Troubleshooting
Bad Times AI Face Consistency
Bad Times AI: Prepare a clear photo and review faces, hands and movement in the finished video. AI appearances can vary.
Written and maintained by Bad Times AI · Updated
Why can a Bad Times AI face change?
AI synthesizes new frames from photo identity information and a motion reference. Fast head turns, occlusion, limited detail and complex group arrangements can make consistency harder. The generated video can differ from the uploaded photo even when the prompt requests preservation. This variability should be considered before purchasing; an example or reference preview is not a promise of identical features in every frame.
What should I inspect?
Check the eyes, nose, mouth and hairstyle across the full clip, including turns and moments when hands cross a face. In a group, check that the same identities remain assigned to the same subjects throughout. Review hands, clothing and extra limbs as well as faces. Listen to the complete soundtrack. Keep the original task link so a problem can be discussed using a task ID rather than a vague description.
Does an imperfect result count as failure?
A provider-confirmed failed task and an imperfect completed video are different states. Confirmed generation failures restore the attempt’s allowance for a retry. An aesthetic concern about a successful video is not automatically a confirmed system failure or a cash refund. Contact support with the existing task ID if the result needs review, and check the published terms before submitting another paid attempt.
Current Bad Times AI creation rules
Bad Times AI accepts one JPG, PNG or WebP photo up to 10 MB. The five templates span 13.09–15 seconds and request a 480p, 9:16 portrait MP4. The free-generation allowance is 0 videos. Guest checkout needs no account. Guest packs: 1 video $7.99, 3 videos $19.90, or 8 videos $49.90. Signed-in credit packs: 50 credits $9.90, 105 credits $19.90, or 160 credits $29.90. Each 480p video uses 20 credits. One-time payment; purchased credits never expire. Photos are processed by the configured AI provider when generation is submitted, and identities and motion can vary. All five current templates are available for selection and generation. These product facts describe the current creation flow, not capabilities inferred from older template code.