Back in 2007, just before I set up the microbakery, we spent three months in France writing. Problem was, it took two months to get going. The following was supposed to go in what I called my Big Book which was a sort of autobiography + bread. I eventually put the first draft round a number of friends for an opinion. Some loved it, some hated it. But they all asked, who is this aimed at? So I very grumpily shelved it and wrote what was clearly a beginners book, Sourdough Made Simple.
The Big Book is still on the shelf, but here are a few chapters of starter experiments using various bakers’ methods (me, Basile Kamir, Frank Deldaele, Dan Lepard) for your entertainment.
Chapter One: Creating a Starter
Oh dear, the problems starters cause. People are afraid of them, lie awake worrying about them, they become obsessed with them, cherish them, nurture them like pets or first-born babes, their demise is cited in divorce proceedings. Myths develop around starters, they become revered for their age and their origin, bakers argue over the way they should be made. Words are spoken, feuds go down generations.
The thing is, all we are talking about is a fermenting paste of flour and water. Go somewhere else for the science but the important elements in a starter from the baker’s point of view are the yeasts it contains, which produce the gas to make the eventual dough rise, and acid-creating bacteria that give the bread flavour.
A surprising amount of the science still isn’t understood, so if the scientists are still arguing about the processes involved, why should the rest of us worry? It calls for a what works approach and the use of a lot of “probably”s.
Yeasts are all around us and one of the predominant myths about starter creation is that you need to capture the yeasts in the air to start off the fermentation process. This in turn makes people think that a lush countryside environment will beget a purer, more natural starter. But, given that the largest element of solids in a starter is flour, it is much more likely that the yeasts which eventually get the starter active are going to come from the flour and not from the air around. John Thorne in “Outlaw Cook” tells a story of how he once threw out a perfectly good starter when he was moving from a grotty, polluted urban environment to the rolling hills of somewhere thinking he would obtain a far superior starter when he arrived only to find that he was totally unable to produce any sort of active starter in his new rural paradise. One thing that can be a negative factor in starter development is your water supply. Generally tap water is fine unless it contains harmful chemical residues.
One problem is, if you are a sourdough baker, the more experienced you are, the less likely you are to know the precise steps to making a starter (if there can be said to be “precise steps”). You create or obtain a starter early in your career and maintain it over the years so pretty soon you forget how you made it in the first place.
Broadly speaking there are two schools of thought when it comes to starter creation. The first says that, as the main element you want to ferment when making bread is flour, the yeast in your starter may as well come from that as anywhere else so take the simple route and make your starter with just flour and water. The other says there are more reliable ways to get a starter active and call for the use of items such as organic grapes, raisins and yoghurt. I like the simple approach – I don’t have any problem with those who take the second line except that there is now a myth going around that says you can ONLY create a starter by using grapes with white mould on their skins.
Anyway, being a simple soul and having forgotten the detail of how I made my current starter at the end of 1999 ready for the new century, I thought I’d try a little experiment.
Freefall Starter Creation
Unencumbered by recipes, I set out to make a starter while staying in France.
No tripod, no studio lighting (no detachable flashgun for that matter) and definitely no safety net.
Most of the photos were taken on the garden table. Unfortunately most of the time the weather wasn’t like this but we battled on.
Morning: Mixed 50g T65 organic flour with 50g tap water and placed in empty and washed Bonne Maman jar. (French flours are graded according to the proportion of bran that has been extracted in the milling process. T55 is usually used as baguette flour, T65 is more general purpose bread flour, T150 is full wholemeal.) The weather was cold throughout the week – a spell of northerly winds in South West France in March with snow in inland areas.
10.00 a.m. Nothing visible from the side view, a few gas bubbles on the top which might just be air bubbles from the mixing.
4.00 p.m. Very small bubbles on side (like black specks), more, but smaller, bubbles on the top.
7.00 p.m. The layer at the top of the jar was noticeably more grey in colour than the rest and the mixture was beginning to look active. I gave it a stir then took 40g of the mixture and added to it 40g water and 40g T65 flour – the remainder I discarded.
