StarTools Launch: Professional Video Enhancement and Image Processing Engine Now Available
VeyoLabs releases StarTools — the StarOxigen 1.0 Engine powered by the AI engine with 24 video workflows and image processing modes, bringing professional-grade video enhancement and image generation to creators at every tier.
StarTools Launch: The AI Engine for Professional Video and Image Creation
VeyoLabs is releasing StarTools — the creative engine powering the StarVision Tools suite. Built on the proprietary StarOxigen 1.0 Engine, StarTools is a professional-grade video and image processing platform that combines AI-powered generation with precision control, designed for filmmakers, content creators, and production studios.
StarTools is not a generic AI image generator. It is a production pipeline with 24 discrete video workflows, 4 image processing modes, and 3 quality tiers (Ultra, Pro, Express) that map directly to professional deliverables. The system is designed for creators who need reliable output, not experimental results.
Video Workflows: From Still to Motion
StarTools video workflows are built around a single principle: start with what you have, generate what you need. The engine accepts images, video, or text prompts as input and produces cinematic output across 24 distinct workflows.
Input-to-Video Pipeline
| Workflow | Input | Output | Use case |
|---|---|---|---|
| Image-to-Video | Single image | 5-second cinematic clip | Product shots, fashion, portraits |
| Merge Video | Two images | Blended transition video | Before/after, storytelling |
| Video-to-Video | Existing video | Stylized or enhanced video | Re-dubbing, style transfer, enhancement |
| Text-to-Video | Prompt only | Full video from description | Concept-to-clip generation |
Input Extension Pipeline
| Workflow | Input | Output | Use case |
|---|---|---|---|
| Video-Extend | 5-second clip | 10-second continuation | Extend scenes, add ending shots |
| Video-Blend | Two clips | Single continuous video | Combine takes, smooth transitions |
| Video-Frame | Video + frame | Inserted frame | Add missing frames, freeze moments |
| Video-Trim | Long video | Precise cut | Edit in time, remove dead space |
Motion Enhancement Pipeline
| Workflow | Input | Output | Use case |
|---|---|---|---|
| Video-Interpolate | Standard video | 2x frame rate | Smooth slow motion, fluid action |
| Video-Stabilize | Shaky footage | Stable footage | Handheld fix, drone correction |
| Video-Deblur | Motion blur | Sharp frames | Sports, action, fast movement |
| Video-Deband | Banded footage | Smooth gradients | Color grading, sky scenes |
Resolution and Quality Pipeline
| Workflow | Input | Output | Use case |
|---|---|---|---|
| Video-Upscale | 720p/1080p | 2K/4K | Archive restoration, re-release |
| Video-Denoise | Grainy footage | Clean video | Low light, high ISO, noise reduction |
| Video-Enhance | Flat footage | HDR-style video | Dynamic range expansion, punch |
| Video-Restoration | Damaged/old | Reconstructed video | Film preservation, vintage content |
Advanced Production Pipeline
| Workflow | Input | Output | Use case |
|---|---|---|---|
| Video-Inpaint | Video + mask | Object removed | Wire removal, set cleanup |
| Video-Outpaint | Video | Expanded frame | Aspect ratio change, wider shots |
| Video-Relight | Video | New lighting | Day-for-night, mood change |
| Video-Depth | Flat video | Depth-aware video | 3D conversion, parallax effects |
| Video-Segment | Video | Isolated subject | Background replacement, compositing |
| Video-Track | Video | Tracked object | Motion tracking, follow shots |
| Video-Animate | Static image | Animated scene | Parallax, environmental motion |
| Video-Composite | Multiple layers | Final composite | Green screen, VFX integration |
| Video-Color | Raw footage | Graded footage | LUT application, color science |
| Video-Titles | Video + text | Titled sequence | Opening credits, lower thirds |
Image Processing: Four Modes, Precision Control
StarTools image processing is built around four modes that cover the complete professional image pipeline, from upscaling to creative transformation.
BVX4 Image Up-Scaler
The Up-Scaler is a resolution engine designed for maximum fidelity. It uses 4-pass context analysis to reconstruct detail at 4x, 8x, and 16x scale, preserving edge sharpness and texture integrity.
Key specifications:
- Input formats: JPEG, PNG, WebP, HEIF, RAW (Canon, Sony, Nikon, Fuji)
- Output resolutions: 4K (3840×2160), 8K (7680×4320), 16K (15360×8640)
- Upscaling ratios: 2x, 4x, 8x, 16x (16K requires 16x from 1080p source)
- Processing: 4-pass context analysis, edge reconstruction, texture synthesis
- Quality grade: 100% fidelity at 4x, 95%+ at 8x, 90%+ at 16x
- Best for: Product photography, film stills, archival restoration, print preparation
The Up-Scaler is not a generative fill. It reconstructs what is there — it does not invent what is not. This is the critical distinction between professional upscaling and AI hallucination.
BVX8 Creative Mode
Creative Mode is the generative engine for image transformation. It takes an existing image and re-renders it with a new aesthetic, style, or environmental context while preserving the core subject structure.
