Shailesh Shakya

Shailesh ShakyaOct 10, 2026

Most AI filmmakers don’t need more shots. They need the right coverage. A scene can look beautiful shot by shot and still fall apart in the edit because every generation was made in isolation. That’s why I like thinking in 5 simple shot jobs: → Master / Wide = where are we? → Medium / Interaction = what are the characters doing together? → Close-up / Reaction = what is the important emotional beat? → Insert / Detail = what object or action matters? → Cutaway / Secondary reaction = where can the editor cut next? Imagine a diner scene. A woman meets a man. He secretly slides a key across the table. You don’t need 12 random cinematic angles. You need: → Wide: establish both characters and the diner → Medium: show their conversation and the key being slid → Close-up: show her realization → Insert: show the key → Cutaway: show him checking the entrance Now every shot adds information. That’s the part many AI creators miss. A cool orbit shot may look impressive. But if it doesn’t help geography, emotion, action, detail, or editing, it may just be another expensive clip you can’t use. One more rule: You do not always need all five. Use only what the scene needs. But before generating another shot, ask: → Do I have geography? → Do I have interaction? → Do I have the emotional beat? → Do I have the important detail? → Do I have somewhere useful to cut? And keep screen direction, props, character positions, costume, and lighting consistent across them. The takeaway: good coverage is not about generating more shots. It is about generating the right shots to build an editable story. Join my newsletter: https://lnkd.in/gu_6fpTk . .

Shailesh ShakyaOct 17, 2026

Adding “85mm lens, f/1.8” to your AI prompt wont magically make your video cinematic. Most AI creators use camera terms without understanding what they actually change. Heres what you should know: → 24mm lens: Wider view. Shows more of the characters surroundings. → 85mm lens: Narrower view. Useful for close-ups and isolating reactions. → f/1.8 vs f/8: Wider apertures generally blur more of the background. Smaller apertures keep more in focus. → Rack focus: Shifts focus from one subject to another without moving the camera. → 180° shutter at 24fps: A 1/48-second exposure, commonly used for natural cinematic motion blur. But heres what most beginners misunderstand: Focal length doesnt control perspective. Camera position does. Instead of prompting: 85mm, f/1.8, cinematic shot. Try: Medium close-up from several meters away. Eyes sharply focused, distant mountains softly blurred. Camera stays still. AI understands visual instructions better when you describe the intended result, not just technical settings. The secret isnt knowing more camera terms. Its knowing what each one actually does to your shot. . . .

Shailesh ShakyaOct 16, 2026

Here is how to make this: 1. Go to Midjourney 2. Ask it for a character 3. Once you have character, go to Midjoureny and ask it to generate the character 3. Repeat the same process for CAR 4. Next go to ChatGPT >> upload one image and ask it to generate a character sheet. Do the same for both 5. Give a situation to ChatGPT and ask it to create a 3x3 Storyboard 5. Now, go to Seedance 2.0 and upload storyboard and ask it to generate the full video. Done... I hope it helps you! if so, please follow me, repost, leave your thoughts. . . .