On this page

Quality

Measures how good a file actually is, so the poor ones can be caught early.

The quality node measures perceptual quality metrics of an image or video — resolution, blurriness (via a Laplacian measure) and bitrate — so low-quality assets can be flagged or filtered.

Kind

quality

Applies to

Image, Video

Input ports

media

Output ports

metrics (all of them together — this is what the quality filter connects to), blurriness, width, height, fps, frame_count, flag. The frame-rate and frame-count ports carry a value only for video

Requirements

OpenCV / video4j native runtime. CPU-bound. No GPU required.

Persists to

asset_json_comp + asset_node_result ledger

Configuration

The Quality Analysis settings panel: Check Blurriness
Figure 1. The node’s settings in the pipeline editor

Set these in the panel above, or in the node’s options block in a pipeline definition. Each one toggles an individual check:

Option Meaning

checkBlurriness

Compute a blurriness score (Laplacian variance)

checkResolution

Record image/video dimensions

checkVideoBitrate

Measure the video bitrate

checkAudioBitrate

Measure the audio bitrate

Seeing it run

Turn on Debug Mode and every node keeps what it produced, on the card itself. Below is a real run of this node over pexels-jack-sparrow-5977265.mp4.

The quality node after a run

The strip on the card lists what each output port carried — blurriness, width, height, and 4 more.

Use Cases

  • Quality gating — pair with the quality filter to route low-quality assets differently or skip them.

  • Blurry-image detection — flag out-of-focus photos.

  • Library stats — resolution/bitrate reporting across an asset pool.

Looking for something else?

All documentation Feature overview