The friend who gave me the sourdough starter recently gave me a copy of Classic Sourdoughs, Revised: A Home Baker's Handbook by Ed Wood and Jean Wood. This is the book she learned from, she speaks highly of it, and she was tired of having to look things up in it when I asked her questions so she got me my own. :-)
The basic recipe in there (the authors recommend that you get this one down first before moving on to others) calls for feeding the starter to make it active, then feeding (part of) it again to get what they call a "culture proof", and then using that to make the bread. My earlier attempts didn't include that step; I was feeding the starter, waiting for it to expand, and then using that to bake with (and keeping the rest as starter). I'm getting better rise now.
There were two other differences I wanted to test (well three, but I didn't formally test the last):
The book says to feed the part of the culture proof you don't use in bread again before you put it away. That seems wasteful, so I wanted to find out if it makes a difference. Last week I divided my leftover culture proof, feeding half and not feeding the other half. (Remember, it's already been fed twice on the way to getting here.)
The book recommends putting the loaf in a cold oven and then turning it on, baking at 375F. The authors say you'll get a nice "oven spring" that you can watch happen suddenly, except that covering the bread with a bowl for humidity defeats that. (I left an earlier loaf uncovered, trying the pan of water instead for humidity, but saw no sudden spring (though it did expand) and was not happy with the resulting crust.)
The third item was using bread flour instead of all-purpose flour. The book actually uses all-purpose in its recipes, and that's what I used in my previous two loaves with this book.
Today (and yesterday, because sourdough requires time) I filled out a little two-by-two matrix: without the extra feed ("1") and with ("2"), crossed by cold-start ("C") and hot-start ("H") in which you preheat the baking sheet, deposit the dough onto it, and bake at 450F. I made two dough batches (for the different starters) following the same recipe, processes, and timings until we got to the baking stage. I divided each into two at the dough-proof stage. I baked the two "C" variants together and then the two "H" variants, which means that within each pair the dough for one had a little more pre-cook time than the dough for the other, but I'm prepared to call that not significant.
All loaves were made with bread flour. All loaves were brushed with olive oil right before baking. All loaves were covered with inverted metal bowls for the first half of their cooking time.
Also, all loaves held their shape better than in the past. I found myself adding some flour toward the end of the kneading (last night); possibly the bread flour makes a difference here too.
(Yes I am aware I did not lay out the matrix correctly. I was wrangling hot loaves, and then I didn't think to adjust before taking photos.)
- All four rose about the same amount.
- The crumb (size of air bubbles) is pretty much the same for all four.
- The "H" loaves taste more sour than the "C" loaves.
- Dani described the "H" loaves as "more spongy", though I couldn't get a sense of what he meant by that.
- The "H" crusts are darker.
- We observed no taste difference between "1" and "2" variants.
Both temperatures have their uses, depending on whether I want more sour or bread that can be used in other applications (like for sandwiches that might not go as well with sour). I'll skip that extra feeding for the starter and save myself some flour, since it doesn't appear to affect the results when you later bake with it.
I'm pretty happy with these loaves. I finally feel like I'm getting bread rather than bread-like shallow mounds.
A silver lining of the pandemic is that if I were going to the office every day I wouldn't be able to make the timing work for sourdough. Maybe in the winter, when sunset is early, starting after Shabbat on Saturday and baking Sunday evening might work. But otherwise, it doesn't seem like it would work, at least following the techniques in this book. Of course there are other techniques, including suggestions from my readers, yet to be explored.