Transcoding Presets & Quality Optimization: From 360p to 4K

Understanding Transcoding Presets

Transcoding is the process of converting a video stream from one format, resolution, or bitrate to another in real time. Every streaming platform needs transcoding to deliver content at multiple quality levels, adapt to different viewer bandwidth capacities, and ensure compatibility across devices. Choosing the right transcoding preset is a balancing act between video quality, CPU usage, and available bandwidth.

StreamDev provides a complete set of built-in presets ranging from low-resolution 360p up to full 4K output, plus a passthrough mode that bypasses transcoding entirely. Understanding when and how to use each preset is essential for running an efficient streaming operation.

Complete Preset Table

Here is the full reference for all available transcoding presets:

• 4K (2160p): 3840x2160 resolution, 15000 kbps video bitrate, 30 fps — Very High CPU usage
• 1080p 60fps: 1920x1080 resolution, 8000 kbps video bitrate, 60 fps — High CPU usage
• 1080p: 1920x1080 resolution, 5000 kbps video bitrate, 30 fps — Medium-High CPU usage
• 720p: 1280x720 resolution, 3000 kbps video bitrate, 30 fps — Medium CPU usage
• 480p: 854x480 resolution, 1500 kbps video bitrate, 30 fps — Low CPU usage
• 360p: 640x360 resolution, 800 kbps video bitrate, 30 fps — Very Low CPU usage

All presets use H.264 video codec and AAC audio codec by default, which provides the widest device compatibility. Audio bitrate is typically set to 128 kbps stereo across all presets.

Passthrough Mode

Passthrough mode is the most efficient transcoding option available. Instead of decoding and re-encoding the video, passthrough simply copies the incoming stream data directly to the output without any modification. This means the video codec, resolution, bitrate, and frame rate of the output are identical to the input.

The key advantage of passthrough is CPU usage: it consumes only 1-2% of CPU resources compared to 30-50% or more for active transcoding. This makes it ideal for scenarios where:

• The input stream is already at the desired quality and format
• You need to maximize the number of concurrent streams on a single server
• Your source is already encoded in H.264 and does not need format conversion
• You are restreaming content to platforms that accept the source format

When to Use Transcoding Instead

Active transcoding is necessary when you need to change the stream's resolution, bitrate, or format. Common scenarios include:

• Converting a high-bitrate 4K source to multiple lower quality tiers for adaptive bitrate delivery
• Reducing bandwidth usage by lowering the output bitrate
• Changing frame rate (e.g., converting 60fps input to 30fps output)
• Adding overlays, watermarks, or other video processing

Custom Transcoding Settings

Beyond the built-in presets, you can configure custom transcoding parameters for each stream. Custom settings allow you to specify exact values for resolution (width and height), video bitrate, frame rate, and audio bitrate. This is useful when you need a specific output format that does not match any preset, such as a non-standard resolution for a particular display or a specific bitrate required by a third-party platform.

CPU Usage Per Quality Level

Understanding CPU consumption is critical for capacity planning. Here is what to expect per stream on a modern server CPU:

• Passthrough: 1-2% CPU — negligible impact
• 360p transcoding: 5-10% CPU
• 480p transcoding: 10-15% CPU
• 720p transcoding: 15-25% CPU
• 1080p transcoding: 30-50% CPU
• 1080p 60fps transcoding: 45-65% CPU
• 4K transcoding: 80-100% CPU

These percentages are per stream on a single core. Multi-threaded encoding distributes the load across available cores, so a 4-core server handles the total load across all cores.

Concurrent Stream Capacity

On a typical 4-core server, here is the approximate concurrent stream capacity:

• Passthrough: 20-50 concurrent streams
• 720p transcoding: 4-8 concurrent streams
• 1080p transcoding: 2-4 concurrent streams
• 4K transcoding: 1 concurrent stream (may require 8+ cores)

These numbers scale linearly with additional CPU cores. An 8-core server can handle roughly double these capacities. For operations requiring many simultaneous transcoding streams, plan server capacity carefully and use passthrough mode where source quality permits.

HLS Segment Configuration

The output of transcoding is delivered as HLS (HTTP Live Streaming) segments. The default configuration uses 2-second segments with a playlist containing 5 segments. This provides a good balance between latency (approximately 10-second glass-to-glass delay) and reliability. Shorter segments reduce latency but increase the number of HTTP requests, while longer segments improve buffering resilience but add delay.

The segment format creates playlist files (.m3u8) and segment files (.ts) that are compatible with virtually every modern video player, browser, and device.

Codec Details

All transcoding presets use H.264 (AVC) for video encoding, which remains the most widely supported video codec across devices and platforms. For audio, AAC (Advanced Audio Coding) is used at 128 kbps stereo, providing excellent quality at a reasonable bitrate. These codec choices ensure maximum compatibility while maintaining professional-grade output quality.

Recommended Configurations

Based on common use cases, here are three recommended configurations:

Production Passthrough (Maximum Capacity)

Use passthrough mode when your source streams are already properly encoded. This maximizes the number of streams your server can handle and is ideal for IPTV operators rebroadcasting pre-encoded channels.

High-Quality Transcode (Best Visual Quality)

Use 1080p at 5000 kbps with 30fps for the best balance of quality and resource usage. For premium content or sports, consider 1080p 60fps at 8000 kbps. Pair with a 720p or 480p variant for adaptive bitrate delivery to accommodate viewers with lower bandwidth.

Mobile-Optimized (Lowest Bandwidth)

Use 480p at 1500 kbps or 360p at 800 kbps for mobile-first delivery. These presets work well on cellular connections and consume minimal server resources, allowing you to serve more concurrent viewers from the same infrastructure.