Every other recipe on this site tells you what to weigh. This one asks. Put whatever you're actually using into the bench below — flour, water, starter, salt, and anything else you feel like throwing in — and it writes the rest: the true hydration, the baker's percentages, what your particular combination is going to do to the dough, and a method with the numbers already filled in.
It starts you on the Calibration Loaf, which is the plainest useful loaf we know how to describe. Change one thing or change all of it. Water is just the default liquid — pour beer instead, or half beer, and the numbers move with you.
Have at it. Have fun with it. We are, nominally, trying to make something edible here — and the tool will say so when you wander off that path. It will not stop you. Somebody has to find out what happens with the Dr Pepper, and it is not going to be us.
These are the Calibration Loaf's numbers — 450 g of white bread flour, 300 g of water, 100 g of starter, 10 g of salt. They're here as a starting point, not an instruction. Change anything and they stop being greyed out.
Everything in grams. The percentage beside each one is worked out for you.
Water is just the default one. Swap it out entirely or pour half and half — every liquid carries a different amount of actual water, and this keeps track.
Fats, sugars, and anything you fold in. Optional, and where this gets properly interesting.
| Ingredient | Weight | Baker's % |
|---|
These change as you build. They're about what the dough will feel and do — a number can't tell you that a 30% rye dough is going to stick to everything you own.
They are tips, not physics. The flour absorption figures are measured. What happens when you pour a bottle of Dr Pepper into a levain is not — that one is "here's what usually goes wrong", and the only way to know for certain is to bake it.
Written for 500 g of flour at 70% hydration, finishing at about 860 g of dough. Times are counted from the moment you mix, at 74°F / 23°C — this dough runs at about the pace of a plain white loaf. Turn it into a wall clock if you need it out of the oven at a particular time.
- Feed the starterBefore you mix · 4–6 hrs ahead
You need 100 g of starter at peak, so feed at least 120 g to leave yourself a margin and something to keep. A 1:1:1 by weight at room temperature gets you there in roughly 4–6 hrs at 74°F.
Peak means roughly doubled, domed, smelling yeasty rather than sharp. Past peak it has spent its rise and your bulk will crawl.
Don't have a starter yet? Nothing else here works without one. Make one from scratch — flour, water, and about two weeks. - Mix the flour and the liquid+0:00 · 5 minutes, then 1 hr rest
Whisk 450 g of flour and 300 g of warm water together until no dry flour is left anywhere. It will look shaggy and unpromising. That is correct.
Cover it and leave it 1 hr. The flour hydrates fully and the gluten starts organising itself without you kneading anything.
Why an hour, and when to skip it Autolyse is optional and behaves differently with whole grain — the full case, both ways. - Add the starter and the salt+1:00 · 3 minutes
Add 100 g of peaked starter and 10 g of salt straight on top. Wet one hand and squeeze the dough repeatedly, folding it over itself, until there are no streaks. About three minutes.
It will feel slack and slightly wrong. It is supposed to. Do not add flour.
- Bulk ferment, with folds+1:05 · 4–5 hrs · 4 sets
Cover and leave it at 74°F. Do one set of folds at 45 min, 1.5 hr, 2.3 hr and 3 hr, then leave it completely alone.
This blend runs at about the pace of a plain white loaf. The bracket above is where to start looking, not a timer.
New to folding? How hard to pull, when to stop, and why more sets are not better — stretch and fold. - Judge the bulk — the one that matters+5:35 · volume, not the clock
This is the skill. Everything else here is mechanical; this is the part you are actually learning.
Stop when three things are true: the dough has risen about 50%, the edges where it meets the bowl are domed rather than sunken, and the whole mass jiggles as one soft body when you shake it.
If the bracket has run out and it has not done that, wait. If it got there early, move on.
This is the skill Volume, dome, jiggle, and the aliquot-jar trick that turns "50%" from a guess into a measurement — reading the bulk. - Pre-shape and rest+5:35 · 5 minutes, then 20 rest
Tip the dough onto an unfloured counter — you need the friction. Fold the edges in, flip it seam-side down, and drag it toward you a few times to tighten the surface. Leave it uncovered for twenty minutes.
- Shape+6:00 · 5 minutes
Flour the top, flip it, and shape into a boule or a bâtard. What you are building is surface tension — a skin tight enough to hold the loaf up while it springs. Seam side up into a floured banneton.
Tension, not decoration The three shapes and how to know you have enough tension — shaping. - Cold retard+6:05 · 12–16 hours
Straight into the fridge. This buys flavour, makes the loaf far easier to score, and puts the bake on your schedule instead of the dough's.
What the cold is doing Cold slows yeast far more than bacteria, which is why you get flavour without extra rise — the cold retard. - Score and bakeNext morning · 20 min covered, 23 min uncovered
Preheat the Dutch oven to 500°F / 260°C for a full hour. Turn the cold loaf out onto parchment, score it decisively in one motion, and get the lid on. 20 minutes covered, then drop to 450°F / 230°C, lid off, for another 23. Internal temperature should read 205–210°F / 96–99°C.
Take it darker than feels comfortable. Pale crust is the most common self-inflicted wound in sourdough.
Steam is the whole trick Why covered-then-uncovered works, and what to do with no Dutch oven — steam and the bake. - Cool — properlyAt least 2 hours
On a rack, not a board. The loaf is still cooking as it cools and the crumb is still setting. Cutting it hot gums it, and you will blame the recipe.
Notes from the counter
The Calibration Loaf exists to hold everything still so you can learn to read one thing. This page is the opposite, and you should probably bake that one first.
But holding things still forever is its own trap. At some point the useful move is to change one variable on purpose and find out what it does — and the reason people don't is that changing the flour means redoing the water, and nobody wants to do arithmetic standing up.
So that's what this is for. Not to design a perfect loaf. To make the experiment cheap enough that you actually run it.
| What happened | Usually | Go here |
|---|---|---|
| Dough was unmanageable | Too much water for that flour | Sticky dough |
| Spread out flat | Whole grain ferments faster than you expected | Flat loaf |
| Dense and tight | Too much whole grain, or bulk cut short | Dense crumb |
| Never really rose | Fermentation, almost always | No oven spring |
If you built something here and actually baked it, we want to hear about it — the ones that worked and, honestly, the ones that did not. Everything on this page is built out of what usually happens. Your loaf is what actually happened, and that is worth more.
The box below fills in your formula automatically, so you only have to write the interesting part.