Format Selection

Format selection

Format choice trades compression against compatibility. The newest formats are substantially smaller and unsupported somewhere, which is why serving one format is either wasteful or broken for a subset of users.

Method

1. Separate the archival format from the delivery format. Keep a high-quality master and derive delivery variants, so a future format change does not require re-encoding from a lossy copy.
2. Negotiate rather than detect. Content negotiation and source sets let the client pick what it supports, which is more reliable than inferring from a user agent.
3. Always provide a fallback. A widely supported format alongside the modern one, since the alternative is a blank space for some users.
4. Match the format to the content. Photographic, graphic, and animated content compress best under different formats, and one default serves some badly.
5. Consider licensing and patent exposure. Some codecs carry licensing obligations that matter at commercial scale, and open alternatives may be the safer choice.
6. Measure the actual saving on your own content. Published compression figures come from standard test sets that may not resemble what your users upload.
7. Revisit as support shifts. Formats that needed a fallback two years ago may not now, and dropping a legacy variant reduces storage and complexity.

Boundaries

Newer formats save bandwidth and cost encoding time and compatibility risk. Support data is a moving target and needs periodic re-checking. Some clients misreport capability, so fallbacks remain necessary even when negotiation says otherwise.