how to make a brat-style video

why screenshotting does not work

The usual approach is to open a generator, type a word, screenshot it, type the next word, screenshot again. It works in the sense that you end up with files. It fails on every detail that makes the finished video look intentional.

Screenshots capture whatever the browser was showing: your status bar, the address bar, a slice of page background. Crop those out by hand and no two crops land in the same place, so the green rectangle shifts a few pixels every frame and the clip flickers at the edges. The aspect ratio is wrong too, because a phone screenshot is your screen's shape, not the canvas shape, so the editor either letterboxes it or zooms in and eats the text. Resolution is whatever your screen happened to be, which is usually less than the platform wants.

The worst part is the text itself. A preview box resizes to fit whatever is in it, so a word can sit at one size in frame three and a visibly different size in frame four. Exporting properly removes all five problems at once: every file comes out the same dimensions, the same colour, at full resolution, with no interface in the shot.

step one, plan the words

Before you touch the generator, write the line out somewhere plain and break it into the chunks you want on screen. One to three words per chunk. Each chunk becomes one frame, so this list is your shot list and your frame count at the same time.

Keep the chunks short. A viewer scrolling past reads a one-word or two-word frame instantly, without deciding to. A longer frame requires a decision, and at a third of a second that decision does not happen, so the text goes by unread. If a phrase only makes sense whole, split it across two frames rather than cramming it into one.

step two, export the frames

Set the size preset to story / reels 9:16, which is 1080x1920, before you type anything. That is the native shape of a vertical feed, so the frames fill a phone screen with nothing cropped or padded.

1. 1

   Open the generator and choose the story preset. Set your colour and blur once, now, not per frame.

2. 2

   Type the first chunk, download it, then replace the text with the second chunk and download again. Nothing else changes between frames.

3. 3

   Download them in order, so the editor sorts them correctly on import.

Two practical notes. First, the shareable link stores the whole design, so if you get interrupted or want to redo a frame tomorrow, open the link and every setting comes back as it was. That is what keeps the last frame identical to the first.

Second, the text size slider. Left on auto, a one-word frame renders big and a three-word frame renders smaller, because the text is fitting the same canvas. That difference is the natural rhythm of the effect and most people want it. If you want every frame at exactly the same size instead, set the text size manually and leave it alone.

step three, assemble the clip

CapCut is the common choice: free, and already on the phone the files are sitting on.

1. 1

   Start a project and import all the frames at once, in order.

2. 2

   Change the default photo duration in settings before you place them. The default runs seconds long, and fixing it once beats trimming every clip by hand.

3. 3

   Drop the whole batch onto the timeline, add your audio underneath, then nudge the cut points until the words land on the beat.

Premiere, DaVinci Resolve, and the editor built into your phone do the same job with the same steps.

how long each frame should hold

At a typical pop tempo each word lands somewhere between a fifth and a third of a second. Start there, play it back, and adjust by ear. Arithmetic from the tempo gives a number that is technically right and usually feels wrong, because sung words do not sit evenly on a grid.

step four, cut to the beat

Listen to the section once with the timeline open and drop a marker every time a word you have a frame for is sung. Most editors have a beat marker key for exactly this. Then snap each cut point to its marker. Marking first is much faster than dragging clips a few frames at a time and playing it back after every nudge.

Do not add transitions. No crossfades, no dissolves, no whip pans. The hard cut is the effect: one flat green frame is replaced instantly by the next, and any blend between them turns a sharp edit into mush. For the same reason, skip zoom, shake, and every motion preset. A frame that does not move is what makes the change land.

the details that make it look right

Keep the same green on every frame. The lime-green cover look uses #8ACE00, and a frame one shade off reads as a glitch mid-clip. Keep the blur identical for the same reason, since a sharp frame among soft ones jumps out. Keep everything lowercase, including names and the word starting each line. And do not mix aspect ratios: a frame from the square album preset at 2048x2048 or an instagram post at 1080x1080 will not match the 1080x1920 frames, and the editor will either bar it or crop it.

Export the finished video at the shape you built it in, 1080x1920, so the platform does not re-scale it. Frame rate barely matters for flat colour and hard cuts, so 30 fps is plenty. 24 fps works, and 60 fps gives you nothing but a bigger file, since nothing on screen moves.

about the audio

If you add the track you are using from inside TikTok or Instagram, you are covered, because those platforms license the music in their libraries. That licence does not travel with the file. The same video exported and posted to your own site, a client's channel, or anywhere with ads attached is a different situation, so for anything beyond a normal social post, use audio you have the rights to.