Calibrating My Cookie Pans, Step #1

I performed an experiment to calibrate my cookie sheets. "What on Earth does 'calibrating cookie sheets' mean?", you may well ask. Good question. That's may not be the best term to describe the process. Though the idea can't possibly be that original, I've never heard of anybody doing exactly this before and so I'm not sure what phrase others use to refer to the process. I'll just explain what it's about.

My last post described the six different cookie sheets I am starting with:

1. cheap 13 x 9 Good Cook traditional finish
2. large 15 x 14 Good Cook slide-off pan
3. medium 13 x 12 Good Cook slide-off pan
4. steel 15 x 14 Goldtouch(TM) insulated pan
5. aluminum 15 x 14 insulated pan
6. medium 15.25 x 10.25 Wilton pan

I will be trying recipes out with these various pans. Every recipe specifies a baking time. If two pans bake the same cookie to different degrees when in the same oven for the same length of time, then of course the baking times of recipes have to be adjusted depending on which pan you use, if you want your result to match what the originator of the recipe intended. Do you know what pan the originator used? Nope! They never tell you that, except for pan sizes with most bar recipes. In fact, you don't even know for sure how your oven's thermostat and temperature fluctuations compare with the oven originally used to determine the baking time - even if you calibrated your oven! - so even if they told you about the cookie sheet they used it usually wouldn't help much. No matter what, you're going to have to figure out how long you have to bake the cookies to make them come out the way you want them to. But if you don't have any idea how your cookie sheets behave in your oven or how your oven's thermostat is set, you'll spend a lot more time trying to figure out what you ought to do for a batch of cookies, especially if one time you bake a recipe with one cookie sheet and the next time with a different cookie sheet, in which case you'll be highly likely to consistently get inconsistent results and just feel like you're not good at cookies. For myself, I really want to be able to interchange these cookie sheets with a single recipe without wasting a lot of time and ingredients because I over- or under baked.

I certainly had no idea ahead of time how much difference using different cookie sheets would make, but that didn't matter. I did the following experiment to help me to know how to adjust the baking time depending on which cookie sheet I am using. I am calling determination of the necessary adjustments "calibration" of the cookie sheets.

Butter cookies are a type of cookie for which many recipes say "bake X minutes, or until the cookies are lightly browned at the edges." It's great having the real criterion for knowing when the cookie is done, because otherwise you can only go by the baking time, which is too vague since you don't know what kind of pan was originally used. And it's even greater when the real criterion is a simple visual examination - way better than "bake X minutes, or until the cookies are almost ready to explode in your oven," for example. I used a generic butter cookie recipe to calibrate my baking sheets because there is a clear and simple way to know how far along the baking has gone. I made up enough dough for about 50 or so cookies. I formed the dough into a roll and sliced it using a ruler so each cookie was approximately the same thickness. Then I baked from one to three cookies at a time on each pan for varying lengths of time, the time being carefully measured and the predetermined intended times being adhered to within about 10 or 15 seconds, and determined, for each pan, how many minutes it took that pan to produce cookies that were slightly browned at the edge.

The experiment works fairly well because 1) the precise state of baking is easily determined, and 2) it doesn't take long for this kind of cookie to go from no brown to fairly brown. Whether that's always true of butter cookies, I don't know, but it worked with my recipe. I took a photo of the cookies that came out of the experiment, and it's attached to the post. There are six rows of cookies, with each row corresponding to the pans in the same order in which they are listed above. Before I look at the results, here's the recipe in case anybody wants to try this with their own pans. Even if you only have one pan, it's good to know where that pan's baking time fits in the spectrum of baking times of all pans, because it helps you make a first guess at how long to bake a new recipe. Not that I'm using all pans, of course - but I do have some range of pans, and you could compare against the results from my pans, if you wanted to, to get some rough idea how your pan compares to others.

1 stick butter, room temperature
1/2 cup sugar
1/4 tsp baking soda
1/2 tsp vanilla
1 cup flour

1. Cream the butter and sugar in a small bowl until light and fluffy
2. Add the vanilla to the butter and sugar, and continue beating for at least an additional minute
3. Mix the flour and baking soda together fairly well - stir them together in a bowl
4. Sift the flour and baking soda together
5. Add the flour and baking soda to the butter/sugar in thirds, mixing a little by hand after each addition. Make sure it's well mixed in
6. Form the dough into a roll about 1.5 inches in diameter by rolling gently on waxed paper or other flat surface
7. Wrap the roll up in plastic wrap or wax paper and refrigerate for at least four hours
8. When ready to do the baking, slice the entire roll into 1/4 inch slices with a ruler, and then immediately put all the slices back in the refrigerator. Use a thin knife and cut straight down all the way through the roll to minimize the number of cookies that break. Of course, if you're only calibrating a couple of pans, there are more than enough cookies. With six pans, I actually was several cookies short because I wanted to bake three cookies in each batch. You can easily double the recipe to get more test cookies, but it's critical that every cookie have exactly the same composition even if you have two rolls, so mix everything together at the same time. Making up a second batch of dough afterwards if you run out invalidates the results of direct comparison of one cookie to another.
9. Preheat the oven to 350 degrees and leave it on for a good long time (like maybe fifteen minutes after the pre-heat light goes off) before you do your first test batch in order to make sure the oven temperature is as stable as it is going to get.

