How to deliver your AI film across platforms without rebuilding it from scratch

You finished the film. It looks the way you wanted it to look. Then someone asks whether it is going on Instagram, and suddenly you are staring at a 16:9 master that puts every important face in the left third of the frame, a title card that gets eaten by the caption bar on TikTok, and an opening ten seconds that works beautifully in a cinema context but reads as slow and uncommitted on a feed. Most solo AI filmmakers treat platform delivery as an afterthought, something you do after the real work is done. That is exactly the habit that turns one finished film into a week of remedial re-exports.

The problem is not that you made the wrong creative choices. It is that those choices were made for one output, and you need several. A 9:16 crop is not just a letterbox trimmed sideways. It changes what the eye rests on, where the implied horizon sits, and which performance beat carries the scene. If you plan for this before you generate a single frame, the same underlying material can serve every platform without compromise. That planning is what this tip covers.

Map your platform set before you touch a prompt

Write down every platform this project will live on, the aspect ratio each one favours, and the attention window you are realistically competing in. Four columns: platform, ratio, maximum effective length, caption overlay zone (top or bottom, and how deep).

A working example:

```
Platform Ratio Effective length Caption zone
YouTube 16:9 up to full length lower third
Instagram Feed 1:1 60–90 s lower quarter
Instagram Reel 9:16 15–60 s lower third + top logo bar
TikTok 9:16 15–60 s lower quarter (large)
LinkedIn 16:9 60–120 s lower third
```

This table drives every compositional decision you make during generation. It tells you where not to put a face, where not to put a title, and which scenes you will need a vertical crop for before you generate them.

Design your safe zone first, then prompt inside it

A safe zone is the area of your 16:9 frame that will survive a 9:16 centre crop intact. Roughly speaking, that is the middle 56% of the horizontal frame width. Anything critical to the image, a face, a graphic element, a text beat, a key prop, needs to live inside that corridor if you want one generation to serve both ratios.

When you write your scene descriptions, be explicit about this:

```
Subject centred in frame, significant negative space left and right,
no narrative-critical elements closer than 20% to either lateral edge.
Shot framed to read as both 16:9 wide and 9:16 vertical crop.
```

This is a constraint, not a creative punishment. Centred compositions with breathing room on the sides often read as more deliberate, more formal, more cinematic than busy edge-to-edge framing. You are not losing composition, you are choosing one that survives the export.

Plan your hook for the ratio, not just the story

The first three seconds behave differently at 9:16 and 16:9. At 9:16, the viewer's eye is already in portrait orientation. Motion near the vertical centre reads immediately. At 16:9, the eye can scan wider before committing. If your hook depends on a wide establishing reveal, it will land on YouTube and feel sluggish on a Reel. If it depends on a tight face reacting, it will dominate at 9:16 and feel pinched on YouTube.

Solve this by generating two versions of your opening shot: one composed for the wide ratio with a pull-back or reveal structure, one composed for the vertical ratio with subject-forward, immediate, close. They can share the same audio. Switch the video under the same sound design and the same opening line of dialogue or music, and both versions feel purpose-built.

Caption strategy is part of the edit, not an afterthought

Every platform renders captions differently, and the overlay zone your captions occupy is real estate you cannot use for image information. Mark that zone as dead on your composition reference before you export anything.

On TikTok and Instagram Reels, auto-captions are generated over roughly the bottom 20% of the frame. If your subject is standing and their mouth is in that zone, the caption will cover their face. Adjust your framing prompt to push subjects slightly higher in the frame for vertical outputs:

```
Subject's eyeline at approximately 40% from top of frame,
significant clear space in lower 25% of frame.
```

For YouTube, the lower third is the conventional subtitle position. If you are burning in styled captions, keep them above the 10% bottom margin so they do not get cropped on TVs.

Build a delivery checklist into your project folder

The end of every project should include the same set of exports, clearly named. A folder structure that works:

```
/delivery
/masters
film_v3_16x9_master.mp4
/platform
film_youtube_16x9.mp4
film_instagram_feed_1x1.mp4
film_reels_tiktok_9x16.mp4
film_linkedin_16x9.mp4
/thumbnails
thumb_youtube.jpg
thumb_instagram.jpg
```

Name files with the ratio in the filename. You will thank yourself at midnight when you are uploading and cannot remember which export was which.

The thing to try next: pull up your most recent finished film and do the 9:16 centre crop test on your three most important shots. Note which ones lose critical information and which survive clean. That note becomes the first line of your brief for the next project.