# Speex Audio Encoder for .NET

Video Capture SDK .Net Video Edit SDK .Net Media Blocks SDK .Net

Speex is a patent-free audio codec specifically designed for speech encoding. It provides excellent compression while maintaining voice quality across various bitrates. VisioForge integrates this powerful encoder into its .NET SDKs, offering developers flexible configuration options for speech-based applications.

The Speex encoder in VisioForge SDKs supports:

• Multiple frequency bands for different quality levels
• Variable and fixed bitrate encoding
• Voice activity detection and silence compression
• Adjustable complexity and quality settings

VideoCaptureCoreX VideoEditCoreX MediaBlocksPipeline

Speex offers four operation modes optimized for different frequency ranges:

The encoder automatically adjusts internal parameters based on the selected mode. For most speech applications, Wide Band (mode 2) offers an excellent balance between quality and bandwidth usage.

Speex works with three standard sampling frequencies:

• 8,000 Hz - Best for telephone-quality audio (Narrow Band)
• 16,000 Hz - Recommended for most voice applications (Wide Band)
• 32,000 Hz - Highest quality speech encoding (Ultra Wide Band)

The encoder handles both:

• Mono (1 channel) - Ideal for speech recordings
• Stereo (2 channels) - For multi-speaker or immersive audio

For consistent perceptual quality, use the Quality parameter:

var settings = new SpeexEncoderSettings {
    Quality = 8.0f, // Range from 0 (lowest) to 10 (highest)
    VBR = false     // Fixed quality mode
};

Higher quality values produce better audio at the expense of increased file size. Most speech applications work well with quality values between 5-8.

VBR dynamically adjusts the bitrate based on speech complexity:

var settings = new SpeexEncoderSettings {
    VBR = true,
    Quality = 8.0f  // Target quality level
};

This approach typically saves bandwidth while maintaining consistent perceived quality, making it ideal for streaming applications.

ABR maintains a target bitrate over time while allowing quality fluctuations:

var settings = new SpeexEncoderSettings {
    ABR = 15.0f,   // Target bitrate in kbps
    VBR = true     // Required for ABR mode
};

This option works well when you need predictable file sizes or bandwidth usage.

For consistent data rates throughout the encoding process:

var settings = new SpeexEncoderSettings {
    Bitrate = 24.6f,  // Fixed rate in kbps
    VBR = false
};

Supported bitrates range from 2.15 kbps to 24.6 kbps:

• 2.15 kbps - Ultra-compressed speech (limited quality)
• 3.95 kbps - Low bandwidth voice
• 5.95 kbps - Basic speech clarity
• 8.00 kbps - Standard voice quality
• 11.0 kbps - Good speech reproduction
• 15.0 kbps - Near-transparent speech
• 18.2 kbps - High-quality voice
• 24.6 kbps - Maximum quality speech

VAD identifies the presence of speech in audio signals:

var settings = new SpeexEncoderSettings {
    VAD = true,    // Enable voice detection
    DTX = true     // Recommended with VAD
};

This feature improves bandwidth efficiency by focusing encoding resources on actual speech segments.

DTX reduces data transmission during silence periods:

var settings = new SpeexEncoderSettings {
    DTX = true     // Enable silence compression
};

For VoIP and real-time communications, enabling DTX can significantly reduce bandwidth requirements.

Control CPU usage versus encoding quality:

var settings = new SpeexEncoderSettings {
    Complexity = 3  // Range: 1 (fastest) to 10 (highest quality)
};

Lower values prioritize speed and reduce CPU load, while higher values improve audio quality at the cost of performance.

Always verify encoder availability before implementation:

if (!SpeexEncoderSettings.IsAvailable())
{
    throw new InvalidOperationException("Speex encoder not available on this system.");
}

var encoderSettings = new SpeexEncoderSettings
{
    Mode = SpeexEncoderMode.WideBand,
    SampleRate = 16000,
    Channels = 1,
    Quality = 7.0f
};

var voipSettings = new SpeexEncoderSettings
{
    Mode = SpeexEncoderMode.WideBand,
    SampleRate = 16000,
    Channels = 1,
    VBR = true,
    VAD = true,
    DTX = true,
    Quality = 6.0f,
    Complexity = 4
};

var highQualitySettings = new SpeexEncoderSettings
{
    Mode = SpeexEncoderMode.UltraWideBand,
    SampleRate = 32000,
    Channels = 2,
    Bitrate = 24.6f,
    Complexity = 8
};

// Create a Video Capture SDK core instance
var core = new VideoCaptureCoreX();

// Configure Speex settings
var speexSettings = new SpeexEncoderSettings
{
    Mode = SpeexEncoderMode.WideBand,
    SampleRate = 16000,
    Channels = 1,
    VBR = true,
    Quality = 7.0f
};

// Add the Speex output
core.Outputs_Add(speexSettings, true);

// Create a Video Edit SDK core instance
var core = new VideoEditCoreX();

// Configure and set the output format
var speexSettings = new SpeexEncoderSettings
{
    Mode = SpeexEncoderMode.WideBand,
    SampleRate = 16000,
    VBR = true,
    ABR = 15.0f
};

// Set the output format
core.Output_Format = speexSettings;

// Configure Speex settings
var speexSettings = new SpeexEncoderSettings
{
    Mode = SpeexEncoderMode.NarrowBand,
    SampleRate = 8000,
    DTX = true,
    VAD = true
};

// Create a Speex encoder block
var speexEncoder = new SpeexEncoderBlock(speexSettings);

When implementing Speex encoding, consider these optimization strategies:

1. Match sample rate to content - Use Narrow Band (8 kHz) for telephone audio, Wide Band (16 kHz) for most voice applications, and Ultra Wide Band (32 kHz) only when maximum quality is required

2. Enable VBR with VAD/DTX for speech content - This combination provides optimal bandwidth efficiency for typical voice recordings

3. Adjust complexity based on platform - Mobile applications may benefit from lower complexity values (2-4), while desktop applications can use higher values (5-8)

4. Use ABR for streaming - Average Bit Rate provides predictable bandwidth usage while maintaining quality flexibility

5. Test different quality settings - Often a quality setting of 5-7 provides excellent results without excessive file size

Speex encoding excels in these developer scenarios:

• VoIP applications and internet telephony
• Voice chat features in games and collaboration tools
• Podcast creation and distribution
• Speech recognition preprocessing
• Voice note applications
• Audio archiving of speech content

The VisioForge implementation of Speex provides .NET developers with a powerful tool for speech compression that balances quality, performance, and bandwidth efficiency across a wide range of applications.