To convert a landscape video to vertical without cutting off the subject, you either crop the 16:9 frame to a 9:16 window that follows the speaker, fill the empty space above and below the original frame with a blur or a solid color, stack two crops on top of each other, or use an AI smart reframe tool that detects the face and moves the crop automatically. For talking-head footage, AI reframing produces the best result in the least time; for footage where the whole frame matters, blur fill or a stacked layout keeps everything visible.
The reason this matters is simple arithmetic. A 1920 by 1080 landscape frame cropped to full-height 9:16 keeps only 608 pixels of its 1920-pixel width, which is 31.6 percent of the picture. Anything outside that narrow window is gone, and the remaining 608 by 1080 crop has to be upscaled 1.78 times to reach 1080 by 1920. Every method below is a different answer to the question of which 31.6 percent to keep, or how to avoid the choice altogether.
This guide compares the four methods, walks through converting a talking-head clip to 9:16 with AI, and covers the export settings and caption safe zones that stop the platforms from covering your text. For the platform specs themselves, see video aspect ratios explained, Instagram Reels size and dimensions, TikTok video size, and YouTube Shorts length and size requirements.
Why 16:9 Footage Looks Wrong on TikTok, Reels, and Shorts
TikTok, Instagram Reels, and YouTube Shorts are all built around a full-screen 9:16 canvas of 1080 by 1920 pixels. The aspect ratio is simply the width-to-height proportion of the frame, and 9:16 is a 16:9 frame turned on its side. When you upload a 16:9 video unchanged, the platform letterboxes it into a thin horizontal strip in the middle of the phone, with black bars taking up roughly two-thirds of the screen. Viewers read that as a repost, faces become too small to make out, captions are unreadable, and the algorithm's completion-rate signals suffer because people scroll past.
The platforms do not penalize the aspect ratio directly, but they reward what a native vertical video delivers: a large face, big readable text, and something happening in the center of the screen. That is why converting is worth doing properly rather than just uploading the landscape file. The same logic applies to podcast recordings, webinar screen shares, and 4:3 archive footage, all of which need a vertical version to work as short-form content.
The Four Ways to Convert Landscape to Vertical (Compared)
Center crop is the fastest method: cut a fixed 9:16 window out of the middle of the frame. It works only when the subject never leaves the center, which is rare in real footage. Blur fill and solid bars keep the entire 16:9 frame visible by shrinking it and filling the space above and below; nothing is lost, but the actual content occupies only about a third of the screen. Stacked and split layouts place two or more crops on top of each other, for example the host in the top half and their guest or slides in the bottom half, which is the standard for podcast clips. AI smart reframe detects the subject, usually a face, and moves a 9:16 crop to follow them, so the result looks as if the video was shot vertically.
In practice, the choice depends on what is in the frame. One person talking calls for smart reframe. Two people or a person plus a screen calls for a stacked layout. Landscapes, sports, gameplay, and anything where the composition itself is the point call for blur fill or a reshoot. The sections below explain each method with the settings that matter.
Method 1: Center Crop (Fast, but It Cuts People Off)
A static center crop is what most free converters and the built-in crop tools in TikTok and CapCut do. You choose 9:16, the tool takes the middle 608 pixels of a 1080p frame, and everything else is discarded. The output is genuinely full-screen, which is its one advantage, and it takes seconds.
The problem is that real subjects move. A presenter who steps to the side of the frame for two seconds disappears from the crop entirely, an interview shot with two people on opposite sides of the frame becomes a video of the wall between them, and a product demo where the hands move across a desk loses the hands. Use a static crop only for footage where you have checked that the subject stays in the middle, such as a locked-off talking-head recording, and even then set the crop manually rather than trusting the center. Our free crop video tool lets you drag the window to the right position before exporting.
Method 2: Blur Fill and Solid Bars (Keeps Everything, Wastes Screen)
Blur fill scales the full 16:9 frame to fit the 1080-pixel width, which leaves the video 608 pixels tall in the middle of a 1920-pixel canvas, then fills the empty 656 pixels above and below with a blurred, enlarged copy of the same frame. Solid bars do the same with a flat color, usually black or a brand color. Both keep 100 percent of the original composition and both take one click.
