The Fresh Loaf

A Community of Amateur Bakers and Artisan Bread Enthusiasts.

Experiments with Autolyse

subfuscpersona's picture

Experiments with Autolyse

A heads up to all bakers who use an autolyse in their bread baking -

Teresa Greenway (a home bread baker of consummate skill who has been sharing her knowledge on her blog  - - for many years) has posted two entries exploring the effect of an autolyse (the technique of mixing water and flour from your bread recipe and allowing it to rest for a period of time in order to develop the gluten in the bread dough).

Teresa specializes in sourdough breads. Her two experiments explore the length of an autolyse (from 30 minutes to 2 hours) and it's effect on the outcome of the bread. Her posts are detailed, well written and include many photos.

Here are the links to her two posts on this subject... (post #1 dated October 26, 2011)

and (post #2 dated November 3, 2011)

Definitely worth the read! Thanks Teresa. We owe you.

=== PS === I don't know Teresa and she certainly doesn't know me. I am simply an enthusiastic follower of her blog and thought that these two posts might be of special interest to some of the more advanced bakers on TFL.

cranbo's picture

Thanks for sharing this. A timely and interesting post, as I was just re-reading a recent comment to the  "Effect of Yeast on Autolyse" blog post I made a while back. Teresa's posts (and the comments on those posts)  have got me thinking about factors which have a big impact on the benefits of a rest period:

  1. % of levain in formula
  2. hydration % of levain relative to hydration % of final dough (making sure they match somewhat closely)
  3. length of post-mix rest period (Teresa lets her mixed dough rest with the levain for up 2 hrs before kneading).  

 Terminology is a little tricky, because Teresa calls it "autolyse" when what she does is not a true autolyse according to the traditional definition, but she does demonstrate the benefits of what I call a "post-mix rest period" with her experiments. 

thomaschacon's picture
thomaschacon (not verified)

She's using the term "autolyse" to mean something that it doesn't/shouldn't.

thomaschacon's picture
thomaschacon (not verified)

I could be wrong on this, but autolyse is flour and water only.

It does not include the levain.

Her first "autolyse" uses levain, flour, and water. That's not an autolyse, it's a bulk ferment.

Her second post is two-fold, one experiment with levain/flour/water, one with only flour/water, but she says both are an autolyse. As I undertand it, only the second is an autolyse.

Isn't autolyse flour and water only? 

Have I been wrong on this all along?

Onceuponamac's picture

It looks like Theresa addressed the definition in her experiment #2...



cranbo's picture

Yep, Onceuponamac is correct and so are you. Autolyse is defined as flour & water only: no yeast, salt or other ingredients. Teresa does address this in her 2nd round of experiments. 

That's not an autolyse, it's a bulk ferment

I'm not sure what Teresa does is a "bulk ferment" either. I haven't seen a good formal definition, but to me bulk fermentation has always meant "the period of rest time after completion of kneading of the final dough." In Teresa's case, I think she's applying the rest period right after the initial mix of ingredients.  

I suppose if you define bulk fermentation as "the period of time after combining all ingredients into a final dough", then yes, it is a bulk ferment. 

Onceuponamac's picture

I can't remember and I'm not with my book - does Tartine Bread refer to the first "rest" after you put the leavan, water and flour together (before the salt) an autolyse? Just curios now given this discussion.  I typically get pretty good oven spring when I make that country bread recipe and various derivations - but am going to try with a two hour rest (my recollection is the book says at least 30 minutes - and I don't think I've let it sit for more than an hour before I add the salt) before I add the salt slurry this weekend.

linder's picture

I am not an expert by any means, but the combination of flour, water and yeast that is mixed together and then allowed to 'sit' for a period of time is called a sponge in the Tassajara Bread Book. 


cranbo's picture

This is why terminology is difficult in baking.

I agree, the definition of "sponge" actually seems to fit best, but I think the difference is in the duration of rest time. To me "sponge" means that the the ingredients are allowed to to significantly ferment, i.e., probably at least double in bulk, somewhere between 2-16 hours (so technically Teresa's is at the low end of the spectrum).

At the same time, some definitions of "sponge" represent pre-ferments made with commercial yeast (not sourdough). 

Clear as mud, isn't it? 

proth5's picture

Calvel, Hamelman, shall I go on?  - tell us that when the pre ferment, be it a levain, be it a commercially yeasted pre ferment, contains a significant amount of the liquid in the final dough, it needs to be added to the autolyse.  This must be done for the reasons that the author of the posts gives - that there will be insufficient hydration resulting in little lumps of flour that will never be incorporated.

