At the moment, we extract 10 frames, equally divided, for a duration of a video. For cases like group meeting recordings that can last for an hour, the 10 frames are fast outdated (not corresponding to the time in the video) when you hover over the player.
We need to investigate a better way to increase the amount of frames we extract while keep the storage needed for them minimal.
Claude generated code
https://github.com/CERNDocumentServer/cds-videos/tree/feat/sprite-sheet-thumbnails
At the moment, we extract 10 frames, equally divided, for a duration of a video. For cases like group meeting recordings that can last for an hour, the 10 frames are fast outdated (not corresponding to the time in the video) when you hover over the player.
We need to investigate a better way to increase the amount of frames we extract while keep the storage needed for them minimal.
Claude generated code
https://github.com/CERNDocumentServer/cds-videos/tree/feat/sprite-sheet-thumbnails