Planned bake: Attempt to replicate Larraburu Bros. San Francisco Sourdough
San Francisco Sourdough from Larraburu Brothers
as described in
https://www.aaccnet.org/publications/cc/backissues/1978/Documents/chem55_461.pdf
as interpreted by
David Snyder
February, 2019
Over many years, there has been much interest in reproducing the San Francisco Sourdough bread baked by Larraburu Brothers' bakery that closed in the early 1970's. The article referenced above seems the most likely to accurately report Larraburu Brother's method. The following formula and methods have been extracted from that article, with a very few modifications as noted.
Because of the current interest on The Fresh Loaf in this, I will post this “plan” now and update it with my own results when I have actually baked bread using this method.
Total Dough |
|
|
Ingredient | Wt (g) | Bakers' % |
Bread flour (12% protein) | 924 | 90 |
High gluten flour (14% protein) | 100 | 10 |
Water | 612 | 60 |
Salt | 20 | 2 |
Total | 1656 | 162 |
Note: I used King Arthur Flour AP flour (11.7% protein) and Sir Lancelot flour (14% protein).
Sponge |
|
|
Ingredient | Wt (g) | Bakers' % |
High gluten flour | 100 | 100 |
Water | 50 | 50 |
Active starter | 50 | 50 |
Total | 200 | 200 |
One day before baking the bread (e.g., before going to bed the night before you want to bake)
Dissolve the starter in the water.
Add the flour and mix thoroughly. Knead until all the flour is moistened.
Place in a dry bowl and cover.
Ferment at 80ºF for 9-10 hours
Remove 50g of the fermented sponge and refrigerate for future use.
Final Dough |
|
Ingredient | Wt (g) |
Bread flour | 924 |
Water | 562 |
Salt | 20 |
Sponge | 150 |
Total | 1656 |
Procedure
In a large bowl or the bowl of a stand mixer, mix the water and sponge cut in pieces to soften the sponge.
Mix the salt into the flour and add it to the mixing bowl.
Mix the dough at slow speed to thoroughly mix the ingredients, then at medium speed to obtain medium gluten development. (A medium window pane.)
Transfer the dough to a clean, lightly oiled bowl. Cover and let rest at room temperature for 1 hour. (Note: The article does not specify the temperature for this step. I think room temperature is most likely.)
Divide the dough into two equal pieces. Pre-shape as balls. Cover with a cloth and let rest for 10-30 minutes to relax the gluten. (Note: The 10-30 minute rest after pre-shaping is my addition, but it is “standard operating procedures” in most artisan bakeries.)
Shape the pieces as boules or bâtards and place, seam-side up, in floured baskets or on a linen or parchment couche.
Proof for 4 hours at 105ºF in a humid environment. (Note: I place the formed loaves in bannetons and place the bannetons in food safe plastic bags and clip them shut. Then, I proof the loaves in a Brød & Taylor proofing box.)
One hour before baking, preheat the oven with a baking stone and steaming apparatus in place.
Transfer the loaves to a peel and score as desired.
Bake at 420ºF for 40-50 minutes (per Galal. I baked at 460ºF with steam for 15 minutes, then at 450ºF convection for 25 minutes.) The oven should be steamed for the first 15 minutes of the bake. The loaves are done when thumping the bottom gives a “hollow sound,” the crust is nicely browned and the internal temperature of the loaves is 205ºF.
Cool thoroughly before slicing.
David
Comments
I guess I’m irreverent, because I generally mix everything (flour, water, salt, levain) initially to a shaggy mass and then rest the dough for 20-30 minutes before developing the dough. I realize the levain starts the fermentation and the salt tightens the dough and also inhibits the fermentation a bite. But that doesn’t bother me.
My question; should it? I don’t see the big deal in starting the ferment a little sooner.
I am anxious to get your opinion. I am an eternal student. And I am positive that some of my thinkng is wrong. I don’t mind being told I’m wrong when the speaker is right :-D
Danny
Danny,
Autolyse was "invented" and named by Professor Raymond Calvel, a highly esteemed authority on bread baking and author of "Le Gout du Pain." In the 1950's, French bread making had been transformed into a highly mechanized process that maximized efficiency over flavor. Rapid and prolonged machine mixing produced a whiter baguette crumb and no need for prolonged fermentation to develop gluten. The result was tasteless bread.
