[westurner]

From https://news.ycombinator.com/item?id=44717303 :

> Objective metrics and tools for video encoding and source signal quality: netflix/vmaf, easyVmaf, psy-ex/metrics, ffmpeg-quality-metrics,

netflix/vmaf: https://GitHub.com/netflix/vmwaf

gdavila/easyVmaf: https://github.com/gdavila/easyVmaf

psy-ex/metrics: https://github.com/psy-ex/metrics/

slhck/ffmpeg-quality-metrics: Calculate quality metrics with FFmpeg (SSIM, PSNR, VMAF, VIF) https://github.com/slhck/ffmpeg-quality-metrics

Something like this would be great too:

The Ardour Manual > Loudness Analyzer and Normalizer: https://manual.ardour.org/mixing/basic-mixing/loudness-analy...

[tejaswigowda]

This is a great addition! You can use the raw ffmpeg mode to do this currently if you know the flag syntax

[westurner]

Someday it would be cool for someone to write and maintain:

(Develop a development and branch coverage testing plan for)

A rust WASM GUI video encoding tool with by default just two UI wells A and B. In the "basic encoding" workflow, show in (A) the input video and audio waveform and metrics, and in (B) the output video and audio waveform and metrics. By default, if there is already an output feed, lock the video scrubbers together so that seeking seeks in both A and B in order to make comparing input signal with output encoding(s) easy.

Integrate or just reimplement e.g easyVMAF and other tools for tuning and optimizing encoding parameters. Develop the metrics report schema in YAML-LD with RDFS (and schema.org,) and store that with a filepath derived from the filename.

In an additional YAML-LD file, Store a list of encoding outputs per input video. It probably looks a lot like the output from ytdl, for example.

Given a set of encoding profiles, generate each output and enqueue each in a review queue to display in wells (A, B, [C, [D, ...]])

Like VirtualDub but in Rust with a WASM build.