· Everpop
A Short loops cleanly when its last frame can hand back to its first without the viewer noticing the join: same framing, same direction of motion, same sound floor, same light. Choose both cut points inside one continuous action, keep captions clear of the final frame, and watch the join three times before you post.
Loops used to be a party trick for editors. They now show up in your public numbers, because YouTube counts a Shorts view every time the video starts again.
What makes a Short loop instead of just stopping?
A loop is not a setting you fill in. It is a property of two frames, the last one you deliver and the first one you delivered, judged by the viewer's eye and ear in the same instant. When the two agree, the video appears to keep going. When they disagree, the clip feels finished even though it is still playing, and the thumb moves on.
YouTube's own vocabulary takes loops for granted: its definition of Engaged views for Shorts is "How many times viewers stayed to watch past the initial seconds, not including any loops" (YouTube Help). The platform separates the first pass from the extra passes, which is a useful hint about how to treat them yourself.
A worked example: a clip of someone repotting a plant opens with their hands flat on the table and ends the instant the hands come back down to it. Nothing clever happens at the join. The hands simply come home.
Where do you cut so the last frame hands back to the first?
Work backwards from a rest frame rather than forwards from the start of the take.
- Scrub for a rest frame. Look for a position the action keeps returning to: hands down, face front-on between beats, a static wide before the next move begins.
- Make that your first frame. Whatever the opening needs has to survive starting there; if the line begins two words earlier, choose a different rest frame rather than clip the sentence.
- End one frame before the action returns to that position. The next thing the viewer sees is your opening frame, which is where the shot was heading anyway.
- Judge the audio separately. Picture can be frame-accurate while the sound gives you away: a cut mid-vowel clicks, and a music bed cut mid-bar lands like a door closing.
At 30 frames per second a single frame is about 33 milliseconds: too small for most viewers to name what went wrong at a bad join, large enough that they feel it. So audit the join on six axes.
| Axis | Match this across the join | Symptom when it is off |
|---|---|---|
| Framing | Subject in the same part of the vertical frame at both ends | The subject appears to teleport |
| Motion | Movement heading the same way at the same speed | A stutter, like a dropped frame |
| Sound | Same room tone, same position in the musical phrase | A click, or a swallowed syllable |
| Light | Same exposure, same background state | A flicker at the join |
| Text | Captions and overlays fully cleared before the last frame | A word blinks in and out |
| Crop | Identical framing offset when the clip was reframed from horizontal | The whole picture nudges sideways |
That last row catches people out. Reframing horizontal footage to vertical adds a crop window that can sit differently at each end of the clip, so two otherwise identical frames land a few pixels apart. Perfectly matched hands, shifted window, broken loop.
Do replays count as views on YouTube Shorts?
Yes, and the fine print matters more than the headline. Since 31 March 2025, "Views will count the number of times a Short starts to play or replay, with no minimum watch time requirement" (YouTube Help). YouTube is extending the same first-frame rule to everything else it hosts: "Beginning August 24, 2026, views are counted the moment a video starts to play across all formats, including Shorts, long-form videos (VOD), and live streams" (YouTube Help).
The stricter metric survived under a new name. Engaged views for Shorts counts the people who "stayed to watch past the initial seconds, not including any loops", and YouTube states that Partner Program earnings still run on engaged Shorts views and engaged watch hours, while eligibility still runs on qualified Shorts views and qualified watch hours. So a good loop can lift the public count without touching the number your payouts are calculated from. It is a courtesy to the viewer and a change in one metric, not a monetization trick, and anyone selling it as one is quoting half the documentation. If that split is new to you, what counts as a view on YouTube Shorts has the longer version.
Which moments loop, and which ones refuse?
Loop well:
- Actions that return to rest: a tool set down, a lid closed, a chalkboard wiped.
- A question asked in the opening line and answered in the closing one, so the answer feeds the question.
- A reaction that reads differently on the second pass, because the setup now has a punchline attached.
- Rhythmic work on a locked-off camera: kneading, sanding, drumming, a repeated drill.
Refuse to loop:
- Anything that ends on a fade, an end card, or a held freeze.
