A method for generation of a video montage file from a plurality of media files is disclosed. The method includes selecting media files from a collection of media files based on a selection criterion, accessing audio feature information for an audio track, the audio feature information including timing information related to audio features of the audio track, generating a first video segment for a first time slot from a first media file, generating a second video segment for a second time slot from a second media file, creating a video montage file by adding the second video segment to the first video segment, and adding the audio track to the video montage file. The selection criterion may comprise a location, a time interval, or presences of the media files in a feed posted by the user on a messaging system or social media platform.