DEV Community

Ken Deng
Ken Deng

Posted on

From Fuzzy Memory to Searchable Database: AI for Your Glaze Log

Does this sound familiar? You have a thousand photos of glaze tests on your phone, but finding that one perfect "rutile blue breakout" from last year is impossible. Your notes are scattered, and "cranberry red" in the studio light becomes "burgundy" in the sun. This disconnection between image, data, and recipe kills consistency and wastes time.

The Principle: Structured Visual Logging

The solution isn't more photos; it's better metadata. The core framework is Structured Visual Logging: treating every test tile not as a standalone picture, but as a rich data node linked to its recipe and firing log. This transforms your subjective memory into an objective, searchable knowledge base.

Your Toolkit: Obsidian as Your Digital Studio Assistant

A free digital notebook like Obsidian is ideal for this. Its purpose is to act as your central hub where visual data, descriptive tags, and quantitative metrics coexist and interconnect. You create a dedicated note for each test, embedding the image and populating consistent fields.

See it in action: You're planning a production batch. Instead of guessing, you query your Obsidian vault: "Show me all glazes with a gloss >70 GU that are stable on vertical surfaces." Instantly, you review the visual log, see that the last test had minor pinholes, and note to sieve twice before mixing.

Implementation in Three Steps

  1. Standardize Your Stage. Commit to a simple, non-reflective mid-grey backdrop. Use it for every single test, pre- and post-firing. This eliminates lighting inconsistency and allows AI tools (now or in the future) to accurately assess color and texture.
  2. Enforce a Data Entry Ritual. For every test, create a new note with a unique Test ID (e.g., 250415-Shino01). Embed the image and fill mandatory fields: Recipe ID, Firing Log (cone, atmosphere, etc.), and measured outcomes (gloss, texture, fit).
  3. Tag Relentlessly and Objectively. Add at least five descriptive tags. Move from subjective "cranberry" to objective #iron_amber_base, #rutile_blue_breakout, #matte, #cone10_reduction, #crazing. This is what makes your library searchable.

Key Takeaways

By adopting structured visual logging, you replace chaos with clarity. You move from disconnected, subjective snapshots to a linked, queryable database of your ceramic practice. This method ensures batch consistency, unlocks advanced search of your own work, and turns your accumulated tests into one of your most valuable professional assets. Start your next test with the end—and the search—in mind.

Top comments (0)