The Fresh Loaf

News & Information for Amateur Bakers and Artisan Bread Enthusiasts

tartine starter instructions

ekphrasis's picture

tartine starter instructions

hi everyone, I'm new in here and this is my first post.  I have some quesitons about the Tartine book's directions for building a starter.

The tartine book, and a few things I've read online here are my reference points for artisan SD breads.  I've baked loaves using commercial yeast, and I've baked a handful of loaves using the Tartine method.  

Oddly, my experience seems to be the inverse of what I've read on here.  My first Tartine loaf ever came out perfectly.  It was remarkable.  Everyone that I showed it to was blown away, as was I.  Is it really that easy?  Each successive loaf has been a little less perfect.  Now I'm at the point (maybe 10 tries later) where my dough is the goopy, unmanageable mess that many people describe on this site.

I accept responsibility for the decline in my product quality.  Admittedly, I was far more careful and exact the first time round, so I'm almost surely doing some things incorrectly.  But my main quesitons have to do with starter maintenence.  I feel like I'm proceeding blindly when it comes to keeping the starter hapy, and I also feel that the directions in the Tartine book are vague, at best.  Hopefully you all can give me some advice.  

my main issue is waste: i only bake once a week at best, so feeding the starter every day is both hard for me to keep up with and hard to justify given the amount of flour i'm feeding.  I've tried keeping the starter in the fridge, but then the whole de-hybernation process vexes me.  It seems like it takes an unpredictable amount of time to get back into action, perhaps because I'm doing things wrong.  Then I end up using it simply because my schedule demands that I start baking, and it probably isn't ready.

the other issue has to do with the specific values for feeding the starter.  The tartine book simply says to start with 20% of your existing starter: "Replace the discarded portion with equal amounts of water and the 50/50 flour blend" (46). (but how much is 20% of "it"--the book is so frustratingly vauge here--I've just started out with "handfuls" of flour and a "container" of water), then to feed this with equal parts water and the 50/50 flour mix.  But it never says what ratios to keep between 1) the old starter and 2) the new flour & water.  am I keeping a tablespoon of the old starter (about 30g) and then a tablespoon of 50/50 flour, and a tablespoon of water?  Also, some people on this board seem to favor feeding more flour than water for the "mother" starter (not a term Robertson uses).  The Tartine book never specifies if it is equal parts by weight or volume.  I had pretty good results for a while keeping 60g of the old starter, then adding 120g water and 120g 50/50 flour (300g total, with 80% getting tossed each feeding).  But again, I didn't like how much flour I was burning through every day.  Even if you follow the Tartine basic country loaf recipe, you make 400g of leaven but only use 200g of it, which then means 200g to start with.  Why so much leftover, if you only really need 60g or so for the next feeding?  Seems wasteful.  

The other question that I have is what is the difference between starter and leaven?  The Tartine book makes it seem like leaven is just starter that you intercept halfway through its cycle, and feed again with larger percentage of flour and water.  THe book says to use one tablespoon of starter that is about 12 hours old (I'm assuming, because it is a little vague: it seems like you feed starter in the morning but start the levain at night).  My problem is this: my starter seems perfectly alive--if I leave it for a day it gets very puffed up and sour smelling.  But when I follow the directions and intercept it in the evening, take a Tbs and add 200g water + 200g 50/50 flour, the next morning it never passes the "float test."  I started just using the puffed up starter as my leaven and totally skipping what the Tartine book describes as the leaven stage.  However, I get bread that rises during the bulk fermentation, but it often is too wet and doesn't hold its shape when it comes to shaping.  I'm guessing that I'm causing the problem by not having my starter/levain cycle dialed in, but could somebody explain to me the rationale behind this?

so what I'm really wondering is how do you know that the starter is at the point that it can be used to make the leaven (and does Tartine describe a 5% innoculation for the leaven?  Again he mentions this but it isn't totally clear to me)?  And then, what's the most economical quantities to use so that the starter is ready to make a leaven once a week, say, on Friday night?  Finally, why does my leaven never pass the float test as quickly as Robertson suggests that it should (e.g. 2 hours after rejuvenating it)?

sorry for the long post!  thanks for your thoughts



cranbo's picture

 am I keeping a tablespoon of the old starter (about 30g) and then a tablespoon of 50/50 flour, and a tablespoon of water? 