Because the recipe is small and dough for small recipes tends to get caught up in beaters in one big intransigent mess - at least in my mixer! - and because the dough stuck in the beaters doesn't get mixed all that well, it really works much better for this to mix the flour in by hand. It's very important that everything be mixed in together pretty well, or else one cookie won't have a close enough compostion to another cookie and the results of the calibration may not be anywhere near accurate. This isn't rocket science, but if we don't do enough mixing for this we're just wasting our time. You'll know it is mixed well enough if, when you form the roll(s) in step 6, there are no problems with the roll wanting to divide itself into two or three pieces. The whole lump of dough should adhere to itself with no "weak" point where it wants to break into separate lumps. If it wants to break apart, mix it some more with your hands. Meaning, in this case, your actual fingers, not a spoon or spatula.

Incidentally, the refrigeration in this case is specifically because otherwise the dough is kind of gooey and it's hard to get very regular cuts with the ruler. We need every cookie to be as much the same as possible without knocking ourselves out. The length of refrigeration is to make sure the temperature of the dough has stabilized so every time you start a new batch you are starting from the same point. We can either stabilize the temperature at room temperature or stabilize it at the temperature of the the refrigerator. Dough doesn't conduct heat all that well, so it takes a good long time for the temperature of a roll of dough to be completely stabilized all the way to the center. (Not that you will be able to do this perfectly without executing with mind-boggling precision which you, perhaps, but not I, might have the ability to do - but it's just cookies after all. We can't do this perfectly, but we can do our best). If you were really wanting to make regular butter cookies this refrigeration would probably be too long because the dough will sometimes crack when you cut it when it is so thoroughly chilled, but here the stability of temperature is the important thing, not whether a few cookies get broken before baking. And finally, to my mind there is too much sugar in this recipe so I personally wouldn't want to eat a bunch of these cookies, but it just isn't relevant for this. It works for the experiment.

If you don't have any idea how many minutes it will take with the sheet(s) you are calibrating for the cookie to turn brown around the edges, take a first guess based on my results - start with 7 minutes for a dark pan with non-stick finish or 10 minutes for a light-colored pan with traditional (not non-stick) finish. If it's neither of those, just take a wild guess somewhere in the range of 7 to 10 minutes. Set up a little table on paper. Figure you're going to bake one batch for that amount of time, and one batch for one minute less than that, and one batch for one minute more than that. Set up columns, and also number each batch. Your paper might look like this:

1. 7 min
2. 6 min
3. 8 min

If you are timing with your watch, you can record the "done" time in the appropriate row, so you don't forget.

Set up your first test batch - take two or three presliced cookies from the refrigerator and put them in the very center of your cookie sheet, spaced about one inch apart. Leave them there for about five minutes. After five minutes or so, quickly stick them in the oven without leaving the door open any longer than absolutely necessary and immediately either set a digital timer or note the exact time to the second and calculate at exactly what time the batch will be done. (A hint, if the latter, it's easiest to wait until the second hand is on the 12 to put them in the oven. :-) ) After the allotted time for this test batch has passed - 6 minutes for batch one in the above example - instantly pull the cookies from the oven. Let them cool for 30 seconds or so -the exact amount of time isn't important at all because the browning doesn't seem to continue after they are out of the oven - and put them in a small group on a cooling rack. Take a scrap of paper and write the number of the batch on it or the amount of passed time, and put it on top of the group of cookies if there's the remotest chance you'll forget which group is which. For me, if there is more than one group, I might forget which is which. I used batch numbers because it was easiest when calibrating a bunch of different pans. I just numbered all the batches and sorted out the groups of cookies by pan and time after everything was done using my record of pans, times, and batch numbers.

