Count total plays instead of added copies
The Total plays setting includes the first playback. Selecting four plays for a 12-second source creates about 48 seconds of video, not 60. Choose an integer from two through ten and multiply that count by the source duration before exporting. The source can be up to 90 seconds, while the repeated result must stay within three minutes. For example, a 30-second source fits six plays within the output limit, but ten plays would make it too long for this tool.
Choose endpoints that make a clean join
Each copy starts immediately after the previous one, using a straight cut from the source ending back to its beginning. There is no crossfade or generated transition between copies. Look for a source whose opening and ending have similar composition, subject position, and movement if you want the join to be less noticeable. For a demonstration, a visible restart may be useful instead. Watch at least one repeat boundary in the result, since the middle of a single playback cannot show how the join feels.
Listen to the repeated soundtrack
Supported source audio repeats with the picture and is encoded as AAC in the exported MP4. Speech, background noise, and music therefore restart with each copy. A visually similar opening and ending can still produce an obvious sound change at the boundary. Listen across that boundary before using the result as a background clip. If the recording should be silent, remove its audio before repeating it. A silent input stays silent; this tool does not generate a music bed or smooth an audio transition.
Distinguish a repeated file from endless playback
The download contains a fixed number of copies encoded as H.264 video. It ends after the chosen number of plays, even if you intended to use it on a continuously running display. Endless playback requires a repeat setting in the player showing the file. Encoding the copies into one MP4 makes the sequence part of the download, but it also creates a new file whose quality and size can differ from the source. Keep the original when testing different repeat counts.