When Dormant Pre-ferments Wake Up: A Baker's Honest Guide

You open the fridge, grab that mason jar labeled 'starter—feed Sat,' and wonder: Is it still alive? The dark liquid on top says maybe not. But here's the thing—dormant pre-ferments are weird. They're not dead, just sleeping. And waking them up is part science, part art, part luck.

I've killed more starters than I'd like to admit. Left them too long, fed them wrong, or just forgot. But over time, I learned what's actually happening in that cold jar. Spoiler: it's not magic. It's biology, with a side of patience. This guide is for anyone who's stared at a sluggish starter and wondered if it's worth saving. (It usually is.)

Why Bother With Dormant Pre-ferments Right Now?

The sourdough boom and fridge neglect

You bought a starter in 2020. Named it something embarrassing. Fed it like a newborn for three months. Then life happened — a work trip, a heatwave, a sudden disinterest in 6 a.m. feedings. The jar slid to the back of the fridge. You told yourself you'd feed it next week. Next week became next month. Now you're staring at a layer of dark liquid and a crust that looks geological. Most people pour it down the drain. That's a mistake — and an expensive one. A healthy starter costs maybe a few dollars in flour, but the time investment? Weeks of daily feedings, the anxiety of that first levain, the trial runs that produced frisbees instead of boules. Tossing a dormant starter isn't just discarding flour; it's throwing away two months of patience.

The catch is that most 'dead' starters aren't dead. Not even close. I have seen jars that sat undisturbed for six months — separated, smelling like acetone mixed with old socks — that bounced back within three feedings. The yeast and bacteria didn't die; they entered a metabolic holding pattern. Slowed down. Recalculated. That thin grey liquid on top? Hooch, a natural preservative that actually protects the culture from unwanted molds. Wrong color? Sure — but not a death certificate. The real cost of throwing away a 'dead' starter is the flavor you haven't tasted yet.

Why flavor depth demands time (and cold)

Here's the part most quick-bread tutorials skip: flavor in sourdough isn't about the flour you use or the hydration percentage. It's about age. A week-old starter produces bread that tastes vaguely tangy, like mild yogurt. A starter that has cycled through cold dormancy for a month, then revived? That bread carries notes of roasted grain, subtle cheese, sometimes even a faint tropical fruitiness. The cold dormancy period forces the microbial community to shift — lactic acid bacteria dominate, acetic acid builds, and complex esters develop. You can't rush this with warm water or extra feedings. You have to wait.

That sounds fine until you realize what it means for your schedule. A dormant starter needs more time to wake up than a fresh one — two or three refresh cycles before it doubles reliably. We fixed this by planning backward: revive on Wednesday for a Saturday bake. The trade-off is real: you lose a day of spontaneity. But the flavor payoff is undeniable. I've side-by-side tested a young levain against a revived dormant culture — same flour, same hydration, same bulk time. The revived loaf had a crumb that stayed moist three days longer. That's not opinion; that's how mature acid profiles retard staling.

“The best flavor I ever got from a loaf came from a starter I found behind a bag of quinoa, forgotten for four months. It smelled like a barn. I almost rinsed the jar.”

— A baker in our community discord, describing the exact moment we all realize patience isn't passive.

The real cost of throwing away a 'dead' starter

Let's be blunt about the economics. A cup of flour costs maybe thirty cents. A dedicated sourdough starter from a bakery runs fifteen to twenty-five dollars shipped. The math favors neglect — but only if you know what you're looking at. The common pitfall is misreading hooch as rot. Dark liquid, sure. Smell it: if it reminds you of nail polish remover or overripe pineapple, you're fine. If it smells like a wet dog that found a dead fish, then you have actual spoilage. That's the edge case. But 90% of the jars I have rescued just needed a pour-off, a scrape, and a fresh feeding. Not a funeral.

Most teams skip this — they see the separation layer and panic. Don't. Wrong order: pour off the hooch, discard all but a tablespoon of the old paste, then feed at 1:5:5 ratio (starter:flour:water). That dilution shocks the culture back into activity. By the second feeding, you'll see bubbles. By the third, you'll be back to your normal rhythm. The hard truth is that dormancy has limits — six months is usually survivable; two years is a gamble — but within that window, you're not reviving a corpse. You're waking up a sleeping partner. One that costs you nothing to keep, and everything to replace.

