# Detailed Convert

## Introduction

The **Detailed Convert** feature is an advanced image-to-3D generation tool that allows users to transform product images into fully textured 3D models with granular control over the generation process.

{% embed url="<https://youtu.be/wkbMJLZpeB4>" %}

### Overview

Detailed Convert provides a comprehensive modal interface for converting 2D product images into 3D models using AI-powered generation. Unlike the basic "Convert to 3D" option, this feature offers extensive customization options for professional-grade 3D model creation.

### Access

The Detailed Convert feature is accessible through:

* Product card dropdown menu → "Detailed converting" option
* Available for products in the commerce section
* Requires existing product images to work with

### Key Features

#### Image Upload & Management

* **Multiple Image Support**: Upload 1-4 reference images (JPG/PNG format, max 10MB each)
* **Drag & Drop Interface**: Intuitive file upload with visual feedback
* **Image Validation**: Automatic format and size validation with clear error messages

#### AI Model Selection

* **Meshy 4**: Legacy model for simple meshes and faster generation (lower quality)
* **Meshy 5**: Stable model for most use cases with higher quality geometry and multi-image support
* **Latest (Meshy 6 Preview)**: Experimental model with improved quality (single-image only)
* **Smart Restrictions**: When multiple images are uploaded, only Meshy 5 is available

#### Geometry Controls

* **Target Polycount**: Adjustable slider from 100 to 300,000 polygons
  * Higher values = more detailed models but longer generation time
  * Default: 30,000 polygons
* **Symmetry Mode**:
  * **Off**: No symmetry applied
  * **Auto**: Automatically detects and applies symmetry if suitable
  * **On**: Enforces symmetry throughout generation

#### Texture Generation Options

Two methods for controlling model textures:

**Text Prompt Method**

* **Texture Prompt**: Up to 550 characters describing desired texture appearance
* **AI-Generated**: Creates textures based on text description
* **Priority**: Takes precedence over image prompts when both are provided

**Image Prompt Method**

* **Texture Image Upload**: Single reference image for texture guidance
* **Surface Detail Control**: Generates colors and surface details matching the reference
* **Geometry Matching**: Works best when reference geometry matches the original asset

#### Generation Process

* **Real-time Progress**: Visual progress bar with percentage completion
* **Background Processing**: Non-blocking generation allows continued app usage
* **Status Updates**: Live feedback on generation stages
* **Error Handling**: Comprehensive error reporting and recovery options

### User Interface

#### Modal Layout

* **Responsive Design**: Adapts to mobile and desktop viewports
* **Tabbed Interface**: Organized sections for different configuration options
* **Form Validation**: Real-time validation with clear error messages
* **Reset Functionality**: One-click reset to default or initial values

#### Controls

* **Generate Button**: Initiates the 3D model creation process
* **Reset Button**: Restores form to initial state with tooltip guidance
* **Progress Indicator**: Shows generation progress and estimated completion

### Technical Implementation

#### File Handling

* S3 policy-based secure file uploads
* Client-side file validation and preprocessing
* Automatic image optimization and format conversion

#### Generation Pipeline

* Integration with Meshy API for AI-powered 3D generation
* Serverless handler architecture for scalable processing
* Real-time progress tracking via WebSocket connections

### Use Cases

#### E-commerce Applications

* **Product Visualization**: Convert product photos to interactive 3D models
* **Shopify Integration**: Direct publishing to Shopify stores
* **Marketing Assets**: Create engaging 3D content for product pages

#### Professional Workflows

* **Rapid Prototyping**: Quick 3D model creation from concept images
* **Asset Generation**: Bulk conversion of product catalogs
* **Quality Control**: Fine-tuned generation parameters for consistent results

### Best Practices

#### Image Selection

* Use high-resolution, well-lit product images
* Ensure clear object boundaries and minimal background clutter
* Multiple angles provide better geometry reconstruction

#### Parameter Tuning

* Start with default settings and adjust based on results
* Higher polycount for detailed products, lower for simple shapes
* Use symmetry for balanced objects like furniture or accessories

#### Texture Optimization

* Combine text prompts with reference images for best results
* Keep texture prompts specific but concise
* Test different AI models to find optimal quality/speed balance

### Integration Points

#### Commerce Workflow

* Seamless integration with existing product management
* Automatic session updates and shop revalidation
* Progress tracking tied to specific product instances

#### Authentication & Permissions

* Requires an authenticated user session
* Respects shop-level permissions and quotas
* Secure file upload with policy-based access control