IdeaWell

Project ideas from Hacker News discussions.

Show HN: Ez FFmpeg – Video editing in plain English

Original Article

Hacker News Discussion

📝 Discussion Summary (Click to expand)

1. LLMs Excel at Generating FFmpeg Commands

Users praise AI chatbots for iteratively crafting complex commands, outperforming static wrappers.
"HelloUsername: The one good usecase I've found for AI chatbots, is writing ffmpeg commands. You can just keep chatting with it until you have the command you need."
"MattDaEskimo: Now, I can simply ask any LLM to write the command... It really isn't that hard anymore."
"tgsovlerkhgsel: LLMs are a great interface for ffmpeg... creates complex commands much more quickly than manual work."

2. FFmpeg's Complexity Justifies Wrappers or Helpers for Rare Use

Infrequent users avoid memorizing syntax, favoring tools/AI over manuals for efficiency.
"serial_dev: There is no universe where I would like to spend brain power on learning ffmpeg commands by heart."
"rolfus: I'm never going to memorize the commands or syntax for those [tools used 0.3 to 3 times a year]."
"BeetleB: It's not hard - just not a good use of our time... ffmpeg is not a vital tool."

3. Wrappers Risk Hiding Complexity and Suboptimal Defaults

Critics note tools like ezff oversimplify (e.g., unnecessary reencoding), urging verification or learning basics.
"qbow883: 'ff convert video.mkv to mp4' maps to ffmpeg -i video.mkv -y video.mp4 here, which does a full reencode (losing quality...)."
"skydhash: You can’t verify LLM’s output. And thus, any form of trust is faith, not rational logic."
"stevage: [LLMs throw] in extra options... if you don't... check them all, you get... unchecked cruft."

🚀 Project Ideas

FFPrompt: The Human-Readable Spec for Multimedia

Summary

  • A "specification language" and compiler that bridges the gap between natural language and dry FFmpeg flags, providing a better middle ground than unreliable LLMs or arcane manuals.
  • Solves the "hallucination vs. manual" dilemma by using a strict, typed, but human-readable syntax that explains why specific flags (like -c:v copy vs. re-encoding) are chosen.

Details

Key Value
Target Audience Developers and Power Users who use FFmpeg 1-5 times a year.
Core Feature Validated prompt-to-command compiler with "Explain" mode and "Lossless-first" defaults.
Tech Stack Rust (for standalone binary security/speed) or Python.
Difficulty Medium
Monetization Hobby

Notes

  • HN users expressed frustration with LLMs hallucinating arguments and "simplified" wrappers hiding important details like re-encoding.
  • "I believe the way to go should be educating users... not by hiding complexity." FFPrompt would offer a --verbose-explain flag that maps the simplified command back to the manual concepts.

Node-less CLI Tool Builder (Metatool)

Summary

  • A framework for creating standalone, secure CLI wrappers for complex tools (FFmpeg, OpenSSL, Git, GPG) without requiring Node.js or heavy runtimes.
  • Addresses the specific complaint that modern CLI helpers often require "massive krakens like node.js" which are seen as security and efficiency risks.

Details

Key Value
Target Audience DevOps, Security Researchers, and minimalist SysAdmins.
Core Feature DSL to define "Recipes" that compile into single, static Go or Rust binaries.
Tech Stack Go or Rust with an LLM-provider API for "Recipe" generation.
Difficulty Medium
Monetization Hobby

Notes

  • Multiple commenters mentioned they would use wrappers if they weren't npm-based: "A CLI utility dependent on Node.js is not a good thing frankly... could’ve been a Rust program or a Go program."
  • This fits the need for "wrappers that don't go away" and are "easy to replicate and verify."

FFStudio Open: Node-Based Video Pipeline Builder

Summary

  • An open-source, local-first visual node-based editor for FFmpeg command generation, similar to the now-defunct "ffmpeg.guide."
  • Allows users to visualize complex filter chains (split, concat, palettegen) as a graph, making the "spaghetti code" logic of filter_complex visible and debuggable.

Details

Key Value
Target Audience Video Editors, Creative Technologists, and Data Scientists.
Core Feature Drag-and-drop nodes (Input -> Filter -> Output) that real-time generate a CLI string.
Tech Stack React/Svelte with a canvas library (like React Flow).
Difficulty High
Monetization Revenue-ready: SaaS for cloud rendering or Free/Open Core.

Notes

  • HN users mourned the loss of visual tools: "There was an ffmpeg drag-and-drop GUI... Unfortunately it looks like they tried to monetize it but then later shut down."
  • This solves the "bounce effect" and "palette-per-frame" complexity mentioned in the discussion, where text-based commands become unreadable.

Read Later