"Equal parts" means "equal parts by weight", not equal parts by volume, if you want to be accurate about your baking. This assumption is true 99.9% of the time in bread baking recipes/formulas.  

This means if you keep 30g of starter, I recommend 60g 50/50 flour mix and 60g water. Keeping 30g starter is good. 60g of flour is between 1/3 and 1/2 cup, which should be plenty for keeping your starter in maintenance mode, and will keep your waste down. When it's alive and you're getting ready to bake, then bump up the feed to 30g starter, 100g flour and 100g water, which should leave you just enough to leaven your dough (200g) while keeping just enough (30g) to refresh your starter.  

My problem is this: my starter seems perfectly alive--if I leave it for a day it gets very puffed up and sour smelling.  But when I follow the directions and intercept it in the evening, take a Tbs and add 200g water + 200g 50/50 flour, the next morning it never passes the "float test."

If it fails the float test, it's probably either not ripe yet, or overripe. You will have to adjust your time accordingly. Next time, wait 30 more minutes, try float test again. If it fails, wait 30 more. If after several hours this continues to fail, my guess is that it's overripe, and your starter has overdeveloped overnight, so start again and this time reduce the amount of time by a few hours, and try the float test every 30 min until it passes the float test. Make careful notes of your overall elapsed time, air temperature, and location of starter storage so you can replicate and test more in the future. 

Hope this helps!

All at Sea's picture
All at Sea

Ekphrasis ... mind if I ask what temperature your kitchen is? It will vary of course - but an average figure?

I ask, because starter activity - and indeed all yeast activity times will always be dependent on temperature.

So, for example, when you make your levain, which is in the evening, temperatures will be probably be cooler and, possibly during the wee small hours, will drop substantially, compared to daytime temps. The effect will be to slow up the wild yeast activity. However, should you make your levain during the day, it is likely to be ripe far sooner - and may even pass the float test in the 2 hours Robertson suggests. Your kitchen may be a lot cooler anyway, than Robertson's. That's why it would be helpful to know what temperatures you're working in first.

Cranbo's excellent suggestions above should save you a lot of discard. I'm not surprised you want to reduce the wastage - it does sound a huge amount.

All at sea

ekphrasis's picture

both of your suggestions are helpful.  currently my kitchen is in the mid 70s, which is, I believe, what Robertson says his is.  it sounds like cranbo is talking about keeping the starter at a 20% innoculation, but then reducing this down closer to 10% (30g/230g) when it is time to make the leaven.  Robertson suggests a much more diluted leaven though.  I'm just wondering what the reasoning is behind reducing the leaven, and how one can guestimate the time it takes to be ready to use based off the percentage of innoculation.

perhaps the bigger question that I still have is how do you know when the mother starter is ready to use in the leaven.  because temperature is such a variable--one among many--i'd rather know exactly what i'm looking for rather than simply mix up the leaven at night time before i intend to bake in the morning, and then hope it passes the test the next day.  for example, I fed my starter with equal parts old starter, water, and flour (30g:30g:30g) around 11am this morning and now, at 6:30 it is about 30-40 percent bigger and bubbly.  can I mix the leaven up right now, or should I wait longer, if my goal is to make bread around 7am tomorrow?


Red5's picture

That's the one thing about the Tartine book I had an issue with, was the vague description on ingredients for the starter. When I did the mix to the instruction of "thick batter with no lumps" it didn't work out, ended up with a pretty good layer of hooch. What my idea of "thick batter" is versus his seems to be completely different. For as detailed as the rest of the instructions and explanations are, it's wierd he was so non-chalant about this part of it. 

How do you know when a leaven is ready? As long as it's fresh and active, it can happen pretty quick, that much can depend on your room temp and the temp of leaven itself. I've taken to using it after 2 hours, for both my white starter and 100% whole wheat starter. I think what he was going after by instructing to wait about 4 hours/till it passes the float test is to keep consistant performance for the rest of the recipe, and for your own future loaves. A starter may be ready at 2, 2.5, 3 hours, but it may take longer to bulk it properly, the Tartine book is looking to get that done in 3 to 4 hours, which can be better done with a starter that is at the float test stage....when I do bread with 2 hour starters, I know I'm looking at 6-8 hours of bulk. 

