nathanm

Canola oil is netural and can be used very successfully. This is especially true if you have some duck fat because the flavor will come from that and the canola is just a way to extend it. Sous vide is another method - this is discussed up thread - that requires only a small amount of fat, but you need to have a food saver.

Polyscience makes great water baths. They were the first water bath company to take the sous vide market seriously - most of the other vendors have focused exclusively on the laboratory market. However, that said, any of the major brands of laboratory water bath are built to very high standards. Pharaceutical companies and other lab users are very demanding. That is in part why the price of water baths are so high - they have very demanding, price insensitive, small volume buyers. Over time cooking oriented water baths will emerge as a distinct product line from the laboratory models. The price points for the lab versions are too high to sell very many to for consumer kitchens. Meanwhile the wattage is (in general) too low for commercial kitchens. So over time we will see water baths designed for kitchen - both consumer and professional.

Lack of oxygen during cooking is the main reason. The outside of sous vide cooked meat typically looks grey, or even greenish. Not very appetizing, but that's how it is. As others have mentioned, searing is the answer.

It makes a pretty significant difference. I will be measuring this for my cookbook -I don't have good quantitative results yet. Previous results are ~10% by weight effect in juice retention. Many steak houses use jaccard. They have larger versions of the machine that have higher throughput. Some butchers do it routinely. You have to look VERY closely to tell.

There are refridgerated water baths and those are great ways to cool things. Ice water is a simpler way to do it - not quite as good circulation, but cheap and effective.

True for both asparagus and broccoli - if not shocked in ice water immediately after cooking they will not only soften but will lose their vibrant color and visual appeal. Both texture and visual appeal are important on a plate. I found the scientific discussion fascinating and learned a great deal from it but we must not forget the ultimate purpose. ←

You are correct!

If your goal is making the meat more tender this is exactly right. Tenderizing the meat is going to be very slow below these temperatures. But, this is really a personal preference. If you like your steak rare, then you need to cook it less than 55C/131F. if you have tender meat (like fillet mignon), and/or if you are jaccarding, then you probably don't care if the cooking is not doing any tenderization.

Just to make this very clear, there are three approaches: 1. Sear the outside to 145F/63C, then cook SV at lower temp (any temp you want). 2. Cook SV at very low temp - 38C/102F or higher, but for less than 4 hours total from refridgerator to plate. 3. Cook SV so that core temperature reaches 130F/54.4C and then stays there for 112 minutes (or high temp for lower time - see fda tables reproduced in various posts or Douglas Baldwin's site). Cooking at higher temp than 145F/63C means that the time requiremnt is so short that by the time you reach temp you are done. #1 is FDA approved for "intact" beef steak. However, as a practical matter it ought to be fine for any intact muscle food, because the inside of muscle is sterile until and unless you penetrate it. #2 is within FDA rules for any food so long as you warn on the menu. Carapaccio and sushi are not heated at all, nor is salad etc. By heating this food for less than 4 hours you are serving raw food warm or "lightly cooked" as far as FDA is concerned. #3 is cooking to "sterilization" or "pasteurization" (many of these terms are used loosely and interchangably. It is the FDA guideline for cook-chill sous vide. Meanwhile #1 and #2 are not approved for cook chill - they are only for immediate service and consumption. This is a very interesting question. If you use approach #1 - sear first, then jaccard, then cook SV, it ought to technically meet FDA rules, and ought to be safe so long as the cooking time is reasonable (4 hours or less). By searing first you are sterilizing the outside, so when the Jaccard blades pentrate into the meat you are probably not carrying anything with them. If you use approach #2 above - just heat for a short time - then again it ought to be OK, for the same reason that it is OK if you chop your salad and serve it raw. If you use approach #3 then it ought to be safe too, because you are heat sterilizing the interior. What I would NOT do is Jaccard, then marinate or otherwise store for a while then use #1 or #2. If you are going to marinate, brine or otherwise introduce things into the interior of the meat (including by injecting marinade) then you are best off using approach #3.

