July 12, 2026
Small image, big comment energy
Handsum: An LQIP Image File Format
Tiny preview images just got a new rival — and commenters immediately started redesigning it
TLDR: Handsum is a new ultra-tiny image placeholder format meant to make websites feel faster by showing a quick preview first. Commenters were instantly split between “make it bigger and more flexible” and “nice, but can it help load the real image too?”
A new image format called Handsum has entered the chat, promising to make those blurry little preview pictures you see while websites load even tinier and easier to store. The pitch is simple: instead of shipping a bigger image right away, a site can show a super-small placeholder almost instantly, then swap in the real thing later. Handsum’s creator says the files stay at a fixed tiny size — from 48 to 147 bytes — which is catnip for anyone who loves neat, predictable systems. It’s basically a new contender in the race to make loading screens look less sad.
But the real action is in the peanut gallery. One commenter immediately went full backseat engineer, asking why everyone is obsessed with 32x32 and demanding to know whether 48x48 or 64x64 could steal the spotlight instead. Translation for normal humans: “Cool idea, but why stop there?” Another commenter came in with the more practical grenade: if this tiny preview doesn’t help load the final full image faster, is it really doing enough? That’s the mood in a nutshell — half the crowd is intrigued by the cleverness, the other half wants more than a cute placeholder and is already asking for a sequel.
The vibe is classic internet tech drama: one side loves the tidy minimalism, the other side sees an unfinished product wearing a fancy new name. It’s less “wow, magic” and more “okay, but can it do one more thing?”
Key Points
- •The article introduces Handsum as a custom image file format for low-quality image placeholders (LQIPs).
- •Handsum uses fixed-size encodings, ranging from 48 bytes at the lowest quality to 147 bytes at the highest.
- •The format is based on the Discrete Cosine Transform, the same fundamental technique used by Blurhash, Thumbhash, and JPEG.
- •A web demo for Handsum uses a decoder written in C and compiled to WebAssembly.
- •The article compares Handsum visually and by byte size against PNG, Thumbhash, WebP, ETC2, and JPEG using resized versions of well-known images.