A release used to need one video, and most releases did not get one. It now needs four, and they are not variations of each other. They live on different surfaces, they are watched under different conditions, and three of the four are watched with the sound already coming from somewhere else.
Treating them as one deliverable in four export sizes is the most common and most expensive mistake in release marketing. Here is what each one is actually doing.
Spotify Canvas: texture, not a trailer
Canvas is the short vertical loop that replaces static cover art in the Spotify Now Playing view while a track is playing. The specification is narrow and worth memorising: 9:16 vertical, 3 to 8 seconds, between 720 and 1080 pixels high, MP4 or JPG, under 8 MB.
The important property is not in the spec sheet. A Canvas plays silently, on a loop, to somebody who is already listening to the song. It is not competing for attention and it is not explaining anything. It is the visual identity of the track while the track is happening.
That changes the design brief completely:
- The loop seam matters more than the content. A visible cut every four seconds reads as broken. A loop that returns cleanly reads as intentional.
- Text almost never works. It is small, it is moving, and the viewer is not reading.
- Motion should be slow. Fast cutting fights the music instead of sitting under it.
Spotify has published engagement figures for Canvas across streams, shares and playlist adds, and the direction is consistent even if any single number is hard to attribute cleanly. The more useful framing is simpler: a track without a Canvas has a blank space where its identity should be, on the one screen where every listener is guaranteed to look.
Apple Music Album Motion: the record as an object
Album Motion is animated artwork on the Apple Music release page. The key difference from Canvas is the level it attaches to. Canvas is per track. Album Motion is per release.
That makes it a different kind of asset. It is not soundtracking a listen, it is establishing that this is a record rather than a file. Apple Music leans harder on editorial presentation than most platforms do, and a release page with motion artwork sits differently next to a major-label release page than a static JPEG does.
For a single, this is optional. For an album or EP that you are pitching for editorial, it is part of looking like you meant it.
Lyric videos: the search surface
This is the one that gets undervalued because it is treated as a nice-to-have rather than as distribution.
A very large share of people looking for a specific song do not go to a streaming service. They type the title into YouTube, or they type a half-remembered line of the chorus into a search box. The result they land on is very often a lyric video, and if it is not yours, it belongs to whoever uploaded one first.
Two practical points:
Word-level sync is the whole product. Lyrics that appear on the beat feel produced. Lyrics that appear a beat late feel like a slideshow, and the viewer notices in the first ten seconds.
Lyrics are licensed content. Song lyrics are copyrighted works. Putting them on screen is a reproduction, and pulling them from a free lyrics site is not a licence to do that. This is the least understood risk in the format and the easiest one to get wrong at scale.
Artwork videos: claiming your own result
An artwork video is the full track over an animated version of the cover. It is the least ambitious of the four and it is the one most catalogues are missing.
Its job is coverage. A track that exists on streaming but has no video presence has an empty result where it should have a canonical one. Artwork videos fill that space quickly, at length, for every track on a release, without a shoot and without a concept.
For a back catalogue this is the highest-leverage asset of the four, because the alternative is not a better video. The alternative is nothing.
The four-in-one mistake
The failure pattern is always the same shape. An artist commissions or generates one video, exports it at 16:9, and uploads that file to YouTube, to Spotify, to Instagram and to Apple Music.
What happens next:
| Destination | What was needed | What arrived |
|---|---|---|
| Spotify Canvas | 9:16, 3 to 8 seconds, silent loop | 16:9, three minutes, sound-designed |
| Apple Music | Album-level motion artwork | A track video |
| YouTube | Full length, searchable, lyric-synced | Fine, if it is the right length |
| Reels and TikTok | 9:16, hook in the first two seconds | Letterboxed with bars top and bottom |
One of the four works. The other three are technically live and practically invisible.
A sensible order of operations
If you are running a release schedule rather than a single moment, this sequencing costs nothing and fixes most of it:
- Before release day: artwork video for every track, so nothing goes live with an empty video surface.
- Before release day: Canvas for the focus track, because it is the asset every listener sees.
- Release day: lyric video for the focus track, because that is when people start searching the title.
- For albums and EPs: Album Motion, before any editorial pitch goes out.
- Ongoing: Canvas for the rest of the tracklist as they get pushed.
None of these are big productions. All four are now generated from the master audio and the cover art you already delivered, which is the actual change worth reacting to.
Full detail on how the four formats are produced inside the distribution dashboard: Promo Videos with Rotor Videos by LyricFind.