Levain or Lots of Culture
I noticed that most of the sourdough formulas in "Bread" use a levain (or sponge or sourdough if you prefer) of 15-20% of the total flour weight with white flour and 30-40% in the sourdough rye chapter. The levain builds are usually 20% (of the levain) mature culture.
"Baking and Pastry", the CIA baking text (at least the old one I have) writes most of the sourdough breads as straight doughs with 30-40% (of the final dough) mature sourdough culture.
This seemed odd until I realized that if you elaborate out your starter the day before to the quantity that goes into the CIA final (only) dough, that's essentially the same as building a levain. The difference is just that you call it elaborating the culture/starter rather than building a levain.
Is there any real difference between these approaches, and their results, beyond what you call it? I suppose you could use a levain with different flour(s) or hydration than the starter for some effect. But most of Hamelman's formulas remind the reader to pull out a piece of the levain to propagate the starter (suggesting that the levain is the same as the mature culture).
Thanks
Yes, Louis, there is a real difference between these approaches.
Theoretically, sponge&dough with a little yeast and direct methods with a lot of yeast, long fermentation and zero time dough, and other methods of bread making might (or should) produce the same bread, let's say, a baguette, and can be chosen freely just because one schedule is more convenient to the baker than another. But in practice, they do produce from slightly different to distictly different breads and certain methods are prescribed as preferred or even obligatory for those breads.
Starter is a carrier of microbial culture, it's a certain number of cells of yeast and bacteria per gram of starter. It usually has a certain acidity and aroma, if it is flour based, but it does not necessarily have flour, or water, or its dough being in "edible" condition, or its flour being the same flour as in the bread dough, or its hydration being the same as bread dough or sponge hydration. Just like baker's yeast, we can use a little of it or a lot of it, from half a gram to up to 100g of yeast per 1 kg of flour in bread, we can use a little or a lot of starter (grams of culture inside it) in direct or sponge-dough methods of bread baking.
Levain is a part of bread dough in the sponge-dough method (sponge=levain), it is made with flour that is taken to make bread. It is seeded with a starter (or baker's yeast) and its goal is to modify a part or all flour in the recipe and produce flavors which define that kind of bread made with that flour or blend of flours. Levains are always flour-based and a particular attention is paid to their dough, so it is in optimal condition (i.e. its flour is properly prefermented, no damaged gluten or an appropriate amount of modified proteins and starches, etc).
Just like commercial yeast is produced in a different from bakery place, in bakeries and bread factories starter production and maintenance is a special job, sometimes done by designated specialists, at home - in a special jar, we keep it alive and feed regularly to maintain its yeast and bacteria concentration per gram of starter.
Taking a pinch of bread dough to keep its microflora as a seed culture for the next batch of bread dough is a special case. It was done in the past just as often as using dried remains of bread dough in a trough as a seed for the levain in the next batch of bread dough. You can see that both are not part of bread dough. If you use 50g of it to seed the dough, you will take 50g out in the end, so it never ends being the part of baked bread. Levain, on the other hand, is always baked and eaten. :)
Levain is always produced ad hoc by bakers and only if they choose the sponge and dough method of making bread. These days, starter is not even always produced by bakers themselves. It can be bought, for example, just like we buy yeast. In the past, both yeast and sd starters were produced and maintained by bakers/homeowners.
Thanks for the detailed discussion of various preferments including sourdough levains.
I'm still not seeing the difference between these two cases:
There would be a difference if the levain flour(s) and hydration did not those of the starter culture.
When you refresh a starter, you optimize it for the microflora thriving, multiplying, and the starter being at its optimal leavening and souring power by the time you need it.
When you create a preferment, you do it to suit your baking schedule of rest and work and to create a certain bread aroma profile, bread looks, and its keeping qualities. By the time a sponge is ready, in many cases, its microbial activity has actually ceased or is in decline. The preferment is no longer bubbling, or at its peak volume, etc. It looks completely deflated and calm.
Of course, sometimes things overlap, at least on the surface they might look the same.
Some breads can be made using the direct method with the same amount of flour in the starter as in that bread made by the sponge&dough method, in its sponge (+starter). Only their timing (bread production schedule) will differ.
