@Jaen I like to have excess starter for naan flatbreadd and pasta, so I can relate to your Einkorn waffle situation. Lol’ing at forgetful vs. busy.
@susanmcc99 That spelt bread looked amazing and it’s great to know the micro levain approach works well with that wheat as well.
@Ninerfantom yep, in such small amounts (5g flour in 10g of 100% hydration starter), you’re not changing the bread much using AP vs. Rye etc.
@PatSalvant I’ve made naturally leavened dosas and GF buckwheat bread, but am not an expert. You might check out this book on gluten free baking, if you haven’t already. It look really comprehensive.
@jacklongfellow Sounds like you have worked out a good system.
@jmandel I’ve been thinking of a large amounts of unfed starter type experiment too…I suspect that would be more noticeably problematic. More experiments needed for sure!
@pallecuador So neat. Thanks for sharing. I love how bread baking has attracted so many people in STEM and we all benefit!
That’s a great experiment. I would be interested to publish the results in BROT, our German bread magazine. Any chance to get in touch about that? My email: [email protected]
It seems you have two variables here. Fed/unfed and 20g/80g.
Perhaps if you arranged your bake so that one would compare the effects of the starter size and another the effects of the nature of the starter used.
Just a thought.
Thanks for sharing your finding.
Very nice article. I love baking experiments AND challenging convention. A few years ago I was baking 2 or 3 loaves every morning and taking them into work. One was a no-knead sourdough recipe that comports with the notion that sourdough and starter is quite flexible.
I tuned Eric’s no-knead sourdough recipe to bulk ferment for 24 hours in a modified wine fridge at 65°F. That way, each day I could turn out and shape one dough for baking, and at the same time start another fermenting (I went for months at a time not washing that mixing bowl ). At the time, I didn’t realize how flexible starter could be regarding regular feeding, but did come to realize how flexible starter could be regarding consistency.
I grew tired of the thick glop and mess and stirring thick stuff and trying to measure it accurately. So, I experimented with making the starter thinner and thinner until I realized I could keep it in a condiment squirt bottle. Going this route, I only ever mixed in new flour and water by adding it in and then shaking the bottle (finger over nozzle). It was so slack and easy to accurately dispense.
At the same time, not knowing better, I kept it fed with religious fervor, where I dispensed a certain amount for the dough and a certain additional amount to exhaust approximately half of the total starter, then exacting amounts of water and “flour” were added. One other flexibility item to note (knowt), I wrote that I fed my starter with “flour” because I used to reserve bench flour and dough scrapings from the table after forming all the loaves each day, and that would become some portion of the flour used to feed the starter.
Looking back now, I wonder if my starter could have been some part bacterial and some part fungal? I wasn’t careful to not use dough bits from yeast breads. It definitely tasted like sour dough, though.
As I type this, there’s an old starter in my regular cold fridge that I haven’t touched for months. Now I’m curious to see how nice of a dough I can get going from it without a revival.
You’re correct! @homebreadbaker and I discussed the two variable issue and decided in the end to go with a test of his method vs. sorta-conventional feeding/amount. But I agree that we don’t have all that we need yet.
I suspect that the unfed starter dough in this experiment doesn’t break down with the long ferment because we are using a small amount of unfed starter, but if we were to use a large amount of unfed starter, we might end up with soup. And maybe that is only a risk in a dough made of weaker gluten flour? I have heard of dough becoming proteolytic with past-ripe starter, but I’m not sure I’ve had the experience myself.
@wendyk320 and I were testing tiny amounts of ripe starter a while back and getting nice results. This is a method put forth by the French baker Yohan Ferrant and some other flour ambassadors.
Sounds like some fantastic method innovations: squirting in liquid starter at mixing time and reusing the bowl indefinitely. Love it. I don’t waste bench flour either.
If you have your rise time down to a science, then I can see wanting to have your feeding schedule also very precise. Predictability may be worth religious fervor lol
My strong suspicion is that you would not get a soup unless the ratio of unfed starter to dough was something like 75:25 (yeah, I mean 75% starter which is obviously ridiculous) and even then you’d have to ignore the rising dough for a while so that it rose and fell. But I also think that if you used even 10% unfed starter that it might have a pretty strong effect on the flavor of the resulting bread and that effect would be in the direction of more sour and even maybe kind of bitter.
The theory I work under with the tiny amount of unfed starter is that even though there is not a lot of active microbial replication and metabolism going on (due to lack of food), there is still a totally viable population of the right kinds of yeasts and bacterial in there and as soon as there is food available (i.e. all that fresh new flour for the bread loaf), that’s going to kick-start activity. The Google tells me that there are billions of yeast / bacteria per gram of starter. I’m totally confident that you could use a tenth of a gram of cold, unfed starter and it would leaven a loaf just fine. It would just take a little longer to get going.
