Fixes
Almost every problem with a render has one of five causes: the photo, the moderation check, the account balance, the queue, or an expectation the template was never going to meet. This page goes through them in that order.
The photo is the only input you control, so it is the first thing to check when a render is refused or looks wrong. One adult, one person in frame, front-facing, even lighting, no sunglasses and no heavy filter. Full body or half body beats a head-only crop.
If the photo is fine, the next filter is moderation. An image that trips the automated check is rejected before any render starts, and the rejection sticks to that image.
Automated moderation flagged the image, or the service could not fetch it. A rejection is stored against that image URL, so re-uploading the same file will hit the same verdict.
Fix: Use a different photo of one adult, with no nudity, no minor and nothing that looks like sensitive content. Re-upload the new file rather than retrying the old one.
A render reserves 120 credits up front. The balance can also change between the page load and the submit if a task just started.
Fix: Check the balance shown in the header. Buy a one-time pack, or wait for a pending refund to land if a task just failed.
The generation provider did not return a clip: a queue rejection, a timeout, or an input it could not process.
Fix: Retry once with the same photo. If it fails again, change the photo — a clearer full-body image resolves most repeats.
Video models queue under load. The page polls while it waits, but the provider decides when the task runs.
Fix: Leave the tab open. A task that never returns is marked as failed and refunded; starting a parallel task is blocked while one is active.
The model is reading a low-information photo: a very dark image, a head-only crop, a heavy filter, or the subject turned away.
Fix: Use a recent, front-facing, evenly lit photo with the whole body or half body visible. Avoid sunglasses and beauty filters.
Those are fixed by the template. It is about 12 seconds, keeps the original knock audio, and uses one prepared living-room scene.
Fix: Film your own version instead. A different room, song or length is outside what this template does, and buying more credits will not change it.
The result was still being copied from the provider to permanent storage when the page returned it.
Fix: Reload the history page and download again. Results are kept after the copy completes; the history entry is the reliable source.
One active generation per account at a time.
Fix: Wait for the current task to reach a finished or failed state, then start the next one.
Credits are reserved when a paid task starts, not when the page loads. If the provider fails to return a clip, the reserved credits are refunded to the balance automatically, and the refund is keyed to the task so it cannot be applied twice.
A render that never finishes is also settled: the task is marked as failed and the reservation is returned. That is why a retry does not cost you twice. The credit breakdown explains what one render costs and what a refund restores.
For anything the fix list above does not cover, the FAQ answers the photo, credit and publishing questions in one place.
Retry once after a provider failure. If the second attempt fails the same way, the problem is the input, so change the photo rather than the button.
If the clip works but is not what you wanted — different length, different song, different room — stop. Those are not faults; they are the boundaries of a prepared template. The comparison page covers the routes that do give you that control.
Email support@toctocdance.app with the account email, the approximate time of the attempt, and the task reference if the history page shows one. That is enough to check the generation record and the credit ledger.
For a rejected photo, say which photo and roughly when you uploaded it. Do not resend the image itself unless support asks.
The automated moderation service flagged it. Common reasons are nudity or sexual content, a minor in the frame, or an image the service could not fetch. Try a different, non-sensitive photo of one adult.
Each render reserves 120 credits before the task starts. If your balance is lower, top up with a one-time pack or wait for a pending refund to land.
The provider did not return a clip, which can happen under load or on an input it cannot handle. The reserved credits are refunded automatically, so retrying does not cost you twice.
Video models queue. Most renders finish in a few minutes, but peak hours stretch that. A task with no result is eventually marked as timed out and refunded.
The model carries the face, outfit and silhouette as far as it can read them, but motion rendering can drift. Better input photos reduce this; they do not eliminate it.
Only one generation can be active per account at a time. Wait for the current task to finish, fail or time out.
Contact support@toctocdance.app with the task reference and the approximate time. Refunds are keyed to the task, so the reference is enough to check the ledger.
Ready to make the clip from one photo? The prepared template carries the move and the original knock audio.
Go back to the generatorWhat it is
What is this dance, this song, this name?
How to make it
How do I get a clip, with or without filming?
Free & pricing
Can I do this without paying, and what does it cost?
Compare
Template, general model, another app or filming it myself?
Troubleshooting
Why did the render, the photo or the credits go wrong?