After the first batch is done, you may adjust the times you wrote down for your next batch. For example, if the first batch was for 7 minutes but there is no brown around the edges at all, then there is probably no point in doing a 6-minute test batch. You want to find exactly how many minutes it takes for the cookies to be lightly browned around the edge, without it becoming golden or browned across the entire cookie. In the end you will ideally have one cookie that looks exactly right, and one that has no brown around the edge, and one that is more browned than the first, all timed one minute apart, for example 7, 8, and 9 minutes, or 9, 10 and 11 minutes. If you have more cookies at times beyond that range of three it doesn't matter - you need to determine exactly when the cookie first looks brown around the edges as the center one of three batches spaced one minute apart. We can't keep opening the oven door to check, because that causes the temperature of the oven to fluctuate and it's not a good result to go by. We have to cook the batches individually for exact amounts of time, leave the door shut, and take the cookies out at exactly the right time.

When you start a new batch, make sure the cookie sheet for this next batch is approximately at room temperature before you put the new slices of dough on it, and make sure you give the cookie a few minutes on the cookie sheet before you put it in the oven - ideally the same amount of time for each batch, but it starts to get pretty crazy if you measure that also, if you are overlapping batches as I did in order to save time. This took me about five hours for six pans because I kept getting distracted and didn't get batches out on time, so I had to do that batch over again. I'm definitely not doing it again tomorrow! It was really a pretty aggravating exercise for six pans and if I'd needed to equalize the on-pan wait time before baking I simply couldn't have done it in a reasonable period of time. But I had done something a little like this with three of the pans a month earlier, so at least I had a good idea where to start the timing from.

OK, so let's look at the results of my calibration experiment. Hopefully the picture shows up well for you.

The point of this experiment, again, is not to find how long it take to make the perfect butter cookie. The point is to find out exactly how different the pans are from each other. So I'm picking a "standard" appearance to look for in my cookies, and am choosing it to be the appearance of the first cookie in the bottom row. There is just a touch of browning around the edges. The time for that cookie's batch was 7 minutes.

I want to mention that for most of these batches in the picture, I have two or three cookies stacked on top of each other, so that messes the picture up just a bit because the edges of cookies below may show around the cookie on top. Sorry about that - but otherwise it's too hard for me to keep the groups together.

Look at the fourth row. That row shows perfectly what I hoped to see - the cookie on the far left is completely white. There is a shadow on the cookie beneath it that looks like it might be brown, but don't be fooled by the picture, actually the cookie is perfectly white. The next cookie is slightly brown around the edges, and matches the "standard" pretty well. The third cookie in the row is noticeably darker, and the fourth cookie is darker yet. The times for those cookies were 10, 11, 12, and 14 minutes, and the one that matches the standard was from 11 minutes. Thus, the sheet for the fourth row produces the same cookie in 11 minutes that the sixth row sheet produces in 7 minutes.

I actually ran out of test cookies, not to mention got really tired of the process, which is why the fifth and sixth rows only have cookies for two baking times instead of three. But the second cookie in row 5 appears to match the standard cookie very closely, and is also an 11-minute cookie.

In the third row, the cookies are for times 6, 7, and 8 minutes. The 6-minute cookie actually has a tinge of brown around the edge, less than the standard, while the 7-minute cookie has no brown at all. That's a backwards progression, isn't it? Well, it's not a perfectly controlled process, that's all I can say. I'd do it again if I was a total masochist. Then the 8-minute cookie matches the standard cookie just about perfectly. I would have done a 9-minute cookie to confirm 9 minutes is too dark, but again, I ran out of cookies (thank goodness, or I'd still be doing this). I'm calling the third row pan an 8-minute pan.

The cookies in the second row are not in a strict time progression. Because I forgot to note the exact starting time for one of the batches, I was a little uncertain I had timed things right, and so I wound up doing two "7-minute" batches and two 8-minute batches. Both of the 8-minute batches are darker than the standard. One of the 7-minute batches is too light, while the other matches very closely. I'm calling the second row's pan a 7-minute pan.

