Rachel Barker
LightNews LightNews
rachelplusplus.me.uk
Rachel Barker
@rachelplusplus.me.uk
54 followers 77 following 59 posts
Computing environmentalist, amateur musician, general-purpose nerd.
Previously worked on AV1+AV2 @ Google, now consulting @ Monocot. She/her.

Blogs at https://www.rachelplusplus.me.uk/
Posts Replies Media Videos
rachelplusplus.me.uk
You know, I can't quite put my finger on it, but I think there might be something wrong with my graph

(correct version coming soon!)
A graph which is supposed to compare the performance of four different video encoders, but with wildly incorrect numbers: instead of expected numbers like +20% file size or 0.1 times a reference runtime, it has +3200% file size and 0.<forty zeros>1 times a reference runtime
July 18, 2025 at 12:29 AM
rachelplusplus.me.uk
Lil' teaser for an upcoming blog post, building on my last post (www.rachelplusplus.me.uk/blog/2025/03...): This graph compares compressing an image at a single fixed resolution (dotted curve) vs. encoding at multiple resolutions, netflix style (solid curve):
A graph comparing the size of a compressed image file against the image quality. There are three pairs of curves, comparing libaom version 3.12.1, jpegli 0.11.1, and tinyavif 1.1. Each pair has one solid curve and one dotted curve, with the dotted curve representing the size vs. quality tradeoff for encoding at a single size, and the solid curve representing netflix-style multi-resolution encoding. The curve for libaom is the leftmost, and the curves for jpegli and tinyavif cross over as each one does better at different quality ranges. The curves for multi-resolution encoding break away from the single-resolution curves and go further to the left at lower qualities, representing better compression due to switching to a smaller resolution.
May 21, 2025 at 3:25 AM