About¶
What is the xBloom Studio?¶
The xBloom Studio is an automated pour-over brewer: it grinds the beans, weighs the dose and pours water at a set temperature on a programmed pour schedule. That makes a recipe repeatable — the same numbers give the same cup.
Before you brew¶
💧 Always rinse the paper filter before you start, to wash out any papery taste — seat the filter in the Omni dripper, pop the dripper off the machine, and rinse it under the tap in cold water, then tip the water out and re-attach the dripper. (Every recipe here assumes a rinsed filter.) A quick hot-water rinse done in place tends to under-rinse and can leave the cup papery/bitter — the cold, full tap rinse works better here.
🚰 All recipes here are brewed with filtered water. Water is ~98% of the cup, so it matters — very hard or heavily chlorinated tap water will taste different (and scales the machine faster).
How to read a recipe¶
Each coffee page lists:
- Dose (g) and ratio (e.g. 1:15 → 16 g coffee to 240 ml water),
- Grind (the lower the number, the finer),
- Temperature and grinder RPM,
- Pours — the bloom (initial wetting) and the following pours, with a pattern (spiral / centered),
- The dial-in story — how the taste was reached.
The golden rule of dialing in¶
- Sour / flat / watery → under-extracted → grind finer (lower number).
- Bitter / astringent / dry → over-extracted → grind coarser (or lower the temperature).
- Change one thing at a time and re-taste. Drink light roasts hot — they turn bitter as they cool.
Brew them from the command line 🤖¶
There's an open-source companion tool — xbloom-ble — that sends a recipe to an xBloom Studio over Bluetooth. It loads the recipe onto the machine; you add the beans and approve on the machine to start it (it never auto-brews). Reverse-engineered protocol, fully documented. Built with Claude Code, same as this site.
Each recipe was dialed on one particular grinder and palate — treat it as a launch pad, not gospel. This is a non-commercial hobby ledger, unaffiliated with xBloom or any roaster.