What a Dormant Pre-ferment Actually Is

Yeast and bacteria in suspended animation

Picture a jar of flour-and-water paste sitting in your fridge at 4°C. The yeast cells—Saccharomyces cerevisiae for most of us—stop reproducing. They slow their metabolism to a crawl, entering something closer to torpor than full sleep. The lactic acid bacteria that give sourdough its tang? They follow suit, but not before dropping the pH slightly lower than usual. That acid shift is the first clue: dormant doesn't mean dead. A starter I forgot for six weeks once came back after three feeds. Smelled like acetone for two days, then settled into its familiar winey funk. The catch is that enzymatic activity—those amylases and proteases breaking down starch and protein—keeps humming along even when the microbes have clocked out. So the flour in your jar is still quietly fermenting, just at a glacial pace.

The role of temperature and pH

Temperature is the off-switch. Above 10°C, most yeasts rouse themselves. Below 5°C, they drift into dormancy but the enzymes don't get the memo—they keep snipping starch chains into simple sugars. This is why a refrigerated starter can develop a dark liquid layer on top (hooch) after two weeks. That dark liquor is the metabolic waste of enzymatic work, not microbial activity. Wrong assumption: many bakers see hooch and think their culture has died. Usually it hasn't; the yeast is just sitting with its thumb up, waiting for warmth. The pH meanwhile continues a slow drift downward—another reason why a dormant pre-ferment can taste more acidic than a freshly fed one. That's normal, unless the pH drops below 3.8 for extended periods. Below that threshold, the bacteria themselves start struggling. I've lost one starter that way—left it three months, came back to a greyish paste that smelled like gym socks and never recovered. So dormancy buys you time, but not forever.

Not every baking checklist earns its ink.

Not every baking checklist earns its ink.

"A dormant pre-ferment is not a pause button. It's a dimmer switch—the lights stay on, just barely."

— overheard at a baking conference, 2023, where a panettone specialist was explaining why she feeds her mother dough every five days even in cold storage.

Why 'dormant' doesn't mean 'inactive'

The biggest pitfall is treating your cold jar like a museum piece. It isn't. Enzymes are still chewing through the food supply. The starches that would have taken eight hours to break down at room temperature now take three weeks—but they do get there eventually. That means a pre-ferment left dormant for a month will run out of fermentable sugars. The yeast and bacteria then start cannibalizing each other's waste products. The pH climbs slightly, weird volatile compounds accumulate. When you wake it up, you get sluggish bubbles and a flat loaf. Common fix: discard all but 10 grams, feed with fresh flour and water at 28°C, and give it two feedings twelve hours apart before using it. That usually flushes the metabolic trash. But here's the hard trade-off: the longer the dormancy, the more feeding cycles you need. A one-week-old dormant starter revives in one feed. A two-month-old one might take three or four. The microbial community composition also shifts—some strains outlast others in the cold. Your revived starter might lean more acetic (vinegary) than it did before. That hurts if you're chasing a neutral crumb for brioche. The fix is to feed it small, frequent meals at warm temperatures for a few days—rebalance the bacteria-to-yeast ratio. Most bakers skip this step. Their first loaf after revival comes out tight and sour. That's the signal: your dormant pre-ferment woke up a little different than it went to sleep.

The Cold War Inside Your Jar: How Dormancy Works

Temperature’s effect on metabolic rates

Put your jar in the fridge, and you're effectively hitting pause on a brawl. The cold slows everything—yeast cells slog along at maybe ten percent of their room-temperature pace, and lactic acid bacteria (LAB) grind down even harder. I have watched a vigorous 100% hydration starter go from doubling in four hours to barely stirring for three days. That seems peaceful. But the metabolic machinery doesn’t stop; it just shifts into a low, grinding gear. Yeast still burps CO₂, LAB still excretes lactic and acetic acids, only now the waste products accumulate faster than the microbes can buffer them. The catch is that cold favors tough, acid-tolerant strains. Your cherished saccharomyces cerevisiae might fade while a grittier lactobacillus hangs on. Wrong order? Not yet—but you're letting the jar’s internal chemistry drift toward sour dominance the longer it sits.