12.00 noon Started to get quite excited. Definite signs of activity.
7.00 p.m. Look at that – one in the eye for the yoghurt and grape brigade. Repeated yesterday evening’s refreshment, i.e. 40g from the jar + 40g water + 40g T65, discarding the rest.
9.00 a.m. Activity still very weak. Decide to take drastic action. Make a 5:1 refreshment (the previous ones had been 2:1) using T110 flour – i.e. higher proportion of bran – 20g starter, 40g water, 40g flour.
10.00 p.m. Switched back to T65 flour and made a refreshment: 50g starter, 100g water, 100g flour, discarding the rest of the starter.
Saturday8.00 a.m. Very active – as active as I would expect my regular starter to be at this stage. Made a refreshment with equal weights of starter, water and T65 flour at 100g each discarding the remainder.
4.00 p.m. Very lively. I used 200g to leaven a mix of Dan Lepards’ White Leaven dough which produces a very good baguette (four actually). I mixed another 20g with 40g water and 40g T65 flour to continue the starter for future baking. At 8.30 p.m. I shaped the baguettes and put then in the fridge for an overnight rise.
At 6.00 a.m. (the clocks went forward that night so it was really 5.00 a.m. This wasn’t heroism – I just happened to be awake.) I took the dough out of the fridge. At 6.45 a.m. the oven was turned on and at 7.00 a.m. the baguettes went in – out at 7.30 a.m.
There you have it: from nothing to producing bread in a mere eight days. You have to admit there was nothing complicated in that even allowing for the odd moment of panic.
The show must go on. Last night’s refreshment at 10.00 a.m. Sunday morning. At this stage I would continue to refresh the starter morning and evening for a few days to build up its strength with a 5:1 mix – something like 20g starter, 40g water, 40g flour. Then start storing it in the fridge.
Because the process is so imbued with mystery and newcomers fear the slightest deviation from the text will cause disaster, I am including a list of factors that are not relevant to the events above:
- The type of container (so long as it’s non-corrosive). The Bonne Maman jar was available and transparent (for photographic purposes only). After Dan Lepard published “The Handmade Loaf” people were convinced that they had to use Kilner jars because that’s what Dan used in his illustrations!
- The lid (and the fact that it miraculously changes colour). The lid is only there to keep foreign bodies out of the starter. It was left on loose to allow a little air circulation + the process (if successful) produces gasses so there is the slight possibility that the jar could explode if the lid was tightly closed. But you don’t have to have it uncovered to allow the wild yeasts in the air to gently settle in the mixture and take root (or whatever they do).
- The type of flour. Generally speaking you can use any flour to make a starter. T65 flour is what I was using most of in France so I used that. Some people swear by rye flour to get a starter active. The higher the bran content the more food there is for hungry yeasts.
Note: The jar is 75mm in diameter and the bowl 145mm. This also is not relevant when it comes to making starters; it’s just information to judge the photos by.
These are not instructions for making a starter. They are an illustration of how a starter can be made. If you want to experiment with just flour and water use them as a general guideline (as previously outlined in Beginning Number One at the start of the book):
- Mix a small amount of flour with water to form a paste.
- Use a non-corrosive container – glass is useful because you can see what’s going on but not essential.
- Cover the container (lightly or tightly) to stop foreign bodies falling in and to stop the mixture drying out. Keep at an ambient temperature.
- Within 48 hours there will probably be gas bubbles forming. If there are no signs of activity after three or four days throw it out and start again.
- Once there are definitely signs of activity, “refresh” the starter by taking a manageable amount of the mixture (50g?) and stirring in the same weight of water and the same weight of flour. Throw out the remaining mixture.
- Wait until your new mixture is producing bubbles (give it at least 12 hours and don’t worry if it is longer) repeat the previous “refreshment”.
- Repeat for two or three days until the activity in the starter is vigorous.
One concept which is hard for the beginner to grasp is that when you add large quantities of flour and water to your starter, you are not diluting it and making it weak, you are presenting the yeasts with large amounts of food which allows them to reproduce rapidly making your culture strong and vigorous. Always add at least equal weights of flour and water to you starter (i.e. one of starter, one of flour, one of water). Generally ratios are considerably higher than this, e.g. 1 starter: 2 flour: 2 water.