Calvel found that a mix of just the flour and water that was allowed to rest for 20 to 60 minutes before salt and yeast (always fresh yeast) and a preferment (originally a pâte fermentée) were added permitted good gluten development with a shorter and more gentle machine mix. The problem with "intensive" mixing was that the yellow, carotenoid pigments in the flour that contribute significantly to the flavor of white bread got oxidized. I have experimentally subjected a 100% whole wheat dough to intensive mixing. The result was a much more open crumb, but also much paler dough and a flat flavor. Yuck!
I have found that an autolyse has other benefits, even with hand mixing and especially with doughs containing significant amounts of whole grain flours. It permits better hydration of the bran and over-all better gluten development with less work. The result is a more open crumb. A 1 to 2 hour autolyse is beneficial for whole grain breads, in my experience.
Now, the remaining questions is: When do you add the levain? The answer is: With liquid levains, you add them to the autolyse. With firm levains, you add them with the salt after the autolyse. The reason is, with liquid levains, the levain contains a significant portion of the total water.
Hope that answers your questions and isn't TMI.
David
Not TMI at all. But the question remains. What are the cons to adding all ingredients into the mix initially, including the levain and salt?
What would be the downside to starting the fermentation a little earlier by including the levain in the initial mix?
Danny
Oh! I also dislike the flavor of bread that has been over oxidized by too much mixing/kneading. The bread is bland and disappointing. I posed this question a short while back, but I don’t remember getting an satisfactory answer. Is it possible to over oxidize a dough utilizing slap & folds? As many as 300 sets in 2 sessions of 150 each. That is my maximum.
Okay. Enough puns.
Salt is primarily a flavoring agent in bread, but it has a couple more effects which are undesirable early in the fermentation process. The most important one is that salt is hydroscopic. It completes with flour for water. Water is necessary for gluten formation. The purpose of the autolyse is hydration of the flour and initiation of gluten formation. Therefore, it works better without salt present. Salt slows down fermentation and other metabolic processes. That is not relevant to a mix with an autolyse, because there is no salt added to the dough yet, but is another reason to withhold salt for a while after the initial mix.
Levain and other pre-ferments have several advantages. They provide leavening, of course. They also provide acid. Now, that has advantages and disadvantages. Acid strengthens gluten bonds, which is a good thing. But, paradoxically, it actually interferes with gluten formation. So, you don't want it there when gluten is just starting to form.
Can you make good bread without doing an autolyse? Of course you can! But I hope you can see some of the reasons (I haven't touched on all of them.) you might want to include an autolyse in your sourdough bread baking.
David
Maybe I need to revisit the true autolyse. If have enjoyed good results mixing everything at once, but I am forever looking for the very best method. Thanks, David.
And you are correct, I am a gluten for punishment <LOL>
I don't want to hijack the post and amazed by your experiments David and Dan......BUT my ears pricked up on the information on autolyse as I am a big fan of it...I often mix up to 4kg dough and the AL is a great way to get the dough softer and extensible as I have no mixer and also a way to get the gluten formation on the way before I add leaven and then salt. I often AL the same time as when I get my leaven ready in the morning, approx. 5 hours.
However, I am pleased that you also confirm that an AL also helps with flavour development because of the Amylase, which it not so often mentioned as a benefit of an AL. So much learning to do... Kat
I have been told and believe that you cannot over-oxidize dough with hand kneading or Slap & Folds.
I cannot recall the numbers offhand, but the number of dough hook revolutions in 5 minutes of machine kneading (a moderate amount) is in the thousands. Typical slap and fold numbers are in the low hundreds or less.
David
That is great information to know.
a question I've had for quite a while, but never investigated.
I hardly ever use a mixer, save for some enriched doughs and the occasional ciabatta. Since the ciabatta dough goes through a severe beating in the mixer for just about forever before it reaches that sweet spot of development...Why do you think it is that the resultant ciabatta winds up having such a lovely and sweet taste, agreeing that the carotenoids have had the heck whacked out of them.
All things being equal, although they hardly ever are, shouldn't ciabatta be completely devoid of taste and in the ballpark of those Wonder or Chorleywood delights? My ciabatta is 100% white flour, whether from IDY/biga or my recent levain venture. Even so, shouldn't the maturity of the preferment having been introduced, be negated by the extended intensive mix?
BTW, I'll be making my annual pilgrimage to Mike in late April, so we shall see. It's been since Nov. that I stepped off a plane at SMF, and there is always that lovely refreshing feel of true central valley weather that hits me the moment I step out of the terminal. Of course I anticipate that the precipitation. will be mitigated by then. Snow in L.A., who'da thunk it?
