You shot great footage in 16:9 and now need it in 9:16 for TikTok, Instagram Reels, or YouTube Shorts. Manually cropping means losing your subject in frame. AI reframing detects the main subject and repositions it perfectly for vertical formats, preserving quality and composition automatically.
You filmed a stunning sunset hike, a client interview, or a product demo in widescreen. Then you open TikTok or Instagram and realize none of it fits. The algorithm rewards vertical content. Your footage is horizontal. That gap used to mean hours of painful cropping, guessing where subjects would move, and publishing video that either cuts off heads or drowns in black bars.
AI changes that equation completely.
The numbers are not subtle. Over 75% of video content is now consumed on mobile devices held vertically. TikTok built an empire on it. Instagram Reels and YouTube Shorts followed. Even LinkedIn now promotes vertical clips in its feed. The 9:16 aspect ratio is not a trend reversing itself anytime soon.
Brands that still post 16:9 widescreen videos on platforms built for vertical viewing see lower completion rates, lower engagement, and lower algorithmic reach. It is not a stylistic preference anymore. It is a structural disadvantage.
For creators working with existing landscape footage, this created a painful fork: reshoot everything in vertical, or spend hours manually cropping clips and hoping the subject stays in frame.
When you manually crop a 16:9 video to 9:16 without smart tracking, you are choosing a fixed crop window. The moment your subject moves to the edge of that window, they fall out of frame. The moment a wide shot pulls back, your crop becomes a closeup of irrelevant background. The moment you cut to a different angle, the composition breaks entirely.
Viewers notice immediately. A talking head that drifts half out of frame looks amateurish. A landscape clip with the horizon cut off looks broken. These are not recoverable errors in post. They erode trust in the content before a single word is heard.
AI-powered reframing is fundamentally different from cropping. Cropping is static. Reframing is dynamic.
When you run a landscape video through an AI reframing tool, it does not just pick a fixed 9:16 window and hold it. It analyzes every frame of the footage to identify the primary subject, tracks that subject as it moves through the scene, and adjusts the crop window in real time to keep it centered and well-composed.
Modern reframing AI uses computer vision to identify faces, bodies, objects in motion, and areas of visual interest. It assigns priority weights to each detected element. In a talking head video, the face gets the highest weight. In an action clip, motion vectors drive the tracking. In a landscape shot, the AI finds the visual anchor of the composition and locks to it.
This tracking happens frame by frame, but with smoothing applied so the crop window does not jump erratically. The result feels like a camera operator was panning to follow the action, not like a crop that happened to move.
The AI also handles transitions between subjects. If your video cuts from a wide shot to a closeup, the reframing adjusts the crop window to match the new composition rather than keeping the old settings. This produces vertical footage that feels intentional, not salvaged.
Before running any footage through a reframing tool, it helps to understand exactly what is different between the two formats at a technical and experiential level.
| Factor | 16:9 Landscape | 9:16 Vertical |
|---|---|---|
| Aspect ratio | 1.77:1 | 0.56:1 |
| Primary platform | YouTube, Vimeo, desktop | TikTok, Reels, Shorts |
| Screen coverage on mobile | ~30% | 100% |
| Average completion rate | Lower on mobile | Higher on mobile |
| Composition focus | Wide context | Tight subject |
| Black bar risk on mobile | None on desktop | Present without reframing |
The jump from 30% to 100% screen coverage is the single most important stat for anyone making content for mobile audiences. A video that fills the screen commands attention in a way that a letterboxed clip simply cannot.
PicassoIA has a dedicated tool built for this exact problem: Reframe Video by Luma. It takes your landscape footage and intelligently reframes it to any target aspect ratio, including the 9:16 vertical format required by all major short-form platforms.
Here is how to use it, step by step.
Go to Reframe Video on PicassoIA and upload your landscape video directly from your device. The tool accepts common formats including MP4, MOV, and MKV. There is no software to install and no account setup beyond PicassoIA itself.
💡 For best results, use footage with at least 1080p source resolution. The reframing process crops into the frame, so higher resolution source material preserves more detail in the final output.
Select 9:16 as your output ratio for TikTok, Instagram Reels, and YouTube Shorts. Reframe Video supports multiple output ratios, so you can also create 1:1 square versions for feed posts in the same workflow without uploading the file again.
The AI analyzes your footage and generates a reframed preview. Watch through the preview to check that the primary subject stays well-composed throughout. For talking head footage, face tracking is typically very accurate. For action or multi-subject footage, review cuts closely since rapid subject changes can sometimes lag the tracking.
Once satisfied with the preview, export your vertical clip. The output is ready to upload directly to TikTok, Instagram Reels, or YouTube Shorts without any additional processing. If you plan to post to multiple platforms, export separate versions for each. Platform compression algorithms differ, and a file optimized for TikTok may look noticeably different on Reels.
💡 Batch your exports when processing multiple clips. Reframing several videos in one session saves time compared to processing them one at a time across different days.
