Why Your AI Videos Look Amateur (And How to Fix It)
Shows filmmakers how to build a cost‑effective, repeatable AI video pipeline that avoids the typical “AI slop” and delivers professional results.
What this lesson covers
In this tutorial Yaroflasher walks through a practical AI motion workflow using FlashBoards and the Nano Banana tool, showing how to maintain character consistency and upscale assets efficiently. He emphasizes generating low‑resolution variations first, selecting the best, then upscaling to save credits and achieve production‑ready quality.
Key takeaways from the creator
AI-extracted notes, not independently verified product claims. Timestamp links let you check each point in the original video.
- 03:05 ↗
Using two reference images (close‑up face and full‑body) with Nano Banana keeps character and outfit consistent across generations.
- 05:33 ↗
Upscaling from 1K to 2K/4K dramatically improves sharpness, texture and facial detail, making images production‑ready.
- 08:08 ↗
The optimal workflow is: generate many 1K variations, pick the best, then upscale – saving time and credits.
Workflow outlined in the video
- Log into FlashBoards and open the Nano Banana workflow.
- Prepare a close‑up face shot and a full‑body reference image of the subject.
- Generate images at 1K resolution, creating multiple variations.
- Review results, deselect all, then select the preferred image.
- Use Nano Banana’s built‑in upscaler to increase the image to 2K or 4K.
- Export the upscaled asset for use in downstream video generation.
Before you use this workflow
These notes describe the source video at its publication date. Model access, pricing, connectors and interfaces may have changed. Check the original source and the provider’s current documentation before installing an add-on, connecting an account or spending credits.
ReelStack has not independently tested this workflow. Preview one representative shot and check motion, continuity and output quality before applying it to a full production. No result, cost saving or model capability is guaranteed.
How to interpret the numbers
188,399 views captured 24 October 2026. ReelStack recommendation score: 61/100. Formula: radar-v5. This is ReelStack’s calculation, not a YouTube rating or a measure of factual accuracy.
Show the calculation
- 65% performance: 3.65× views vs 3 other channel videos observed at a similar age. View-evidence factor 99% (views / (views + 1,000))
- 25% freshness: 0/100 with a 14-day half-life
- Momentum pending: collecting daily snapshots; its 20% weight goes to observed performance, not free points
- 10% engagement: 100/100; likes + 4× comments, smoothed with a 500-view neutral prior