As this book is supposed to tell you how I make bread I was going to finish this section on starter creation at this point but while I was in France I found a book by Basile Kamir in which he described a very different approach to making a starter (and one that I felt a little dubious about) so I thought I would try it out. A couple of weeks later I found another bread book (if you’re are in Bordeaux, avoid going into Mollat, a seriously good bookshop which can devastate your budget and leave you with excess baggage), this time by Frank Deldaele with yet another method. So I tried that by which time I thought, sod it, I’ve been having gentle dig at the yoghurt and raisin approach, it’s time I gave Dan Lepard’s way a go. The results follow:
Chapter Two: Basile Kamir’s Levain – Will it work? Will it produce a baguette?
Basile Kamir is one of the great bakers responsible for restoring the reputation for quality of French bread. In 1975 he established “Le Moulin de la Vierge” and now has a number of boulangeries in Paris.
In 1999 he published quite a modest little book for the home baker – compared with the production numbers of some other well-known bakers – “La Journee du Pain”. This was revised and republished in 2006 with the slightly bolder title, given that it only has 80 recipes, “Tous les pains”. But for 9.90€ it’s extremely good value.
It contains a two stage starter; a levain chef which when active is used to seed the final levain.
According to the book the levain chef should inflate with gas little-by-little during 12-24 hours – a picture shows it about doubling in size. This was my first attempt after 36 hours:
It hadn’t risen in the slightest and, unlike the sun, it didn’t rise the next day either so, after an autopsy that revealed not a single sign of activity, it went in the bin.
The first stage – the levain chef – is much firmer than my starter, double the weight of flour to water and he uses rye flour.
Mix together 50g rye flour, 25g water at 27°C and a teaspoon of honey. Knead for a few minutes, form a ball and place in a container covered with a damp tea towel.
My second attempt started on Tuesday morning:
We have great faith that this time it will work
Well, I didn’t pack my micrometer but I can’t see any difference in the three shots. You can tell more about the weather from the pictures than the levain chef.
There are only three elements involved: the water has already been used to successfully create a wheat starter so that should be all right; the rye is good quality T130 organic flour; the honey is finest organic Pyrenean single-producer honey from individually named bees costing and arm and a leg. I don’t have a thermometer to measure the 27°C Basile Kamir wants the water to be but I carefully mixed some lightly lukewarm water and the temperature outside has been low 20°s so there should be no problem there (he calls for an endroit tempéré which I take to mean not cold rather than warm).
I have discovered later in the book that by “endroit tempéré” he means a place at 25-30°C so we’ve been operating at a lower temperature than he intends. Some people insist on fermenting dough at around 30°C and produce scientific charts to demonstrate that this is the optimum temperature for yeast activity. But (1) I don’t have anywhere at home where 30°C is the ambient temperature, (2) I prefer to operate in conditions that are as “normal” as possible, i.e. without having to devise incubator boxes, etc., and (3) I produce perfectly good bread at lower temperatures; that means between roughly 16°C at night and 22°C during the day.
By 6.00 p.m. Wednesday I’m convinced this is another failure BUT, come Thursday morning the dough has definitely moved! (I never warned you it would be this exciting.)
HOWEVER, this is but the first stage. In stage two we use the levain chef to make the actual levain for the final dough.
Basile wants you to mix all of the levain chef with 80g of water at 27°C, 75g of rye flour and 75g T55 wheat flour and to knead it for five minutes. I only had 50g rye and T65 flour so I made do with that. Now he wants you to cover it with a damp tea towel and let it ferment for four hours. I did this at 9.00 a.m.
The recipe calls for the levain to ferment for four hours before use. My experience says this is not long enough and I allow six and even here I’m cautious. The levain is increasing in size and, if I push gently into it with a floured finger, the indentation pushes back and slowly closes showing it is active.
