this could be done on the client using drawImage on canvas or in a small service calling ffmpeg
scaled down and jpeg compressed the images are probably small enough (270p@20kb) to cache in the dht
maybe webms with a trailing set of frames sampled from several minutes of video?
this could be done on the client using drawImage on canvas or in a small service calling ffmpeg
scaled down and jpeg compressed the images are probably small enough (270p@20kb) to cache in the dht
maybe webms with a trailing set of frames sampled from several minutes of video?