Check One Check Two Photo & Outfit Tips
Choose a clear photo for your Check One Check Two AI video. Learn how framing, light and visible clothing affect face and outfit replacement.
For a Check One Check Two AI video, choose one sharp photo showing a single person's face, shoulders and the outfit you want the performer to wear. Keep the eyes and mouth visible. Your photo provides the appearance reference; the preset supplies the performance, crowd and scene. You do not need to photograph yourself at night or hold a microphone.
A practical photo checklist
| Check | Choose | Avoid |
|---|---|---|
| People | One clearly visible person | Several faces competing for attention |
| Face | A front-facing or near-front view | A strong side profile or hidden mouth |
| Light | Even light across the face | Deep shadows or blown-out facial details |
| Detail | An original, sharp photograph | A blurry thumbnail or compressed screenshot |
| Clothing | Visible neckline, shoulders and upper body | A face-only crop when the outfit matters |
| Editing | A natural-looking reference | Heavy beauty filters that obscure features |
The upload accepts JPG, PNG and WebP files up to 10 MB. A larger image is not automatically better: extra pixels cannot recover a blurred face. If the file is too large, export a smaller copy while checking that the facial details remain clear.
Why did the face change but the clothes stay similar?
The model can only reference clothing that is visible in your photo. A tightly cropped headshot gives it facial details but little information about the shirt, jacket or neckline. If clothing is important, use an upper-body portrait that clearly shows the outfit's main colors, shape and sleeves while keeping the face recognizable.
Avoid making the person so small that the face loses detail. A full-body photo with a distant face can be less useful than a clear upper-body image. The preset asks for face, hairstyle and outfit replacement, but fabric patterns, logos, accessories and exact garment details may still vary. Visible clothing improves the reference; it is not a guarantee of an exact copy.
Can I use a group photo?
Crop the image to the one person you want to use before uploading. Remove other faces, including partial faces at the edges, while retaining useful shoulder and clothing detail. Check the cropped image at normal viewing size rather than judging it only from a tiny preview.
The generator targets the main performer in the preset clip. It does not let you choose a supporting person or replace the entire crowd with separate uploaded identities. If you want a friend to be the main performer, use their photo with permission in a separate generation.
What should I change if the face is hard to recognize?
Check whether the reference has sunglasses, a hand over the mouth, motion blur, strong shadows or a heavy filter. Choose another photo where the main facial features are visible and the head is directed toward the camera. A clear input can reduce ambiguity, but exact likeness throughout an AI-generated performance is not guaranteed.
Change one thing at a time when comparing results, such as the lighting or the crop. Keep the original photo so you know which image produced each result. Each new generation uses one video from your allowance; uploading a different photo to the preview does not change an already completed video.
Can a different download ratio fix the composition?
If the person looks right but you want a different frame, try the existing result's download options before generating again. Download video keeps the original square frame. Download other ratios offers a centered landscape crop and a portrait crop positioned slightly left of center.
Cropping only removes edges. It cannot recover missing clothing, repair a face or reveal details outside the generated picture. Check that the chosen crop keeps the face and gestures visible. Downloading another available ratio does not use another generation allowance.
Before you submit
Open the file once to confirm it is readable, check the face and outfit, and use a photo you have permission to transform. Selecting the photo creates a local preview; submission uploads it for review and generation. For deletion requests, use the contact information in the Privacy Policy and include your task ID.
Continue with the step-by-step creation guide, or open the generator when your photo is ready. If the saved clip has a sound problem, use the audio and download checklist.