The Fresh Loaf

A Community of Amateur Bakers and Artisan Bread Enthusiasts.

foodforthought's blog

foodforthought's picture

Been on something of a viennoiserie jag lately brought on, of course, by the holidays, visitors and deliciousness. With 7 dozen croissant, 4 dozen kouign-amann and one chausson aux pommes under my belt (gonna need a new one if this keeps up), a few brainworms have been nibbling on my thoughts.

First, I have to thank Benny (@Benito), TFL’s magellanic bread explorer, for his testimonials to the Ankarsrum roller and bread doughs. I’d been kind of stuck on the Ank hook since mine landed last June-ish and have revisited the roller for breads. It turns out that incorporating levain, poolish, etc. into the liquid components of the dough require less management (or dough herding as I think of it) with the roller. I also like the ability to adjust the roller depth during kneading of lower hydration doughs of yeasted viennoiserie. 

Secondly, It occurs to me that the pâte fermentée I have been using in my croissant dough bears a moderate resemblance to Benny’s sweet, stiff levains. At 60% hydration, it’s quite a bit wetter than several of Benny’s levains that I’ve reviewed, less sweet (at a mere 5%), contains a bit of butter (7.5%) and I’ve been using ADY for leavening it. I generally add my standard sourdough levain to the dough at mixing, more as a pre-ferment than as the sole leavening agent, so more ADY is added there. I’m not sure there are any direct parallels to Benny’s doughs here as the dough barely gets any bench time until final proofing. The pâte gets at least 12 hours of bench fermentation and the final dough gets at least 12 hours in the refrigerator after a short bench rest. But the dough definitely turns out sweetish with little or no sour notes, I think it’s safe to say that the sweet, stiff path works well in this context.

Pâte fermetée formula

T45 or pastry flour     100%

Salt                                2%

Sugar                             5%

ADY                               2%

Water                           60%

Butter                          7.5%

 I shoot for the flour in the pâte to represent ~17% of the total dough flour, or more simply, to equal 20% of all the other flour in the dough. So for 2 dozen croissants, 200g of pâte fermentée containing 114 g of flour was added to 685 g flour for a final dough of ~1400 g requiring 470 g of butter for lamination.

 Pâte feuilletée is a related but unyeasted laminated dough. Just for laughs, I used some I had stashed in the freezer to fancy up a chausson aux pommes (apple turnover) for Christmas dinner.

Been having a great time at it. Thanks to Benny and the many TFL folks who inspire me and make me think about what makes our bakes better.

foodforthought's picture

Finally got around to making bialys based on the King Arthur recipe. I've always loved the bialys they used to make at the Mendocino Bakery which has since changed hands several times and had sadly disappeared from the menu on our last visit early in the Covid epoch. I will be looking for them when we visit again in a few months. In the meantime, these came out just great so we won't go without.

Had some sourdough levain on the counter and added a poolish just because. At 60% hydration, the dough seemed quite stiff but my newish Ankarsrum mixer handled it well though I had to do a bit of dough herding with a silicon spatula as it (the dough) really wanted to climb up the hook and into the arm that positions the hook or roller in the bowl. The results were quite good, so will be in the repeat often lineup. 

I deviated from the King Arthur recipe in a few ways, including the levain and poolish and using a 50/50 mix of bread flour and AP instead of the 100% bread flour specified. Levain and poolish were both built using AP flour. The results were good enough that I will probably do the same in future iterations.

  • 261 g Sourdough Levain @100% hydration
  • 244 g Poolish @100% hydration
  • 253 g AP Flour (420 total with levain and poolish)
  • 420 g Bread Flour
  • 256 g Water (510 g total with levain and poolish)
  • 5 g Yeast (0.6%)
  • 17 g Salt (2%)
  • 2 g Onion Powder (0.2%)


  1. Mix levain, poolish, water and yeast. Rest 10 minutes.
  2. Combine flours, salt, onion powder. Add to wet ingredients. Mix to shaggy. Rest covered for 20-30 minutes.
  3. Knead at medium speed for 8 minutes.
  4. Refrigerate overnight or longer.
  5. Next day, divide into 110-115 g portions. Preshape into balls. Counter rest 1-2 hours or until puffy.
  6. Sautee 1 finely chopped onion until lightly carmelized in 1 T olive oil with 1 tsp of poppy seeds, 1/8 tsp salt, 1/4 tsp black pepper. Let cool.
  7. Final shape into flattish disks with centers thinned almost to a windowpane. Place a scant teaspoon of onion mixture into each center. Rest on counter for 10-30 minutes.
  8. Bake at 425 with steam.