Then, the recipe specifies that you mix 400g of T65 flour with 225g of water at 27°C and knead for five minutes. Allow this to stand for 35 minutes. You then mix 200g of the levain into the dough and knead for five minutes. 5g salt is then added and the dough kneaded for a further five minutes. By now it was 4.50 p.m. whereas on Basile Kamir’s schedule it should be 2.50 p.m.
Now you should let the dough stand in an (altogether now) endroit tempéré for 90 minutes. Should be 6.30 p.m. but it’s 6.45 p.m. Divide the dough into five (I made four) form them into balls, cover and rest for 15 minutes.
Form these into baguettes, place them on a baking sheet, allow to rise for an hour, slash with a razor and bake in a preheated oven at 240°C.
The result is the worst set of baguettes I have produced since I started baking.
Basically the dough was underproved at every stage of development which I was aware of but I wanted to follow the recipe at least to some extent. It took me back a dozen years or more to when I was struggling to produce naturally leavened bread from Joe Ortiz’ “Village Baker” and other books before Dan Wing turned me on to a more liquid starter. I’m not blaming these books or Basile Kamir’s recipe but I don’t want newcomers to sourdough to have to go through the same struggle to achieve something which is fundamentally so simple. It has, on the other hand, made me interested to have another look at French levain methods.
Even the birdies walked away from it.
Chapter 3: Frank Deldaele – Another Starter
I wasn’t going to do this but I just bought another bread book and found a very different way of making a starter.
The book is “Pain Plus” by Frank Deldaele – haven’t even worked out if he’s Belgian or Dutch yet, but he is professor of the bakery school at the Ter Groene Poorte and won a bronze medal at the Coupe du monde de boulangerie at Paris in 2002.
He calls for you to mix 150g raisins, 10g honey and 350g water at 25°C and to hermetically seal them in a jar. You are then to keep them for 4-5 days at 23°C until small gas bubbles form on the surface. The liquid is then drained off and mixed with 500g flour in a spiral mixer for 5 minutes at slow speed and 5 minutes at high speed. The dough is kept for 24 hours at 23°C and constitutes your levain chef.
I halved the quantities but otherwise did as I was told:
The first picture shows the starter after it was mixed at noon on Saturday – the murkiness is just because it hasn’t settled. The second picture is noon on Sunday. The mixture has cleared but there is no sign of activity.
The following two photos show the mixture on Monday morning and the enlargement shows that it has now clearly become active. I did not have a thermometer with me (or a spiral mixer for that matter) but the weather was warm and the temperature probably in the range called for.
I’m supposed to leave this for 24 hours but by 9.00 p.m. it has more than doubled. (I don’t apologise for the quality of the second photo. I had to press the camera against the bottom of the cooker hood to keep it steady to get it. This is live action writing taken down as it happens.)
As you can see, by the next morning, only the lid is keeping the monster in there. I’m going to rename this section “Deldaele’s Rocket”.
After 24 hours (we are only at 14 hours) M. Deldaele calls for the levain to be made by mixing (in your spiral mixer at low speed) 500g of water with 1k flour, and then adding to the dough (in your spiral mixer at high speed) 500g of levain chef and 10g salt. This refreshment is then supposed to be repeated every 24 hours for 3-4 days until it doubles in size.
A couple of observations: It’s a bit hard to see who “Pain Plus” is aimed at. It covers a good range of breads, including some naturally leavened breads (well, only four, and three of these also contain yeast), some using preferments and plenty of couques. The “A-Z of French Food” says a couque is a gingerbread cake (Northern France and Belgium); the big dictionary says it’s Belgian French for cake. Here it seems to mean the opposite of “savoury”.
The quantities generally suggest it’s for the amateur but then you are expected to have a spiral mixer and there’s a section on using proving cabinets.
The quantities used for making a leaven are weird. A kilo of flour a day for 3-4 days just to get your starter going?
Salt is not usually used in a starter except in small quantities by professional bakers to slow down its development in hot weather. Is its use here because the fermenting grape liquor is the baking equivalent to rocket fuel?
Your starter needs to be active but long fermentation is needed for better flavour development.
