The thermometer reads the right 31 °C, the method was followed, and ten minutes later the test smear on parchment is still tacky and matte. So it is not the temperature: temperature is necessary but not sufficient. Cocoa butter crystallisation also depends on moisture, cooling speed, the composition of the chocolate itself and what is happening in the room.
The seven reasons below are ordered by frequency. The good news comes at the end: chocolate is almost impossible to ruin for good.
One: water got in
The commonest and the most galling. Chocolate is a fat system with no free water, and even a few drops make the sugar clump together. The mass thickens as you watch, turns matte and grainy and loses all fluidity.
The sources are always domestic: a bowl that was not dried, steam from a bain-marie, a drop of condensation off a lid, a damp scraper. Water left on the slab after washing counts too.
A seized mass cannot be tempered. It can only be rescued in another form — as the base for a ganache or a filling, where water belongs. For shells it is lost.
Two: overheating
Dark chocolate above 50 °C, milk and white above 45 °C, begin to change irreversibly. In milk and white the milk proteins scorch and the mass goes gritty; in dark, severe overheating separates the cocoa butter and leaves an oily film on the surface.
Overheated chocolate can look perfectly normal while hot and refuse to crystallise whatever you do to it. The microwave is the main source of this mistake: the centre of the mass runs past 60 °C while the edges are still warm.
Three: it was never fully melted
The opposite error. If unmelted microcrystals of unstable forms remain, they become growth centres for the wrong crystallisation. The mass looks homogeneous and lump-free, but the result is matte streaking and a soft set.
This is why the first heating always goes to the top of the range and holds there: it is not enough to reach 45 °C and take it straight off, the chocolate needs a few minutes at that temperature, stirred.
Four: undissolved callets when seeding
Seeding only works if the added callets dissolve completely. A piece that settles at the bottom is not only a lump in the finished piece but a local heat imbalance around it: while it melts, the chocolate next to it gives up heat unevenly.
The tell: some of your mouldings came out glossy and some matte, though they were poured from the same bowl. That is almost always uneven stirring rather than temperature.
Five: the room is too warm, or too cold
The working range for the room is 18–22 °C. At 26 °C and above the chocolate cannot pick up crystals: it stays fluid for longer, the stable form does not develop, and the piece sets wrong even in the fridge.
The other end is no better. Below 15 °C, or poured into an over-chilled mould, the surface sets with a jolt: a skin forms, and under it the chocolate cools slowly and crystallises at random. The result is matte patches and pale streaks on the finished bonbon.
Moulds should be at room temperature, not out of the fridge. A gap of more than 10 °C between chocolate and mould is already enough to cause the fault.
Six: high air humidity
The reason almost nobody thinks about. Above 60 % relative humidity, moisture from the air condenses on the cool surface of the chocolate — the same effect as a drop of water, only spread across the whole area. As it dissolves it draws out sugar, and once it evaporates the surface carries sugar bloom.
Moving a piece from the cold into a warm humid room is especially risky. That is exactly why bonbons are not taken out of the fridge straight into the kitchen but left to come up to temperature inside sealed packaging.
Seven: old or unsuitable chocolate
Cocoa butter ages: in chocolate that has been stored for over a year, or has ridden temperature swings, part of the fat has already moved to form VI. It tempers less readily and gives a weaker shine.
Compound coating is a separate case. If the ingredients list substitute fats instead of cocoa butter — palm, lauric — it does not need tempering and cannot be tempered: it sets on its own and by different rules. The only check is the label.
What to do with a failed batch
If no water got in, everything is recoverable. Take the chocolate back up to the top temperature in the chart, dissolving every crystal, and temper again. You can repeat this as often as you like — re-tempering does not damage chocolate, it only costs time.
If the mass seized from water, there is no way back. It goes into a ganache, a filling or a batter, where water is no obstacle.
And one more case: if the bonbons are already moulded, set wrong and covered in a pale film, the film itself tells you what went wrong. Fat bloom and sugar bloom look different and mean different mistakes; that is a separate piece.
The tempering scheme with three methods and a temperature chart is covered separately, and the practice in the masterclasses on callets and stone.