This was an easy prep. Less than an hour to build the dough, though the final kneading in the Ankarsrum required some attention as previously mentioned. The final shaping will take some practice as most of my bialys failed to stay depressed in the center even though I docked and severely compressed the center dough with fingers then a small bowl. Will try picking them up and stretching the centers thinner next time. Never mind that they were most excellent spread with a little (ok...a lot of) cream cheese. Next batch will need lox, red onion and ripe tomatoes from the garden.

Light browning doesn't show too well here (lighting?), but the finished goods were lightly browned with a good chewy crumb.

foodforthought's picture

Back in February, when @DanAyo posted about his experiences with an unexpected (by me, anyway) style of enriched sourdough posted by @txfarmer in 2011, I decided to give it a try. My own experience with the formula after quick scans of Danny's and @txfarmer's posts was unsettling, but ultimately produced 2 loaves of a very light, white and shreddable sandwich bread. I had difficulty and no small amount of anxiety with the relatively dry levain, a super wet dough and sluggish gluten development.

With the last slices of the February loaves finally disappearing from the freezer, I thought it was time to revisit the bread. This time I had the patience to thoroughly read through Danny's and @txfarmer's original posts and subsequent comments. Suspicions I had during my first attempt at the bread were confirmed by a more careful reading of the audience dialogues that focused on ambiguities of @txfarmer's original post.

I developed the stiff levain (71% hydration) over 3 12-hour generations starting with 5 grams of my 100% hydration standard starter. My previous attempt resulted in a rather dry, crumbly lump that didn't really change much overnight. This time I hand-kneaded the levain at each generation to ensure complete hydration and the dough rose into surprisingly puffy masses at rest. I'm still a bit uncertain with using milk in a levain destined to 36 hours on the kitchen counter, but I noticed no unusual smells or tastes, so I'm guessing LAB and yeast action are to thank.

Reviewing my notes from the first attempt, I had spotted something that sent me back to the original posts. I had not included the egg whites (24% of flour weight) in my hydration calculation. Dopeslap (facepalm?) time. So I had actually produced a 95% hydration dough. No wonder I never got a windowpane and nearly threw the whole mess out.

This time aroundI let the KitchenAid gently moosh up the levain in most of the formula's milk for a few minutes. Mixing the dough seemed much less unusual and I actually got to a windowpane (though I question whether it was a @txfarmer stage 3 windowpane) after 20 minutes of KitchenAid 2-4-6 thrashing. The dough nearly cleared the bowl in the final minutes.

After 2.5 hours of bulk with 4 sets of coil folds and 17 hours at 37º, I scaled and shaped16 75 gram balls setting them into standard loaf pans in the style of brioches nanterres. The rose nicely over a 4 hour proof at 80º. This in contrast to the 7 hour proof of my 95% hydration first attempt.

I'm still amazed that I accidentally got a decent result with my February 95% hydration dough. The crust on that batch of bread seemed exceptionally crisp but delicate. This current batch has a less crunchy crust, so the excess hydration in the February loaves must have contributed in some way.

A few takeaways

  1. Stiff levain is kind of nice to handle.I t's actually less messy the wetter types.
  2. Egg and egg whites are mostly water. Include 90% of their weight in hydration calculations.
  3. Sourdough levains do not have to always yield chewy, crusty breads.
  4. Re-engineer spreadsheet templates to adjust for levain hydration. Current model hard-coded at 100% levain hydration.
  5. Practice patience. Don't always be in such a hurry. In technology we often took a macho kind of pride to infrequently or never "read the freaking manual". Maybe not a great practice on a first attempt at a new style of bread.
  6. It's all right to totally botch a bread recipe. You learn a lot. And most of the time yeast and gluten will cover your backside.
foodforthought's picture

A few days ago, @pul posted some lovely pics of creamy, yellow semola loaves that reminded me of the community bake of a year or so ago and my good results from that exploration. ...which got me ruminating and scheming on following @pul's lead.

