Pick the shot. Skip the guessing.
Example footage. Change the folder to see its available subjects.
Search the log you built
Every useful field gives you another way to find the footage. Search a name or note, a phrase from the transcript, or the labels you reviewed. ClipLogger searches its own log natively; you do not need to export it or ask a model to identify the same thing again.
When you have a looser clue, use semantic search for meaning, a visual description to find matching imagery, or Find Similar from a promising clip. Semantic and visual discovery need their relevant models and indexes prepared.
Let the directory show you your options
Open a shoot’s folder and enable Drill-down to include nested camera folders. In the filter picker, ClipLogger shows the fields and values found in the indexed media for that directory, with counts.
You can see that Shrimp scampi and Tajín-rim margarita are in the shoot before typing either name. Pick a subject, then add shot size, action or rating. Technical properties such as frame rate and duration can narrow the same result.
Changing the folder changes the available choices. The list reflects indexed records in that scope, so finish indexing before treating a missing value as proof that a shot does not exist.
Combine the things you know
| Selection | What it asks for |
|---|---|
| Shrimp scampi or Tajín-rim margarita | Either subject can match |
| Add Close-up | The clip must also be logged as a close-up |
| Add rating of four or higher | The clip must meet that rating too |
Multiple choices within a field match any of those values. Choices across different fields must all match. This lets you narrow a whole shoot without guessing a perfect search phrase.
The same approach works for Maya Torres + celebration + four stars, or interview + a named speaker + a topic. Your palette supplies the vocabulary.
Use a different clue when you need one
| Your clue | What to use |
|---|---|
| A name, filename, note or spoken phrase | Text and transcript search |
| An idea with different possible wording | Semantic search over prepared text and logs |
| A description of what the shot looks like | Visual text search with the supported image model ready |
| A frame that looks right | Find Similar |
| A subject, action, rating or technical property | Metadata filters |
If a known clip disappears, clear the filters and check the current folder, project or bin. A visual match is a reason to inspect a clip; review it before treating a person or item as confirmed.
Keep a useful search
Save a recurring request as a smart bin. It can collect matching clips as the log grows, with any items you explicitly pin or exclude. The bin refers to the files where they already live.
Play the results, mark the useful moments and make virtual clips when a range needs its own log. You have gone from an archive you barely remember to selects the next editor can use. Work with ranges and markers.