At 8.00 a.m. Tuesday I made the first refreshment halving the quantities given in the book (even this seems too much). After four hours its expansion is becoming a danger to shipping:
I’m supposed to refresh this monster every 3-4 days until it doubles in size? The only sensible thing is to go out to lunch.
Note the blotches on the dough in the second photo where the salt hasn’t properly dissolved indicating poor mixing.
For the purpose of comparison: the glass bowl above in 23cm as opposed to the red-rimmed plastic bowl at 14cm.
At 8.00 p.m. I used some of the levain in a mix of Frank Deldaele’s Pain du Moulin. Haven’t seen this before but he specifies the flour in a recipe like this:
1kg de farine 12/680
Which I take to mean Type 680 flour with 12% protein (an indication of the gluten content). Anyway, I use French T65 flour which should be similar.
His recipe: Mix 1k flour, 600g water, 350g levain, 22g salt, 20g fat (optional) in aforementioned spiral mixer for 5 minutes at low speed and 5 minutes at high speed. Two hours bulk fermentation. Divide in two and form two boules. Rest for 30 minutes then reform the boules and place in bannetons. Prove for 10 hours at 12°C. Bake at 260°C reducing to 220°C for 50 minutes. He also calls for steam to be injected at the start of baking and for the steam vents to be opened 10 minutes before the end of baking. He hasn’t seen this oven.
I mixed the dough at 8.00 p.m. opting for the ten prong mixer, omitting the optional fat but otherwise following the recipe. At 10.00 p.m. I roughly shaped the two halves of the dough into bâtards because two boules would not fit in the oven. Half an hour later I reshaped them, placed them seam side up in couches improvised from a tea towel – which is where things started to go wrong. I put them in the fridge intending to bake them at 6.30 a.m. because, the way the levain had been performing, I thought that they would over-prove well before their allotted 10 hours was up. Stumbling to the loo at 3.30 a.m. I had a little rethink and decided that, as it was the levain’s first outing, it might not after all prove that fast. By this time the dough would be well cold; a gentle rise to ambient temperatures couldn’t do it any harm – so I took it out of the fridge.
When I got up at 6.30 a.m. not only had the dough over-proved but the two halves had risen higher than the dividing seam and stuck together. At least I don’t panic any more. I gently prised the two halves apart and, as they equally gently deflated in my hands, struggled to hold their shape while I transferred them to the stupid, flexible, silicone baking sheet I am currently lumbered with. Then I made matters worse, helping the dough to deflate more by slashing it in close diagonals instead of doing something simple and delicate.
I tried glumly to avoid the oven window as I drank my morning coffee and the crust formed preventing any additional rise beyond the very modest reinflation that had taken place.
The only thing I don’t accept the blame for is the burnt ends – unburnt is not an option in this oven.
And, for all its appearance, it made a delicious breakfast, which is the great thing about breadmaking – even your mistakes taste good.
PS: It was better than I thought. Come Friday we were still eating it untoasted for breakfast. Like most sourdoughs it kept extremely well and the flavours developed with the keeping.
Chapter 4: Dan Lepard’s Starter
Well, I most definitely was not going to do this but I’ve been and bought the yoghurt. It is the finest organic sheep’s yoghurt but there’s no getting away from the fact: it is YOGHURT. It has turned out to be fascinating trying out these different methods.
Dan Lepard is a very fine baker, teacher, communicator, photographer, generous with his time and knowledge, always looks for the best in situations, a pretty good all round human being. Makes you sick. The only thing I can do better than him is grow hair.
I have a long-standing joke at his expense (it’s the best I can do) about sparkling water. Like the Kilner jars that people started to think were essential to starter creation after he used them in the photos illustrating his step-by-step method, he called for sparkling water to be used in one of his Italian bread recipes and panic set in across the baking classes that the bread could not be made without this vital ingredient. If he had been Delia Smith there would have been a nation-wide shortage of San Pelligrino.
The reasons Dan gives for using raisins and yoghurt in his mix are because simply using flour and water can be unreliable whereas raisins are likely have yeasts on the surface and because the yoghurt should contain lactic bacteria. Lactic acids give the bread that sharpness of flavour characteristic of sourdough.
