This blog covers practical SVG and browser-graphics writing from the person building SVG Lab: how the editor gets built, what changes, and what actually works when you're making vector graphics without a design background.
Three ways to make an SVG with AI: a chatbot writes the markup, an image model gives you pixels to trace, or a vector generator returns real paths. What each outputs and how to get one you can actually edit.
ChatGPT writes valid SVG markup, but it draws blind: no preview, guessed coordinates, broken curves. Where it works, where it fails, and how to get an AI vector you can actually edit.
Resizing an SVG edits the root element's width, height, and viewBox attributes. The paths inside do not change. Here is what each attribute controls and what to do when viewBox is missing.
Converting SVG to JPG renders the vector to pixels and applies JPEG compression. Transparent areas become white. Here is what actually changes during the conversion and how to control the result.
How to convert SVG to PDF without rasterizing it, using the browser's print pipeline or a vector toolchain, plus what breaks along the way.
An SVG has no fixed resolution, so a blurry PNG is a sizing choice, not a quality loss. Here is how to export sharp, transparent PNGs from SVG.
What an SVG path editor actually changes in the d attribute, the path commands you need to know, and when to edit by hand versus visually.
How to change an SVG's colors with fill, stroke, currentColor, and CSS, plus why colors in an img tag won't respond to CSS at all.
Why a browser SVG editor got built instead of getting better at prompting AI for graphics: a story about vocabulary, not talent.