How to make a video from photos that does not look like a slideshow

The default way to make a video from photos is to put them on screen one after another with a crossfade. Everyone recognises that format instantly, and nobody watches it twice.

There is a better version, and the difference is not more transitions. It is that some of the photos move, the order tells a small story, and the music carries the cuts rather than decorating them. Here is how to build that.

Choose which stills move

You do not animate all of them. Animating everything is exhausting to watch and expensive to produce.

A good ratio is roughly one moving shot for every two or three stills. The moving ones should be the images that carry the most weight — the opening, the emotional centre, the last frame.

Pick for animation the photos that have somewhere to move: depth from foreground to background, something plausibly in motion, margin around the subject. Our guide to animating a photo covers what makes an image a good candidate and what makes it a bad one.

The rest stay still — but a still on screen for three seconds is not dead. A slow push on a static image, done on the timeline rather than by generation, gives it life for free.

Order for a shape, not a chronology

Chronological order is the least interesting available option and the one most photo videos use.

Three orderings that work better:

Wide to close. Open on the widest, most contextual image. Move progressively closer. End on the smallest detail. It reads as attention narrowing, which is how people actually look at things.

Question and answer. Open on something incomplete or ambiguous. Spend the middle establishing context. Resolve on the image that explains it.

Rhythm. Alternate wide and close, calm and busy. Two similar shots in a row flatten; alternating keeps a viewer's eye moving.

Whichever you pick, decide the last image first. The final frame is what a viewer is left holding, and building backwards from it makes every other choice easier.

Match motion across the cut

The thing that separates a video from a slideshow is that motion continues across cuts.

Two rules do most of the work:

• Keep direction consistent within a sequence. If one clip pushes in, the next should push in or hold — not pull out. Reversing direction on every cut reads as indecision.
• Cut on motion, not between motions. A cut that happens while something is moving is nearly invisible. A cut that happens during a still moment is a noticeable event, which you should save for the moments that deserve one.

Our camera movement reference covers which moves are reliable; for a photo video, slow push-ins and static-camera-with-environmental-motion are the two you will use most.

Make the transitions boring

Crossfades, wipes, page turns and star wipes are what make a photo video look like a photo video.

Use hard cuts almost everywhere. Use a short crossfade only where you genuinely mean "time passed". Nothing else. The restraint is what reads as professional, and it costs nothing.

Let the music decide the length

Pick or generate the track before you finalise the edit, and cut to it.

Generating an original track in the Music Studio costs 25 credits and takes a couple of minutes, and it means the music fits your piece rather than the other way round. Prompt for the mood and the shape you need:

Warm nostalgic instrumental, slow build, gentle piano over soft strings, resolves clearly at the end

Then place your cuts on the beat. If you are working by ear, the BPM tapper will give you the tempo so you can space cuts in bars rather than by feel. Cuts on the beat are the second thing, after motion matching, that makes a photo video feel edited.

A working sequence

1. Select twelve to twenty photos. Fewer than you think. Every weak image costs you a viewer.
2. Choose the last one. Build toward it.
3. Mark four or five to animate. Draft each at 360p, 6 seconds, for 25 credits.
4. Generate the music while the drafts render.
5. Lay the stills and drafts on the timeline in your chosen order, cutting to the beat.
6. Watch it once without sound. If it works silently, the visual order is right.
7. Re-render the keeper clips at 1080p (95) and swap them in.
8. Export at the ratio you need — 9:16 for vertical feeds, 16:9 for wide. Framing for the target beats cropping afterward; the vertical video guide covers why.

Sizing and export

Generate and export at the aspect ratio you will actually publish in. A 16:9 video cropped to 9:16 throws away a third of the frame and usually the composition with it. The aspect ratio calculator has the dimensions, and social media sizes has the current per-platform matrix.

One limit to plan around

Photographs of identifiable real people cannot be animated — image-to-video refuses them. For a family or event video this matters: those photos can appear as stills, and everything else in the piece can move around them. Generated scenes, products, and characters can carry the motion instead.

Other ways to get from an idea to a finished video: How to make a cinemagraph with AI, How to make a book trailer with AI, animating AI art, How to make an animated poster with AI, and How to make an explainer video with AI.

Frequently Asked Questions

How many photos should a video use?

Twelve to twenty for a piece under two minutes. The instinct is to include everything, and it is the main reason photo videos drag. If a photo does not do a job — establish, advance, or land the ending — cut it.

Can I animate every photo in the set?

You can, but it is worse. Constant motion has no contrast, so nothing stands out and everything feels restless. One moving shot for every two or three stills gives the moving ones their impact back.

What music works for a photo video?

Something with a clear shape — a build and a resolution — rather than a loop, so the piece has an ending rather than a fade-out. Generating an original track for 25 credits means it fits your cut; our music generator guide covers what to ask for.

Can I make a video from photos of my family?

Yes as stills, no as animation — image-to-video refuses photographs of identifiable real people. Build the piece with those photos as stills and let generated or non-person imagery carry the motion.

What is the right export size?

Match the destination: 9:16 (1080×1920) for vertical feeds, 16:9 (1920×1080) for wide players. Decide before you edit rather than after, because framing for the target is what makes a composition survive.

Start with five photos

Pick five images, animate one of them, generate a track, and cut it to thirty seconds. That small version teaches the whole method, and it is finishable in an evening. A free account includes 150 one-time credits and no card, which covers the drafts and the music.