Dan has had this method used under sterile conditions by the National Collection of Yeast Cultures in Norwich. After a week at 25°C most yeast was found in the raisins followed by the wheat flour. No yeast was found in the rye or the yoghurt, only bacteria.
On the other hand Paul Merry recently told me that he no longer bothers to keep a rye starter because he has discovered that rye/water solutions become active so quickly it is just as easy to create a new one whenever it is required.
Dan also had a starter tested that he had brought back from Denmark and reputed to be over 100 years old. It was found to contain many different types of yeasts and bacteria. This, Dan suggests, may mean that starters become very stable once established and are able to withstand other yeasts and bacteria they come into contact with. There is another school of thought which says it is pointless to collect different types of starter because they quickly become taken over by whatever yeasts and bacteria predominate in the environment where they are stored.
Lots of scope for opinion and experimentation.
Day one of Dan’s method calls for you to mix 150g water, 2 tsp rye, 2 tsp strong white flour, 2 tsp raisins or currents and 2 tsp yoghurt in a Kilner jar. Giant killer that I am, I used a clear plastic bowl. I also substituted T65 flour for the strong white.
Left, just after mixing. Right 24 hours later – just as Dan says, no perceptible change except that the surface will look shiny as the solids separate and sink. Must be doing something right.
Day two, stir in 50g water 2 tsp rye, 2 tsp strong white.
Day three. Dan says that the raisins will have started to break down resulting in a coffee coloured ring appearing round them. He also says there may be the odd pin-prick of fermentation on the surface. I probably had the shakes carrying the bowl out to be photographed eliminating any brown rings but – look at those bubbles on the left
On day three 100g water, 4 tsp rye and 4 tsp strong white are stirred in.
Day Four. According to Dan there should be the “froth of fermentation” beginning. Well, we had the froth of fermentation yesterday afternoon during the photographer’s day off and what we have now is a scummy swamp (above right) – the solids have separated and sunk apart from the raisins floating under the surface. You can see on the sides of the bowl that the mixture has foamed up and then subsided. The white spots are bubbles that float to the surface when the bowl is disturbed. I’m going to assume it’s OK and proceed.
We now have to throw away three quarters of the mixture and stir in 100g water, mixing well. Then we strain out the raisins and mix 125g strong white flour into the remaining liquid.
Day Five: I managed to get the photographer to come in for an hour in the afternoon of Day Four because the fermentation is peaking about eight hours after the refreshment of flour and water has been added. As you can see from the right hand photo the activity has died down after twenty four hours. Again we throw away three quarters of the mixture and add 100g water and 125g strong white flour.
This is where we depart from the script for two reasons: (1) because I think the starter is ready for use; (2) because I’ve only got page one of Dan’s instructions with me. And while we are confessing our little incompetences, I have to admit that on Day Four I forgot to discard three quarters of the mixture before straining out the raisins and adding the refreshment ingredients. So I then had to discard most of that and then add flour and water until I achieved what I thought to be the right consistency. This explains why the mixture in the photos above is more liquid that the ones below. But there is a lesson here: situations are redeemable if you don’t panic.
Dan wants you to let the Day Five refreshment stand for a further 24 hours but I let it stand for about 10 hours until 6.00 p.m.
The mixture is back to the consistency specified by Dan and is very active but has fallen slightly as you can see from the right hand picture.
At 6.00 p.m. I used 200g in a mix of Dan’s White Leaven dough (with the addition of sesame seed), took 50g and refreshed it with 100g water and 100g flour for baking later in the week.
Dan says the reason he uses more flour than water in the refreshment is because it slows the fermentation process down which prevents the leaven from rising and falling to fast. Traditionally in French baking leavens were quite stiff doughs, as in the Basile Kamile example although some modern French bakers, such as Eric Kayser (100% Pain), have taken to using liquid leavens. I still keep mine at equal weights of flour and water because discovering this way of dough making from Dan Wing (The Bread Builders) transformed the quality of my bread. Perhaps it’s time I started experimenting with different ratios.