Regarding the diluted thick or thin the leaven you use for the bread will affect the final texture and crumb, and bulk time.  There is a difference between breads made with bigas, liquid levains, and poolish, etc. The reason to keep controls on the timing of the leaven is that you are controlling the flavor, how sour or not sour the final loaf will be. The bread you would get from a loaf made with leaven that is 2, 4, 8, or 24 hours old will all be different, all take different times to bulk to the same level...even with the same ingredients. 

And you do not need to keep more than 100-150 grams of starter, and you don't need to feed it everyday if you keep it refigerated between feedings. I keep both my starters at 6 oz total (about 150 grams), 2oz each of levain, flour, water. The white one is almost 2 years old and the 100% wheat is about a year. Never had any issues with raising bread with them, I gave up on waiting for the float test to pass, started using them at 2 hours old and never looked back. 

In the end, you're looking to find your bread, there are many ways to get there. The Tartine book is just explaining one way and giving you the tools to find another. 

ekphrasis's picture

thanks, red5, you answer so many questions in your very informative response.  I like the tartine book a lot, I just was a little vexed by what you refer to as its non-chalance around the starter issue.  that seems to be the cornerstone of the whole endevour.  

i'm interested to try using the leaven after only a few hours, even if it doesn't pass the float test.  in fact, i prefer leaving the dough for a long bulk fermentation.  i'd actually like to get started at night and then let it do its bulk overnight.

 i'm wondering, though, if i'm starting to build the leaven too late in the growth cycle of the master starter.  today, for example, I took out a tablespoon of the starter around 9:30 PM and made leaven following the tartine instrucitons.  I had fed the starter in the morning.  It was pretty warm here today--in the high 70s/low 80s, and the starter had already puffed up and started to drop a little by time I used it.  I tried using slightly cooler water than the directions called for (78F) but I'm guessing that by the morning the leaven will be more developed than the tartine book intends for the recipe, just because it's warm here.  it sounds like, from what you say, that I should adjust for the maturity of the leaven with the time of the bulk.  I'll see how that goes in the morning!


All at Sea's picture
All at Sea

i'm interested to try using the leaven after only a few hours, even if it doesn't pass the float test. in fact, i prefer leaving the dough for a long bulk fermentation. i'd actually like to get started at night and then let it do its bulk overnight.

Are you meaning to bulk ferment the main dough overnight, or to leave the levain to ripen overnight?

The levain can be used while it is still on the rise. Rather than risk overproofing the levain, I add it to the main dough ingredients when only 2/3rds to 3/4 risen. This also avoids the problem of overproofed, hungry yeast producing enzymes that can play havoc with the gluten structure of your main dough. Only trial and error will show you (roughly) how long this will take. And the timing will be hugely dependent on how much starter you use to make your levain, and the temperature of your kitchen during its development.

For example, I live in a very warm environment, so if I want to allow a levain to ripen overnight, I must use very little starter, a generous portion of flour, and very cold water. Only then can I be sure it will be ready by morning, but not over-proofed. However, if your temperatures drop significantly overnight, then the yeast in the starter will ferment the flours of the levain more slowly, buying you more time before you need to use it.

i'm wondering, though, if i'm starting to build the leaven too late in the growth cycle of the master starter [...] and the starter had already puffed up and started to drop a little by time I used it.

I would certainly try to build a levain with starter that is on the rise rather than sinking, yes.

I tried using slightly cooler water than the directions c, yeslled for (78F) but I'm guessing that by the morning the leaven will be more developed than the tartine book intends for the recipe, just because it's warm here.

Yes - see comments about overnight levain proofing above. My understanding is Robertson likes to use starter and levain that is young(ish) to avoid too much sourness in the finished bread. I don't have his book, so forgive me if I've misunderstood, but that was the gist I've picked up from here. If that is so (and I empathise with his aims) then you really do need to keep an eye on your starter and levain, to use them before they ripen too much.