So here is info on pressure and vacuum. Pressure is measured in many units. Imperial units are psi - pound per square inch. The metric unit is Pascal (Pa) which is 1 Newton per square meter. 1 Newton is the force equivalent of 102.041 grams. Bar is another pressure unit, which is designed so that 1 Bar is approximately equal to atmospheric pressure. 1 Millibar = 1/1000 bar - abbreviated mbar Torr is another vacuum pressure unit. 1 Torr = 51.715 psi = 133.322 Pa The point of a Torr is that 1 Torr is approximately = 1 mm of Mercury, written mmHg The reason for this is that barometers, and many other old fashioned pressure gages would use mercury in a tube to measure pressure. 1 mm height difference corresponds to a certain pressure. In Imperial units, you have inches of Mercury, written inHg Ok that is units. The pressure due to the atmosphere depends on height above sealevel and the weather. The standardized value is Atmospheric pressure (atm) = 14.7 lbs / square inch (psi) = 101325 Pascal (Pa) = 760 Torr = 760 mmHg = 29.92 inHg Technically a 1 Bar = 100,000 Pa, so 1 Bar is slightly different than actual atmospheric pressure. ---------------------- OK, so now on to your watermelon question. 200 mbar = 20% of atmospheric pressure. So the net pressure on the outside of your bag during compression is about 12 lbs per square inch (80% of 15). Of course the bag comes into equilibrium quickly, so when you put a bag with interior vacuum of 200 mbar into the air at 1000 mbar, it crushes down until the force is equalized at atmospheric pressure. But during the compression/suction cycle there is 12 psi net force on the outside of the bag. A perfect vacuum would get you an additional 3 lbs/square inch. So, the question of whether the extra vacuum is worth it depends on whether you need the extra 3 lbs/square inch of pressure. ----------------------- The compression is not really at at a cellular level. The cells do not get smaller. It is actually about compressing space between cells - watermelon is a foam-like structure . ----------------------- When you seal food in a vacuum bag for storage the difference between 200 mbar and 20 mbar is that you are removing more oxygen from the bag. At 20 mbar there is a factor of 10 less oxygen then at 200 mbar. So if you buy supermarket cheese, it comes vacuum sealed to the much lower vacuum, because oxygen makes the cheese rancid and every factor of 10 counts. This difference is not important for most sous vide cooking however.

No plans to make equipment - we are just working on a book! But we have tested, and will describe, a bunch of different kinds of sous vide equipment.

This is a very interesting point. It is very valid for bacterial or viral pathogens. People with compromised immune systems need an extra level of caution. This includes people who are: - HIV positive - In the middle of chemotherapy - Transplant recipients on transplant anti-rejection drugs - Infants - Other conditions that compromise the immune system In addition to virsuses and bacteria, it may be valid for some microscopic parasites. However, the immune system is basically irrelevant for anisakid nemotodes (or for dyphllobothrium tapeworms, for that matter). These are large scale parasites that are in your gut and thus not subject to the immune system at all. Watch the fascinating (and disgusting) video in slkinsley's post. The anisakid is about 1 mm wide and probably 25mm long. I recently found one in halibut bought at an organic supermarket chain that was even longer. It lives in the gut, which puts it beyond the reach of the immune system. It is basically like swallowing a bone. The worm acts as a physical irritant because it tries to poke its way through your stomach wall. It can't do that successfully in humans (its built for seal guts), but it hurts when it pokes you. The immune system isn't going to stop that, anymore than the immune system could stop a swallowed bone from poking your gut. I am sure that somebody with a compromised immune system would find these worms annoying. And if you are on death's door anyway, you don't need one extra problem. However, they shouldn't be any more medically dangerous for them than other people. While we are on the topic, the immune system state is also pretty irrelevant for food poisoning - i.e. eating food that has bacterial toxins in it. If you eat food that has botulism toxin, or toxin from Bacillis cereus, or Staphlococcus aureus, then you will get sick regardless of your immune system, because the problem is a chemical toxin. Basically you NEVER want to consume botulism toxin! This is unlikely to be a surprise...

There is an interesting thread (which originated as a discussion in this thread) on the safety aspects of cooking salmon mi-cuit. Since most people who cook salmon or other fish sous vide use low temperatures, I think it is an important topic for sous vide. It can be found here.

I do not model temperture dependent diffusivity - which combines both conduction effects and specific heat effects. People have done the experiments and at these temperatures it is not a significant effect. It does get to be very significant when you either totally dessicate the food, or when you freeze it. Frozen food has about 4X the thermal conductivity as non-frozen. Specific heat makes a huge change upon freezing due to latent heat of fusion.

