song-to-image

Each line of lyrics converted into an image using generative AI (chat + image)

How it Works

$$\text{Cost of one song} \approx $2.50 \text{(A chocolate chip cookie)}$$

• Add sketch (in goes song title...)
• Add Link to YouTube Video (embedded)
• Insert screenshots & gifs

---

Features

– Estimated cost tracker – "Enter a song"

• CSV of Json
• SHow the API costs – use PIP to like add the actual lyric text to the thing… – go line by line w/ a loading bar and image emoji – To get more bang for buck, pass through prompt optimiser (chatGPT) [show compariosn] [Sketch Flow] – Each line of lyrics converted into an image using generative AI (chat + image)
• Automation

---

End Goal / Motivation

• Started with a whacky YouTube video idea (my first good idea since (link stickman parkour)), and realised I’d need a python script to make it, a common realisation (link to girlfriend GPT)
• So end goal is a YouTube video... QUICK. NEED YOUTUBE CLOUT $$$

Also:

– First large scale application.

• First real practice of:
  • Classes
  • argparse
  • True modular programming
  • OpenAI’s function-calling API
  • PIL

---

The Pipeline

`ask user for song -> genius API -> parse lyrics w/ Beautiful soup -> convert into list of lines -> iterate through list -> for each line of lyrics -> pass to ChatGPT to get image prompt -> image prompt to image -> PIL to add text add text to an image? ``

---

Requirements

• Genius API – https://lyricsgenius.readthedocs.io
• 
• Tkinter
• DaVinci Resolve (to piece it together in the end)
• OpenAI
• 

Difficulties and Learnings

raise self.handle_error_response(
openai.error.InvalidRequestError: Your request was rejected as a result of our safety system. Your prompt may contain text that is not allowed by our safety system.