Key specifications:
- Input: Any image (JPEG, PNG, WebP)
- Output: Transformed image at original or scaled resolution
- Style library: 80+ preset styles (Cinematic, Anime, Oil Painting, Watercolor, Cyberpunk, Film Noir, 80s Retro, Wes Anderson, Wong Kar-wai, Ghibli, Pixar, Claymation, and more)
- Custom style: Free-text prompt-based style control
- Structure preservation: 90%+ subject fidelity with style transfer
- Seed control: Lockable seeds for reproducible output
- Best for: Mood boards, style exploration, campaign concepts, portfolio variation
Creative Mode is not a filter. It is a full re-render with structural understanding. The subject stays the subject. The world around it becomes what you describe.
ReSkin (Style Transfer)
ReSkin is the character and object consistency engine. It takes a source image and applies a new visual skin — clothing, material, environment, lighting — while preserving the identity and pose of the subject.
Key specifications:
- Input: Character portrait, product shot, or object image
- Output: Same subject with new visual properties
- Control depth: Full body, upper body, face-only, or custom mask
- Material library: 50+ presets (denim, leather, silk, metal, glass, wood, stone, etc.)
- Custom material: Free-text material description
- Identity preservation: 95%+ facial recognition match, 90%+ body structure match
- Batch processing: Up to 20 variations per source
- Best for: Fashion catalogs, product visualization, character design, material testing
ReSkin is not face-swapping. It is identity-preserving transformation. The person remains the person. The look becomes what you need.
Expand Mode (Outpainting)
Expand Mode extends the canvas beyond the original frame. It generates new image content that matches the existing style, lighting, and perspective, allowing you to change aspect ratios, add environment, or create panoramic compositions.
Key specifications:
- Input: Any image with defined edge
- Output: Expanded canvas at custom dimensions
- Expansion: Up to 200% in any direction (left, right, top, bottom, or combinations)
- Seam matching: 100% edge continuity, zero visible seam
- Perspective matching: Automatic or manual perspective control
- Content awareness: Subject-aware expansion that preserves focal points
- Best for: Aspect ratio conversion, social media formats, panoramic creation, environmental extension
Expand Mode is not simple mirroring. It is context-aware generation that understands what should continue beyond the frame.
Quality Tiers: Ultra, Pro, Express
StarTools operates on three quality tiers that map to production requirements and credit budgets.
| Tier | Credits | Resolution | Passes | Best for |
|---|---|---|---|---|
| Ultra | 25 credits | 4K (3840×2160) | Full pipeline | Final deliverables, commercial, broadcast |
| Pro | 10 credits | 2K (2560×1440) | Standard pipeline | Production drafts, client review, portfolio |
| Express | 5 credits | 1080p (1920×1080) | Optimized pipeline | Quick iteration, concept testing, social media |
Ultra tier runs the full StarOxigen pipeline: 4-pass analysis, edge reconstruction, texture synthesis, color science, and final quality control. Output is broadcast-ready.
Pro tier runs the standard pipeline with 2-pass analysis and optimized synthesis. Output is production-quality for most use cases.
Express tier runs the optimized pipeline with single-pass analysis and fast synthesis. Output is ideal for rapid iteration and preview.
The StarOxigen 1.0 Engine Architecture
The StarOxigen engine is not a wrapper around existing models. It is a proprietary orchestration layer that coordinates multiple AI systems — GPT-Image-2, Gemini 3.1 Flash, and Seedance 1.5 Pro — to produce consistent, professional output.
Pipeline Architecture
- Input Analysis — The engine analyzes the input (image, video, or prompt) and classifies content type, quality, and processing requirements
- Model Selection — The engine selects the appropriate model tier (Ultra/Pro/Express) and workflow pipeline
- Parameter Mapping — The engine maps user settings to model parameters with production-quality defaults
- Generation — The selected model executes the generation task with optimized prompts
- Post-Processing — The engine applies quality control, color science, and format optimization
- Output Delivery — The final asset is delivered in the requested format with metadata
Content Policy Integration
StarOxigen injects a centralized content policy into all system prompts, automatically replacing celebrity names, politician names, and trademarked characters with generic descriptive archetypes. This ensures compliant output without user intervention.
Continuous Render Optimization
StarTools includes a Continuous Render Optimization system that monitors output quality in real-time and adjusts parameters dynamically to maintain consistent results across batches.
Key features:
- Seed locking — Lock a seed for reproducible output across sessions
- Batch consistency — Generate up to 20 variations with locked parameters
- Style memory — Save and recall style presets for brand consistency
- Quality gating — Automatic rejection of output below quality threshold
- Progress tracking — Real-time progress with SSE and polling fallback
How to Access StarTools
StarTools is available now at /starvision-tools to all VeyoLabs users. The engine requires:
- A VeyoLabs account with active subscription or pay-as-you-go credits
- Credits are consumed per generation based on tier (Ultra: 25cr, Pro: 10cr, Express: 5cr)
- All generated assets are stored in the user's VeyoLabs asset library with full download rights
What's Next for StarTools
The StarOxigen engine roadmap includes:
- Video-to-Video workflows — Full scene re-dubbing and style transfer (Q3 2026)
- Audio-visual generation — Joint audio-video generation with voice control (Q3 2026)
- 3D scene generation — Depth-aware video with camera control (Q4 2026)
- Batch processing API — Programmatic access to all 24 workflows (Q4 2026)
StarTools is the engine. The work is yours.