Bring Old Photos to Life with Seedance 2.0: A Complete Guide

A photo, a sealed memory.

>

Now, AI can make the people in old photos blink, curl their lips, let the wind ruffle their clothes, and set the still background back into motion.

>

This isn't science fiction — it's one of the most captivating applications of Seedance 2.0.

---

1. What is Seedance 2.0?

Seedance 2.0 is the next-generation AI video generation model from the Doubao team, specializing in generating coherent video from a single image.

Unlike simple "image shaking" or "panning effects" — Seedance 2.0 truly understands what's in the image: it recognizes faces, hair, leaves, flowing water, and generates motion that follows real-world physics.

In plain English: You give it an old photo. It gives you a video that might just make you emotional.

Tomato AI has already integrated Seedance 2.0 — just open the website and start creating. No coding required.

---

2. How Does "Photo Revival" Work? (In Simple Terms)

Why can a static photo "come to life"?

Think of it this way: Seedance 2.0 has been trained on massive amounts of real video footage. It has learned "how still scenes typically move in reality."

• A photo of grandpa sitting in the yard → The model understands "people have subtle breathing movements" and "wind makes leaves sway"
• A group photo at an old train station → The model knows "people on the platform make small movements" and "clouds drift slowly in the distance"
• A candid childhood shot → The model knows "clothes flutter in the wind" and "hair has natural movement"

The magic of "photo revival" is AI's "reasonable imagination" after understanding the scene — not adding things that don't exist, but "awakening" elements already present in the photo according to real-world logic.

---

3. Choosing the Right Old Photo

Photo Type	Effect	Notes
✅ Clear frontal/side face shot	⭐⭐⭐⭐⭐	Best results, ideal for character revival
✅ Scenes with wind (clothes, hair, plants)	⭐⭐⭐⭐⭐	Natural motion, strong visual impact
✅ Water, smoke, clouds, etc.	⭐⭐⭐⭐	Rich dynamic effects
✅ Group photos	⭐⭐⭐⭐	Each person adds to the story
❌ Blurry or over/under-exposed photos	⭐	AI can't understand the content well

Tip: If your old photo is damaged, faded, or blurry, try restoring it first with a photo editing app before using Seedance 2.0 — the results will be much better.

---

4. Three Steps to Animate Your Old Photo

Step 1: Open Tomato AI

Go to the Tomato AI website and enter "Image to Video" or "Seedance 2.0" mode.

Step 2: Upload Your Old Photo

Upload your scanned or photographed old photo. Supported formats include JPG, PNG, WEBP. Max file size is 20MB.

Step 3: Write Your Prompt

This is the most critical step. Your prompt is the "director's instruction" — telling AI how the scene should move, in which direction, and how fast.

The "Old Photo Revival" prompt formula:

[Scene atmosphere] + [Character micro-movements] + [Environmental changes] + [Camera feel] + [Overall style]

A complete example:

"A faded old photo, an elderly man in a Zhongshan suit sitting in a traditional courtyard. He slowly blinks, the corners of his mouth lift slightly, his shoulders rise and fall with each breath. Wind passes through, plane trees rustle, light and shadow shift gently on the ground. The scene is warm and nostalgic, like a forgotten home video re-emerging from memory."

Write your prompt, click generate. Wait a few dozen seconds, and your "revived" video appears.

---

5. Ready-to-Use Prompt Templates

Don't want to start from scratch? Here are templates you can use right away.

📸 Template: Family Elders

A faded old photo, [who] sitting/standing [where].
They slowly blink, take a breath, and show a gentle smile.
The wind gently blows [clothes/hair/surroundings].
The scene feels warm and nostalgic, like a memory.

Filled in:

A faded old photo, grandpa sitting in his favorite rattan chair at home.
He slowly blinks, takes a breath, and shows a gentle smile.
The wind gently blows the collar of his shirt and the leaves of the potted plant beside him.
The scene feels warm and nostalgic, like a memory.

📸 Template: Old Street / Building

A faded old photo of [place].
Leaves sway gently in the wind, light and shadow flow on the ground,
someone seems to be walking slowly in the distance.
The scene has a warm sepia tone, as if traveling back in time.

Filled in:

A faded old photo of the historic Bund in Shanghai.
Plane tree leaves sway gently in the wind, light and shadow flow on the ground,
a pedestrian seems to be walking slowly with an umbrella in the distance.
The scene has a warm sepia tone, as if traveling back to that era.

📸 Template: Childhood Photo

A faded old photo, [who] at [place].
They blink curiously, tilt their head slightly, showing an innocent smile.
[Plants/things] by the door sway in the breeze,
warm afternoon sunlight pours down, the scene is precious and tender.

Filled in:

A faded old photo, young me squatting on the steps at the old house's entrance.
I blink curiously, tilt my head slightly, showing an innocent smile.
Morning glory vines by the door sway in the breeze,
warm afternoon sunlight pours down, the scene is precious and tender.

📸 Template: Family Group Photo

A faded old photo, a family at [scene].
Each person has the subtlest movements —
someone blinks, someone's mouth twitches, someone adjusts their stance.
The wind blows through [background], the scene warm and precious, like that afternoon in memory.

Filled in:

A faded old photo, a family of five standing on the drying ground in front of the old house.
Each person has the subtlest movements —
father slightly adjusts his stance, mother gently purses her lips, the children blink curiously.
The wind blows through the rice paddies in the distance, the scene warm and precious, like that afternoon in memory.

---

6. Tips for Better Results

1. How Long Should the Video Be?

Photo Type	Recommended Duration	Reason
Single person	5–7 seconds	Subtle movements, short time is enough
Group photo	7–9 seconds	Need time for everyone to show movement
Street/architecture	9–11 seconds	Environmental motion is slow, needs more time
Nature/landscape	11–15 seconds	Clouds, water, light need time to unfold

2. Choosing Aspect Ratio

Most old photos are nearly square (6×6 camera) or 4:3. If unsure, select "Smart Fit" and let AI recommend automatically.

3. Audio On or Off?

Recommended: On. Subtle sounds of wind, breathing, and leaves instantly create immersion. It won't feel like AI-generated content — it will feel like real family footage.

4. Want More "Cinematic Feel"?

Add these keywords to your prompt:

Desired Effect	Add These Words
Cinematic	Cinematic quality, shallow depth of field, soft visuals
Nostalgic	8mm film grain, warm tones, faded effect
Realistic	Hyper-realistic, rich details

5. Close-up Not Working Well?

Shorten the duration to 4–5 seconds and add "keep facial features stable" to your prompt.

---

7. A Real Story: From One Photo to a Memory

Let me share a true story.

Xiao Wang found an old photo of his grandmother when she was young — wearing a floral dress, standing in front of the old house, smiling beautifully. The photo had yellowed, its corners creased.

He took a photo with his phone, did some basic restoration, opened Tomato AI, uploaded the photo, and wrote this prompt:

"A faded old photo, a young girl in a floral dress standing before an old house, with a gentle smile. She slowly blinks, the wind blows her dress and the loose strands of hair on her forehead, the locust tree leaves behind her sway gently in the sunlight. The camera slowly zooms in, the scene warm and beautiful, as if that summer never really ended."

A few dozen seconds later, the video was generated.

Grandma blinked in the video — just that one moment, and Xiao Wang's mother next to him burst into tears.

That's the most moving thing about Seedance 2.0. It's not just "making videos" — it's letting those frozen moments in time flow again, if only for a little while.

---

Open Tomato AI, pick a treasured old photo from your family album, and see what it looks like when it "comes alive."