The best photo for a Bad Times AI dog or cat video shows one whole pet, sharp and well lit, facing the camera, with all four legs visible against a plain background. Bad Times Dance AI turns that single photo into a 15 second video in which your pet stands up, raises both paws and argues with the camera, so the photo decides almost everything you get back.
Key Takeaways
- Use one clear photo with the whole animal in frame. The model needs to see legs, paws and tail to build a believable standing pose.
- Choose Dog or Cat for one pet and Pet Duo for two. Both pets must be fully visible in the same photo, not overlapping.
- Coat colour, markings, ears and any outfit carry over. Very fluffy coats can lose their shape, so sharpness matters more than anything.
- If the result is wrong, change the photo first, then the tier (Pro or Ultra), then render again. Failed renders return credits.
Why the Photo Matters More for Pets
The pet version is the one that made the Bad Times AI trend spread. The original take was a French TikTok creator's close-up lip-sync, first posted in late August 2026, and the AI remakes began when pet owners swapped the performer for their own cats and dogs. The song is "Bad Times" by Imael Angel; the site you are reading is an independent fan-made tool and is not affiliated with the artist, the label or TikTok.
A human photo is forgiving. We all stand upright, so the model already knows roughly where a person's arms, legs and shoulders go. A pet is different. The dance asks a dog or cat to rise onto its hind legs, lift both front paws, wave one off to the side, push a paw toward the lens and finish nose to camera. The video model has to invent a standing pose from a photo that probably shows the animal on all fours, and everything it needs to do that comes from your image: leg length, body proportions, ear shape, coat texture, markings.
So the rule is simple. A photo that gives the model more truthful information about the animal produces a more believable animal. A photo that hides the legs, blurs the coat or crowds the frame forces the model to guess, and its guesses are where the odd results come from.
One Pet or Two: Choosing Dog or Cat or Pet Duo
Bad Times Dance AI has two pet dances, and they are not interchangeable.
Dog or Cat is for a single animal. It runs 15 seconds, follows the same reference motion as the Original, and has the pet stand on its hind legs in a warm kitchen with a tiled floor, the phone at the pet's eye level. It is the better choice when you want the face to carry the close-up, because the entire frame belongs to one animal.
Pet Duo is for two animals in one photo. It also runs 15 seconds and places both pets side by side on their hind legs on a clean light grey studio backdrop, pawing at the camera in sync. It works with two dogs, two cats or one of each, and it keeps their markings apart. Each pet keeps its own coat and size, though animals of very different sizes may be scaled closer together in the dance.
A few practical points follow from that.
- Two separate photos cannot be combined. If you want both pets in the video, they need to be in the same picture.
- Two pets mean smaller faces. If your two animals look nothing like each other, Pet Duo is a delight. If one of them is the star, use Dog or Cat with a photo of that pet alone.
- Pet Duo is built for two animals. If a household photo has a third animal in the background, crop it out or take a new picture.
Framing: Full Body Beats Face Only
The most useful single instruction is to show the whole pet. The Dog or Cat template works best when the photo includes the legs and tail, because the model needs them to make the standing pose believable.
That rules out the most common phone photo of a pet: a close-up of the face. A face-only picture is lovely, but nothing in it tells the model how long the legs are or whether the paws are white, black or spotted. The video will invent them. Sometimes the invention is fine. Sometimes it is not, and one of the typical problems in AI pet videos is a paw that turns into a human hand, which is far more likely when there are no real paws in the photo to copy.
A good frame has these qualities.
- The whole animal is inside the picture, with a little space around the ears and paws.
- The camera is at the pet's eye level or slightly above, not looking steeply down from a standing human's height, which shortens the body.
- The pet is facing the camera or at a gentle three-quarter angle. A strict profile hides half the face and gives the end close-up less to work with.
- The front paws are visible and not tucked under the body. Sitting upright with paws showing is excellent, since it is already halfway to the dance pose.
Coats, Markings and Fluff
Coat information carries over very well when the picture is sharp. Clear markings and coat colour are copied faithfully: a white blaze on a dog's chest, tabby stripes, a black ear, a ginger tail. Distinctive patterns are an advantage, since the model has a lot of detail to hold onto from one second to the next.
Fluff is the exception. Very fluffy coats can lose their shape, because soft edges blur into the background in the source photo and give the model no clean outline. If your pet is a long-haired cat or a heavy-coated dog, pick the sharpest photo you have, ideally taken in good daylight with the pet standing out from the background.
Dark pets need a little extra care. A black dog on a dark sofa in a dim room becomes a silhouette, and details such as the eyes, the nose and the paw pads get lost. Light the animal, or photograph it against something lighter, so the edges and the face show.
Pets of different colours help in Pet Duo. One dark and one light pet stay distinct more reliably than two identical ones. If your two cats are the same shade of grey and the same size, try to find a photo in which something sets them apart, a collar, a bandana or a different pose.
Outfits, Accessories and Backgrounds
Outfits in the photo carry into the video. A sweater or bandana stays on during the dance, which is one of the reasons Pet Duo is so popular for matching costumes. One of the example clips on the site shows two corgis in outfits doing the dance.
For outfits, keep these ideas in mind.
- Snug, simple outfits work best. A sweater or bandana is easy for the model to hold. Loose capes, long ribbons and elaborate hats are harder.
- Show the outfit clearly. If it is barely visible in the photo, it may disappear.