Acid buildup and microbial shift

Here is the cold war: every day in the fridge, pH creeps downward. A fresh starter sits around pH 5.0–5.5. After a week of dormancy, I have measured pH 3.8. That's sharp enough to suppress the very yeast you need for lift. What usually breaks first is the gas production—your starter looks alive, bubbles faintly, but fails to double because the yeast population has thinned. The LAB, happy in low pH, keep producing organic acids. That explains why a neglected starter smells vinegary or like cheap nail-polish remover: the acetic acid fraction spikes. One baker I worked with kept a rye starter dormant for six weeks. When she revived it, the dough tore like wet cardboard. We fixed it by discarding 90% of the acidic liquid and feeding twice a day for three days. The pH recovered; the yeast bounced back. But the clock matters—every extra week in the cold accelerates the microbial shift.

‘Cold slows metabolism unevenly—LAB limp along at 8% speed, yeast at 12%. That gap rewrites the population.’

— observation from a bakery trial, not a peer-reviewed stat—just what we saw over ten cycles.

Most teams skip this: the ratio of acetic to lactic acid rises as temperature drops. Your fridge’s typical 4°C favors acetic acid production, which tastes sharper and inhibits yeast more than lactic acid does. That's why a starter that sat for three weeks may need two extra feedings before it leavens a loaf again. You're not nursing a sleepy pet; you're rebalancing a community that drifted into extremophile territory.

Why some starters survive longer than others

High-hydration starters (100% or above) buffer acid swings better because the water dilutes the toxic load. A stiff starter at 60% hydration crashes faster—less water means acid concentrates in the micro-zones around each cell. I keep a liquid levain at 125% hydration for that exact reason; it has survived a four-week vacation twice without full resuscitation drama. Conversely, whole-grain starters carry more buffering minerals (ash, phytic acid) and a broader microbial starter pack—they often limp through two months cold while a white-flour starter flickers out at week five. The trade-off is that whole-grain dormant starters smell like barnyard and old cheese when they wake up. That's fine—the smell usually clears after two feedings. The hard limit hits when the LAB outgrow the yeast so completely that no amount of warm feeding restores gas production. You lose a day, maybe two. Or you lose the whole jar. That hurts.

One rhetorical question worth asking: does your jar have enough food reserves for the long standoff? Flour in the fridge still converts starch to sugar, but the cold retards enzymatic activity. A starter fed just before refrigeration will have a sugar supply that runs dry around day ten. After that, the microbes start cannibalizing dead cell matter. That's survival mode—not dormancy. You're not pausing; you're running a slow-motion collapse. I have seen starters revived after eight weeks, yes, but they needed a week of intensive feeding and smelled like a stable for the first five days. Not everyone has that patience. If you want a truly dormant pre-ferment that rebounds cleanly in two days, keep the cold period under three weeks and give it one last feeding with twice the usual flour weight before chilling. That buys you margin. Skip that, and the cold war inside your jar tilts toward acid—and you lose the war before you even start the bake.

Reviving a Neglected Starter: A Step-by-Step Walkthrough

Assessing viability: smell, hooch, mold

The jar has been tucked in the back of your fridge for six weeks. Maybe eight. You open the lid and immediately recoil—that sharp, alcoholic punch is pure acetone, and the surface swims in a grey-brown liquid layer. That's hooch. It's not rot, not yet. What you're smelling is the starter's metabolic exhaust: yeast producing alcohol, bacteria generating organic acids. The trick is distinguishing between "fermenting hard and then sleeping" and "actually dead." A dormant starter should smell sour, boozy, maybe a little cheesy—not putrid. If you catch ammonia or something that reminds you of compost bin regret, that's a red flag. Mold is the only hard stop: fuzzy green, pink, or black colonies on top or embedded in the grain mean discard everything. No salvage. But that grey hooch layer? Pour it off, scrape the top 3-5mm of discolored paste away, and check what's underneath. If the remaining starter looks normal—creamy, with small bubbles clustered near the sides—you're likely fine. Most people panic at the smell alone and throw out a perfectly revivable starter.

The feeding schedule that works (and doesn't)

Don't double the feeding. I see this mistake constantly: hungry baker sees hooch, assumes the starter needs more food, and dumps in twice the usual flour and water. That overloads the system—the yeast wakes up to a buffet they can't process, pH drops too fast, and you get a sticky, sluggish mess that takes an extra two days to stabilize. Instead, start lean. Day one: remove all but 20g of the hooch-cleaned starter, feed it 40g flour and 40g lukewarm water. Cover loosely, keep at 75-78°F (24-26°C). Twelve hours later—check for bubbles. You might see nothing. That's normal. Day two: same ratio, no increase. What usually breaks first is the bubble activity: tiny pinprick holes across the surface by hour 14, then a gentle dome at hour 18. Don't chase the peak yet. The catch is that dormant starters hit a false peak around day three—they double quickly but lack the microbial balance for proper fermentation. You'll think you're done. Not yet.

