Kojiology

Vacuum-bag SSF · backslop + external agitation

Agitation Optimizer

For miso or shoyu (soy sauce's mash stage, called moromi) sealed in a vacuum bag and started with backslop — a scoop of mature miso or moromi saved from a previous batch, used to seed this one instead of starting from spores alone: how often external kneading is actually worth doing, and the point past which more doesn't help.

Why agitate at all?

To finish months of fermentation in weeks. Miso and shoyu mature because koji's enzymes cut the soy and grain apart, and an enzyme only works on what it is physically touching. Sealed in a bag, nothing moves on its own: each pocket of enzyme works through the paste around it, then idles while unconverted material sits a centimetre away. Kneading the bag puts enzymes back against fresh substrate — the same batch, the same salt, the same flavour target, reached sooner. It also spreads the salt and the backslop culture evenly, so no corner ferments at a different speed from the rest.

From outside the bag
the seal never opens. Air is what lets spoilage moulds and film yeasts in, so all the work is done through the plastic.
Why there is a ceiling
once everything is in contact there is nothing left to redistribute, and a stiff paste stops flowing however hard you squeeze. Past that point you are working, not fermenting.
Backslop
a scoop of finished miso or moromi stirred into the new batch — it carries the salt-tolerant yeasts and bacteria, so the batch starts populated instead of waiting.
The bag puffs up
that is CO2 from those yeasts, and it is normal. The timeline below says when to expect it and when to burp it.
kg
°C
40%
Not sure what's safe? Check Salt Compass →
20% of substrate
mm, coarser = slower
5% of total mass — starter from a past batch
Recommended
Every 8.7 h
~25s per session, from outside the sealed bag only · 2.75×/day
Near the rheological-lock ceiling.The recommended frequency is capped by dry-paste rheological lock, not by diminishing returns — going more often than this is modelled as actively counterproductive (gumming the starch network), not just wasted effort.
Backslop source matters.This model only accounts for how much backslop you add, not what's actually in it. Use a known-good, well-characterised starter batch — an uncharacterised “previous ferment” is a real food-safety variable the salt/WPS model doesn't cover.
Achieved WPS
16.7%
a_w ≈ 0.88
Salt rated for
~38 wk
target duration — see the FAQ on week N+1
Agitated window
~13–25 wk
wide range — batch-checked, not lab-precise

Effort vs. yield

More kneading helps — until it doesn't. The marked point is where extra effort stops paying for itself.

lock risk →0/day8/day100%0%

Phase timeline

What to expect and when, at these inputs.

What the three phases are
0 d (backslop bypasses lag)
Lag phase

Backslop inoculum is already active — no adaptation wait, unlike a cold start with no starter.

7.2–21.6 d
Active hydrolysis & gas

Gas production ramps toward its onset around day 12.0. Check bag tautness and burp through a small pinhole (re-seal immediately) during this window — never open the bag.

~13.2–24.5 wk
Slowing / maturation

Gas and local-depletion cycles slow down; agitation can taper off. Watch colour (Maillard browning) and the diminishing-returns curve above, not the clock.

The mechanism