There are good reasons to use liquid pre ferments, so the technique will, from time to time, be applied with their addition.

It is not a bulk ferment as insufficient dough development has occurred (although this becomes subtle with some mix techniques).  Ever after a 2 hour autolyse some additional mixing (or folding) will be required.

Calvel did work with varying autolyse times and he concluded that sometimes even salt should be added.

I'm not going to say Calvel is wrong as the whole codification of the technique is his.  Your understanding is correct to a point, but limited. 

Please let us someday put this to rest on these pages.  It was an an interesting experiment using correct terms in common usage among well qualified bakers.

thomaschacon's picture
thomaschacon (not verified)

This is really quite the head-scratcher.

I don't have Clavel, but Hamelman does say that.

He also adds...

"Since the percentage of yeast is quite small in both poolish and liquid levain, however, the inclusion of either of these does not significantly affect the eventual autolyse benefits." p. 09.

There's a disconnect here that my brain is rejecting.

Hamelman says the liquid levain has a small percentage of yeast, so shouldn't significantly affect the process. When I read over Teresa's process, however, it sounds like she's using a very active levain in large quantity. That's the disconnect, as how can her levain not contribute a ton of yeast at mixing and, as such, cause fermentation, which is not what you want during autolyse?

She also folds it several times during autolyse, which usually isn't done, and admits that Dough #1 is two hours ahead in fermentation time.

To me, it's like she's comparing apples to oranges, bulk fermentation to autolyse. Alas, I see your point re: insufficent development, so it's not really bulk.

I rather think the whole experiment just confuses the autolyse issue rather than clarifies it. The high-hydration pre-ferment exception case adds a variable that makes me question if anything was learned or established in the trying.

proth5's picture

you look at the sourdough formulas in the book you cited you will see that additional commercial yeast is generally added.

I think one of the issues for many of us left brained folks who take up bread baking as a hobby (and maybe have done so in an attempt to balance an overly rational life with one full of sensation) is that they search for the "one true answer."  The more I study bread baking, the more I have the chance to hear one celebrated baker declare one thing and then hear another celebrated baker declare the exact opposite.  From time to time I get to hear one celebrated baker tell me one thing one day only to tell me the opposite a few weeks later.  I've learned to make peace with this.  Bread is a living organism and without assigning it complex personalities and motives, things vary from day to day, place to place, and hands to hands.  Some things are nearly always true, but a lot vary.  One baker's experiments at a point in time are things to be considered - no more, no less.  Logic is the beginning of wisdom.  I think about this from time to time.

Abe put it best below in terms of the most significant reasons for using autolyse.

Mini Oven's picture
Mini Oven

slowing fermentation than anything else.  

I will often delay salt to shorten a long fermentation.  Just as I add salt early to slow it down.  Rye sourdoughs show this dramatically and I can cut proofing times exponentially with the length of salt delay.  I believe the doughs #1 and #2 in the first experiment when given more proofing time, would have equaled #3.  #2 needing more proofing time than #3 and #1 needing more proofing time than #2.  The delay in baking was not enough time for the proofs to equal one another.   

The experiment needs to be repeated and this variable investigated, the one of kneading the dough at different intervals.  

All of the dough was folded three times during the six hour bulk ferment with no other kneading taking place except to incorporate the salt.

In my opinion, each time salt was added to one of the doughs, the other two lumps needed the same handling (kneading) even tho no salt was added.  I find when yeast or rapid fermentation is involved as opposed to just wet flour, the autolyse will be faster, without salt,  fermentation boarders on lack of control and loses predictability resulting in over-fermented dough.  So to repeat myself from another angle, I will say that adding salt to #1 gave a great deal of control in fermentation time when compared to dough #3.  The baking time was determined by the fermentation rate of dough #3 and not #1.

My question would be how did dough #1 compare to the predicted rate of fermentation when the experiment was set up?  As it turned out, was #1 under fermented? (yes, it was)   All doughs were folded for gluten development (even tho tied to fermentation) resembling many typical dough progressions so letting the dough ferment longer and proof fully is the solution to dough #1 and #2.  I predict that if #1 loaf was baked first, then the other two would have been over-proofed.  It would be interesting to see the dough kneaded at the 30 min and 2 hr marks and each baked when ready to be baked.

Erik Henchal's picture
Erik Henchal

I'm sorry to be late in asking this question but I just found your post.  

For rye sourdough you reported shorter fermentation times when the addition of salt is delayed.  

