Convert video to vertical, square or horizontal
TubePull Video Aspect Ratio Converter reshapes a video for the screen or app where you plan to use it. Choose vertical 9:16, square 1:1 or horizontal 16:9, decide what goes in the space the original picture does not cover, look at the preview and press Start. You get an MP4 with the full length and the sound of the original. Your device does the conversion in this tab, so the footage stays with you and is never uploaded. There is no price, no account to make and no watermark stamped on the output.
Three shapes, two size tiers
| Shape | Output if the shorter side is 720 px or more | Output for smaller videos |
|---|---|---|
| Vertical (9:16), selected when the page opens | 1080 × 1920 | 720 × 1280 |
| Square (1:1) | 1080 × 1080 | 720 × 720 |
| Horizontal (16:9) | 1920 × 1080 | 1280 × 720 |
The size tier depends on the shorter side of the picture as it is shown, after any rotation flag. A large source converted to vertical becomes 1080 × 1920, the square choice becomes 1080 × 1080, and the horizontal choice becomes 1920 × 1080. Smaller sources use the 720-pixel tier. Phone videos stored sideways with a rotation flag are read upright, and the new MP4 has no rotation flag of its own.
YouTube's help center, for one, treats an upload of up to three minutes with a square or vertical shape as a Short. The horizontal choice produces a standard widescreen video instead. TikTok, Instagram and other services set their own recommendations and revise them, so look up the current requirements in the app where you are posting.
Three ways to fill the frame
When the source and output have different proportions, the picture does not naturally cover the new frame. The Fill setting decides what happens to the rest.
- Blurred background (the default). The whole picture is scaled until it touches two sides of the frame, then centered. Behind it, a second copy of the same footage is enlarged to cover the frame and blurred. The large output tier uses a 40-pixel blur and the smaller tier uses 27 pixels. The bands move and change color with the video, and no part of the original is cut off.
- Solid color. The picture sits in the same place, and the bands are one flat color. A color picker appears when you pick this mode. It starts on black and takes any color it can show.
- Crop to fill. The picture is enlarged until no gap is left, and whatever sticks out is cut away. A slider from 0 to 100 picks the part that stays. At 50, where it starts, the middle is kept. The edge labels show whether the slider moves left to right or top to bottom for the current source and output shape.
Here is the difference on a 1920 × 1080 video turned into 1080 × 1920. With a blurred background or a solid color, the picture becomes a 1080 × 608 strip across the middle, with 656 pixels of fill above it and below it. With Crop to fill, it is enlarged to 3413 × 1920, and only a 1080-pixel-wide window survives, about a third of its width. Made square instead, the same video keeps that 1080 × 608 strip with 236 pixels of fill above and below. Choosing horizontal keeps the source's 16:9 shape and places it on an exact 1920 × 1080 canvas.
Crop to fill enlarges the picture more than the other two modes, so it can look softer when the source and output shapes differ sharply. Its position is fixed for the whole video and does not follow a person who moves.
A preview before you convert
Once the file has been read, a preview shows your shape and fill on one still frame, taken a tenth of the way into the video. Changing the shape or the fill, moving the slider or picking another color redraws it at once, so every option can be tried before anything is encoded. A line under it states the output size in pixels. Until the frame arrives, or if none can be taken from the file, gray boxes mark where the video will go, so the layout is still visible, and Start does not wait for the frame.
When the video already has the shape you picked, to within 1%, a note says so. Converting it is still allowed, which is useful for producing an exact 1080 × 1920, 1080 × 1080 or 1920 × 1080 file. The shape, fill, color and crop position you set are kept when you open the next video.
What the MP4 contains
- Picture: H.264 at the new size, encoded again from the first video track. Every frame is kept, each at its original moment, so the video runs as long as the source. Like every re-encode, it costs a little detail.
- Sound: the first audio track. AAC sound is copied unchanged. MP3, Opus and any other codec become AAC at 128 kbps, with the source's sample rate and number of channels, so a mono recording stays mono. When the source is silent, the MP4 is silent as well, without any audio track.
- Left out: additional video and audio tracks, subtitles and chapters.
- Tags: the title and similar tags of the source come along, and they can include where it was filmed. The Video Metadata Remover can strip them later.
- Name: your file name plus "-vertical" for 9:16, "-square" for 1:1 or "-horizontal" for 16:9. For example, trip.mov becomes trip-vertical.mp4, trip-square.mp4 or trip-horizontal.mp4.
The result then lists the output size in pixels, the original video codec with a note that it was re-encoded to H.264, and whether the sound was copied or converted to AAC.
Files the converter opens
Nine containers are accepted: MP4, MOV, MKV, WebM, MPEG-TS, 3GP, AVI, FLV and WMV. Because every frame is redrawn, the codec inside the file counts as much as the container.
Your browser decodes what it can. The compatibility engine covers H.264, HEVC, VP8 and VP9 as well as Windows Media video and the MPEG-4 Part 2 codecs Xvid and DivX. AV1 is only possible where the browser can decode it. Sound-only files cannot be used, even with album art, since there is nothing to reframe, and DRM-protected videos are refused as well. The converter takes a single video per job.
Where the conversion runs
Nothing about the video is sent to TubePull: not the footage, not its name, not the details the tool reads. Apart from loading the tool's own code, the page makes no network request while it converts.
Usually the browser does all of it. The open-source Mediabunny library reads the file, each frame is drawn onto the new canvas, and the browser's WebCodecs encoder writes the H.264. The compatibility engine, FFmpeg built for WebAssembly (ffmpeg.wasm), takes over if that route fails, if the browser cannot decode the video or encode the new size, and from the outset for AVI, FLV and WMV files. Getting it means a one-time fetch of around 32 MB, which your browser then holds in its cache. Mediabunny and FFmpeg each have their version, license and source link listed on the licenses page.
Both routes produce the same layout. Where a browser's canvas cannot apply a blur filter, the tool blurs the pixels itself, and the compatibility engine runs three box-blur passes at a matching radius, so the bands look much alike whichever route ran. Away from the picture, a solid color comes out very close to the one you chose. In the narrow strip right beside the picture, video compression can shift it more.
Limits
- The compatibility engine counts on an output about the size of the original and will not begin above roughly 1.5 GB. That applies to AVI, FLV and WMV files and to any job it takes over.
- Beyond that, your device's memory is what limits a long video. On a phone, a file above 500 MB triggers a caution, though you may go ahead.
- The available ratios are 9:16, 1:1 and 16:9 at the two size tiers in the table. There is no custom ratio, blur-strength setting or zoom control.
- The converter does not track faces, add captions or shorten the video. To keep only part of it, cut it first with the Video Trimmer. For a smaller file afterwards, use the Video Compressor.
Clips that are still on the web
Only files stored on your device can be opened. To reshape a clip that is still on YouTube or another site, download it with the TubePull downloader beforehand and pick the saved copy here. For a looping clip without sound, try Video to GIF, and for everything else, browse the free video tools.