Kojiology

Before the spores · soak, steam, cool

Substrate Prep

Soak 2-4 hours, steam 15-45 minutes, aim for 25-35% moisture. Both ranges are wide because they depend on things nobody writes down — how much you are cooking, how cold it went into the pot, what shape your lid is, how you cooled it. This works out where in them your batch actually lands, and tracks every gram of water on the way.

New to this? Start here

Koji is a mould, and moisture is the single thing it is fussiest about. Too dry and the starch never gelatinises, so there is nothing for the enzymes to work on. Too wet and the grains stick together, the mycelium sprawls across the surface instead of digging in, and Bacillus gets a warm nutritious bath. The surprise is where the water comes from. Soaking is the obvious step, but steaming adds around 100 g of water for every kilo of dry grain you started with, by condensing on the food, and cooling removes a chunk of it again — so how you cool matters as much as how long you soak.

Wet-basis moisture
water as a percentage of total weight. 30% means 300 g of water in a 1 kg batch. Every figure here is on this basis.
Gelatinisation
starch absorbing water and softening under heat. It needs both, which is why an unsoaked grain stays chalky inside no matter how long you steam it — and why amylase can barely touch it.
Condensate
steam turning back into water on the cold food. It is how a steamer delivers heat at all, and it is water you did not measure out.
Al dente
the texture target — resistant to the bite, never mushy, never raw. It is a moisture reading in disguise.

Your batch

The reference case behind the familiar numbers: a 2-4 h soak and a 15-45 min steam. Both ranges are load- and temperature-dependent, which is what this tool resolves.

kg
°C

Soak & drain

3.0 h
3 h of 20°C-equivalent exposure
°C
min

Steam

cm → 4.1 cm deep
W · sets come-up only
m · sets the steam temperature
Lid

Condenses over the whole underside and drops it straight onto the food.

Cool down

How you cool it

The most underrated moisture lever here. Spores need the substrate below 40°C either way — but how you get there decides whether the batch sheds 46 g of water or almost none.

Finishing

34% · oven step only

Set it below where your process naturally lands and the tool works out how much water has to be driven off, and how long that takes.

Spores go on at
36.3%
1382 g, 1.38× what you started with
Harvests around
32.1%
inside the 25-35% band
Steam for
26 min
6 min to come up, then 20 min held
Where the batch sits, and where it drifts
10%20%30%40%50%60%70%
into the steamer
32.4%
spores go on
36.3%
harvest
32.1%
25-35% target, during incubation30-39% the same band, at loading
In the steamer

A steamer has no thermostat. It runs at whatever temperature water boils at where you are, delivers heat by condensing water onto the food, and loses some of that heat through the lid — which comes straight back down as drips.

Steam temperature
100.0°C
101 kPa at 0 m
Goes in at
21°C
a 79 K climb
Come-up energy
239 kJ
c_p 2.32 kJ/(kg·K)
Condensed on the food
106 g
energy ÷ 2257 kJ/kg latent heat
Dripped off the lid
21 g
over 26 min
Bed depth
4.1 cm
in a 24 cm basket
Cooling down
Cool it to
32°C
spores go on below 40°C
Evaporated on the way
46 g
45% of the heat, over ~35 min
In the 35-50°C band
13 min
no spores on it yet
Optional: oven-dry to reach 34%
Water to drive off
49 g
at 100°C
Roughly
12 min
on the same tray area
Steaming does not sterilise.Bacillus cereus and B. subtilis ride in on raw grain as endospores and survive to about 120°C — killing them takes hours in a pressure cooker at 121°C, which would destroy the substrate. Prep lowers the microbial load and opens the starch. It does not clear the field, so the batch still has to be won on inoculation dose and speed.
The lid alone drips about 21 g back onto the food.A flat metal lid condenses steam over its whole underside and drops it straight down. The traditional fix is a cloth: cotton or boiled cheesecloth under the lid absorbs the condensate and releases it as vapour instead of as drips, which is why traditional steaming is done through cloth rather than in spite of it. A domed lid runs most of it to the rim instead.

Every gram of water, step by step

Through every prep step the dry solids block stays exactly the same width — nothing you do to a grain adds or removes solids. Everything else is water moving in and out, and each step is a lever you control. Only the last row breaks it: the mould respires some of the substrate away as it grows.

  1. As bought1000 g · 12.0%

    12% moisture, 880 g of dry solids that stay constant from here on

  2. Soak · 3.0 h at 20°C+463 g water · 1463 g · 39.8%

    239 g absorbed into the grain — 93% of what it can hold, and the curve is nearly flat past here — plus 223 g clinging to the outside

  3. Drain · 30 min161 g water · 1302 g · 32.4%

    161 g shed, 62 g still clinging to the outside — this is the water that turns a steam into a boil

  4. Steam · 26 min at 100.0°C+127 g water · 1429 g · 38.4%

    106 g condensed bringing it up to temperature, 21 g dripped off the lid

  5. Cool to 32°C · ~35 min46 g water · 1382 g · 36.3%

    45% of the heat released leaves as vapour, taking water with it

  6. Incubation · 48 h, ~10% mass lost103 g water · 1244 g · 32.1%

    Respiration and sweating over the cycle — 1000 g in, about 900 g out. Mostly water, so the substrate dries as it grows

dry solids — fixed through prep, and only the mould's own respiration touches them water

Why soaking is not about softness

Starch needs water as much as it needs heat. Where the substrate sits on this axis when it goes into the steamer decides whether amylase has anything to eat at all — yours goes in at 32.4%.

Chalky core
below 25%

Starch needs water as much as it needs heat. Short of about a quarter moisture it cannot fully gelatinise at any steaming temperature, and un-gelatinised starch is close to invisible to amylase. The koji still grows — it just has far less to work with, which is why an under-soaked batch can look perfect and taste of nothing.

Firm outside, cooked through
25-45%

The window called al dente: resistant to the bite, never mushy, never raw. Gelatinised on the inside where amylase needs it, dry enough on the outside that the grains stay separate and the mycelium is forced to go inward rather than sprawl across a wet surface.

Your batch
Sticking together
above 45%

Grains gelatinise into one another and free water sits on the surface. This produces "lazy" koji — flawless white fluff on the outside with no penetration and very low enzyme yield, because the mycelium never had a reason to dig. Wetter still and it is a Bacillus culture with a mould problem.

FAQ

Next: how much starter this batch needs is Spore Inoculation, the chamber it grows in is the Incubator Simulator, and steering the result toward sweetness or umami is Enzyme Balance.