I have trouble getting the salt to adequately mix into the dough late in the bulk fermentation stage.  Have you experienced this as an issue?

Mini Oven's picture
Mini Oven

when the dough has a high hydration like 83% or  more for the rye flour portion of the dough.  Sprinkle on the surface and fold in, repeat until all the salt is in. Then lightly knead or fold more.  The dough oftens gets a slippery wet feeling but just keep folding past that stage to incorporate.

Bread Breaddington's picture
Bread Breaddington

Something that's not clear to me is, is it that each dough was autolyzed (ignoring any terminilogy nitpicks here) for a certain time, -then- bulk fermented for 6 hours? Or, over a total 6 hours of fermentation, the salt was incorporated at different times?

If it is the first of those two, then I would think some of the difference could be attributable to the last dough simply having fermented for longer than either of the others.

jcking's picture

I know there are a LOT of theories about the prime purpose of the Autolyse.... But I believe the primary purpose is that flour simply takes longer than the 2-4 minutes of the incorporation phase of the mix to fully hydrate. And gluten forms all on its own without mixing. As a general rule you want to mix most doughs as little as needed for proper development, so its not so much a chance for the gluten to form prior to yeast or salt being in the mix. Its just giving the mix time too form gluten bonds passively during the hydration of the flour, prior to the final mixing of it.

So why be mixing the stuffing out of your dough trying to develop it and creating all sorts of deleterious affects of over mixing, like whitening of the dough, destruction of flavor and toughening of the final product, when you can just wait 20 minutes, have the flour fully hydrate, and then do your higher speed phase of the mix for less time on a fully hydrated flour.

Also, even though yeast might theoretically be in there robbing the oxygen necessary for the gluten bonding/crosslinking process most of us now use dry instant yeast which takes 20 minutes in itself to fully hydrate and activate, so we include it in the mix prior to autolyse anyway. And even without an autolyse it would have a delay before it went for the oxygen.

In some cases the whole preferment (teaming with yeast) is chucked in as well prior to the autolyse if it is very high hydration pre-ferment, so the final dough needs the moisture from the preferment or the final mix would just make dry little hockey pucks. But even in this case, the autolyse step still lets you do less mixing.

--Abe Faber
Clear Flour Bakery
178 Thorndike Street
Brookline, MA 02446
(BBGA Yahoo Group 4/23/06)

proth5's picture

Thanks, Abe.

I had it in mind to discuss these things, but you put it very well....


copyu's picture

Jim...or Abe! Heheheh!

Although I'm a true 'Calvel method' fan, as I indicated to you a couple of weeks ago, on another thread, I wanted to say what you said, so all I can add is a big've saved me a lot of typing!

My methodology is now wild yeast (sourdough) Biga; main flour(s) and water mixed for autolyse; add salt to the biga and then incorporate with autolysed flours for a short, second mixing. (Works a treat, for me, but other methods will doubtless work, too—in someone else's kitchen. Heheheh!)




jcking's picture

If we go back to the source on this matter, that is, the writings of Professor Calvel himself, a lot is revealed. Interestingly, he initially began experimenting with "inert" dough (that is, no yeast but at first he did include salt) in the 1950s, when the French were required to import a lot of flour from Canada due to extensive cold weather having killed much of the winter wheat stocks in France. He used this forerunner to autolyse to try to tame the too-strong North American flours, and it worked, but then things returned to normal in terms of availability of French flours, so he put the experiments on the shelf ("just as the ocean's surf erases the tardy fisherman's footprints on the beach")-until 1974. That's when he did extensive trialing, using different percentages of pre-ferment, different lengths of time for the autolyse, and now always leaving out salt and yeast.

Some of his conclusions were:

~reduced mixing time by 10-15% (in practice, we find at the King Arthur
Bakery a reduction of second speed mixing of 50% and often more)

~better loaf tolerance, particularly during final fermentation;

~better opening of the cuts;

~always better loaf volume ("at times to unexpected levels");

~better cell structure and color;

~better flavor

jeffrey hamelman|certified master baker and bakery director|king arthur
BBGA Yahoo Group, 12/9/10

copyu's picture

Thanks for this info, Jim. It's always good to read what Messrs. Calvel and Hamelman have to say on these issues...(Should that be singular? 'This issue: Autolyse')



jcking's picture

Professor Calvel didn't invent the word autolyse. If we look at the dictionary;

Autolysis ~ Destruction of tissue or cells of an organism by autogenous substances , as enzymes.

Autolysin ~ A substance that results in autolysis.

Autolysate ~ An end product of autolysis.