- A slow push-in, unless you cut before the lens has travelled anywhere.
- Someone walking out of frame, or into it.
- Light that changes mid-clip: a cloud crossing, a monitor going bright, a stream alert firing.
- A punchline that needs silence to land. Looping over the laugh steps on it.
That last one deserves respect. Some moments earn their ending, and a clip that stops cleanly can beat one that loops.
How do you check a loop before you post it?
- Play the last second, then the first second, back to back, three times. If you can point at the join, so can everyone else.
- Watch it once with the sound off, then once with your eyes shut. Most failed loops fail on one channel, and separating them tells you which fix to make.
- Confirm the caption track clears before the final frame, and that the final frame is real footage: not black, not a held freeze, not a logo.
- Watch it on a phone, at arm's length, the way it will actually be seen.
- After it publishes, read Viewed vs. Swiped Away in Studio, which YouTube describes as "the percentage of times that viewers viewed your Shorts versus swiped away" (YouTube Help).
Checking joins is work that belongs before anything goes public. In Everpop, every proposed cut point is shown to you first and nothing publishes until you approve it, with three free re-renders per clip when a join needs nudging, and an export to FCPXML, EDL or SRT if you would rather finish it in your own timeline.
What can a loop actually not fix?
A loop cannot rescue a weak moment. It extends one that was already working, which is a different job.
The order of events is unforgiving: a viewer has to survive the opening before they ever reach your carefully matched final frame. YouTube's Shorts analytics puts that opening stretch on a card of its own, Viewed vs. Swiped Away, and the Engaged views definition stops counting at loops. Both instruments point at the front of the clip.
So the sequence is: pick a moment worth watching, give it a hook that earns the first second, and then, if the footage allows it, let the ending hand back to the beginning. In that order. A loop on a clip nobody stayed for is a very tidy door on an empty room.
Frequently asked questions
- Do replays count as views on YouTube Shorts?
- Yes. YouTube Help states that "Views will count the number of times a Short starts to play or replay, with no minimum watch time requirement", a change effective 31 March 2025 ([YouTube Help](https://support.google.com/youtube/answer/10059070?hl=en)). From 24 August 2026 the same first-frame rule applies across long-form video and live streams too.
- Do looped replays count toward monetization?
- No. YouTube keeps a second metric, Engaged views, which for Shorts counts "How many times viewers stayed to watch past the initial seconds, not including any loops", and it says Partner Program earnings are based on engaged Shorts views and engaged watch hours, while eligibility is based on qualified Shorts views and qualified watch hours ([YouTube Help](https://support.google.com/youtube/answer/12220281?hl=en&co=GENIE.Platform%3DDesktop)). Loops move the public view count, not that one.
- Does looping help the Shorts algorithm?
- There is no public YouTube documentation describing loops as a ranking signal, so treat any specific claim about that with suspicion. What is documented is narrower and more useful: replays add to the public view count, and Engaged views excludes them. Write the loop for the viewer, and judge distribution from your own analytics rather than from a rule someone repeated on a podcast.
- How long should a looping Short be?
- Shorts can run "up to 3 minutes long" ([YouTube Help](https://support.google.com/youtube/answer/10059070?hl=en)), and YouTube's recommendation guidance says "There's no universal 'ideal' length for YouTube videos. It depends on your content, audience, and the video's purpose" ([YouTube Help](https://support.google.com/youtube/answer/16559651?hl=en)). Practically, a loop asks the viewer to pay for a second pass, so the shorter the clip, the cheaper that decision is for them.
- Why does my loop still jump when the first and last frames look identical?
- Usually one of three things: the audio was cut mid-word or mid-bar, so the ear hears a click the eye cannot see; a caption or overlay was still on screen at the final frame; or the clip was reframed from horizontal footage and the crop window sits in a slightly different place at each end. Check those three before re-cutting the picture.
- Can you make a loop from horizontal footage?
- Yes, with one extra check. Reframing to vertical adds a crop window that can drift during the clip, so a pair of frames that match in the original can be misaligned in the vertical version. Pick a stretch where the framing holds still, then verify the first and last vertical frames against each other rather than against the source.