File Transcription

LansonAI's file transcription service converts your audio files into accurate, searchable text using advanced AI models.

Overview

File transcription is an asynchronous process where you upload an audio file and receive a complete transcription after processing. This is ideal for:

Podcast episodes

Interview recordings

Meeting recordings

Lectures and presentations

Voice memos and notes

How It Works

The Transcription Pipeline

1. File Upload - You upload your audio file - File is validated and stored securely

2. Queue Management - File enters processing queue - Position depends on system load

3. AI Processing - Advanced speech recognition models analyze audio - Audio is segmented intelligently - Text is generated with timestamps

4. Post-Processing - Speaker diarization (if applicable) - Punctuation and formatting - Confidence scoring

5. Delivery - Results available in dashboard - Notification sent on completion

Supported Features

Audio Analysis

Automatic language detection

Multiple language support

Accent recognition

Background noise filtering

Speaker Identification

Speaker diarization: Automatically detect different speakers

Speaker labeling: Assign names to speakers

Speaker tracking: Maintain consistency across segments

Output Quality

High accuracy: 85-98% depending on audio quality

Punctuation: Automatic sentence structure

Timestamps: Precise timing for each segment

Formatting: Organized into readable paragraphs

Audio Quality Guidelines

Best Results

For optimal transcription accuracy:

Recording Environment:

Quiet location with minimal background noise

Controlled acoustics (avoid echoing rooms)

Stable recording setup

Equipment:

Good quality microphone

Proper microphone placement (6-12 inches from mouth)

Pop filter to reduce plosives

Recording Settings:

Sample rate: 44.1 kHz or higher

Bit depth: 16-bit or higher

Mono or stereo (mono preferred for speech)

Speech Quality:

Clear articulation

Consistent speaking volume

Natural speaking pace (not too fast)

Minimal overlapping speech

Common Issues to Avoid

Audio Problems:

Heavy background noise (traffic, fans, crowd)

Music playing during speech

Multiple simultaneous speakers

Poor microphone quality

Excessive echo or reverb

Recording Problems:

Too quiet (low volume)

Clipping (too loud, causing distortion)

Inconsistent volume levels

Sudden loud noises

Processing Times

Estimated Duration

File processing typically takes:

Factors affecting processing time:

Audio file length

Audio quality and complexity

Number of speakers

System load and queue position

Selected AI model

AI Models

LansonAI uses state-of-the-art speech recognition models:

Model Options

Standard Model (Default)

Balanced accuracy and speed

Best for general use cases

Supports multiple languages

Good with various accents

Premium Model (Higher Accuracy)

Maximum accuracy

Better with technical terminology

Superior noise handling

Longer processing time

Language Support

Supported languages include:

English (US, UK, AU)

Spanish

French

German

Italian

Portuguese

Chinese (Mandarin)

Japanese

Korean

And many more...

Advanced Options

Custom Vocabulary

Add specific terms for better recognition:

Technical terms: Industry-specific jargon

Names: Person names, company names

Acronyms: Uncommon abbreviations

Proper nouns: Places, brands

Formatting Preferences

Customize output format:

Timestamps: Include or exclude

Speaker labels: Enable or disable

Paragraph breaks: Automatic or manual

Punctuation style: Customize rules

Processing Options

Speaker diarization: Identify different speakers

Noise reduction: Enhanced cleaning for noisy audio

Profanity filter: Option to censor explicit language

Punctuation: Automatic or minimal

Usage and Costs

Credit System

File transcription uses credits based on:

Audio duration: Minutes of audio processed

Model used: Standard vs. Premium

Additional features: Speaker diarization, etc.

Example Costs:

10-minute standard transcription: ~10 credits

10-minute premium transcription: ~20 credits

Additional speaker diarization: +20% credits

Free Tier

New users receive:

Welcome credits for testing

Limited monthly free credits

See Free Credits for details

Paid Plans

For regular usage:

Monthly subscription plans

Pay-as-you-go options

Enterprise custom pricing

See Pricing for current rates

Best Practices

Before Upload

1. Review audio quality - Listen to a sample - Check for major issues - Clean up if necessary

2. Prepare metadata - Name file descriptively - Note speaker names - List any unusual terms

3. Optimize file - Convert to supported format - Compress if too large - Trim unnecessary sections

During Processing

1. Monitor progress - Check dashboard periodically - Watch for error notifications - Verify queue position

2. Plan next steps - Decide on export format - Prepare for editing if needed - Organize workspace

After Completion

1. Review results - Quick accuracy check - Verify speaker labels - Note any issues

2. Edit if needed - Use Episode Editor for corrections - Adjust timestamps - Refine formatting

3. Export and use - Download in preferred format - Archive for future reference - Share with team if needed

Troubleshooting

Poor Transcription Quality?

Check these factors:

Audio quality (noise, clarity)

Speaking pace (too fast?)

Accents (strong or unusual?)

Technical terms (uncommon words?)

Solutions:

Try premium model

Add custom vocabulary

Improve audio quality

Contact support for tips

Processing Stuck?

If processing doesn't complete: 1. Wait 10-15 minutes 2. Refresh the page 3. Check your credits 4. Contact support if > 30 minutes

Unexpected Results?

Common issues:

Wrong language detected

Speaker confusion

Missing sections

Formatting problems

What to do:

Re-process with correct settings

Use manual editing

Report persistent issues

Integration

API Access

For developers and automated workflows:

REST API for file submission

Webhook notifications

Status polling

Result retrieval

See Trans API Documentation

Workflow Integration

Connect with your tools:

Zapier integration (coming soon)

Webhooks for automation

Batch processing support

Custom integrations available

Security and Privacy

Data Protection

Encrypted uploads: TLS/SSL encryption

Secure storage: Files encrypted at rest

Access control: Only you can access your files

Automatic deletion: Old files removed per retention policy

Privacy Policy

We don't share your data

Files processed privately

GDPR compliant

See full Privacy Policy

Next Steps

Learn about Real-time Transcription

Explore the Episode Editor

Check Usage Statistics

Read Trans API Documentation

Support

Need help with file transcription?

Email: support@lansonai.com

Include file ID for faster assistance

Response time: Usually within 24 hours