“I once revived a starter that had been dormant five months. Day four it doubled in six hours. Day five it collapsed and smelled like old socks. Day seven it made a loaf that earned a spot in my permanent rotation.”

— personal note from a baker who learned patience the hard way

Odd bit about baking: the dull step fails first.

Odd bit about baking: the dull step fails first.

How long until it's ready to bake?

Optimistic answer: four to five days, if you're feeding twice daily by day three. Realistic answer: a full week. The key visual cue isn't height—it's consistency of rise. A revived starter that triples one day, then barely crests 50% the next, isn't stable yet. That yo-yo behavior means the yeast and LAB (lactic acid bacteria) populations are still fighting for dominance. You want three consecutive feeds showing similar doubling times—within an hour window—before you risk a loaf. Worth flagging: the first bake from a revived starter often spreads more than rises. That's not failure; it's the starter's gluten-weakening enzymes being overactive from the long rest. Add 10g of commercial yeast for that first loaf if you need a safety net. Or embrace the pancake—some of my favourite flatbreads came from those transitional bakes. The real threshold? Your starter should float in water (a pinch dropped into a cup of cool water) and smell fruity, not sour-bitter. Until then, keep feeding, keep noting the time-stamps. That jar is teaching you patience in grams and hours.

When Dormancy Goes Wrong: Edge Cases

Mold invasion and when to toss it

You pull the jar from the back of the fridge and see it: a fuzzy gray island floating on top, or worse—pink or orange streaks bleeding into the hooch. That pink hue? Not a funky sourdough mutation. It's Serratia marcescens or a related opportunistic bacterium, and it means the culture has gone septic. I have seen bakers try to scrape off the top layer and feed the rest. Don't do this. Mold sends roots deeper than you can see, and those pigmented streaks signal toxins that no feeding cycle will fix. Toss the whole jar. Wash it with boiling water and start fresh. The one exception: a thin layer of clear hooch with no discoloration—that's just your starter being lazy, not dangerous.

Over-acidification and 'hunger'

Acetone smell. Nail polish remover hitting your face when you crack the lid. This is your culture screaming—it has run out of food and the yeast are producing volatile organic acids as they die off. The catch: you can sometimes fix this, but only if you catch it early. Dump the liquid, take a tiny tablespoon of the remaining paste, and feed it at a 1:5:5 ratio (starter:flour:water) every 8 hours for two days. Worth flagging—if the acetone persists past the third feeding, the yeast population has crashed below recoverable levels. I had a jar of rye starter once that smelled like a chemistry lab for a week before I admitted defeat. That hurts. But over-acidification isn't always a death sentence; the real killer is when the bacteria outpace the yeast completely.

What usually breaks first is the bacterial-to-yeast balance. In a healthy dormant pre-ferment, the lactobacilli produce lactic acid slowly during cold storage. When you neglect it too long—say, three months unfed—the bacteria keep cranking acid while the yeast starve. The pH drops below 3.8, and the yeast simply can't function. You end up with a culture that bubbles but never rises. That's split culture territory: live bacteria, dead yeast. The fix? Not the usual feeding routine.

Separating yeast from bacteria in a pinch

Most teams skip this: you can actually isolate a small yeast population from an over-acidified starter by playing with pH. Take a tablespoon of the soured paste and mix it with a high-ash whole wheat flour—the buffering capacity of the minerals will raise the pH just enough. Feed it warm (28°C) at 1:10:10 for three consecutive feeds. The yeast, given a less acidic environment and abundant food, will outcompete the bacteria temporarily. I have used this trick twice—once on a 14-month-old dormant rye starter that smelled like pickle brine. It worked on day four, but the starter was weak for another two weeks. That's the trade-off: you might save the lineage, but you lose a week of reliable baking.

'If the hooch smells like a hospital and the starter looks like oatmeal with glitter—throw it. Some things don't come back.'

— overheard at a baker's swapping table, after someone tried to revive a jar that had been in a moving box for six months

The hard boundary? If the culture has visible mold, any pink or orange pigmentation, or a smell that makes you gag—not just sour, but putrid—it's gone. Acetone you can fight. Split cultures you can rebalance. But mold and Serratia are non-negotiable toss conditions. One rhetorical question worth asking yourself before you start: how many days are you willing to spend nursing a jar that should have been composted last month?

