The Fresh Loaf

News & Information for Amateur Bakers and Artisan Bread Enthusiasts

Extremely small percent starter, no autolyse, and long ferment?

jmoore's picture

Extremely small percent starter, no autolyse, and long ferment?

Have any of you experimented with a very small inoculation percentage of starter/levain? I'm curious if I can create as good of a loaf with 1-2% levain as I can with my normal 10-20%. I expect that the required long fermentation time will degrade the gluten structure due to excessive enzymatic activity.

A related question, why not just get rid of the Autolyse and do a longer ferment with smaller inoculation (maybe not 1-2%, but maybe 5-10%)?

I'm testing this today. Doing two trials: one a standard batard and the other a pizza dough. Both at around 1.5% innoculation with ~20% freshly milled spelt.

Let's see what happens. :) I'm projecting pancake for the batard. 

Lechem's picture
Lechem (not verified)

The 'Do Nothing Bread' by Yohan Ferrant of which Teresa Greenway's San Francisco sourdough is based on. 

I always get a delicious loaf good oven spring. 

Suggest you start off simple before adapting but here is an example


DanAyo's picture

I use 2% Levain all the time. You are right. It will take longer to ferment and if left long enough the dough will degrade. Buy boy is the bread good.

I think think Paul, aka Pmcool, has baked bread using only a single gram of starter. 


pmccool's picture

My levain percentages are in the more typical 10-20% range.  


DanAyo's picture

I’m not sure who posted about using a single gram of starter to bake a bread, but I read about it this week on TFL.



Doc.Dough's picture

As already pointed out, you can make bread with very little starter (no levain at all really since the levain is mostly to multiply up the starter so that you can start a batch of bread and not have to wait for a long BF/proof). 

However, there are a few things to be aware of:

First, flour is not sterile so you do run a small risk of allowing competing critters in the flour to dominate the mix if you start with too small an amount of known-good culture.

Second, since the bacteria and yeast spin off water as a byproduct you might want to mix the initial dough a little stiff to start with.

Third, since the growth process is exponential in time with a temperature dependent term, you will have to allow for a fairly large uncertainty in when it will be ready to divide/proof/bake. Initiating a batch with a larger fraction of pre-fermented flour means that you have only a few doubling times between mix and divide.

As for getting rid of autolyse, yes you can.  It is my opinion that autolyse was introduced to make the bakers job easier by letting time substitute for work.  It allows the flour to fully absorb the liquid without any expenditure of energy, after which mixing/kneading is easier and shorter.  There is a claim that autolyse also allows the yeast and LAB to build the cellular machinery that they need to replicate before the salt is added and slows that down, though there is a clear benefit to doing an autolyse without having the yeast in the dough, and in the absence of salt the biological "getting ready to replicate" happen faster (thus autolyse is done before salt is added though perhaps without even yeast or levain).  Do what you want and just deal with it.  If you are still playing with these variables you are not close to optimizing your process for predictability and maximization of throughput per unit of your capital investment.  As a hobbyist you don't care so much and mostly need iteration which becomes the engine for learning.


dabrownman's picture

NAd ANother