GifText

Add animated text to GIFs for meme creation. Track text labels on moving subjects with keyframe-based animation and smooth interpolation.

The only actively maintained desktop app for animated GIF text editing.

Features

• Video Import - Load short video clips (MP4, WebM, AVI, MOV) with FPS, trim, and resize controls
• Keyframe Animation - Position, size, opacity, rotation, and color all animatable per-frame with smooth ease-in-out interpolation
• OpenCV Motion Tracking - Generate forward keyframes from the selected layer position with CSRT/KCF when available and optical-flow fallback
• Bezier Path Animation - Draw a four-point path and generate editable position keyframes over a chosen frame span
• Motion Presets - Generate editable Bounce, Wiggle, and Shake emphasis keyframes
• Easing Curve Picker - Choose per-keyframe cubic Bezier timing curves for interpolation
• Staggered Text Reveal - Reveal lines, words, or letters over time for animated captions
• Subtitle Import - Import SRT/VTT captions as timed editable text layers
• Unicode Export Fallbacks - Non-Latin captions use broader Windows font fallbacks during export rendering
• Separate Stroke & Shadow Styling - Tune stroke opacity plus shadow color/opacity independently
• Responsive Long Jobs - GIF loading, tracking, and export run in cancellable background workers
• Structured Diagnostics - Load, project, recent-file, tracking, and export failures write timestamped logs and appear in the in-app Diagnostics panel
• On-Canvas Editing - Click to select text, drag to reposition, drag corner handle to resize
• Onion Skinning - Ghost previous frame to track moving subjects
• Multiple Text Layers - Color-coded with individual timing controls
• Layer Timeline - Visual bars with keyframe diamonds and playhead
• Range Tools - Apply, copy, paste, delete, and visibility-limit keyframes across selected frame ranges
• Fade In/Out - Per-layer entry and exit animations
• Custom Font Import - Load any TTF/OTF font from disk for both preview and export
• Meme Presets - Classic Meme, Modern Clean, Subtitle, Bold Impact, Neon
• Background Box - Semi-transparent subtitle-style background
• Undo/Redo - Full 50-level undo history (Ctrl+Z / Ctrl+Y)
• Project Save/Load - Resume work later with .giftext project files
• Template Library - Save and load multi-layer setups as reusable .giftemplate files
• Autosave & Recovery - Project state saved every 30 seconds with crash recovery prompt on restart
• Versioned Project Schema - Project loads validate schema, paths, frame ranges, colors, opacity, and keyframe fields before applying state
• Multi-Format Export - GIF, WebP (with alpha), APNG, PNG sequence, MP4, WebM
• Size-Target Export - Iteratively re-encode until output fits within a target file size (e.g., Discord 10 MB)
• Trim Frames - Cut start/end frames from the loaded source
• Resize Source - Downscale loaded frames for smaller output files
• Zoom & Pan - Ctrl+wheel to zoom, middle-click to pan
• Drag & Drop - Drop GIF files directly onto the canvas
• Playback Speed - 0.25x to 4x preview speed
• Recent Files - Quick access to previously opened GIFs
• Multi-Line Text - Full text area with line break support

Quick Start

python -m venv .venv
.venv\Scripts\python -m pip install -r requirements.txt
python GifText.py

Dependencies:

• Python 3.10+
• PyQt6
• Pillow
• OpenCV
• NumPy
• imageio + imageio-ffmpeg (for video import)

Workflow Example: Labeling People in a GIF

1. Load your GIF or video (or drag & drop it)
2. Add text layers - one per person/object
3. Type names in the text panel
4. Drag each name above its subject on frame 1
5. Mousewheel on canvas to step forward ~10 frames
6. Drag names to follow movement (keyframes auto-created)
7. Use Track Forward to generate motion keyframes when the subject is easy to follow
8. Use Draw Path when a label should follow a deliberate arc or swoop
9. Import .srt or .vtt captions when you want timed subtitle layers
10. Use Bounce, Wiggle, or Shake to add emphasis over the selected span
11. Repeat until the end - interpolation fills the gaps smoothly
12. Export as GIF, WebP, or PNG sequence

Controls

Action	Input
Step frames	Mousewheel on canvas
Zoom	Ctrl + Mousewheel
Pan	Middle-click drag
Select text	Click text on canvas
Move text	Drag text on canvas
Resize text	Drag bottom-right corner handle
Undo / Redo	Ctrl+Z / Ctrl+Y
Save project	Ctrl+S

Accessibility

All interactive controls expose accessible names and descriptions for screen readers. The canvas and timeline accept keyboard focus. High-contrast text on the dark theme passes readability checks.

Tech Stack

• Python / PyQt6 for the GUI
• Pillow for GIF I/O and export rendering
• OpenCV and NumPy for forward motion tracking
• Modular architecture: models, animation, rendering, project I/O, diagnostics, workers, and UI

Build Windows EXE

.venv\Scripts\python -m pip install -r requirements-build.txt
.venv\Scripts\pyinstaller --clean GifText.spec

The executable is written to dist\GifText.exe.

Release Build

The release script cleans build artifacts, runs PyInstaller, generates a SHA-256 checksum, and runs a smoke test:

python build_release.py

Options:

• --skip-smoke — skip the launch smoke test
• --sign CERT.pfx — code-sign the exe with a certificate

Outputs: dist/GifText.exe and dist/GifText.exe.sha256.

Troubleshooting

Failures are shown in the status bar, written to the right-side Diagnostics panel, and appended to timestamped logs under %USERPROFILE%\.giftext\logs\. Use the Export Diagnostics Bundle button to save a text file with app version, dependency versions, OS details, project state, and recent log entries for bug reports. Project files are validated before loading so invalid .giftext files do not replace the current document.

Command-Line Rendering

Render a .giftext project file without opening the GUI:

python GifText.py --render project.giftext -o output.gif
python GifText.py --render project.giftext -o video.mp4 -f mp4

Options:

• --output / -o — output file path (default: <project>_rendered.gif)
• --format / -f — output format: gif, webp, png, mp4, webm

License

MIT