My point is that a sponge is not specifically designed to be a nursery for yeasts and bacteria, a place where they reproduce. A sponge is a place where they work, producing gas and acids and modifying flour's proteins and starches. Thus, a sponge can be prepared in refrigerator or at elevated temperatures in a proofer or in a bread machine where yeasts or bacteria won't reproduce, whereas a starter needs a certain temperature range for its mcroflora to reproduce.
In extreme cases, there are 100% starter breads, when a ripe flour-based starter is salted and baked into a loaf of bread. There is one such bread on TFL, I love it. There are 100% starter discards recipes for pancakes and waffles, etc. And there are 100% sponge breads, where all flour is prefermented in a sponge (+some of it will come from its starter) which is then salted and enriched if necessary and then baked into bread.
Most importantly, for learning purposes, is to get used to distinguishing between starters and preferments, so that when you see some authors calling all of them leavens/sours or all of them "starters" or saying that they are the same, you know that it is their own way of calling them, to simplify things a bit and it comes from their personal experience with baking.
I was taught to distinguish between, let's say a 30% starter direct rye bread dough and a sponge+dough method of preparing that same rye bread dough where 30% of flour is prefermented in a sponge seeded with a certain amount of starter. Their schedules of bread production would be different and bread will be a little or remarkably different, and that allows a baker to decide which one suits them better.
I take your point, that the baker can choose to do a lot of different things with a preferment.
What struck me when I looked at the CIA's vs Hamelman's formulas was that in many (not all) cases Hamelman seemed to be using the levain just to elaborate the starter; in several formulas he reminds the reader to save a part of the levain to propagate the culture. In other words, the CIA's straight doughs with 20-40% starter were very similar to Hamelman's final doughs, if you think of Hamelman's levain as the CIA elaborating the starter the night before.
That's what you have in your last paragraph. The schedules are actually pretty similar if you count elaborating the starter as analogous to mixing the levain.
The formulas I looked at, not in detail, would likely have had very similar results (BTW, Hamelman and Ginsberg of "The Rye Baker" taught at the CIA at the same time).
I cannot compare those two books, because their recipes are not necessarily about the same starters (i.e. their ingredients differ) and probably not about the same breads (they look and taste different).
I haven't baked anything from the CIA book yet, but Hamelman's breads are unlike anything I have baked or tasted before, they are very diverse in formulas and methods from at least six different bread baking schools/traditions that I baked from his collection (American, Irish, French, German, Swiss, Jewish breads). Hamelman's steps to maintain or elaborate his starter(s) are described in a special section of his book amd they are quite different from his sourdough preferments in the recipe section ( levains and sours). By the way, he calls his stiff rye culture "starter" and his liquid white all purpose flour culture "levain".
To me a starter is just an ingredient from the list of ingredients. It does not have to be elaborated the night before mixing preferment or bread dough. It can be a day old, a week old, or a month old. Different starters have different "use before" dates. For as long as it is performing as expected it is fine. My refrigerated starter, for example, will not be good for use in bread after one overnight refreshment. It usually needs a series of three over 24 hrs period before it looks and smells and behaves properly and then it is good to use during the following three days. It has do do with its peculiar microflora.
A preferment is not an ingredient, it is a step in preparing bread dough. Sometimes it is a series of steps as in panettone or multistep German sours. They are different in nature from accumulating a certain amount of starter by a series of refreshments or feeds.
To me a starter is just an ingredient from the list of ingredients.
I think we have an unfortunate set of terms in this area. To myself, I tend to use the term "culture" to mean the "mother culture", "seed culture", or what have you to indicate the resource from which everything else is made. "Starter" would be what I mix into a dough. But people have different ideas about what "starter" means and don't commonly use the term "culture", so I usually just use "starter" instead of "culture".
It so happens that most of my breads use a refreshed portion of my culture as their starter. I usually try for a balance between simplicity, handling, and flavor, and this way works very well most of the time. Not so much for rye breads, I am learning. If I were to elaborate the starter in more stages, I'd tend to call the result a "sponge" or "levain" and only use "starter" to refer to the initial stage of building.
TomP
Tom,I agree with you and I think it is actually easy.
If the 'term' is in the list of ingredients, then it's a starter, even if it is called any other name (chef, levain, levain-chef, barm, mother dough or mother culture, lievito madre, etc.)
If its preparation is described in the text under the list of ingredients, as a distinct step of preparing bread dough, then it is a preferment - a step that describes how to preferment a portion (or even all) of bread flour. No matter how it is called - a sour, a levain, a a poolish, a sponge, a 100% sponge, biga, etc.