The cost is that the actual video occupies 31.6 percent of the screen, the same fraction that a crop discards. Faces are small and text inside the original frame is hard to read. Blur fill is still the right choice for footage where cropping would destroy the meaning: wide landscapes, full sports plays, dashboard screen recordings with information at the edges, and cinematic b-roll. It also becomes far more useful when you put something in the empty space: the title of the clip at the top, animated captions at the bottom, and a logo or channel name in a corner. Vidpal's Aspect tool, for example, offers blur fill and bars as one option and reserves the top and bottom bands for captions and titles automatically, which turns wasted screen into a layout.
Method 3: Stacked and Split Layouts (Great for Podcasts and Reactions)
A stacked layout divides the 9:16 canvas into two or three horizontal bands and fills each with a separate crop from the source. For a two-person podcast, the host gets the top half and the guest gets the bottom half, each cropped tightly on their face. For a screen-share webinar, the presenter's face sits in a smaller band and the slides fill the rest. For reaction content, the original clip sits above and the reactor below. This is the format almost every podcast clipping tool produces by default, and it is the only method that keeps two speakers large on screen at the same time.
The trade-offs are that it needs two well-separated subjects in the source, it takes more setup than a single crop because each band needs its own framing, and captions have to sit between or below the bands rather than over a face. If you are clipping long recordings, an AI clipping tool that detects the active speaker and switches the layout automatically saves the most time; see our guides to turning a podcast into short clips and the best AI tools for turning long video into shorts.
Method 4: AI Smart Reframe (Follows the Speaker Automatically)
Smart reframe uses face and subject detection to decide which part of each frame to keep. The tool scans the video, finds the speaker's face in every frame, and positions a 9:16 crop so the face stays centered, either as a single best-fit position for the whole clip or as a crop that moves with the subject. The output is a full-screen vertical video with the person large and centered, with no bars and no manual keyframing, which is why it has become the default for talking-head, interview, and vlog footage.
There are two flavors. Static smart reframe analyzes the clip, works out where the subject spends most of the time, and locks one crop position, which gives a steady, professional look and is ideal for a presenter who stays roughly in place. Dynamic tracking moves the crop frame by frame as the subject walks, which keeps them in view but can feel restless if the tracking is not smoothed. Good tools smooth the motion, only pan when the subject really moves, and cut rather than pan when there is an edit in the source.
Vidpal applies face-aware reframing in two places. In AI Clips, every clip extracted from a long recording is reframed on the speaker automatically. In AI Editing, the Aspect tool's Smart Reframe option analyzes your upload, centers the crop on the detected face, and falls back to a centered cover crop if no face is found, so the export never shows bars unexpectedly. Both are covered on the AI clip maker page, and for a broader comparison of tools with this feature see the best Opus Clip alternatives.
Step-by-Step: Converting a Talking-Head Video to 9:16 with AI
Start with the highest-quality source you have, ideally 1080p or 4K, because a 9:16 crop from 1080p has to be upscaled 1.78 times and a crop from 4K does not. Upload the file to your editor and trim it to the segment you want before reframing, since analyzing a full hour of footage to reframe a 45-second clip wastes time. If you are extracting several clips from one recording, let an AI clipping tool pick the segments and reframe them in one pass.
Choose the aspect setting. Select 9:16 and pick Smart Reframe or the equivalent face-tracking option rather than a center crop. Play the result and watch for two things: the face should never touch the edge of the frame, and the crop should not drift when the speaker gestures. If it does, most tools let you nudge the horizontal position or choose a static position for that clip. For a two-person source, switch to a stacked layout instead.
Add captions after reframing, not before, so the text is positioned relative to the final frame. Use a preset with a highlighted active word, place it in the lower-middle of the screen, and keep it out of the bottom 20 percent where TikTok and Reels draw their own interface. Then add a title in the top third for the first two seconds, check the safe zones described below, and export at 1080 by 1920. In Vidpal this is the standard AI Editing flow: upload, Aspect to 9:16 with Smart Reframe, captions, hook title, export, and optionally schedule to Instagram, TikTok, and YouTube from the same screen. Our complete guide to AI captions covers the styling choices in depth.
Export Settings for TikTok, Reels, and Shorts
Export at 1080 by 1920 pixels, which is 9:16 at full HD and the native resolution of all three platforms. YouTube publishes its recommended upload resolutions and aspect ratios and 1080 by 1920 is the vertical equivalent of its 1080p recommendation. Use MP4 with H.264 video and AAC audio, 30 frames per second for talking-head content or 60 for gameplay and sports if the source was 60, and a video bitrate between 8 and 12 megabits per second for 1080p. Keep the file under a few hundred megabytes for fast uploads. Higher resolutions such as 2K and 4K are accepted by YouTube and TikTok and can sharpen a 9:16 crop taken from 4K source footage, but they do not help a crop that was upscaled from 1080p.
