How Binary Search Solves Exact Image File Size Compression in the Browser

A developer has detailed a browser-only technique for compressing images to an exact target file size, such as under 100 KB, without relying on server-side tools. Instead of a simple step-down loop that can require up to 18 encoding passes, the approach uses binary search over quality levels, converging on the optimal result in just eight iterations. The algorithm tracks both the best-fitting and smallest-output results separately, ensuring graceful handling of edge cases where no quality level meets the target. For PNG files, which lack a quality parameter, the search is adapted to vary the color count instead, using a synchronous encoder wrapped in a Web Worker to avoid freezing the browser tab. The architecture separates pure encoding logic from DOM-dependent canvas operations, making the CPU-intensive work offload-friendly and reusable.