# How to log footage faster for a post-production team > Log less by hand, and only log as deep as the footage will actually be used. Three changes do most of the work. Canonical URL: https://vivu.ai/guide/how-to-log-footage-faster-for-a-post-production Log less by hand, and only log as deep as the footage will actually be used. Three changes do most of the work. Fill in a short, fixed log template at ingest. Use transcripts for anything spoken. And rely on search for visual content instead of writing shot-by-shot descriptions. A detailed log has always been a search tool for later, and once footage can be searched by describing the shot, most of that log no longer needs to be typed. ## Decide what the log is for A footage log can do three different jobs. The ingest record says what came in: client, shoot, date, camera, card and usage rights. The editorial log says what's good: circle takes, the line the director liked, the take the client approved. The archive log describes what's in the footage for someone who will need it next year. Writing archive-depth descriptions for footage that will be cut once and then shelved is the most expensive way to handle any of these jobs. Split them, and each one gets faster. ## Make the manual part smaller Start with a fixed set of fields and don't add more. Client, shoot, date, camera, card, content type (interview, b-roll, event) and one free-text note are enough for the ingest record. Name files at ingest with the same pattern every time, so the name alone tells you which shoot a file came from. Get notes from set: a camera report or a short message from the shooter about what's on each card saves a watch-through. Log at the clip level, not the shot level, unless the clip is going into an edit soon. For footage that is going into the edit, drop markers during the first watch instead of writing a separate log afterward. None of this needs new software, and all of it makes the next two steps easier. The machine steps depend on consistent names and a clear content-type field. ## Let transcripts log the spoken material Interviews, testimonials and panel recordings don't need a hand log. A timecoded transcript is more complete than anything someone would type, and you can search it by the exact words. [Working from a transcript](https://vivu.ai/guide/how-do-i-use-a-transcript-to-find-good-video) covers how editors pull selects from one. What a transcript can't do is describe pictures, so everything without speech still needs another route. ## Let search handle the visual material For b-roll, product shots, event coverage and on-screen graphics, there are two machine options. Auto-tagging tools attach keywords to each clip when it comes in. There's a separate look at [how automatic tagging works](https://vivu.ai/guide/what-software-can-automatically-tag-video-files). Tags are fixed at ingest, so you're stuck with whatever vocabulary the tagger chose. A search layer indexes the footage once and waits for a question. The description gets written at the moment someone needs the shot, in their own words, and the answer comes back as time ranges. [Vivu](https://vivu.ai/platform) is one search layer built this way: each client's footage goes into its own project and is indexed once on upload, so a question like "where does the price breakdown appear on screen" can wait until an editor actually asks it. Each range comes back with a one-line reason, and when that reason paraphrases on-screen text, check the frame before anyone copies a figure from it into a brief. Both machine routes cost an upload step for each client and usage on a plan. And both only find footage. They don't pick takes or cut anything. ## When hand logging is still the right call Some logs can't be replaced. If a client contract requires a shot list as a deliverable, if a documentary needs rights and release notes on every clip, or if scripted work comes with a script supervisor's notes, a person has to write things down. Small shoots that go straight into one edit don't need an archive log at all, by hand or by machine. The edit project is the log. ## Where to cut first Work out which of the three logs takes most of your team's time. If it's the ingest record, fix the template and the naming. If it's the editorial log, switch to markers during the first watch. If it's archive descriptions of footage you may never open again, stop typing those and hand that work to transcripts and search. ## FAQ ### How detailed should a footage log be? Detailed enough for the next person to decide whether to open the clip. For most footage that means client, shoot, date, content type, rights and a one-line note. Shot-by-shot description is only worth it for footage that's going into an edit soon. ### Can AI log footage automatically? Partly. Transcription handles speech well, auto-tagging adds keywords, and search tools can find shots by description later. Judgments still need a person: which take is best, what the client approved, and what can't be used for rights reasons. ### Should the assistant editor finish logging before the edit starts? Only the ingest record and the selects for the first cut. Logging everything in depth before the edit delays the edit, and much of that footage may never be used. ### Should each client's footage be logged and stored separately? Yes. Keep each client's footage in its own folder or project and use the same template for all of them, so the log structure is shared but the footage isn't. That also means each search covers one client's material at a time.