So down the rabbit hole of @dmsnyder's, @Franko's and @dabrownman's meanderings on Pane Tipo Altamura I dropped, deciding eventually to follow @pul's lead with a few modifications.

  1. Planned for 1.5 kg  of 73% hydration dough to build 2 batards
  2. Prepared 220 g of 56% hydration biga per @pul
  3. Prepared 300 g of semola poolish (because it's what i do)
  4. 1 hour fermentolyze(?) incorporating biga, poolish, 900 g flour and 400 g (to get to 95%) of water
  5. Added remaining water and increased salt to 1.7%
  6. 3 (more?) hours of bulk with 4 stretch and folds
  7. 18 hour retard at 37º F
  8. Preshape with 30 minute rest, then final shape and 1 hour bench rest
  9. Bake with convection/steam 20 minutes at 450º, then convection only 18 minutes at 425º

A few observations of note:

  1. This semola biga showed quite a bit more growth than I expected. The flour really liked liquid though the dough seemed quite easy to work with.
  2. 100% hydration semola poolish, on the other hand, grew but less vigorously than my usual AP poolishes. Suspect that the gluten was just holding on tighter.
  3. Machine mixing the flour went well until the very end. I had to really work on getting the last 100+/- grams incorporated. Once again, gluten seemed quite tight though, visually, an absence of the stringy fibrous gluten matrix you see with white flours.
  4. The dough, on the other hand was very silky and nice to handle. I would hesitate to call it highly extensible, though, in the end, it was cooperative enough on shaping.
  5. Loaves seemed smallish before loading, but they really took off, exposing beautiful ears. I keep relearning that I tend to get better ears on lower hydration loaves like these.
  6. Thicker and crunchier crust then I was expecting. Lovely cream(?) colored crumb. Distinctly different and pleasant flavor

In conclusion, I'm not seeing a lot of reason to handle this dough a whole lot differently than I would for standard sourdough loaves. I think next time around, I'll try getting to 80% hydration with a normal 100% hydration levain as I do for my 50% semola ciabatta.

Thanks to @pul for the inspiration and TFL community (alfanso!) for a million insights.

foodforthought's picture

A few months ago, Jack Sturgess mentioned that a breadstick challenge stumped most of the British Bake Off contestants. I happened to be making a batch of ciabatta at the time, so I thought, why not dedicate some of the dough to breadsticks. The results were pretty good and we love a crunchy snack with our nightly glass or two of wine. Anyway, after 4 passes, I think I finally have a repeatable process that produces attractive, munchable grissini.

Several weeks after starting this journey, I watched the Bake Off challenge show and saw, to my horror, an assignment that resulted in these massive cheese and olive loafs more on the order of a ficelle or even a baguette. The winner did a passable job on a ludicrous assignment. I'm glad to have been inspired by Jack's mention and happy to have missed the initial broadcast which probably would have put me off my feed enough to never look into how to make decent breadsticks.

So…I mixed up a 2 kilo batch of my 50% Durum ciabatta recipe based on the Community Bake and Alfanso's helpful process. Then while my formed ciabatta were rising on the counter, I chopped a kilo of dough into maybe 50 20g batons. The 80% hydration dough is very sticky, so I've started rolling the batons in semolina before stretching them to half-sheet pan length. You fight the gluten a bit, but I've found that leaving one pan of partially stretched batons to rest while I fill another pan, I'm able to return to the initial batch and get a bit more length on the shorter sticks. I proof them at 80-85 dF for up to an hour depending on how many batches I need to bake. Before baking, I spray the shaped sticks with water and dust with Maldon salt and coarsely ground pepper. Have made versions with rosemary or Parmigiana incorporated into dough during the final mix. Definitely going to stay in my repertoire.