My best guess is that, in a very soupy dough like ciabatta, there is more amylase produced and, thus more free sugar which produces the typical sweet flavor of good ciabatta.
David
The Scott MeGee version that has taken my heart recently is only 76% hydration, so it is lower than a lot of Pain au Levains and other doughs floating through the pages of TFL. The funny thing about this dough is that it doesn't mix or feel any different than my former go-to 84% hydration biga version at any stage. Odd.
C'mon man, you're a scientist, sort of ;-) . I need answers. I also need my meds...
alan
could the sweetness derive from a long, cold ferment that permits the flour's own enzymes to break some of the starches down into sugars? My baseline for that thought is Reinhart's Pain a la Ancienne, which uses cold ingredients and a cold ferment. The resultant bread is sweeter than the ingredient list would lead one to expect.
that I let control my timetable, more or less, rather than the reverse. It never hits the retard stage, which is an unlikely scenario in my kitchen. Straight from BF to divide to final proof. Fortunately, the whole process takes a mere ~3 hours before the dough hits the oven, including mix time.
But I know what you mean with the BBA Pain a la Ancienne, a somewhat inscrutable wet dough that resists shaping. But to describe it as sweet is correct. An early go-to favorite when I had no other tricks in my bag and needed to have a delicious treat at the ready. Unfortunately, that bread disappeared quickly around my place. Yum!
But for now, our commercial time is over, and without hijacking Señior Snyder's post, let the Larraburu show resume...
alan
David, this looks like an interesting experiment. I might give it a go too, but have a few questions for you. First, I suspect that the goal is somehow to replicate the taste of a Larraburu Brothers loaf. Do you have any information on what that was? (That is, was it truly sour? Lightly sour? Distinctive in some particular way?)
Now for my questions. [Reference in brackets.]
[Sponge, step 4] Unless I wait for the summer, trying to achieve 80dF is going to be tough. Do you think there is anything special about 80dF and 9-10 hours? Might a cooler temperature and longer time produce the same effect? Or is this the magic combination?
[Procedure, step 2] Given the relatively short total time between mixing and baking (a one-hour bulk fermentation and a four-hour proofing), it seems that there could be a difference between a 20-minute autolyse and a two-hour one. Any thoughts on this?
[Procedure, step 6] By pre-shape as "balls" I assume you mean rounds and not spheres.
[Procedure, step 8] Trying to maintain 105dF (as well as 96% humidity) will be tough too, regardless of the season. Is there an alternative without sacrificing the goal of replicating a Larraburu loaf? (As an aside, I have seen your plastic bags in various photos and always mean to ask where you got your food safe plastic bags. I fashioned some from sheets of plastic but would prefer getting some of the ones that you use.)
[Procedure, step 11] Do you think that the Larraburu Brothers baked at 420dF? Perhaps Galal asked them about this. Perhaps also there is something magic about 420dF for this particular combination.
[General comment] The taste of bread is often heavily influenced by the flour used. Do you think that the Larraburu Brothers had access to some particular flour that produced their brand taste? They might have maintained their starter in some particular way. There is no overnight retarding to develop flavor, and in fact the proofing is accelerated via a warm and humid environment. I am curious about what might make the taste of this bread differ from other loaves that I bake in a relatively short time frame.
Thanks for posting this, and please update the thread when you do actually bake. My questions are thoughts that I had, but also intended to guide me in trying to achieve an accurate replication.
Ted
I don't recall ever eating Larraburu Bros. bread. I may have, but my family always ate Parisian, from another bakery that no longer exists. My understanding is that Larraburu's bread was moderately sour - more sour than French pain au levain but less sour than today's Boudin. (Note: Boudin makes a custom-baked sourdough only for Tadich Grill in San Francisco. That bread is the closest to old-fashioned San Francisco Sourdough than any I know of. It is wonderful - quite sour.)
Re. your questions about temperatures. I have a Brød&Taylor proofing box. I can comply with the protocol. If you can't, you can't. Production of acids and the balance between lactic and acetic acid production is temperature dependent. So, a different fermentation temperature will change the flavor of the bread.
I get bread bags from Breadtopia, King Arthur or Rye Bakers web sites.
I doubt if the flour used makes a significant difference in the taste compared to other variables.
Re. Autolyse: Remember, this is done without any leavening. The difference in timing will impact gluten formation and need for mixing, not fermentation.
