How I keep a starter and never lose it
A starter is a jar of flour and water in which wild yeast and lactic acid bacteria have moved in. It sounds like something complicated, but in practice caring for it resembles caring for any household creature: feed it on time, do not overheat it, do not forget it. Here I have written down my rhythm and my mistakes.
The jar, the flour, and the water
I keep my starter in an ordinary glass jar of about a liter, with a lid that I simply rest on top without screwing it down. The jar must breathe: fermentation gives off gas, and a tightly closed jar will one day remind you of that with a pop and splashes on the cupboard. A rubber band or a marker line on the wall of the jar helps me see how much the starter has risen after feeding: I mark the level right after stirring, and later check whether it has doubled.
I feed my starter with plain whole rye flour and water at room temperature. I tried wheat whole grain, I tried blends — in the end I came back to rye: it ferments predictably, smells pleasant, like bread and apples, and forgives my irregular schedule. My water comes from the tap, left standing for a day in an open jar. The chlorine that tap water sometimes smells of depresses fermentation, and over a day it escapes. If your water is very hard or smells strongly, try bottled drinking water — the difference is sometimes visible at once.
My feeding proportion is simple: one part starter, two parts water, two parts flour, all by weight. For example, I leave fifty grams of old starter in the jar, add a hundred grams of water and a hundred grams of flour, and stir to the thickness of heavy cream. Some people feed one to one to one, some one to five to five — the more fresh flour you add, the longer the starter takes to reach its peak. My proportion gives a peak in about six to eight hours at room temperature, which fits my evening rhythm nicely.
Feeding rhythm and signs of readiness
When I bake two or three times a week, the starter lives on the kitchen counter and is fed every day. When a period comes where no bread is needed, the jar moves to the fridge and is fed once a week. This is not dogma, just two modes I switch between without ceremony: take it out of the fridge, feed it, wait for the peak, feed it again — and after two or three feedings the starter is cheerful again.
I judge readiness for dough by three signs at once. First: the starter has doubled or tripled and stands at its peak or has just begun to fall. Second: the surface is bubbly and a web of bubbles of different sizes is visible inside. Third: the smell has become distinctly fruity, with a light acidity, without sharp notes of acetone or alcohol. If the smell has gone into sharp acid, the starter is overripe and hungry: I feed it again and wait for the next peak, and I do not mix dough with such a starter — the bread would come out sour and heavy.
There is a simple home test I use as an extra check: a spoonful of starter dropped into a glass of water should float. It does not always work, and not with every starter, especially a thick rye one, so I do not rely on it alone. But if the starter sinks while all the outer signs say it is ready, I give it another hour or two. A floating starter is a good sign; a sinking one is not always a bad one.
The fridge and going away
The fridge is a busy baker's best friend. At around four degrees fermentation almost stops, and a starter can live without feeding for a week, sometimes two. I put the jar in the fridge right after feeding, not at the peak: that way it spends its flour reserve more slowly and stays hungry-but-alive longer. Before baking I take the jar out a day ahead, feed it twice with a twelve-hour gap, and mix the dough on the second peak.
If I go away for longer than a week, I make the starter thicker — half again as much flour as usual — and put it in the fridge. A thick starter starves more slowly. Coming back, I do not panic at the dark liquid on the surface: this is the so-called hooch, an alcohol layer, a sign that the starter is hungry. I pour it off, discard the top layer if it has darkened, and feed the rest as usual. The first two feedings may be sluggish; the third usually returns the old strength.
I once rescued a truly neglected starter after a month in the fridge: it smelled of acetone and looked dead. I fed it three days in a row, each time throwing away everything but a couple of spoonfuls, and on the fourth day it rose as if nothing had happened. Wild yeast is tougher than it looks. A starter is worth throwing away only when there is mold — colored fuzzy spots on the surface mean the jar is lost, and there is no room for economy there.
Three of my failures and what I fixed
Failure one: the starter stopped rising after I began feeding it with warm tap water to speed things up. Warmth speeds up not only the yeast but also the bacteria, and the balance shifted toward acid. I went back to room-temperature water and began feeding more often but with smaller portions of fresh flour — within a week the rise returned. Lesson: do not rush a starter with warmth; it does not like hurry.
Failure two: the bread was consistently sour even though the starter looked lively. It turned out I was mixing dough with an overripe starter — one that had already fallen and smelled of alcohol. I began waiting for the peak itself, not just sometime after the rise, and the acid went away. At the same time I cut the share of starter in the dough from thirty percent to twenty: the taste became milder and the rise barely suffered.
Failure three: after changing flour the starter went sluggish for a week. I had switched brands of rye flour, and the new flour differed in grind and in strength. The starter needed several feedings to adjust. I changed nothing in the routine, just kept feeding and waiting — and it adjusted. Lesson: when you change flour, give the starter time before treating it for diseases it does not have.
The short version
If I squeeze this whole page into a few lines, it goes like this: feed the starter predictably, watch the peak and the smell, keep it in the fridge when you are not baking, and do not panic after a trip. A jar of starter forgives almost everything except mold and drying out completely. Everything else is a matter of a few feedings and patience. And once the starter has become predictable, move on to the simple rye loaf — with a good starter it almost bakes itself.