Video trend
This is the how-to page for making the trend yourself, under the name "you fked the transmission". Short version: you give Oakgen one or two photos, it rebuilds the 15-second driveway take with your people in the two roles, and you check the result before posting. In the skit the line is about the transmission, the car part under the hood. What follows is the recipe in order, with the decisions that matter at each step. If you want the story of the joke first, the page titled You F'd the Transmission AI Trend covers it.
The template has two photo slots, and you may fill either or both. Add two photos if you have two people to cast. Add one if you only have one, and the other role keeps the character from the reference. Either way the clip is still the full 15 seconds, and the quote only includes preparation for the photos you actually add. One person per photo, as the studio asks. Use photos of yourself or of people who have agreed to appear.
The slot labelled Coat character is the left side of the frame: the person who leans over the engine and makes the accusation. The slot labelled Suit character is the right side: the person who denies it and then gives the long excuse. The right role has the most mouth movement, from about 8.7 to 12.6 seconds, with the face fully toward the camera, so your sharpest, most front-facing photo belongs there. The left role is often seen bent over or in profile, so a photo showing the hairline and head shape helps more than a tiny face.
You cannot rewrite the dialogue. The source soundtrack is part of the recipe, and your characters' mouths are made to follow it, so the wording, the pauses and the strong language come with it. That also means the timing is fixed: the diagnosis starts the clip, the transmission line lands at about 2.5 seconds, and the clip finishes on the left character's reaction near 15 seconds. Pick the output resolution, check the quote in the studio, and expect the result to keep that soundtrack.
Watch it three times. First with sound off, checking that the coat and suit are on the right sides and that no grid or mesh lines remain on either face. Second at about 2.5 seconds, where the left character should be turning from the hood to speak. Third during the long excuse, where the right character's lips should keep moving across all four seconds and both hands should spread. Look at the car too: the black car, its raised hood and the grey house with the barred window should stay put, because there are no cuts to hide drift.
If lip-sync drifts, try a clearer photo: a face that is turned away, hidden or small gives the tool less to work with. Swap it and run the video again. If a character reads as the wrong person, use a photo with plain, even light, so the hair and clothing are easy to carry over. Costumes may not match exactly. The reference has a bald figure and a long-haired figure, and results with different hair will look different, which is fine as long as the roles are still easy to tell apart.
Say that it is an AI video made from photos, name @ausar.ai as the reference reel if you show or mention it, and do not present the original sound as yours or as having a known creator, because we could not find one. A plain caption works: the car part in question is the transmission, and the joke is that one person blames the other for it. If you show the reference next to your version, label them separately so viewers can tell which is which.
Choose the roles you want to change. Add a photo for each selected role.
Add a clear photo for the coat character, the suit character or both. Use photos of yourself or of people who have agreed to appear. One character per photo.
Choose 480p, 720p or 1080p. The credit quote covers the complete 15-second edit and the preparation of only the photos you select.
Watch who speaks when, check both faces in the yard and review the hood scene at the start before downloading your portrait MP4 with sound.
The coat character leans over the engine, makes the accusation and delivers the deadpan closing question. The suit character stands in the yard, denies it and gives the long explanation. Your photos replace both, or one of them.
The handheld 15-second portrait clip has no cuts: a suburban yard, a grey clapboard house and a black car with its hood up. The camera and body motion follow the reference.
The final MP4 keeps the reference sound, so the diagnosis, the accusation, the denial, the long excuse and the closing question play out in order. The clip ends on that question, as the reference does.
The preview is the credited @ausar.ai reel, a fan-style parody and not an Oakgen result. We have not verified the original sound or its creator, and Oakgen is not affiliated with any franchise.
It is a one-take driveway skit where one character in a long black coat argues over an open car hood and another in a black suit gives a long explanation in the yard. The reference is a fan-style parody built around the Harry Potter villain. Oakgen lets your photos play both roles.
The credited reel's caption asks "Voldemort did what" and uses the Harry Potter tag, and the two figures are dressed in the style of characters from that franchise, but the clip is a parody made with AI. In Oakgen you replace both figures with your own characters.
It is the name people use for this skit, written here with a softened spelling and the car part spelled out. The clip's abbreviated wording is car slang for the transmission: the coated character spots the problem under the hood, and the twist comes from the second character reacting as if he had heard something else. The credited @ausar.ai reel is an AI parody, and we did not find the original skit or the creator of its audio. The accusation and the closing question both come from the coated character.
No. This is a fan-style parody format, not affiliated with or endorsed by the franchise or anyone connected with it. Your photos replace the reference characters.
Only use photos of yourself or of people who have agreed to appear. Do not upload private photos of others or anyone's likeness without permission, and label AI videos clearly when you share them.
Add one photo per role, up to two. A role left empty keeps the character from the reference. Both characters share the yard, so check both faces in the result.
Yes. The final MP4 keeps the reference dialogue, and that original soundtrack contains strong language, so it may not suit every audience or platform. We have not verified the original audio or its creator, so we do not label it with a guessed source. Generated lip sync can vary, so check the long speech and the closing question.
The output is 15 seconds in portrait 9:16 format and covers the whole reference take.
No. The yard, the car, the camera and the dialogue are fixed by this reference. Your choices are the photos and the output resolution.
No. It is the attributed @ausar.ai reel and a source reference. Generated faces, coats and car details can vary, so review your video before sharing.
3,173 credits per 720p video. This two-role price includes both photo preparations. Selecting fewer roles reduces preparation costs; the complete video is still edited. Review your selected quote before generating.
Plan a story, keep your character recognizable, and make the finished clip your own.
Review your finished Snape and Voldemort car skit video by timestamp, spot the photo-choice mistakes behind each flaw, and fix it with one change at a time.
Why the 15-second car-hood take works: its two long pauses, one handheld shot, what the left and right roles do physically, and which photos suit them.
Upload 1–2 photos. Everything else is set.