Keep the length within each platform's limit: Shorts up to three minutes, Reels up to three minutes for most accounts, and TikTok up to ten minutes, although the sweet spot for reframed clips is 30 to 60 seconds. Our guide to the ideal video length for every platform covers the retention data behind those numbers. If you need to shrink a file without re-editing, the video compressor and resize video tools handle that in the browser.
Captions, Safe Zones, and Text Placement After Reframing
Each platform draws buttons, the caption, the username, and the sound name over the bottom and right edges of a vertical video. Text placed there is covered and cannot be read. As a rule, keep captions and titles inside the central 1080 by 1350 region of the frame, which is the 4:5 area that also survives when Instagram shows a Reel as a cropped feed post, and stay at least 250 pixels from the bottom and 120 pixels from the right edge.
After reframing, check that the subject's face is not fighting with the caption for the same space. The most reliable layout for talking-head clips is the face in the upper-middle of the frame, captions in the lower-middle, and the title at the top for the first few seconds only. If you used blur fill, the bands above and below the video are exactly where captions and titles belong. Choose a bold, high-contrast font and keep line length short; our best fonts for subtitles guide lists the ones that stay readable at phone size.
Common Mistakes When Converting Horizontal Video
The most common mistake is converting the whole recording instead of a chosen segment, which leads to a vertical video that has no reason to be short. The second is trusting a center crop on footage that was never checked, so the speaker drifts out of frame halfway through. Others include leaving a low-resolution source to be upscaled until it looks soft, placing captions in the bottom fifth where the interface hides them, stacking two crops with no visual separation so viewers cannot tell who is speaking, and forgetting to re-check text overlays that were designed for the 16:9 version and now fall outside the crop.
A quieter mistake is ignoring what the original framing was doing. A wide shot that shows a product next to the presenter is telling the viewer something, and a tight face crop deletes it. When the context matters, a stacked layout or a brief blur-fill section preserves the information while keeping the rest of the clip full-screen.
When You Should Not Convert (Reshoot Vertical Instead)
Some footage should not be converted at all. If the subject fills the width of the frame, such as a wide product table or a group of people, no crop will keep everyone and blur fill will make them tiny. If the video relies on on-screen text, diagrams, or a screen share with information across the full width, the vertical version will be unreadable. And if the content is evergreen and important to your channel, a native vertical reshoot, or recording both orientations at once with two phones, will always beat a conversion.
For everything else, especially interviews, talking-head explainers, podcasts, webinars, and live-stream recordings, converting is the right call, because one long recording can produce a week of vertical clips. Our guides on repurposing one video into a week of content and repurposing long-form YouTube videos into Shorts cover the full workflow, and the pricing page explains what the AI Editing and AI Clips plans include.
Frequently Asked Questions
How do I convert a horizontal video to vertical without cropping? Use blur fill or solid bars, which shrink the whole 16:9 frame to fit the 1080-pixel width and fill the space above and below, or a stacked layout that shows two crops at once. Both keep the full composition, at the cost of the video occupying about a third of the screen.
What is the best way to convert 16:9 to 9:16 for a talking-head video? AI smart reframe, which detects the speaker's face and keeps a full-screen 9:16 crop centered on them. It produces a native-looking vertical video with no bars and no manual keyframing.
How much of the frame is lost when you crop 16:9 to 9:16? A full-height 9:16 crop from a 1920 by 1080 frame keeps 608 pixels of width, which is 31.6 percent of the picture. The crop is then upscaled 1.78 times to reach 1080 by 1920, so higher-resolution source footage produces a sharper result.
Can I convert landscape video to vertical for free? Yes. Free crop and resize tools, including the browser-based ones on this site, handle static crops and blur fill. AI reframing that tracks the speaker is usually part of a paid editor because it requires face detection on every frame.
What resolution should I export a vertical video at? 1080 by 1920 pixels in MP4 with H.264 video, at 30 frames per second for talking-head content, is the standard for TikTok, Instagram Reels, and YouTube Shorts. 2K and 4K exports are accepted and help when the source was shot in 4K.
Does Vidpal reframe videos automatically? Yes. AI Clips reframes every extracted clip on the speaker's face, and the Aspect tool in AI Editing offers Smart Reframe alongside blur fill and solid bars, with a centered crop as the fallback when no face is detected.