foodforthought's picture

Reverted back to classic recipe after Thanksgiving sourdough buttermilk adventure. So used pate fermentee with whole milk and water in equal parts. Definitely produced perfect croissants after 2 hour proof at 75 df. Loving the proofing function on the new oven. Baked at convection 375 for 15-18 minutes…a change in strategy to reduce burns I’d been getting with 400-425 bakes. 2.5 kilo dough just about brought the KitchenAid to its knees so going to have to scale down a bit…maybe limit batch to a mere 2 dozen :-|
foodforthought's picture

Having always been envious of the incredible loft of the humongous croissants at Pavel's Backerei in Pacific Grove, California, I challenged myself to see if I could approach Pavel's great result for post-Turkey breakfast. I made several deviations from textbook croissant dough including buttermilk (instead of milk) , a poolish (instead of pate fermentee) and liquid sourdough levain in my 2.7 kg dough batch. I've been kicking the tires on the unexpected(!) steam injection function that came with a newly installed German oven so I baked the croissants with an initial burst of steam. The results were quite good though, in retrospect, I could have proofed for 3 or even 4 hours instead of just 2 1/2 to coax maybe another 1/2 inch of loft from the dough. The new oven's proof function was a huge improvement over my usual counter proofing which goes much slower since our kitchen tends to run to 65-68 degrees in cool weather.

Half of the dough went into a dozen croissants so I used an average of about 110 g of dough per croissants. I suspect Pavel is possibly getting as much as 150 g into one of his. Will have to ask next time we visit the Monterey peninsula.

With the other half of the dough I made a dozen Kouign-amann, which are constructed of danish dough rolled out in a lot (as in 25% of dough weight!) of sugar for the last lamination fold. The result was quite good, caramelly and a bit sticky, though the dough browned more than I would have liked. Rose Levy Beranbaum uses no milk in her tediously meticulous version which I have followed in the past...apparently to avoid this browning problem. Not wanting to prep multiple doughs, I took the path of least resistance and still came out with a decent product, but I now better understand her caution. I think this dough could have also benefitted from a longer proof than the 45 minutes suggested by RLB, though I'm not sure that the sugar content does not inhibit dough rise. Weill have to test that next time around.

Altogether a fun Thanksgiving distraction. And if you have the good fortune to visit Monterey, don’t miss Pavel’s for great croissants and beautiful artisan breads.

foodforthought's picture

I have long been a fan of the Semolina version of Jason’s Quick Coccodrillo Ciabatta, which may have been my first TFL recipe attempt. @alfanso’s recent post on 80% PFF Ciabatta got me thinking it had been too long since I’d investigated alternative versions. Thinking back on the Durum community bake, I decided to apply Alfanso’s method to a hybrid ciabatta and I think it worked out ok.

Increased Jason's semolina portion from 30% to 50% and used the finer Semola Rimacinata
Reduced Jason's salt percentage from 3% to 2.5%, splitting the difference with Alfanso
Upped Alfanso's hydration to 83% (including the oil portion)
Retained Jason's 1.3% yeast...twas a lively dough indeed
Pre-fermented flour at 35%

Semola Flour 535 g
Bread Flour 162 g
AP Flour (374 g (all from levain and poolish)

ADY 27 g
Salt 27g
Olive Oil 32 g
Water 857 g (374 g from levain and poolish)

Process (mostly per Alfanso)
Prep ~350 g levain over 3 generations
Prep ~400 g poolish
Mix poolish, levain and 95% of additional water, add yeast, rest 10
Add remaining flour, mix to shaggy, rest 10
3 minutes with hook at 4, rest 10
3 minutes with hook at 6, rest 10
Add oil, salt & remaining water
Mix high speed 8 minutes
Bulk Ferment ~2 hours until 3x (lively dough!)
30 minute folds
Divide, "shape" and rest 45 minutes
Preheat oven to 550
Flip, stretch
Bake 10 minutes @475with steam, then 10 minutes @450 no steam

2 kg batch produced excellent bread. Lovely airy crumb. Crust had that pahoehoe ropy lava texture thing going. Made a second 3 kg batch with similar results, though the one 500 g batch of dough I held in refrigerator overnight did not get quite the loft of the unrefrigerated dough.