Re. pre-shaping: Balls, rounds, spheres. Whatever.
Re. Baking temperature: I assume the 420ºF is what Larraburu used. I prefer a bolder bake than this would produce, personally.
David
My wife and I will be in San Francisco for a couple of days in April. We ate at Tadich over twenty-five years ago, so I have no recollection of the bread, but it was a neat dining spot. Perhaps we should revisit. (I am also playing with the idea of going to Tartine just to see what that place is like and sample some of the famous bread, but we have only a couple of days in SF and there are many things that can be done.)
That proofer looks like a possible acquisition. Do you use it often? I typically just watch the dough (after measuring dough temperature at the conclusion of mixing, which gives me some idea of what to expect) and am at the mercy of the environment du jour. Perhaps the proofer would be useful.
The bread bags at King Arthur look like the ticket, but I am not sure what I would do with a hundred.
In pondering your replies and the recipe, it seems that the proofing temperature (and humidity) might be the key factors in the final result for this bread.
I also prefer strong bakes and higher temperatures than 420dF.
Thanks again.
Ted
Tadich hasn't changed in 25 years ... or 75 years.
The B&T proofer is highly recommended. I use it for almost every bread I bake, since I have taken to bulk fermenting at 80ºF routinely. I also use it for fermenting levains, usually at 76ºF.
I took a protègé and aspiring professional baker from Dubai on a bakery tour of San Francisco a few years ago. We visited 5 bakeries in 6 hours. Then he pooped out from jet lag (and maybe butter toxicity!). (He now is baking amazing looking breads at Tusk Bakery in Beirut.)
Besides Tartine, I always recommend B. Patisserie and Thorough Bread Bakery. You can also visit the Acme presence in the Ferry Building. Each has its unique aspects. I haven't been to Josey Baker's bakery but want to visit there.
David
Wow! I wondered what happened to Khaled (mebake). He was such a regular here when I started on TFL, as were others who seem to either drift in or out, or are just plain gone from visibility. Best of luck to him.
BTW If there is any baker who has stepped away from the site that I'd like to see return, that would be Phil (PiPs) in Brisbane. A true craftsman and a heck of a photographer and documenter/storyteller of his work.
Thanks for bringing Khaled to mind. alan.
A bakery that I admire in the Bay area, with two storefronts in SF, is Arizmendi, one in the Mission and one near the south side of GG Park. They also have a few places spread out locally not far from beyond the city limits. Perhaps not on a par with the more famous SF bakeries, but if you're in the neighborhood... They are also a worker owned cooperative, a big plus in my book.
David, now that your planned bake has become an actual event and moved to a different thread, I'll post my question here. You wrote "I use it for almost every bread I bake, since I have taken to bulk fermenting at 80ºF routinely." My question is whether you cover your dough during the bulk fermentation while the dough is inside the proofer. Is the proofer sufficiently sealed to act the same as a tight lid?
My reason for asking is because I really like using a 12-qt Cambro tub for mixing and the bulk fermentation, but it would barely (if at all) fit inside the stated interior dimensions from the B&T website. On the other hand, a 6-qt Cambro should fit (including the lid). That is what prompts my question.
Thanks in advance for your time in answering.
Ted
Read the OP. Step 7. in Procedures.
OP, Step 7, pertains to the proofing. My question had to do with the bulk fermentation.
The B&T Proofing Box comes with a metal tray that one can fill with water to humidify the box. I don't use it. I always BF in covered bowls, either a 2 liter batter pitcher which comes with a cover or a wood composite larger bowl, covered with an airtight silicon disk. The proofing box will actually hold 2 of the 2-liter pitchers. I like them because I can actually measure volume expansion and can see the bubbles forming in the fermenting dough.
David
Thanks for your reply. I like clear plastic not only to watch the side, but also because I can lift the Cambro tub up in the air and look at what is happening on the bottom. I have found that the accumulation of bubbles visible through the bottom is a good indicator of how the bulk fermentation is proceeding. No bubbles means early in the process. Some bubbles means we are getting somewhere. A good scattering of bubbles means it is time to dump the contents, divide, and pre-shape. That has become as indicative of readiness as the increase in volume (which I have found can vary greatly depending on hydration) and bubbles on the side (which does not have the surface area of the bottom to expose bubbles).