The Hard Truth: Dormancy Has Limits

Why yeast dies before bacteria

Pop the lid on a jar that’s been sleeping six months and you’ll smell it before you see it. Acetone. Nail polish remover. The yeast colony is already in retreat. What most bakers don’t realize is that yeast and lactic acid bacteria don’t age at the same speed. Yeast runs out of glycogen reserves faster—its metabolism demands energy even at 4°C, just slower. Bacteria, especially the lactobacillus strains, can idle on almost nothing. They scavenge dead yeast cells, produce acetic acid, and gradually shift the pH downward. The result? A starter that smells sharp, tastes sour, and takes three days to show a single bubble. I have seen people dump a six-month-old starter after one failed feed, assuming it was dead. It wasn’t dead—just yeast-starved and bacteria-dominant. Wrong order to fix it: you need warmth, whole-grain flour, and patience. Not a sugar splash. Not a commercial yeast pinch.

The catch is that revival time scales non-linearly. Two weeks in the fridge? Fine. Two months? Manageable. Six months? You might be rebuilding the yeast population from single digits. That hurts.

The fridge isn't a pause button

Most teams skip this distinction: refrigeration slows decay but doesn't stop it. Every day in the cold, the starter loses a fraction of its enzymatic power. Amylase activity drops. The microorganisms enter a maintenance-only state, but maintenance still costs energy. They burn through stored carbohydrates, produce metabolic waste, and slowly acidify their own environment. Think of it as a very slow fire, not a frozen snapshot. Freezing, by contrast, actually halts biological activity—ice crystals rupture cell walls, sure, but the survivors enter true stasis. A frozen starter can sit for a year and bounce back in two feeds. A refrigerated starter that sits a year is a chemistry experiment, not a baking ingredient. Worth flagging—commercial bakeries I have worked with rotate their refrigerated pre-ferments every 10-14 days. They don't trust the fridge as a long-term solution. Neither should you.

One baker I know stored his liquid levain for four months, fed it once, and baked a loaf that spread sideways like a pancake. The gluten had degraded. The acid had chewed through the protein network. The fridge had not paused anything—it had just slowed the collapse.

“The fridge delays the inevitable. It doesn't cancel it.”

— conversation with a sourdough consultant who lost a 40-kilo batch this way

Honestly — most baking posts skip this.

Honestly — most baking posts skip this.

How long can you really store a starter?

Here is the honest answer without the internet hyperbole: a well-fed, mature starter kept at 4°C in a sealed jar can go maybe 3-4 weeks before quality drops. After that, you're gambling. The yeast count declines measurably after week two. The bacterial balance tilts toward acid producers. The aroma shifts from fruity to harsh. By week eight, your revival feed might require a 1:10:10 ratio and a 36-hour warm ferment just to wake the yeast. By month four, you're essentially starting over—with the advantage of a more resilient bacterial strain, but at the cost of time. Freezing buys you 6-12 months with minimal loss. Drying flakes buys you years. But refrigeration alone? That's a 30-day window, maybe 45 if you feed every week and tolerate some sourness creep. Pushing past that's not bravery—it's wishful thinking dressed as frugality.

What to do next: label every jar with the last feed date. If it passes 30 days, either refresh it or freeze a portion. Don't assume the fridge is a magic coffin. It's just a colder room. And time still moves there.

Reader FAQ: Dormant Pre-ferments Edition

Can I use the hooch?

Yes — with one grubby caveat. That dark, grayish liquid sitting on top of your neglected starter is alcohol, a byproduct of yeast activity when food runs out. Pour it off or stir it back in? I pour it off. Every time. Stirring hooch back into the mix drops the pH too fast, making your starter acidic before you even start feeding. The result: sluggish fermentation and a sourness you can't control. If the hooch is clear and smells faintly boozy — that's normal. If it's dark brown and smells like a frat house carpet, your starter has been neglected for weeks, not days. Worth saving? Usually. But the first few feeds will smell funky. Toss the hooch, take 50g of the sludge underneath, and start a fresh feed.

Do I have to feed it every week?

