Setting curd is not really cooking. It is keeping a population of bacteria in the conditions where they will work, for long enough that they change the milk. Nearly every failure is a condition problem rather than a recipe problem, and temperature accounts for most of them.
What the culture is doing
The two spoonfuls of curd you stir in are carrying live lactic acid bacteria. Given warmth and lactose, they consume the sugar and produce lactic acid, and the milk becomes steadily more acidic.
Casein, the main protein in milk, is stable while the milk is near neutral. As acidity rises the casein loses that stability, the particles begin to link to one another, and at a certain point the whole volume sets into a soft continuous gel. That is the set. It happens fairly suddenly once the acidity crosses the threshold, which is why curd looks like liquid for hours and then like curd.
It is the same coagulation used to make paneer, arrived at slowly by bacteria instead of instantly by lemon juice.
The window, and both ways to miss it
The bacteria involved are most active somewhere around 40 to 45 degrees C. That is the target, and there is a failure mode on each side of it.
Too hot. Milk much above about 50 degrees starts killing the culture. Add the starter to milk that is still near boiling and you have not inoculated anything; you have pasteurised your starter. This is the most common mistake, because hot milk feels like it should help.
Too cold. Below about 30 degrees the bacteria are alive but slow, and the milk may take a whole day to set or never get there before other organisms spoil it. In a cool kitchen this is the usual culprit.
Without a thermometer, the finger test is reliable enough: you should be able to hold a clean finger in the milk for about ten seconds. Comfortably hot, not painful.
Why boil the milk at all
Two reasons, and only one is about safety.
Boiling clears out competing organisms, so your chosen culture has the field to itself. But it also denatures the whey proteins in the milk, which then take part in the gel network instead of staying dissolved. The result is a firmer curd that holds together better and weeps less. Holding the milk at a simmer for five to eight minutes rather than just bringing it to the boil makes a noticeable difference here.
Do not disturb it
The gel forms as a single connected structure across the whole vessel. Moving the pot, stirring it, or tilting it to see whether it has set will tear that structure while it is still weak, and it does not knit back together. You get a broken, grainy curd sitting in its own whey.
Put it somewhere warm, and then leave it there for the whole six to eight hours.
More starter is not better
It is tempting to add a generous spoonful or three on the theory that more culture means a faster, firmer set. In practice a heavy inoculation tends to produce a curd that is grainy and sharply sour, because acid is produced faster than the gel can form in an orderly way. Two teaspoons per litre is enough.
Elsewhere in the same principle
Acid-set protein turns up again in why paneer goes rubbery, where the problem is the opposite: a network that contracts too much rather than one that never forms.
Join the discussion