From some of your earlier comments in your first post about recent attempts at Tartine Bread being difficult, due to the gloopiness of the dough - it does sound as if overproofing your levain is part of the problem. And possibly overproofing your main dough, too.  The most important thing to ensure is that the yeast doesn't run out of food. When levains or starters start to sink, that's what is happening. If you can arrange to do a trial levain using the quantities of starter, water and flour you want to use for the recipe, but ensuring you are on tap to spot when the levain is ready to use - and for how long - before it begins to collapse, then it will be very, very useful. Based on your results, you can then fine-tune the ingredient quantities, temperature of water etc, and arrange your baking day (and night ferments) to fit in around your schedule.

All at Sea

timko's picture


I had similar concerns around the question of waste and feeding my starter.

I was able to resolve it by reducing the starter volume to around three tablespoons - So I keep a very small amount of my starter each day - around a tablespoon, feed it with about twice that amount of flour (50 % white mixed with 50% whole meal) add water and mix to a paste, cover and leave out in the kitchen.  Feed it around dinnertime. You don’t have to feed it every day – but I like to as I keep it out in the kitchen.

This amount, I have found, offers enough starter to build your leaven. Robertson is specific about the leaven weights and proportions, but you are right, not with the starter.  I think the idea is to get accustomed to how it rises and falls and how it smells at each stage.

Trying to build in baking every three or so days for me means that I prepare a leaven in the morning before work, then prepare dough that evening and bake the following morning.

So the leaven has around 8 hours at about 60º F. (remember the temperature of leaven acclimatizes to that of the exterior temp of kitchen) While I don’t do the float test, the texture of the leaven, the smell and the aeration are the things to look for. Or do a float test. If yours has puffed up and developed an aerated silky texture, its ready to use. Floating it would verify what you can already see.

I have often wondered about float tests myself – and thought that with such long period of time for the leaven, the bulk rise,  final rise that it would probably work without a test.  It has every time.



Doc.Dough's picture

You don't need very much maintenance starter since you can blow up 25g to a lot more than you need in one or two refresh cycles.  I keep about 30g and (in the summer) use 1g to start a 1:14:14 refresh cycle while using the other 24g of starter + 230g of water + 230g of flour to grow (overnight) enough starter for 1400g of dough. The link below shows how I do it.  It is not delicate, but if you don't plan carefully with refresh ratios and temperature, your timing may not be what you would like.  I have let it run 24 hrs with no apparent impact on the refresh time or growth profile for a next-day batch of bread.


ekphrasis's picture

well my first attempt at baking after reading your suggestions was a huge success.  I was more careful with the leaven and built it following Robertson's directions.  the starter that I used I had fed earlier that morning using 30g starter:50g water: 50g flour mixture.  By 8pm it had easily doubled in size, and I made the 30g:200g water:200gflour ratio for the leaven.  The difference was that this time I used King Arthur flour, which I splurged on rather than using the store brand stuff.

Wow...the dough was a like a different substance.  So much more structure even at the early stages.  by the time I got to shaping it was far more workable--wet but it actually let me stretch out and make the folds that the Tartine book illustrates.  Baking it the bread rose a lot and it came out of the oven with a beautiful--but far more delicate--crust.  That's actually the biggest difference that I noticed.  Even a day later the crust seems much softer than when I made it before, even though it is still a dark brown and has a nice flavor.  Could that be the flour?


HappyHighwayman's picture

I've been using the book and I too found it confusing. INstead of a "leaven" I used 200 grams of my starter and it seemed to work out fine. Last night I took a tablespoon of my own starter, fed it with 200 grams of 50/50 and 200 grams water and let it sit overnight (it's pretty cool in San Francisco in my house). It had risen about 20%...didn't bother with the float test, I just used it as the leaven and mixed the remainder back into my refrigerated starter.

Right now it's undergoing the primary rise in my oven...I pre-heated it with a stone and let it drop to around 100F and will let it rise 3-4 hours. 

Also, he recommends 3-4 hours proofing time, but most recipes only call for 45-60 minutes, why is his so much more?

Codge's picture

I know this question will sounds like blasphemy, but what if I use a small amount of commercial instant yeast, say 1/2 tsp in addition to what comes from the starter?  I find it enhances the rise.  Can this do any harm?  Do the two yeasts compete?  Has anyone else tried it?

mrfrost's picture

It's done all of the time(in various recipes).

If you have been doing it, and like the results, that's all the verification you need. In a sense, it becomes "your" recipe.