Backgrounds matter because a busy room can distract the model. A cluttered living room, with chair legs, a patterned rug and a second animal in the back, gives the model more things to mistake for part of your pet. The kitchen set of the Dog or Cat template and the grey studio of Pet Duo are both clean on purpose, so a plain wall, a clear floor or a simple lawn in your photo sits well beside them.
Do not worry about matching the template's background. The scene comes from the dance, not from the photo, so your sofa will not appear in the video. What you want from the background is simply to stay out of the way.
Lighting and Sharpness
Daylight from a window is the best light you can give a pet photo. It shows coat colour accurately, lights the face evenly and makes the edge between animal and background obvious. Avoid harsh flash, which flattens the coat, makes eyes glow and washes out markings.
Sharpness is the second ingredient. Pets move, and a slightly blurred photo is the usual result. Check your picture at full size before you upload. If the ears, the nose and the paws look soft on your screen, they will be soft in the video, and the end close-up (about the last four seconds of the 15 second dance) will magnify the problem. A fast shutter, a treat held beside the camera and a few extra attempts will give you a better photo than any later fix.
Finally, avoid filters, beauty effects and heavy cropping. They change the pet's colours and outline before the model sees them.
What Fails and What to Do About It
Most pet problems fall into a few buckets, and none is solved by writing a prompt, because Bad Times Dance AI builds the prompt for you.
Human hands on paws. The most notable pet failure. The template tells the model to keep paws, pads and claws, but it still happens on a small share of renders. Render again, because each render is a fresh draw. If it persists, use a photo with the front paws plainly visible, then try Pro or Ultra.
The pet's face changes. This is the pet version of face drift. Use a sharper, brighter, front-facing photo. A small pet in a large frame is a common cause, so crop or retake so the animal fills most of the picture.
Two pets merge. Overlapping bodies fuse. Photograph the pair side by side with air between them, or choose a different photo.
The standing pose looks odd. If the animal is photographed lying down, curled up or from above, the model has a harder job. A sitting or standing pose from the front helps.
Wrong number of animals. A second pet half visible at the edge of the frame may appear in the video. Crop it or retake.
Each of these has the same three levers: a different photo, a different dance or a higher tier, or another render. For the full list of nine failure types, see 9 ways a Bad Times AI video goes wrong.
Choosing a Tier and Counting Credits
The default is the Standard model at 720p, which costs 10 credits per video. That is the whole price of a Single pack ($6.99 for 10 credits). For pets, think of the tiers like this.
- Standard (Seedance 2.0 Mini) costs 5 credits at 480p and 10 at 720p. Good for a first test of a photo.
- Fast (Seedance 2.0 Fast) costs 8 credits at 480p and 17 at 720p.
- Pro (Seedance 2.0) costs 10 credits at 480p, 20 at 720p and 50 at 1080p. A sensible step up when paws or faces keep going wrong.
- Ultra (Seedance 2.5) costs 17 credits at 480p, 37 at 720p and 93 at 1080p. Keep it for the video you intend to post.
The Duo Pack gives 50 credits for $29 and the Crew Pack gives 250 credits for $99. Credit packs never expire and are not refundable. If a render fails, its credits come back automatically.
A sensible routine for a new photo is to render once at Standard 720p. If the paws and face are right, you are done. If not, adjust the photo, and only then spend on a higher tier.
Safety and Practical Rules
Your own pets are fine. What is not allowed is a photo of any person who is a celebrity or public figure, a minor, or anyone who has not agreed. Everyone in the photo must be 18 or older. If you are posing with your pet, take the photo in a way that either includes only you or crops the rest of the household out. Your uploaded photo is deleted within 24 hours, and the finished video can be downloaded for 24 hours, so save the MP4 as soon as it is ready.
The file arrives vertical at 9:16 by default with the original song audio mixed in, ready for TikTok, Reels or Shorts. The next step, including how to label AI content, is covered in how to post your Bad Times AI video. When you are ready to try, open the generator and pick the Dog or Cat dance.
FAQ
What kind of photo works best for the Bad Times AI dog or cat video?
A sharp, evenly lit photo that shows the whole pet, legs and tail included, facing the camera against a plain background. Bad Times Dance AI uses the photo to build the standing pose, so visible paws and clear markings give the most faithful 15 second video.
Can I put two pets in one Bad Times AI video?
Yes. Bad Times Dance AI's Pet Duo dance takes one photo with two pets, dogs, cats or one of each, standing side by side without overlapping. Two separate photos cannot be combined, and pets with different coats stay distinct more reliably than two identical ones.
Will my pet's outfit appear in the Bad Times AI video?
Usually yes. Bad Times Dance AI carries outfits from the photo into the video, so a sweater or bandana stays on during the dance. Simple, snug outfits that are clearly visible in the photo work best, while loose capes or barely visible accessories may disappear.
Why does my dog or cat get human hands in the video?
It is the most common failure in AI pet videos. Bad Times Dance AI's Dog or Cat dance tells the model to keep paws, pads and claws, but it can still happen. Render again, use a photo where the paws are clearly visible, or try the Pro or Ultra tier.
Do fluffy pets work in the Bad Times AI trend?
They can, but very fluffy coats may lose their shape. Bad Times Dance AI copies markings and colour well from a sharp photo, so choose a bright, in-focus picture where the coat stands out from the background rather than blending into a sofa or blanket.
Sources
- Songfacts: Bad Times by Imael Angel
- AI Musicpreneur: five AI acts on the iTunes Top 20
- Official Charts: Imael Angel
- AI Miracle: Bad Times AI trend guide