Ted
I often completely seal the container to avoid loss of moisture. But there are occassions when a sealed vessel can work against you. If you have an extremely long (>=12 hr) and temperature sensitive ferment, the inclosed and sealed container will cause the dough temp to rise substantially. In this case, I leave a slght crack in the cover to allow heat to escape and fill the water tray.
Dan
David, I started the levain fermentation last night using a Brod & Taylor. I am off the opinion that a few (+/-3) degrees variance can make a noticeable difference in flavor when the duration of fermentation is long.
I ran a thermal chart on the levain and it appears that even though the proofer maintains ~80F, the levain temp rises 2-4 degrees higher. I ran this test with the assumption that Larraburu set the proofer for ambient temps and not the actual dough temp. What do you think?
HERE is a link to the chart. The chart is interactive. You can zoom, get timeline data (time & temp). I included a screen grab below.
Notice how the ambient proofer temp runs around 81 - 79F (I made adjustments) and the levain is 3-4F warmer. I know this is to be expected because of fermentation activity. But I wanted to bring this temp variance to the attention of those that are participating in the experiment. Like others, I have failed numerous times attempting to reproduce the Larraburu bread. It is possible, IMO, that small variances could make a difference in the flavor. Especially with the extreme temps and long ferments.
Everyone - please share your thoughts and opinions... As David mention, I am a gluten for punishment :-D
Danny
was want ever the bakery temperature was where the dough was being left to bulk ferment before shaping. If it was next to the ovens it might have been over 80 F and if it was far aw from the oven in the winter in might have been 60 F in SF. Since this is quite a difference I think it was some where close to the ovens and that huge proofer they had to hold a huge pile of bread at 105 F. for proofing. Also at 105 F the yeast should be really stressed and not making CO2 like they do at 84 F but the Lab will having a filed day.. I'm guessing the resting temperature of the 1 hour bulk was around 72 F - 75 F. I couldn't get it to work at all well enough to call it a success but hopefully you will get it to work much better.
Changes from my originally posted method:
1. I activated my refrigerated firm mother starter with my usual SD Feeding flour mix at 100% hydrateion and fermented it overnight at 76ºF. Then, I mixed the firm "sponge" as described in the OP and fermented it at 80ºF for 7 hours, at which point it was almost ready to collapse. I then refrigerated the sponge overnight.
2. I decided to forgo the autolyse and followed the procedure in the modified OP.
The loaves are now proofing. It is 1135h now. I plan to bake after the prescribed 4 hour proof at 104ºF at 3 pm. If it looks like the loaves are getting over-proofed, I will bake sooner.
Stay tuned!
David
Dan's was way overproofed earlier today at 4 hours
I just posted those words of wisdom on Danny's topic!
My loaves were really puffy at 2.5 hours, so I baked at 3.5 hours. This short BF and long, hot proof is kind of terra incognita to me, so we'll have to wait to see how the loaves turn out. At this point, they are still in the oven. <Play suspenseful music now> Looking pretty darn good, so far.
Note: My bake is 460ºF with steam for 15 minutes, then 450ºF in a dry oven set to convection bake for another 25 minutes.
David
Dab, for this particular experiment, I attempted to follow the written process from Galal closely. I never checked the dough for rise at any place in the ferments. As you know (you worked with me In the past) other attempts ended in failure. Whaat differentiated this bake from other attempts is the flavor. I found it quite nice. I consider this one a mild success.
HERE is the link to the documentation and images of the bake.
Danny
This has been a good discussion, but I am going to post the results of my bake in a fresh topic.
David
I'm just now seeing this thread.
I grew up on Larraburu. It was at every family gathering and my mother kept several loaves of it in the freezer at any given time. I have a vivid recollection of the flavor.
Sorry to contradict, but I can state with authority that Larraburu was very sour, not merely "moderately sour". It had a strong lactic-acid presence.
I have had both Boudin and Tadich Grill bread in recent years on the above recommendations.
Boudin has an unmistakable vinegary flavor. Tadich Grill is sour but is completely lacking the distinctive lactic-acid tanginess of Larraburu. Both of these data points suggest that the bread is being artificially soured. It is possible that vinegar is extrinsically added to Boudin. Possibly Tadich Grill is soured with the addition of fumaric acid or some other ingredient. I'm not privy to the baking process so I don't know if this is indeed the case, but that is my suspicion. As we all know, sourdough done properly takes hours and hours of proofing time. Using extrinsic souring agents makes sense as it would cut down the proofing times and would result in greater throughput for the bakery.
They seem to be making the bread for tourists, not purists.