Diluting soy milk (using tapwater left to stand in an open-topped bottle overnight) gives curd:whey ratio slightly better than 1:2. Gallery contents left too long on one occasion and entire tube had to be removed and cleaned out using small bottle brush attached to long dowel. Problems with getting (part) fermented milk to leave the gallery – sometimes the spigot has to be left open for the best part of a minute before contents begin to drain. Dilution seems a good answer to problem of curds setting inside the gallery.
Set up a control (same amount of milk mixed with small amount starter culture in a bottle for similar length of time). Most notable feature was a slightly cheesy smell which increased with time, indicating aerobic fermentation. This should have been predictable, but underlines the need for anaerobic conditions, if cheese curds are not the desired outcome.