Have you tried it? Frozen fish gets a bad name for many reasons. It is true that fast freezing is better than slow freezing. However, it is very unclear to me that fish frozen in a home freezer, held for 24 hours to 4 days, then thawed and cooked sous vide would actually suffer much in quality. I will be doing some tests for my cookbook project but this is not done yet. So, try it and find out. The fair comparison would be buy some fish, divide in half, vacuum pack each half. Keep one half in the fridge, freeze the other half for 24 hours, then defrost the frozen one by putting in the fridge. Then cook both and see if you can tell the difference.

I should add that Diphillobothrium tapeworms can infect humans through salmon, as discussed in the Gourmet article but this is MUCH more rare than anisakids. Worrying about Diphillobothrium is pretty futile, it is extremely rare in comparison with anksakids. Freezing to kill anisakids will almost certainly kill Diphillobothrium, but the condition is so rare that the FDA does not even know this for sure - here is the FDA report on Diphyllobothrium.

I agree with Jack about risk versus thinking that it is certain. In addition there are some other factors. First, "tapeworms" is the wrong name - at least for pratical concerns. The primary risk with parasites in fish are Anisakid nematodes. These occur in wild fish caught near shore - rock fish, sea bass, salmon. These are quite different than the tapeworms that infest humans. Those are a totally different creature (different genus), and you can't get them from fish - you get them from contamination with human feces. There are very very rare tapeworms that can come from fish - I will dicuss that in the next post. The Anisakids normally parasitize seals that eat fish. They are unable to parasitive humans, - when a human ingests Anisakid nematodes you get a bad stomache ache. Blue water fish, such as tuna, do not get anisakids. Farmed salmon, or other farmed fish do NOT carry anisakids. In order to get infested with anisakids the fish have to eat anisakid contaminated seal feces. Anisakids are pretty common. For my cookbook project we wanted to photograph some so we went to the local branch of a large organic supermarket chain and took a careful look at the halibut fillets on sale. We spotted one with anisakids and bought it. Got some great photos. Despite the fact that they are common, there are very few cases of anisakid related illness per year in the US - about 10 per year. Japan has many more, but still only about 2000 cases a year. Mostly this is from home prepared sushi. But even 2000 cases a year is quite rare considering that the population is 127 million people, many of whom eat raw fish on a daily basis. Here is a popular article. Here is the FDA report on Anisakids, and another FDA report on smoked fish. Anisakids are killed two ways - by heat, and by freezing. There are very few good studies on heat - one study says 60C/140F for 1 minute, but does not give other time and temperature combinations. I do not trust information given with just one data point like that - I am looking for better data. The other approach is freezing. A standard freezer is good enough - you do NOT need a blast freezer. The US FDA recommends -4F / - 20C for 4 days. Or -31F/-35C for 15 hours. EU regulation is -4F / - 20C for 24 hours. Any household freezer can reach -4F/-20C. I eat salmon mi-cuit even without freezing it. The risk is very low, but it is possible. However, if I was concerned I would simply freeze the fish overnight.

