Why tempering fails

The most common invisible cause. Tempering begins by erasing every existing crystal, which needs 45–50°C for dark and 45°C for milk and white. Chocolate melted only to the point where it *looks* melted — around 40°C — retains stable crystal fragments.

Those survivors compete with the crystals you seed afterwards, and the result is unpredictable: sometimes it works, often it does not, and nothing about the visible process explains why.

The second invisible cause. Correctly tempered chocolate needs an ambient temperature below about 20°C to set properly, and below 18°C for milk and white.

At 26°C it sets slowly and softly; at 30°C it does not set at all. A test strip that never sets in a warm kitchen is not evidence of a failed temper — it is evidence that the test cannot be run there.

Dark, milk and white each have their own curve, and the differences are one to two degrees at each stage. Applying dark chocolate's 31–32°C to white chocolate, whose window is 29–30°C, is above the range and gives a soft streaky set.

This is the most common *visible* cause, and it is the reason a guide that gives one set of numbers for all chocolate is worse than no guide.

The final rewarm is a two-degree operation and the margin above it is small. Dark chocolate warmed past about 34°C has had its Form V crystals melted along with the unstable ones, and is back where it started.

A microwave burst is almost always too much. A few seconds over a bain-marie, or a heat gun at a distance, with stirring and a thermometer between each application.

Wrong product: chocolate chips and compound coating either temper poorly or not at all. Compound has no cocoa butter to crystallise, so nothing about the procedure applies to it.

Added fat: oil, butter or cream added to the melt prevents proper crystallisation permanently. Cocoa butter is the only fat that can be added without breaking temper.