mirror of
https://github.com/vondas-network/videobeaux.git
synced 2026-01-26 08:41:08 +01:00
58 lines
2.7 KiB
Markdown
58 lines
2.7 KiB
Markdown
# extract_frames
|
||
:contentReference[oaicite:1]{index=1}
|
||
|
||
## Description
|
||
Extracts individual frames from a video file and saves them as image files (typically PNG).
|
||
Ideal for analysis, archival, animation workflows, visual debugging, and creating frame-based artwork.
|
||
|
||
## Purpose
|
||
`extract_frames` gives creators a fast and predictable way to output every frame (or selected frames via global videobeaux options) as standalone image files.
|
||
This is useful for:
|
||
- animation and rotoscoping workflows,
|
||
- ML dataset preparation,
|
||
- glitch-art and frame-painting processes,
|
||
- shot-by-shot inspection or QC,
|
||
- archival still-frame extraction.
|
||
|
||
## How It Works
|
||
1. **Frame Decoding**
|
||
Video frames are decoded sequentially.
|
||
2. **Image Export**
|
||
Each frame is exported as its own image file using the globally configured pixel format and image output settings.
|
||
3. **Naming Convention**
|
||
Frames are typically numbered sequentially (e.g., `000001.png`, `000002.png`, etc.), depending on videobeaux output rules.
|
||
4. **Output Directory**
|
||
The destination directory is defined by global videobeaux settings (`--outfile`, mapping rules, etc.).
|
||
|
||
## Program Template
|
||
videobeaux -P extract_frames \
|
||
-i input.mp4 \
|
||
-o output.mp4
|
||
|
||
## Arguments
|
||
- *(No program-specific arguments — this tool relies entirely on global videobeaux settings such as image format, numbering, and frame selection.)*
|
||
|
||
## Real World Example
|
||
videobeaux -P extract_frames \
|
||
-i myvideo.mp4 \
|
||
-o extract_frames_styled.mp4
|
||
|
||
## Technical Notes
|
||
- PNG is the default format due to lossless quality, but other formats may be used depending on global configuration.
|
||
- Large videos may produce thousands of frames; ensure adequate disk space.
|
||
- Frame extraction is decode-limited — higher-resolution videos take longer per frame.
|
||
- If frames are dropped or duplicated in the source video stream, extraction will preserve exactly what the decoder receives.
|
||
- Good for use ahead of per-frame manipulation tools, compositing, or generative workflows.
|
||
|
||
## Recommended Usage
|
||
- Creating frame sequences for animation, compositing, or painting.
|
||
- Building datasets for computer vision or machine learning.
|
||
- Inspecting motion continuity, exposure behavior, or compression artifacts.
|
||
- Generating image sequences for later re-import into editing or FX tools.
|
||
|
||
## Quality Tips
|
||
- Use PNG for archival quality; JPG for lightweight previews.
|
||
- If color accuracy is critical, set a high-bit-depth pixel format globally (e.g., `rgb48le`).
|
||
- Use SSD storage for significantly faster write speeds on large sequences.
|
||
- When extracting for VFX, ensure your project frame rate matches the video’s native frame rate to avoid timing mismatch.
|