Filters

Each filter contains a set of models that were trained for a specific use case.

Summary

Understanding the filters and models within Topaz Video AI will allow you to create the best possible output.

Stabilization

Improve video quality by removing unwanted camera shakes and jitters due to hand jiggling and unintentional camera panning. This model stabilizes footage by moving and rotating the original frames to make subjects visibly stable from frame to frame.

• Input and output files will not match pixel to pixel when using this model.
• Increasing strength will result in a more stabilized output at the cost of increased artifacts (Full Frame) or a more cropped output (Auto-Crop).

Full Frame

Full-frame mode fills the blank spaces at the edges of the video using frame interpolation and deep learning. This method preserves the original resolution but may result in blurriness on edges in some cases.

Auto Crop

Auto Crop mode will crop out blank spaces on the edges of videos, resulting in an image that is always sharp but may have a reduced resolution and frame size.

Rolling Shutter Correction

Reduces wobbly distortion caused when the camera moves too fast.

Reduce Jittery Motions

This toggle corrects parallax distortion and tilting motion by running Chronos Fast for each interpolated pass.

Motion Deblur

Reduces the blurring caused by objects in quick motion.

Themis

Increases the perceived sharpness and detail of videos by removing blur instead of just increasing edge contrast.

Frame Interpolation

Process of synthesizing in-between images from a given set of images to create more fluid movement or to slow down the speed of a video file.

Frame Interpolation options:

• Choose a slow motion factor or to keep the speed the same, select none.
• Replace Duplicate Frames will replace duplicate frames with interpolated frames.
• Sensitivity can be raised to increase the sensitivity for removing duplicate frames. A higher setting may remove frames that are not duplicates but look very similar.

Apollo

Specializes in non-linear motion and slightly blurry inputs.

• Generates 8 interpolated frames per run, and then creates the selected output framerate from this result.
• Faster than Chronos.
• Use when you need exactly 2x, 4x, or 8x frame count multiple. For example, converting from 30 → 60fps at 200% slow motion would require exactly 4x the frames.

Apollo Fast

Typically when interpolating files faster than Apollo, however, may produce a lesser quality output than Apollo.

• Recommended for use for any non-double conversions. For example, 24fps -> 29.97 or 60.
• Use when speed is a factor and you require exactly 2x or 4x frames.

Chronos

General frame interpolation model suitable for FPS conversion or slow-mo. Processing speed decreases linearly with the number of generated frames.

• Recommended for slow motion beyond 8x, or smaller than 4x, or FPS conversion.

Chronos Fast

Specializes in fast motion with larger changes between frames.

Enhancements

Choose the appropriate input footage type for the best results.

• Interlaced: Deinterlace and enhance + upscale footage with a combined model.
• Interlaced Progressive: trained for interlaced videos that were previously transcoded to progressive with the incorrect field order.
• Progressive: Enhance and upscale without deinterlacing.

Interlaced

An interlaced file is an early form of video compression that was used to make videos look smoother. Achieved by breaking up each full frame of a video into alternating lines taken from two separate fields that were captured at slightly different times.

The Dione models are used to deinterlace interlaced files and upscale using a single AI model.

Dione DV

Trained for digital video footage like DV, DVCPRO, DVCAM, 720i, and 1080i.

Dione TV

Trained for analog videos from older TVs, camcorders, DVDs, and SD footage (480i, 576i, etc.).

Dione DeHalo

Dione TV variant that specializes in removing halo and fringing.

Interlaced Progressive

These models should be used when working with interlaced footage that has been previously incorrectly transcoded to progressive.

Dione Robust

Performs best on the following resolutions;

• 480i
• 576i
• 720i
• 1080i
• If results are not as expected, try one of the Interlaced Dione models.

Dione Robust DeHalo

Specializes in removing halo and fringing.

• If results are not as expected, try one of the Interlaced Dione models.

Progressive

Progressive scan video displays both the even and odd scan lines (the entire video frame) at the same time. Examples of this would be 480p, 720p, or 1080p.
There are 4 main progressive models to choose from; Proteus Fine Tune, Artemis, Gaia, and Theia. Each of these filters has a subset of models trained for specific quality issues.

Proteus Fine Tune

General enhancement model that allows fine-tuning several parameters for optimal results.

Excels when reducing noise when working with low to medium-quality footage

• Auto: Proteus will determine the optimal settings based on a 20-frame average.
• Relative to Auto: Proteus will determine the optimal settings and factor in your slider settings.
• Manual: You have full control of the settings. Use Estimate to allow the app to generate settings for you.

Artemis

General enhancement model that offers a good balance of improved detail, noise, and artifact reduction. Includes variants trained for low, medium, and high-quality footage with a variety of issues like haloing or aliasing.

Gaia

Improves already high-quality (HQ) and computer-generated (CG) files.

Theia

Sharpen and enhance detail to files with 4 parameters for greater control over the final result.

Grain

Adds a visual texture that gives an organic, textured feel.