Grams of dry malt extract per litre of starter wort: 100.
The standard homebrew starter recipe is 10 : 1 water to DME by weight.
100 g of DME into a kilo of water is 100/1100 = 9.1 °P, which is
STARTER_GRAVITY. Starter calculators default to this gravity, and
it's close to the ~10 °P (1.040) Kai ran his growth experiments at.
Building a starter at a different gravity doesn't just rescale the model.
Kai tested it and found 5, 7 and 10 °P all plateau together, while 20 °P
grows visibly less yeast and lowers viability. So we plan at the gravity the
model was measured at and report the DME, and don't offer a gravity setting.
Grams of DME versus grams of extract
Braukaiser's axes are per gram of extract. Ours are per gram of DME, and we
treat the two as the same number. The published calculators don't agree on
the conversion either: Brewer's Friend works out 98.5 g of extract in a
litre at 1.037, Homebrew Dad 100.9 g. That's a 2.4 % spread, and our flat
100 sits between them. Kai puts the error bar on his model at "+/- 15 % or
more", so the difference between 98.5, 100 and 100.9 is an order of
magnitude below the noise. What matters is using one convention
consistently, and 100 g/L is easy to weigh out.
Grams of dry malt extract per litre of starter wort: 100.
The standard homebrew starter recipe is 10 : 1 water to DME by weight. 100 g of DME into a kilo of water is 100/1100 = 9.1 °P, which is STARTER_GRAVITY. Starter calculators default to this gravity, and it's close to the ~10 °P (1.040) Kai ran his growth experiments at.
Building a starter at a different gravity doesn't just rescale the model. Kai tested it and found 5, 7 and 10 °P all plateau together, while 20 °P grows visibly less yeast and lowers viability. So we plan at the gravity the model was measured at and report the DME, and don't offer a gravity setting.
Grams of DME versus grams of extract
Braukaiser's axes are per gram of extract. Ours are per gram of DME, and we treat the two as the same number. The published calculators don't agree on the conversion either: Brewer's Friend works out 98.5 g of extract in a litre at 1.037, Homebrew Dad 100.9 g. That's a 2.4 % spread, and our flat 100 sits between them. Kai puts the error bar on his model at "+/- 15 % or more", so the difference between 98.5, 100 and 100.9 is an order of magnitude below the noise. What matters is using one convention consistently, and 100 g/L is easy to weigh out.