Short answer: no. Longer answer: depends how long "dormant" needs to be. A stiff starter in the fridge — 50% hydration or less — can go two weeks without a feed. I have pushed three. The trick is the ratio. Less water means less enzymatic activity, fewer microbes eating themselves into a coma. A liquid starter? That needs weekly attention. Skip two weeks and you're looking at mold or that terrifying orange-pink ring that means Pseudomonas bacteria. That's a hard discard. Not a salvage job. So match your feeding schedule to your starter's hydration. Stiff? Two weeks. Liquid? Seven days, max. Between these extremes? I keep a jar of stiff starter for exactly this reason — it buys me time when life gets chaotic.

Why does my starter smell like nail polish remover?

That's acetone — a sign your starter is starving, not dying. Acetone production happens when yeast metabolizes stored glycogen under stress. It's common. Annoying, but common. The fix: two warm feeds at 12-hour intervals, discarding 80% each time. The smell should vanish by the second feed. If it doesn't — if the acetone lingers through day two — check your water. Chlorinated tap water can suppress the microbial balance that keeps acetone in check. Switch to filtered or let tap water sit out overnight. What if the smell shifts from acetone to vomit? That's butyric acid, a sign of Clostridium contamination. Different problem. Toss the starter, sterilize the jar, and borrow a friend's backup.

I once revived a starter that smelled like a nail salon for six straight days. It baked fine on day seven. The bread tasted clean. Smell is not always death.

— personal anecdote, not a promise. Some starters just take longer to recover.

Here is the hard line: if your starter shows fuzzy mold — green, blue, or that alarming pink-orange — don't try to scrape it off. Mold sends roots deeper than you can see. One gram of infected starter ruins a whole batch. Toss it. Clean the jar with boiling water. Start over. The difference between "dormant" and "dead" is whether the microbes can bounce back with food and warmth. Mold means they already lost the battle. Your next move: feed a small backup from a friend or a dried stash. That's why I keep a dehydrated sample in my freezer. Not paranoid — just prepared.

What to Do Next: Practical Takeaways

A simple feeding schedule for fridge storage

Stick with 1:2:2 — one part starter, two parts flour, two parts water — before you slide the jar into the fridge. That ratio buys you roughly ten to fourteen days of quiet dormancy. I have pushed it to three weeks once, and the jar smelled like acetone and regret. Feed it straight from the fridge; no need to let it warm up first. Grab a clean spoon, scrape off any dark liquid (that's hooch, not a seasoning), and discard all but 20 grams of the old starter. Then add 40 grams flour and 40 grams water, stir until no dry patches remain, and loosely cap the jar. The catch is temperature: a fridge at 38°F (3°C) slows fermentation almost to a stop. Your crisper drawer runs colder — risk a longer dormancy there, but check after day seven. Wrong order. Feed first, then chill. If you feed after refrigeration, the microbes get cold-shocked and the rise flattens.

Signs your starter is ready to bake

You pull the jar from the fridge. It looks flat, maybe a thin grey liquid on top. That's normal. Don't pour it off — stir it back in. Then do one maintenance feed at room temperature (21°C, 70°F) and wait. The real test comes eight to twelve hours later: the starter should double in volume, dome slightly, and smell fruity, not sour. If it barely bubbles after eighteen hours, the microbes are still sluggish. Try a second warm feed. What usually breaks first is patience — bakers toss a candidate starter because it rose only 60% instead of 100%. I have baked with a starter that rose just 70% and the crumb was fine. Not gorgeous, but edible. A rhetorical question for the skeptic: how much spring do you actually need for a focaccia or a sandwich loaf? The answer matters because an underfed starter produces dense bread, but an overfed starter dilutes flavor. Trade-off.

‘If the discard pile smells like old gym socks after two feeds, stop. That starter is dead, not dormant.’

— overheard at a weekend bake-off, bakery owner explaining when to quit

When to start over (and not feel bad)

Three warning signs. One: the starter shows no activity after three consecutive room-temperature feeds spaced twelve hours apart. Two: the smell turns putrid — not sour or yeasty, but like rotten vegetables. Three: you see pink or orange streaks. That's mold or bacterial infection, not a phase. Cut losses. Dump the jar, wash it with hot soapy water, and begin fresh with 50 grams whole-rye flour and 50 grams unchlorinated water. The hard truth is dormancy has limits — some starters simply don't survive a four-week fridge nap. I have lost two jars that way, one because the lid was too tight and gas built up, another because the kitchen power went out for a weekend and the fridge thawed to 15°C. You can't save every jar. Start over, name the new one something funny, and move on. Specific next action: label your jar with the date of last feed and a sticky note that says ‘check in 14 days.’ That returns consistency without guesswork.