The first row is a little bizarre. The times there are for 8, 10, and 10.5 minutes. Odd times due to mistakes. I actually lost the 9.5-minute cookies I'd made with that pan! I swear I didn't eat them. The 8-minute cookie in the first row has no browning whatsoever. The 10-minute cookie is way too dark. The 10.5-minute cookie is actually a near-perfect match to the standard, even though in the picture it looks a little darker. In life, it looks the same. It's odd that the 10-minute cookie was so much darker than the 10.5-minute cookie. In real life, it actually is significantly darker. Notice that the pan for the first row is the very lightest pan I have. The lightness of the metal means that the pan's temperature changes more rapidly as the oven temperature changes. I noted in my "Calibrating My Oven" post that the oven temperature will slowly fluctuate maybe ten degrees either side of the average temperature. The heavier pans require more heat to rise a degree in temperature and need to give off more heat to drop a degree in temperature, simply because they contain more metal. There is only so much surface area on the pan where heat can be taken up or given up, and the amount of heat that can be exchanged in a minute's time is determined partly by that surface area. It gets kind of complicated, but the end effect is that the heavier pans stabilize the temperature more around the average temperature. If the lightweight pan is put in the oven as the temperature is rising past the average temperature, then the pan will more quickly go to the above-average temperature than a heavier pan would, and the bottoms of the cookies on it will be exposed to the above-average temperature more than they would for a heavier pan that doesn't rise in temperature so quickly. That means the cookies in that particular batch will get done faster. If the cookies are put in the oven on a light pan as the temperature is dropping below the average, it's not clear cut what will happen - the temperature of the light pan will initially rise more quickly than for a heavier pan as the pans go from room temperature to oven temperature, but if the temperature continues to fluctuate downwards after the heavy pan reaches the oven's temperature, the cookies will begin to receive less heat on the lightweight pan. Sheesh. Anyway, I would give the effects of the light weight of that pan as the hypothesis for why the 10-minute cookie got done more than the 10.5-minute cookie. Either that, or possibly the 10-minute cookies were placed on a warm cookie pan so got a head start heating (but I didn't do that), or the 10.5-minute cookies weren't out of the refrigerator long enough before putting them in the oven (possible). Oh well. It's my story and I'm sticking to it, since nobody's life or job depends on it. As for calibrating the pan, I am calling it a 10-minute pan, based in part on my foggy memory of the appearance of the 9.5-minute cookie which was pretty close to the standard. That's good enough for me, for now.

So, the results are:

1. cheap 13 x Good Cook traditional finish: 10 minutes
2. large 15 x 14 Good Cook slide-off pan: 7 minutes
3. medium 13 x 12 Good Cook slide-off pan: 8 minutes
4. steel 15 x 14 Goldtouch(TM) insulated pan: 11 minutes
5. aluminum 15 x 14 insulated pan: 11 minutes
6. medium 15.25 x 10.25 Wilton pan: 7 minutes

I can use these numbers to switch pans for a recipe. If I know a recipe works well baking at 10 minutes with pan #3, and I want to switch to pan #5, then I know pan #5 takes about 3 minutes longer than pan #3 so I add that to the 10 minutes, and bake the same cookies for 13 minutes using pan #5 that were done in 10 minutes with the other pan. That's probably way too many numbers in one sentence. Oh well - if you made it all the way to here, I bet you already get the idea anyway. :-)

There are a couple of cautions about these results, though. First, if I had filled the pans with cookies instead of just baked one to three cookies per batch, the extra cookies themselves might have helped even the temperature of the pan, so the absolute times might have been different, especially for the lightweight pan. Second, if the recipe calls for 30 minutes baking instead of 8 minutes baking, these calibrations might not be correct. I think it would be worthwhile for me to recalibrate the lightweight pan with this recipe but using full panfuls and check that 10 minutes still looks right for that pan; and if it doesn't, determine what actually is right, and also then recalibrate one of the heavier pans with a full panful and make sure that the heavyweight pans are not actually so affected by how many cookies are on them that it would change the times. I'll report that in a different post. But as for the extended time issue, I tend to think an extended baking time will reduce the effect of the pan on the actual baking time, but that's just a guess. Someone just told me that more efficient cookware actually reduces the roasting time of meat, so maybe it would still make a difference with longer baking times after all. When I come across a suitable recipe I could use to calibrate at a longer baking time, I'll find out whether my guess is right or wrong. That's where "step #2" of the calibration comes in - see the title of this post. :-) But for times up to fifteen minutes or so, I think this calibration should be pretty close.

One thing worth mentioning here - I used the lightweight pan for most of the cookies I made for last Christmas, and they seemed to come out well following the recipes, which came from several sources, which were primarily fairly old recipe books. So, it may be that a cheap light-weight cookie sheet with a traditional surface with might best yield the right results for older recipes when you use the prescribed times, and if you calibrate your pans against one of those cheap ones and adjust your baking times accordingly you'll be more likely to get the correct baking time right off the bat. I would think recently-concocted recipes from professional bakers or cooks tend to be originally made with non-stick pans, which have shorter baking times, and so if that's what you have, maybe you don't need to adjust as much with modern recipes. But that's pretty iffy - I could be wrong about that non-stick judgement, and it's hard to know if a recipe is a new one or not, even if it's recently published. Cookie recipes have a way of being passed down with few changes. Take the thought for what it's worth.

Oh, and yes, I do actually have degrees in biochemistry and microbiology, in case you haven't already guessed that I have some background in the scientific method. :-) I promise the blog will get away from this kind of thing after a bit - this is just setting up for the fun part I really want to pursue. Meanwhile, this is what I have to write about, so I hope somebody gets something out of it.