The Fresh Loaf

News & Information for Amateur Bakers and Artisan Bread Enthusiasts

Interesting Experiment with Sourdough Starters

hanseata's picture

Interesting Experiment with Sourdough Starters

I just tried my first recipe from Hamelmann's "Bread" - the Mellow Bakers September challenge: Sourdough Rye with Walnuts. Since I had enough mature rye starter I decided to experiment a bit. I made two sourdough starters, one following Hamelmann's instructions to the letter, ripening the starter for 16 hours at 70 F. The other got the 3-steps treatment (3 feedings at falling temperatures, from: M. P. Stoldt "Der Sauerteig-das unbekannte Wesen").

After 16 hours the 3-stage starter looked more developed and hat a very pleasant sweet, almost fruity, smell. The "Hamelmann-starter" looked a little less smooth and its aroma was much less pronounced. Both starters went into the fridge overnight. The next morning they looked about the same, the Hamelmann one smelled stronger, but still less than the 3-stage one.

Mixing the doughs I realized that Hamelmann's instruction pertains only for industrial mixers, no way a regular stand mixer could incorporate a cup of walnuts at the end of the mix, when the dough is almost fully developed, at low speed. Because I wanted to see the difference, I nevertheless followed the instruction with the "Hamelmann-dough", and, as expected, had to mix some more by hand in order to avoid the dough getting too warm. To the other dough I added the nuts slowly and continuously through the feed, and had no problem incorporating them without additional time or hand work.

Both doughs were then treated exactly the same, proofed in bannetons and baked together according to the recipe. When they came out of the oven, they looked pretty much the same. But when I cut them there was a remarkable difference: the "Hamelmann-Loaf" was denser and had an oddly marbled look - the nuts being basically in one layer - , whereas the "3-Stage-Loaf" was less dense and had a more uniform look - the nuts being evenly distributed.

But the most amazing difference showed when we tasted the breads. 3 people found unanimously that the 3-stage Sourdough Rye with Walnuts tasted better than Hamelmann's one stage version!

Both breads looked like this one: Sourdough Rye with Walnuts


Crumb: Sourdough Rye with Walnuts (following Hamelmann's instructions)


Crumb:Sourdough Walnut Rye (3-stage version)


Comparison - the upper slice is the "Hamelmann-Loaf", the lower one the "3-Stage-Version"

nicodvb's picture

Hi Karin, both loaves look  very nice to me.This rye+walnuts combination seem to be recurring very frequently lately, I'm seriously tempted.

I'm curious to know if the threee-stage method you followed is the Detmolder.


wally's picture

Hi Karin,

Thanks for the experiment and for sharing it.  One thing I keep in mind when using Hamelman's wonderful book is that his mixing instructions are indeed basically geared for those using commercial mixers.  So when I use a recipe calling for walnuts or olives or other 'large' objects, I improvise: I'll do the initial mix in my stand mixer, but then take the dough out and do a series of slap and folds and whatever to equally distribute the addition.  With walnuts in particular, I've taken to chopping them up a bit more finely since my mixer is never going to accomplish that.

Good looking breads!


hanseata's picture

Nicodvb, I used the method described in Martin Pöt Stoldt's book: "Der Sauerteig - das unbekannte Wesen (the Sourdough - the Unknown Being). The author is also administrator of a German sauerdough website ( In his book he compares the

  • Classic 3-stage Development: too difficult for the average person, requires unlimited time and strict temperature control.
  • Detmolder Method: feeding with a large amount of flour heightens risk of possible contamination (and spoiling) of the sourdough. Also the less than optimal yeast development often requires the use of additional commercial yeast.
  • "Pöt-Method": the author developed a modified 3-stage method combining the good points of both techniques, the easier use of the Detmolder with the better results of the Classic 3-stage. It can be also used with lower temperatures as long as every stage has a lower temperature than the one before (2 degrees Celsius less = 1 hour longer to develop).


1. Stage: 50-100 g sourdough starter + 100 g rye flour + 100 g lukewarm water                   (6 - 8 hours at 26 - 28 C/79 - 82 F)

2. Stage: all stage 1 starter + 100 g rye flour + 100 g lukewarm water                          (6 - 8 hours at 22 - 26 C/72 - 79 F)

3. Stage: all stage 2 starter + 100 g rye flour + 100 g lukewarm water                          (3 - 4 hours at 18 - 22 C/64 - 72 F)  

I divided the amount of rye flour and water for the starter by three and followed the method as good as possible, running around in our house with my instant thermometer to check out the temperature differences.

The reason why I did this experiment: I had used Pöt's method before in a bread, but had no comparison and therefore no idea whether it made any difference in performance or taste. I actually was quite surprised at the result!

Larry: I suspected as much, I had experienced the same difficulties before with my olive bread (not Hamelmann).

And thanks to you both for your compliments!









hanseata's picture

Mein Keyboard kann leider keine Umlaute. Sorry for the mistake about your website - I have your book. The baking world is small - thanks to the internet!.

Herzliche Gruesse aus Maine,



rodentraiser's picture
hanseata's picture

Thanks, Rodentraiser, for your help - though the Mac has a different way to do it, I found the link to the right website - I never bothered to look before.

Viele Grüße,