MANN-PISHACH | Full Film | Zero-Budget AI Horror Experiment (2026) | Rahi Anil Barve

Описание: MANN-PISHACH — MAKING PROCESS
For every broke, struggling filmmaker dreaming of making a zero-budget film.

“Can one person make an 80-minute film alone on a home PC?”

Yes, I could spend money, hire people, and create better visuals.
But then the experiment fails.

The real question was:

Can a broke filmmaker do this alone?

They would need:

• writing ability
• shot imagination
• a basic laptop
• Photoshop
• After Effects or Premiere
• patience and passion

THIS IS A ZERO-BUDGET EXPERIMENT.

If it helps even one broke, frustrated artist create something out of nothing — even with almost empty pockets — and if they enjoy the process,

that’s a win for me.

— Rahi Anil Barve

MANN-PISHACH Synopsis:
June, 1960.
A junior archaeology officer, Sadashiv Rao, is sent to the remote village of Hadamgaon to survey a mysterious stone dome revealed after the monsoon.
He stays in the house of Savitri, a young widow living under the suffocating rules of the village. Hadamgaon feels frozen in time: doors reinforced with iron, windows sealed, streets empty, and every night the entire village locks itself indoors before dark. No one will explain why.

As Rao investigates the buried structure and the strange silence surrounding it, he begins to uncover an ancient warning. The dome beneath the hill is not a temple of worship, but a mechanism built to contain something buried deep within the human mind.

Because what is suppressed does not die.

MAKING PROCESS
1. Script / Screenplay

The screenplay was written and locked by Zaai Gulmohar, keeping AI limitations in mind from the very beginning.

Instead of conventional talk-heavy scenes, the film relies on:

• long visual actions
• silence
• voice narration

We did not try to force AI to behave like traditional cinema.

Instead, we wrote cinema that AI could realistically survive.

2. Storyboards — the real backbone

After the script was locked, every shot was roughly hand-drawn on paper.

Without storyboards, maintaining 80 minutes of continuity with cheap AI tools is almost impossible. (free-cheap tools like chatgpt- kling -grok were used for this zero budget experiment. )

Shots were designed around AI duration limits:

• 5 seconds
• 10 seconds
• 15 seconds (rarely used)

Many sequences were solved at the storyboard stage itself.

Example: the ancient dome reveal through spiders crawling inside statue cavities was solved on paper first, not discovered during generation.

Without storyboards:

• costs explode
• continuity collapses

3. Real actors

After script and storyboard lock, performances were recorded on a simple phone setup.

Actors:

Yaaneea Bhardwaj
Deepak Damle

This step is critical.

AI characters feel alive only when guided by human performance intention.

Without actors, AI gives movement.
With actors, AI can produce emotion.

4. Locking the world (production design)

Before generating shots, the film world was locked using high-resolution still images:

• village exteriors
• Savitri wada interiors and exteriors
• the ancient dome hill
• props
• costumes
• lighting moods across time

AI cannot maintain visual continuity on its own.

So the world had to exist first.

Only after that did animation begin.

5. Shot creation workflow

Pipeline used:

ChatGPT → Photoshop → Kling → After Effects

Method: first-frame / last-frame control

For each shot:

• first frame created
• last frame created
• refined in Photoshop
• generated in Kling
• refined again in After Effects

After Effects work included:

• masking
• focus/defocus
• corrections
• occasional face fixes

Expensive generators were deliberately avoided.
The goal was a mass-affordable pipeline.

Numbers

• ~2500 final PSD images created
• discarded attempts ≈ 9–12× more
• Kling credits used under 30K

A rule was followed:

Never press generate until everything is ready.

Otherwise AI wastes both time and money.

Learning curve

Beginning:
1 usable shot after 12–20 attempts

End of film:
almost 1:1 success

About 75% of shots were made in Kling 2.5 Turbo.

Without planning, the same output could easily cost ₹7–9 lakh.

6. Post-production

The entire film was assembled in After Effects 2024.

Rules followed:

• each scene a separate project
• final assembly after scenes were locked

Sound and music mostly came from free sources and limited AI audio.

7. Why this method

Better visuals could be made with larger budgets.

But then the experiment fails.

The real question was always:

Can a broke filmmaker do this alone?

They need:

• writing ability
• shot imagination
• a basic computer
• Photoshop
• After Effects or Premiere
• patience and passion

Six months ago I knew almost nothing about AI filmmaking.

Yaaneea Bhardwaj helped me start.

MANN-PISHACH is a zero-budget experiment.

If this process helps even one struggling filmmaker create something from nothing — even with empty pockets — and if they enjoy the process,

then the experiment has already succeeded. - Rahi Anil Barve