/blend operations.
• Timing accuracy: converts frame delay units to milliseconds and reports total animation duration and per-frame delays.
• Lossless handling: preserves original compression and color/alpha data when possible.
• Reassembly support: optional output of a JSON/INI manifest to allow rebuilding the APNG in the same order with original metadata.
• Batch processing: handle multiple APNGs in a folder and generate uniform output structure.
• Validation: detects malformed APNG chunks, missing sequence numbers, or nonstandard chunk ordering.

Common uses

• Animation debugging and QA.
• Converting APNG to sprite sheets or video.
• Extracting frames for editing or analysis.
• Reverse-engineering frame timing and disposal/blend behavior.
• Creating fallback static PNGs or GIF conversions.

Inputs and outputs

• Input: .png file containing APNG chunks.
• Outputs: frame_N.png files, a manifest (JSON/CSV) with per-frame metadata (delay_ms, sequence, width/height, blend_op, dispose_op), and an optional rebuilt .apng.

Implementation notes (practical)

• Parse PNG chunks sequentially; identify acTL (animation control), then fcTL/fdAT for frames.
• fcTL contains width/height/x_offset/y_offset, delay_num/delay_den (compute ms = 1000delay_num / delay_den), dispose_op, blend_op.
• fdAT chunks carry frame image data with a sequence number; combine fdAT/IDAT data appropriately.
• Preserve IHDR/color/chunk ordering for standalone frames; reconstruct valid PNGs by inserting fcTL-derived dimensions into IHDR if different.
• Handle missing delay_den (treat as 100) and very small delays (floor to 1–20 ms depending on target).
• Consider using existing libraries (libpng, pngcrush, apngdis, apngasm) where available to avoid reimplementing low-level PNG parsing.

Output manifest example (fields)

• frame_index, filename, delay_ms, seq_num, width, height, x_offset, y_offset, blend_op, dispose_op

Limitations and gotchas

• Some APNGs use nonstandard or corrupted chunks—tool should warn and skip or attempt repair.
• Accurate timing depends on delay_den; some creators omit it or use unconventional values.
• Reconstructed frames may differ visually if ancillary chunks (sRGB, gamma) are dropped—preserve relevant chunks when exporting.

If you want, I can:

• Provide a short Python script that extracts frames and writes a JSON manifest.
• Recommend existing open-source APNG disassembler tools.