How to Turn a Photo Into a Video on iPhone

Animating a photo on an iPhone takes about a minute of work and a few minutes of waiting. The part that decides whether the result is good happens before you press generate — which photo you pick and how you describe the motion.

Here is the whole process, plus the Live Photo variant, which behaves differently enough to be worth its own section.

What you need

• An iPhone running iOS 16.6 or later.
• A photo with a clear subject — a portrait works best to start.
• An app that does photo-to-video generation. We make Rigel, so that is what the steps below describe; the structure is similar in most tools in this category.

Step 1 — Pick a photo

Choose a Live Photo, a portrait, or any shot from your camera roll. Then apply four filters before you commit to it:

• Sharp. Motion blur and soft focus in the source become mush in motion.
• Well lit. Even light beats dramatic light for a first attempt. Heavy backlighting is the most common reason a result disappoints.
• Subject large in frame. A face that occupies a small part of the image has little detail to animate. Crop in before you upload.
• Uncluttered. Crowds, busy reflections and readable text are the hard cases for every model on the market. Fewer of them means a cleaner clip.

If you are choosing between two photos, pick the sharper one, not the more interesting one. You can add interest with the prompt; you cannot add detail that was never captured.

Step 2 — Tell it the vibe

Type a line describing the scene, pick a mood, or add a script if you want the subject to speak with lip-sync.

The rule that matters: describe motion and light, not adjectives. Compare these two prompts for the same portrait:

• Weak: "cinematic, beautiful, 4K, amazing quality"
• Strong: "She turns slowly toward the camera and smiles; warm late-afternoon light from the left; hair moves slightly in the breeze"

The second one tells the model what happens, where the light comes from, and what moves. The first one tells it nothing it can act on.

Three things worth naming explicitly when they apply: the camera (static, slow push in, slow orbit), the environment (rain, dust, sunset, studio), and the emotional beat (she laughs, he looks away). Keep it to one or two sentences — longer prompts are not better prompts.

Step 3 — Hit generate

Generation takes a short while and produces a 5–30 second clip, vertical or square, ready for TikTok, Instagram Reels, YouTube Shorts and Stories. In Rigel, the credit cost of a generation is shown before you confirm it, so you always know what an attempt costs.

Watch the result once at full attention before deciding anything. Then check three specific things:

1. Does the face stay the same face for the whole clip?
2. Does the motion obey physics — hair, cloth, water, the way weight shifts?
3. Does the light stay consistent from the first second to the last?

If one of them fails, regenerate with a more specific prompt rather than a different photo. If two of them fail, the source image is usually the problem.

Step 4 — Save and post

The clip saves to your camera roll. From there it goes straight into whichever app you post from, or into an editor if you are cutting several clips together.

If you are building something longer than a single clip, that is a different craft: plan the beats, generate scene by scene, keep the character consistent. The full method is in our multi-scene guide and summarised on the long AI video page.

The Live Photo variant

Live Photos are a genuinely different starting point, because they already contain real motion — roughly a second and a half of frames captured around the shutter press.

That gives an animation a natural anchor: the way the person was actually moving, the actual direction of the light, the real micro-expression. Clips generated from a Live Photo tend to feel less invented than clips generated from a flat still, especially for candid moments.

To use one:

1. In the Photos app, confirm the shot is a Live Photo — the icon appears at the top left when you open it.
2. Pick it as your source the same way as any still. Rigel reads the motion.
3. Describe what should happen next, not what already happened. The captured motion is your first beat; the prompt extends it.

Details and examples are on the Live Photo to video page.

Common mistakes

• Uploading a screenshot of a photo instead of the photo. Compression artefacts get amplified by motion.
• Asking for too much in one clip. A location change, a costume change and a dialogue line in five seconds will produce a mess. One idea per generation.
• Judging a result at full volume on a laptop. Most of the audience sees it muted on a phone. Check it that way.
• Regenerating with the same prompt and hoping. Change one thing per attempt so you learn something from each.
• Using someone else's face without asking. Get consent. Every platform's rules and basic decency both point the same way.

FAQ

Does this work on older iPhones? The app requires iOS 16.6 or later. Generation happens on servers, so the phone's chip is not the limiting factor — the connection and the source photo matter more.

Can I make the person speak? Yes. Add a script and the clip is generated with lip-sync. The talking photo page covers what works and what does not.

How long does a generation take? Minutes, not seconds, depending on mode and load. Longer or higher-realism modes take longer.

Does it cost money? Rigel is free to download with starter credits included; credit packs and a subscription are available in-app. The credit cost of each generation is shown before you confirm.

What about old family photos? They work, and they are one of the most rewarding uses. Scan at the highest resolution you can and read the notes on animating old photos first.

Why did my result look wrong? In order of likelihood: soft or badly lit source image, a prompt made of adjectives, or an ask that involves hands, text or crowds. Fix them in that order.