Such recipes of course assume that the readers maintain their starters in mint condition, always ready to be used in baking. At most they need to be warmed up if kept refrigerated.
Yeast manufacturers and bread technologist even consider a water brew a preferment, even though it has no flour, just yeast, water and 3% sugar (used to activate yeast before mixing dough), because it introduces organic acids to bread dough.
The whole confusion about 'levain' as a term stems from French tradition where
Levain naturel is the first spontaneously fermented mix of water and flour
Levain-chef is a ready to use starter (mother dough, masa madre) with a certain number of sourdough yeasts and sourdough bacteria in it.
Levain is a preferment, a portion of bread flour is inoculated with the yeasts and bacteria found in chef.
Pain au levain thus means more or less the same as baguettes with poolish, for example. It's a loaf of bread prepared by the 2 step method.
I've noticed the difference too, not just Hamelman, CIA, but bakers influenced by Robertson who quote everything as Starter as % of dough flour. And then German's who do the same but talk "inoculation" % rather than ratio. (Your use of the term "straight dough" confused me though, can you define it? I think of straight dough using commercial yeast with no prefermentation.) There is also the difference in whether you make extra starter to save some, or assume all of it is consumed by the elaboration.
My conclusion is that these are all essentially the same thing, described a different way, and you can choose whichever better fits your process and easily convert between them. Your method of elaboration can affect the bacteria of course; but that's always under your control.
All these methods start from a "ripe" starter/chef/madre. All these methods do at least one elaboration/build to get the final levain/starter. All the methods have some method of keeping some of the ripe levain for next time. (As part of the recipe or outside it.) Refreshing a culture, letting it ferment, and then taking a part of it to leaven a dough is basically building a levain
The French tradition has historically used three or four builds to get to the final levain, Whereas most recipes today (even Pain au Levain) do the simpler "one step" elaboration. (Similarly, German rye breads that used to be three-stage are mostly one or two stage these days.) But all these, Hamelman included, assume you already have a ripe chef! If you store yours in the fridge like I do you'll end up needing at least two-builds anyway, so even there the different methods collapse.
As I've gotten more knowledgeable I now mostly focus on the amount of PFF in the final levain from the recipe as my independent variable, and make other adjustments around. As one example, I baked Hamelman's "Levain with two Starters" yesterday. Hamelman calls for his standard 83% rye starter and a liquid 100% wheat starter. But because both starters had to be ready at the time of mixing, I needed a way to have the widest time windows so they'd sync up. Instead of Hamelman's "poolish" sourdough wheat levain (which has a narrow window), I made my "standard" stiff 65% desem wheat levain. It stays viable longer at R. To get the 32g of wheat flour Hamelman's formula requires, I elaborated 6g fridge desem : 32g wheat flour : 22 g water for 12 hours. (And took 6g off to use as next week's chef.) And then for the rye starter, instead of Hamelman's "one stage Detmold" at 83%, I used a "monheim salt-sour" method for building the rye: I used 6g of refreshed rye starter with 0.6g salt and 32g rye and 34g water. I had to adjust the overall hydration and salt in Hamelman's final recipe. And I'm assuming my flavor profile was likely different than his but hopefully you see what I mean by "adjusting to suit your tastes around the same PFF amount.
I think of formulas that have one mix step followed by bulk fermentation as "straight doughs".
Many simple commercial yeast formulas are straight doughs - Mix flour, water, salt, yeast, knead, and bulk ferment.
Most sourdough formulas I have encountered are not straight doughs - there is a preferment with the sourdough culture, flour, and water. This is similar but not identical to preferments like a poolish or biga.
Many of the sourdough formulas in the CIA text look like straight doughs - there is just one mix step, no preferment, However they do use an amount of culture that's comparable to a levain in "Bread".
I guess in the kind of commercial environment that the CIA text is oriented to, somebody in the bakery is doing a big feeding/refresh on the culture every day. Some distinctive breads may use a levain with flour or hydration that is different from the regular house culture. But many just use a large quantity of culture in what looks like a straight dough formula.
Trouble is, I think most people use the term "straight dough" to mean dough leavened with commercial yeast, and most likely a lean dough as well.
TomP
In "Bread" 3rd edition, Hamelman says that straight doughs are those with all ingredients mixed at once, no sourdough or levain build, no preferments like poolish, biga or pate fermantee.