I use a small amount both because I want to prolong the bulk proof and also because I don’t want the “old” flavor of the unfed starter to influence the flavor of the bread. With a tiny amount of starter, I think the whole flavor of the loaf is given by the “fresh” interaction between the microbes and the fresh flour. I then modulate the flavor I get by which flour I use and by directly varying the length of the bulk proof.
I have been using minimum starter for some time and having good results. My procedure is 1000 gm flour (which can be 100% home milled or store bought AP and WW) 1-2 % salt 1-2% starter 70% (store bought) 80% (home milled) water. Room temp bulk ferment approx 13 hrs. Cold bake in pans 60 min at 425 F. Starter can be cold, straight from fridge, or fed to increase volume. The bread is always excellent,
Thank you so much for publishing and performing such a well run study! You are a true scientist and the environment thanks you as well for decreasing unnecessary flour consumption!
Question- why do you do the cold to warm to cold bulk? What endpoints were you using to determine when to switch from one temp to another?
So far I’ve tried your micro-starter twice with great success. Thanks again - I always had a suspicion that successfully going from “Starter” to finished product is really just a continuum with endless variables.
I completely agree with this but with one question. As it really is hard to kill a sourdough starter, there will almost always be active yeast and Lactobacillus even after three weeks. My question is whether the yeast:Lactobacillus ratio in an old starter like this is essentially the same a recently fed starter? My assumption was always that we were feeding starters before preparing sourdough bread not just to have it be active (i.e., both to aid in absolute timing and timing predictability) but also to have a consistent flavor based upon predictable starting yeast:Lactobacillus ratios. My trick knee tells me the old starter will have a lower yeast:Lactobacillus ratio (i.e., a higher percentage of Lactobacillus) and that this may contribute to a more acidic environment resulting in a stronger sour taste and potentially a weaker gluten structure.
Yes, from what I’ve read as the starter ages and the food disappears, the ratio tilts in favor of the bacteria. But my fading microbiology memory tells me that the ratio of different microbes in a culture at any given time is a dynamic equilibrium based on their replication and metabolic activity and the characteristics of the environment. And that balance changes fairly quickly with changes in the environment.
IOW, I think that when you put the old starter into a mixing bowl with a relatively giant amount of flour and water you are feeding it and the same exact thing happens as if you had fed it “before preparing sourdough bread”. The ratio in the dough quickly normalizes to the dynamic equilibrium that obtains under conditions of plentiful food. My belief is that it makes no difference at all whether this happens in the starter jar before you inoculate the bread dough or in the mixing bowl afterwards - especially when using a tiny amount of starter.
That’s all theory. In practice, I can just say that since I started using tiny amounts of old starter vs. the more conventional practice of using well-fed starter in larger quantity, the differences that I have observed are that the bulk proof takes longer to get going (providing something that I think of as a period equivalent to autolysis) and the flavor seems to me to develop more depth.
The side benefit is that the starter “management” becomes a non-event; it allows me to be lazier, more efficient, and less wasteful, all of which nourishes my mind while the bread is nourishing my body.
@Locksley Someday I’m going to try baking from cold. The time that the oven is on it’s not much less then if I preheat for only 30 minutes, bake for about 35. But I am curious to see the differences and it might be nice to not handle a super hot pan.
@happidural That’s cool you’ve already done some micro starter bakes and they worked well! The in and out of the refrigerator (often over and over) is pure scheduling of life and sleep, not fermentation strategy
The bread flour loaves were baked in a preheated (500F for 30 minutes) Challenger bread pan.
15 minutes at 500F lid on
10 minutes at 450F lid on, baking sheet directly under, same shelf
10 minutes at 450F lid off
The whole grain loaves were baked in a preheated (500F for 30 minutes) Breadtopia clay baker (oval).
15 minutes at 500F lid on
10 minutes at 450F lid on
8 minutes at 450F lid off
(one of the loaves looks lighter because it had a little more flour on the skin at the start of baking)
ok i got the idea in part … but if you want to compare unfed bread with well feed bread should not you use same quantity of starter … maybe i missed something here my english fails sometimes.
You’re correct and not alone in seeing this : )
The ideal experiment would have been four doughs: small unfed, small fed, large unfed, large fed. Here is my earlier response and you may want to read that of @homebreadbaker afterward.
Crumb of this week’s loaf leavened with about 1/2 a teaspoon (< 10g) of unfed (~ 4 weeks now) starter straight from the fridge. This is 100% whole grain home milled flour, 2/3 red fife and 1/3 hard white (all the berries from here at Breadtopia). Bulk proof was about 17 hours at something close to 75 F room temperature. Then after shaping the basket went into the fridge for another ~ 18 hours before baking.
Another week of neglect has not hurt my starter’s ability to leaven bread. This week I used about 1g of starter from the jar in my fridge that has not been fed in 5 weeks.
My email: 
). At the time, I didn’t realize how flexible starter could be regarding regular feeding, but did come to realize how flexible starter could be regarding consistency.
It definitely tasted like sour dough, though.