Reframing crops into your original footage, which means it always involves some quality reduction. A 1920x1080 landscape video cropped to a 9:16 vertical at full height yields a lower-resolution output than the original. For footage you want to look sharp on modern high-resolution mobile screens, upscaling after reframing is worth the extra step.
PicassoIA offers strong options for this:
Video Upscale by Topaz Labs uses AI to push footage to 4K while sharpening detail, reducing noise, and recovering texture that compression removed. It is the strongest option for footage that started at 1080p and needs to look premium after a tight crop.
Crystal Video Upscaler provides similar upscaling capabilities with a focus on cinematic texture preservation, making it a good fit for narrative or cinematic footage.
Video Increase Resolution by Bria can push footage up to 8K resolution, which gives enormous headroom for quality on any screen size and future-proofs your content for higher resolution platforms.
Vertical video is frequently watched without sound. Studies consistently show that 85% of social media videos are watched muted, and the pattern holds across short-form platforms. Captions are not optional for vertical content that needs to perform at scale.
After reframing, run your clips through Autocaption by Fictions AI on PicassoIA. It automatically transcribes your audio and places styled captions directly onto the video, positioned for the vertical format. For talking head content especially, well-placed captions can meaningfully increase completion rates by keeping viewers engaged even when they are in a sound-off environment.
💡 Position captions in the lower third for TikTok but consider the center for Reels, where the platform UI overlays the bottom edge on some devices and can obscure text that sits too low.
Not all footage benefits equally from AI reframing. Knowing where the tool excels sets expectations correctly and saves review time.
This is where AI reframing is most reliable. A single face as the primary subject gives the tracking algorithm a clear, stable anchor. The crop window rarely needs to jump or adjust dramatically, and the output almost always looks intentional. Client testimonials, tutorials, and vlogs fall into this category and reframe with very little manual review needed.
Outdoor footage with a clear focal point, a person hiking, a landmark, an animal in a natural scene, reframes well. The AI picks up the person or object as the visual anchor and tracks it through the shot. Wide establishing shots with no clear subject are harder, but the AI typically centers on the most visually interesting area of the composition.
Conference talks, performances, and presentations translate well to vertical because there is usually one speaker or performer as the consistent subject. Multi-camera event footage with frequent cuts requires more preview review, but single-camera recordings reframe cleanly in most cases.
Fast-moving subjects challenge all tracking systems. Reframing works better for slower-paced action than for rapid athletic movements. For high-speed sports footage, review the output frame by frame at transitions and be prepared for some clips to need manual composition adjustment.
If your video has multiple people in frame, AI reframing picks the subject it identifies as primary, typically the most centered or largest face. If you want a different person tracked, or if you are prioritizing a product over a presenter, check the preview carefully. Do not assume the AI always picks the subject that matches your editorial intent.
Vertical video platforms overlay UI elements on the footage: like buttons, comment sections, username labels, and hashtag links. These overlap the bottom and right side of the frame on TikTok and Reels. If your subject or any critical visual element lands in these zones, it gets covered by the platform UI. Most platforms publish safe zone templates. Verify your output against them before publishing to avoid having important content obscured.
Reframing without upscaling is leaving quality on the table, especially if your source footage was 1080p. A cropped 1080p video displayed on a modern OLED screen will show compression artifacts that were invisible in the original widescreen version. Running the clip through Video Upscale by Topaz Labs or Real ESRGAN Video takes a few minutes and the quality difference on mobile screens is visible.
Different platforms have specific requirements that affect how you prepare reframed footage. Knowing these before you export saves time and prevents needing to re-process clips.
| Platform | Ratio | Max Duration | Recommended Resolution |
|---|---|---|---|
| TikTok | 9:16 | 10 min | 1080x1920 |
| Instagram Reels | 9:16 | 90 sec | 1080x1920 |
| YouTube Shorts | 9:16 | 3 min | 1080x1920 |
| Pinterest Video | 9:16 | 15 min | 1080x1920 |
| LinkedIn Video | 9:16 | 10 min | 1080x1920 |
All five major platforms converge on 1080x1920 as the recommended vertical resolution. This means your reframing workflow should always target that output regardless of which platform you are posting to, which simplifies batch production significantly. Export once at 1080x1920 and the file works everywhere.
Every piece of landscape footage you have already shot is an asset sitting unused on short-form platforms. AI reframing removes the barrier between the footage you have and the formats the algorithm actively rewards.
The workflow is straightforward: upload to Reframe Video on PicassoIA, set your output ratio to 9:16, review the AI tracking, add upscaling with Video Upscale by Topaz Labs if needed, and add captions with Autocaption. Four steps, and your widescreen footage is ready for every major short-form platform without reshooting a single frame.
PicassoIA puts all these tools in one place, so you are not switching between five different apps to finish one clip. Pick a video from your archive that has been sitting unused because it was in the wrong format, run it through the reframing workflow, and see what comes out. The gap between your landscape footage and a vertical-first content strategy is smaller than it looks.