So I guess that a sourdough formula that looks like a straight dough because it doesn't explicitly include elaborating out the culture isn't a straight dough - it just looks like one.
It's a communication issue, really. Nothing of substance involved. If you get right down to it, refreshing a culture, letting it ferment, and then taking a part of it to leaven a dough - doesn't that amount to building a levain and using it?
TomP
I agree with you and Hamelman. The fundamental distinction of straight doughs is no preferment including no starter culture.
So a sourdough formula that does not include elaborating out the culture is not a straight dough formula, but looks like one.
The formula looks like a straight dough but tastes like a levain bread.
Mariana has written a comprehensive yet general answer, and I would like to put down some thoughts as an experienced but not expert home baker. As context, I am always looking to streamline and simplify my procedures but not at the expense of bread quality.
One consideration is that the longer that flour is hydrated, the more flavor gets developed. And the longer it is fermented (within limits, of course, like anything else) the more flavor as well. With that in mind, why ferment part of the flour - as a sponge - for a longer time rather than all the flour? So I usually mix 20 - 35% starter with the rest of the flour and water and ferment the whole mass.
If that's a sound approach, why do anything else, like make a sponge first? I can see two main reasons. First is that too long a fermentation can degrade flour - and how much depends on the particular flour you use - and you might want to add fresh flour to your sponge before that happens. One extreme case of this is where a starter has been left unrefreshed too long and becomes soupy. You probably want to refresh the starter instead of using the soup in case the soup would degrade your dough.
Another reason for wanting to use a sponge is if, as you mentioned, you want to develop certain specific qualities in your preferment, either for flavor or for dough conditioning. This shows up especially in making rye breads. Rye breads need to be well-acidified to prevent too much starch breakdown (also called "starch attack"). So many rye bakers build up a rye sponge that is very acidic so that the overall dough will start out acidic. They might also build up their sponge through several stages to also develop more yeast or flavor in addition to acidity. Even with wheat breads there could be some cases where you'd want to do something like this.
But I suspect that in the majority of cases where a sponge is used in a wheat-based dough, the baker does so because that's how he/she was taught, or is using a recipe that was written by someone who was taught that way, rather than because it's actually the best way. IOW, habit and tradition. It's like kneading dough for 15 or 20 minutes vs developing the gluten with rests and brief Stretch&Fold sessions. You can do either but, except for special cases, why wouldn't you take the easy way with the S&Fs?
In addition, as Peter Reinhart came to see (e.g., in Artisan Breads Everyday), cold retardation can often achieve the same or better improvements in bread flavor and crumb qualities, as earlier sponge methods. Basically, time is your friend and cold storage lets you use more of it.
About reserving a piece of fermented dough as the seed for next time, I think that comes from commercial bakery practice, where the same breads are made all the time on the same schedule. My own home baking is much more sporadic and unpredictable, and it's better and easier for me to just refresh a starter from time to time.
TomP
Good insight on the difference for rye breads.
Do you always feed your starter so that there is that much available? I follow the directions in "The Rye Baker" and refresh with 7 g starter, 70g flour, and 70g water for a total of 147g,
For rye breads I haven't settled down to anything standard yet - early days for me. For my white wheat starter, I keep it at 100% hydration and feed it 3oz (90g) each of water and flour, usually once very three or four days (it stays mostly refrigerated). I would usually use 2 - 4 oz per bake, so I do get some discard.
Hi Louis,
Are you comparing in the same format? I don't have Baking and Pastry to compare, but in Bread, all of Jeffrey's formulas specify the percentage of the flour that is prefermented, which is not the same as the quantity of levain. (However, his levain builds are weight of starter to fresh flour.)
My best,
dw
I didn't run the numbers but a quick glance suggested that Handelman's levains were pretty similar to the CIA big starter weights in straight doughs.
I think I see what you are questioning.
Is there any real difference between these approaches, and their results, beyond what you call it?
They are about the same or similar rates of preferment when you extend the math, i.e., the levain in many of Hamelman's white levain breads is about 32-43% compared with the weight of the final dough flour depending on the specifics. So in answer to the question, there isn't a big difference in the proportion of starter/levain to final dough flour. The difference, when there is one, is in how you condition that prefermented flour whether you call it starter or levain. The "levain" step is a more conscious opportunity to align the starter with the type of bread you're making, but at the end of the day, it's semantics (and mathematics). Same with straight doughs. Since sourdough is always prefermented before incorporating into dough, to me, it is not a straight dough regardless of whether you think of it as starter or levain.
