Quantum Minds Media Operators

Introduction

Media operators in Quantum Minds enable you to work with non-text content types including images, audio, and speech. These operators allow your minds to process and generate multimedia content, expanding the capabilities beyond text-only applications to create rich, multimodal experiences.

Available Media Operators

ImageGenerationUrl

The ImageGenerationUrl operator creates images based on text descriptions and returns accessible URLs for the generated images.

Inputs

Outputs

Example Usage

Model: "dall-e-3"
Prompt: "A futuristic smart city with sustainable architecture, flying vehicles, green spaces, and solar panels, in a photorealistic style"
Quality: "hd"
Size: "1024x1024"

Output: Markdown with an embedded image URL showing the described futuristic city

Best Practices

• Provide detailed, specific descriptions
• Mention style, perspective, lighting, and mood
• Include important details about what should and shouldn't be included
• Be aware of content policies that may restrict certain types of images
• Use quality and size parameters appropriate for your use case

ImageGeneration64

The ImageGeneration64 operator creates images based on text descriptions and returns them in base64-encoded format for direct embedding.

Inputs

Similar to ImageGenerationUrl:

Outputs

Supported Models

Both image generation operators support:

Choosing Between Image Operators

SpeechToText

The SpeechToText operator converts audio content to text transcriptions.

Inputs

Outputs

Supported Models

Example Usage

File: [Meeting recording audio file]
Model: "whisper-large-v3"

Output: Complete transcription of the meeting with speaker identification

Best Practices

• Use high-quality audio when possible
• Consider pre-processing noisy audio
• Choose appropriate models for your language needs
• Be aware of length limitations for audio files
• Review and correct transcriptions for critical content

TextToSpeech

The TextToSpeech operator converts text to spoken audio content.

Inputs

Outputs

Example Usage

Text: "Welcome to our quarterly financial review. In this presentation, we'll cover the key performance indicators, revenue highlights, and our outlook for the coming quarter."
Language: "en-US"

Output: Audio file containing the spoken version of the provided text

Best Practices

• Break long text into natural paragraphs
• Use punctuation to influence pacing and intonation
• Consider including pronunciation guides for uncommon terms
• Test with smaller segments before processing large content
• Be mindful of total character count for efficiency

Combining Media Operators

Media operators can be combined to create powerful multimedia experiences:

Text-to-Speech-to-Text Verification

{
  "operator": "TextToSpeech",
  "input": {
    "text": "Important announcement regarding the system upgrade scheduled for next weekend."
  }
}

↓

{
  "operator": "SpeechToText",
  "input": {
    "file": "$TextToSpeech_001.output.content"
  }
}

↓

{
  "operator": "TableToTextSummary",
  "input": {
    "prompt": "Compare the original text with the transcription and identify any discrepancies",
    "dataframe": { "original": "Important announcement...", "transcribed": "$SpeechToText_001.output.content" }
  }
}

Image Generation with Audio Description

{
  "operator": "ImageGenerationUrl",
  "input": {
    "model": "dall-e-3",
    "prompt": "A visualization of global supply chain networks with highlighted routes and distribution centers",
    "quality": "standard",
    "size": "1024x1024"
  }
}

↓

{
  "operator": "TextToSpeech",
  "input": {
    "text": "This visualization shows our global supply chain network. The red lines represent primary shipping routes, while the blue dots indicate major distribution centers. Note the concentration of activity in Southeast Asia and North America."
  }
}

↓

{
  "operator": "CardGenerator",
  "input": {
    "prompt": "Create an interactive card with the image and audio narration"
  }
}

Integration with Other Operators

Multi-Modal Content Creation

Media operators work seamlessly with other operator categories:

1. Data Visualization with Narration:
   SQLExecution → TableToGraph → TextToSpeech

2. Automated Report Generation:
   PandasAi → TextToSpeech → CardGenerator

3. Image Generation from Data:
   TableToTextSummary → ImageGenerationUrl

4. Audio Transcription Analysis:
   SpeechToText → TextToNoSQL → MongoExecution

Using with MultiModal Operators

Media operators complement the MultiModal operators:

• Use SpeechToText to prepare audio for GeminiMultiModal analysis
• Use ImageGenerationUrl to create visuals based on ClaudeMultiModal insights
• Process TextToSpeech output with GeminiMultiModal for secondary analysis

Advanced Use Cases

Accessibility Enhancement

Create accessible versions of content:

{
  "operator": "RAGSummarize",
  "input": {
    "prompt": "Summarize the key points from our annual report",
    "collection": "company_reports"
  }
}

↓

{
  "operator": "TextToSpeech",
  "input": {
    "text": "$RAGSummarize_001.output.content"
  }
}

Content Localization

Translate and voice content for multiple languages:

{
  "operator": "OpenSearch",
  "input": {
    "prompt": "Translate the following product description to Spanish, French, and German: [product description]"
  }
}

↓

{
  "operator": "TextToSpeech",
  "input": {
    "text": "$OpenSearch_001.output.content",
    "language": "es-ES, fr-FR, de-DE"
  }
}

Interactive Tutorials

Create rich learning experiences:

{
  "operator": "ImageGenerationUrl",
  "input": {
    "prompt": "Step-by-step illustration of how to configure the system settings",
    "model": "dall-e-3",
    "quality": "hd",
    "size": "1024x1024"
  }
}

↓

{
  "operator": "TextToSpeech",
  "input": {
    "text": "In this tutorial, we'll walk through the system configuration process..."
  }
}

↓

{
  "operator": "CardGenerator",
  "input": {
    "prompt": "Create an interactive tutorial card with both visual and audio guidance"
  }
}

Best Practices for Media Operators

Performance Considerations

• Be aware of processing times for media generation
• Consider asynchronous processing for long audio files
• Optimize image sizes for your specific use case
• Cache frequently used media when possible

Content Quality

• Provide detailed prompts for image generation
• Use clear, well-paced text for speech synthesis
• Ensure audio files have good quality for transcription
• Review generated media for accuracy and appropriateness

Technical Limitations

• Be aware of file size limits for audio processing
• Consider format compatibility across systems
• Understand resolution constraints for image generation
• Plan for potential failures in media processing

Ethical Considerations

• Ensure generated content adheres to appropriate guidelines
• Consider bias in image generation and speech recognition
• Be transparent about AI-generated media when appropriate
• Respect copyright and intellectual property in prompts

Next Steps

Explore how Media Operators can be combined with Excel Operators to create rich, data-driven presentations and reports.

Overview | Operator Categories | LLM Operators | Excel Operators