This project leaves me with one rumination. Why do ciabatta recipes insist on Biga instead of poolish? I know...Italian tradition, but poolish is so much more straightforward and the math is easier. Does anyone have any objective observations about qualitative differences between doughs resulting from the two pre-ferments. As with Hammelman's 125% poolish, the difference is interesting but I can't tease out a great reason to go there...


foodforthought's picture

Blog post in progress

As of Mix Day ( February 8)...

Deviated slightly from recipe on percentage levain and included a Semola poolish.

Levain Build, start 36 hours before mix

  • 3 generations x 12 hours, all Bread Flour, 125% hydration 

Poolish, 12 hours before mix

  • 200 g water, 200g Semola Rimacinata, 1/4 t ADY
  • beautiful creamy color, less bulk, denser than AP poolish

Mix target 2000 g dough for 2 batards

  • 286 g levain
  • 392 g poolish
  • 346 g Bread Flour
  • 514 g Semola Rimacinata
  • 470 g water

Machine mix 2 minutes then scrape, cut, cajole in stray flour remaining in bowl, then 2 minutes machine mix. Rest 30 minutes. Lovely butter color dough.

  • 24 g salt

Add salt then machine knead 5 minutes

Yield 2020 g dough @70dF, room @65dF. Rest 30 minutes

4 Stretch and folds @ 30 minute intervals

  • Nice loft and gluten development after just 2 S&Fs

Retard at ~6:00 PM after 4.5 hours total Bulk on counter

First bake after 11 hour retard

  • Dough nearly doubled overnight 3 liters to 5.5 liters with Benny-recommended amount of jiggle (“like jello on springs”?)
  • Pre-shaped, 30 minute rest, shape then 45 minute rest (very easy dough to handle)
  • Bake with steam @475dF for 17 minutes, then 450dF for 25


Wish I had some control of image spacing??? 

  • pretty gorgeous batard, great oven spring, nice ear
  • soft crumb with a creamy tinge
  • crispy crust
  • fairly delicate but distinct flavor
  • On toasting, crust is reminiscent of deli sour rye, crisp but thin

Will bake another from same dough batch in a few days to see if flavor develops, dough retains it’s excellent qualities. After 96 hour retard, dough is on counter rising.

Finally got around to posting results of 96 hour retard. Pre-shape, shape durations, room temperature pretty much the same as for previous batard. 

  • Dough seemed harder to handle, wetter, slacker(?”)
  • Less oven spring
  • Bake with steam @475dF for 18 minutes, then 450dF for 28
  • Good crumb, thicker crust
  • Delicate, but more complex flavor

Conclusion: 96 hours maybe a tad too long for this dough, but overall a nice bread which I will bake again as more normal 800 g batards and maybe baguettes.

foodforthought's picture

I decided I should be more than an occasional commenter on forum posts so thought a formal introduction would be a good way to kick off my TFL blog. I most especially would like to reduce the bloat of the community bake threads so will post a link once to CBs if I participate but try to keep any discussion channeled back here. Worth a try anyway.

I’ve been lurking and learning here for the last few years most notably from the posts of @dmsnyder and the ongoing coaching, support, encouragement of frequent posters Danny @DanAyo, Benny @Benito and TFL’s cast of thousands of friendly bakers.

Oh yeah, the introduction. Grew up at the foot of Julia Child. Discovered real cheese, bread and vegetables in the seventies Berkeley food scene. Twelve years as a practicing ecologist mostly planning vegetation system rehabilitation in National Parks all over the west. Dabbled in catering and even had a restaurant for a year. Then started a family, rediscovered computers and made a second career writing software, building online learning and reference systems for a large semiconductor manufacturer. All along the way retained my interest in food and entertaining which brought me to bread construction and TFL (many thanks @FloydM) after I retired a few years ago.

 A bientot,


Subscribe to RSS - foodforthought's blog