You're right of course that there is a world of difference between a straight dough with commercial yeast and a sourdough formula that starts with mixing the final dough and only implies elaborating the starter the day before to the weight of a levain.
I wonder if professional bakers would call those big starter % formulas "straight doughs". In one sense they are (only one final dough), in another they have preferments (elaborating the starter) but they don't call that a preferment.
An argument can be made that a sourdough formula never starts with mixing the final dough because it always starts with the starter. Whether it's fed 4 hours ago or 3 days ago, or whether the final feeding is the same or different than routine maintenance, it needs to be fit for duty. And it brings with it flour and water that become part of the bread.
In the flower experiment I followed up with a series of test bakes using Jeffrey's Vermont Sourdough formula adapted to the experimental starters which were maintained at different hydration (100% and 60%) from the specified 125% that the original formula calls for. I used the formula and process from the 3rd edition. -- 90% "bread" flour, 10% whole rye, 70% water, 1.9% salt, with 15% of the total flour prefermented in the levain. After adjusting for the various starter hydrations here are the original stats next to my working formulas:
OF 125 100 60
KAB ap 90 88 88 88
BRM wr 10 12 12 12
Water 70 60 65 72
Salt 1.9 2.2 2.2 2.2
Levain 40 35 28
171.9% 202.2% 202.2% 202.2%
OF = Overall Formula also referred to as the Total Formula in BBGA format. This is the constant for the other three formulas.
125 = the final dough formula when utilizing Jeffrey's 125% hydration starter. It's not calculated in the book, but if you do the math, this is what you get. And the 100 and 60 columns are final doughs when calculating for starters/levains of those hydrations. Regardless of how much the water and levain percentages change in the final doughs, the overall formula and 15% PFF is the same for all.
The question to ponder: If I (or Jeffrey) only gave you the final dough formula, does that make it a straight dough? It's not the total formula, it's just the working formula.
You can take a "straight dough" formula, and with the starter hydration, calculate back to the total formula (which I highly recommend doing). That is the best format to compare between breads, especially to get a sense of the overall hydration, saltiness, and how much of the total flour is already fermented (or even spent if it's very overripe).
A sourdough bread can always be represented by at least 3 columns, whether they are all presented or not (I didn't present the sourdough preferment columns here for simplicity's sake but I still had to compute them). There will be 1) an overall/total, 2) some sort of sourdough preferment as inoculant, and then 3) the difference between the previous two (the measurements that actually go into the mixing bowl).
Jeffrey gives you all three, albeit without percentages on the final dough, whereas it sounds like the CIA gives only the final dough column in what I think of as a working formula. It's a helpful format for production bakers who are scaling batch sizes to the day's orders and making one big batch of levain to inoculate multiple doughs. They only need to see what goes into the mixer. Each column tells you something about the dough, but it helps to consider which one you are looking at.
Hopefully that gives you some perspective on how to look at these formulas and recognize the different formats for what they are.
My best and happy baking,
dw
I don't tinker much anymore, beyond things like trying olives instead of walnuts (10% green olives in the Hamelman's Workday 100% WW don't have much impact; someday I'll try kalamatas). And I alternate the Workday 100% WW with a few of the simpler formulas from "The Rye Baker".
Once I got Ginsberg's ("The Rye Baker") whole rye @ 100% hydration culture going, I use that in all the sourdough breads. At low room temps, 70-72 degrees, it's a little slow but has a nice flavor.
The other day I met another amateur sourdough baker, an Italian lady, who had no idea what I was talking about when I mentioned sourdough discard. What she does is this:
This is same approach that most of the CIA Baking & Pastry textbook formulas use for sourdough breads, It has the advantage of creating no discard.
So why refresh a fraction of your starter, putting the rest in the discard container and using a small amount of the refreshed starter (~10-20g) the next day (or that evening) for a levain?
With the levain method, you can change your starter from stiff to hydrated, or white flour to whole wheat or rye. And you can make nice pancakes with the discard.
The CIA text is aimed at training commercial bakers. Commercial bakeries that use different starters regularly probably keep those starters separate and maintain all of them, rather than converting them in the levain. And unless the bakery also sells sourdough muffins, etc, the discard from the levain method would be wasted.