Table of Contents

Enum MaxineArtifactReductionEffectMode

Namespace
VisioForge.Core.Types.VideoEffects
Assembly
VisioForge.Core.dll

Specifies the AI processing mode for NVIDIA Maxine Artifact Reduction effect.

public enum MaxineArtifactReductionEffectMode : uint

Fields

HighBitrate = 0

High Bitrate mode optimized for higher bitrate video sources with moderate compression artifacts.

This mode applies gentler artifact reduction that preserves subtle gradients and low-contrast details better. It is designed for videos encoded at higher bitrates where compression artifacts are present but not severe. The AI model uses a conservative approach to avoid removing legitimate fine details while still addressing visible blocking, ringing, and banding artifacts.

This mode removes fewer artifacts compared to LowBitrate mode but provides superior preservation of:

  • Subtle color gradients (sky, skin tones)
  • Low-contrast textures (fabric, surfaces)
  • Fine spatial details (hair, foliage)
  • Film grain and intentional noise

Recommended for:

  • Video encoded at 10+ Mbps for 1080p
  • Near-lossless compression (CRF 18-22 in x264/x265)
  • Professional broadcasts and high-quality streams
  • Sources where detail preservation is critical
LowBitrate = 1

Low Bitrate mode optimized for lower bitrate video sources with severe compression artifacts.

This mode applies more aggressive artifact reduction to combat the severe compression artifacts typical of low-bitrate video encoding. The AI model is tuned to recognize and eliminate strong blocking artifacts, mosquito noise, ringing, and color banding that are characteristic of heavily compressed video. This mode sacrifices some fine detail preservation in favor of more comprehensive artifact removal.

Best suited for videos where compression artifacts are highly visible and detract from viewing experience. The aggressive processing helps restore a cleaner appearance to heavily compressed content, though some subtle details may be smoothed in the process.

Recommended for:

  • Video encoded below 5 Mbps for 1080p
  • Web streaming with aggressive compression
  • Low-quality video downloads
  • Legacy media with visible blocking and banding
  • Sources where artifact removal is the primary concern

Remarks

The Artifact Reduction AI model adapts its processing strategy based on the expected severity and types of compression artifacts in the source video. Different bitrate scenarios produce different artifact patterns, and these modes optimize the neural network's behavior accordingly.