There are two ways to answer - the official FDA answer, and the best available scientific information. They are different. For an "intact beef steak" the FDA requires that the outside be seared to 145F. There is NO REQURIEMENT on internal temperature whatsoever, and no requirement on the cooking temperture. There is also a general FDA rule that food should not spend more than 4 hours between 40F and 130F. So, if you want to go by the letter of the ruling, you can cook a beef steak at any temperature you want so long as the exterior is seared, but you can't cook it for more than 4 hours if the temp is below 130F/54.4C. For a beef roast, the minimum FDA temperature is 130F/54.4C. You might ask "how is a roast different from a steak?". Most steaks are cut from muscles that could be roasts. The answer is that there is NO definition of the terms in the FDA Food Code. Steaks are given a special case because of tradition support for rare steaks and politics of the beef industry. 130F is also the lowest temperature the FDA recommends for effective food safe sterilization/pastereurization - and they give time-tempertaure tables fors this. So, the FDA answer is that it is definitely OK to cook at 130F, if you cook for at least 90 minutes. For a beef steak there is no lowest temperature, but you can only cook below 130F for 4 hours. Scientifically speaking the issue is quite different. Intact muscle is sterile inside. If the outside is well seared before cooking in a clean, sealed sous vide bag, and very good hygiene is observed, then it is almost certainly safe to cook at lower than 130F. How much lower is an interesting question. There are numerous scientific papers documenting the killing of pathogens at various temperatures. The lowest temperatures commonly reported as being effective for killing Salmonella, Listeria, Camplyobacter and other common food pathogens is 120F/48.89C. A few papers suggest even lower temperatures for a few pathogens (E. coli, for example). The key point is that 120F is found in dozens of papers. So, it is likely the case that 120F/49C is a safe temperature as long as you cook for long enough. The next question is how long? Many people have the wrong intuition here - they say "how long can I cook it at that low temp". The real question is the OPPOSITE - what is the SHORTEST time that I can cook it and be safe? The longer you cook something the more likely you kill the pathogens. The simplest approach to this is to extend the FDA food safety table. I won't repeat the whole table, but the last two temps are: 133F .... 56C .... 1.0 hours 131F .... 55C .... 1.5 hours (this is the lowest in the FDA table) extrapolating mathematically from the same curve, we get 129F .... 54C .... 2.33 hours 127F .... 53C .... 3.5 hours 126F .... 52C .... 5.3 hours 124F .... 51C .... 8.0 hours 122F .... 50C .... 12.1 hours 120F .... 49C .... 18.4 hours These numbers are rounded a bit for convienence. Note that these figures are NOT a guarantee on my part that your food will be safe. That depends on many factors and you must be responsible for your own actions here. Below 130F is definitely below FDA recommendations. However, my personal (and extensive) survey of the scientific literature suggests that beef brought to the temperatures above, and held there for at least the time in the table above, will likely render it just as safe as the FDA time-temperature tables. But again, this is not a guarantee! Normally I like to sear meat after sous vide cooking. However when cooking at lower temperatures than 130F, I would sear first, as an added precaution. Finally, note one more thing - the main reason to slow cook meat at low temperature is to keep the meat rare to medium rare, while at the same time making it tender through denaturing collagen into gelatin. The lower the temp, the more rare it will be - however at some point the temperature will get so low that the collagen will be very slow at denaturing into gelatin. If you go too low it might be self defeating. Unfortunately it is not known what the minimum temperature is for collagen denaturing, so we don't know exactly how low is too low, you just have to experiment.

One point that I should clarify in the posts above. I used two WATER BATHS - one at 70C and the other at 1C, 10C, 20C, 30C. This is NOT the same as taking the food out and leaving it in air in the kitchen! The reason is that heat transfer from air to a solid, via natural convection, is quite slow. It is about 100X less effective than stirred water. Natural convection in air (i.e. not forced with a fan) will transfer about 20 watts per square meter, per degree C of temperature difference. A stirred water bath will transfer about 2000 watts per square meter per degree C of temperature difference. So, the cooling curves for leaving the steak or roast out in the kitchen air at 20C are NOT the same as for putting into a water bath at 20C. The air will be MUCH slower.

There are several issues to consider here: - If your goal is to cool the food down to refriderator temperature, then doing it quicker is going to help food quality (a little) and help food safety (a little). - Ice water is pretty cheap and available in the kitchen so it is not a big deal. - You avoid overheating other things in your refriderator. Most fridges are designed to maintain temperature - putting hot food into the fridge strains the cooling capacity, and will warm up other food nearby. - Technically speaking, the FDA rules allow 4 hours of time in the so called "danger zone" between cooking temperature and 5C. You are correct that if you extrapolate the acutal amount of time depends on temperature, and you can hold much longer at 20C than at 44C. However, that is "just" scientific reality. The actual health inpspector is taught 4 hour maximum holding time. In the case of the 125mm cube roast that I posted above it does not actually make it within 4 hour time limit - but is close enough.

Here is a graph for a cube of meat 125mm on a side. This is like a roast. I know that most roasts are not perfect cubes, but this shows a case that is much thicker than the 30mm thick steak in previous. Once again the meat starts at 5C and is in a bath at T1 = 70C. It takes much longer to reach temperature - about 25,000 seconds or 7 hours to reach 70C. At 7500 seconds (2 hours and 5 minutes) we switch the roasts from the first bath to second bath at T2 = 1C, 10C, 20C, 30C. Here is detail near the peak. The core temperature at the time of the bath switch is 45C. The 30C bath peak is 53.16C, the 1C bath peak is 52.19, for a spread of just under 1C. This is still insignificant for any cooking application. The total overshoot from the time the meat is taken out of the first bath to the peak is quite significant - it is 8.26C for the 30C bath. If this was a beef roast, it would be warm but raw at (45C, or 113F) with no color change, but would be cooked to rare at 53C / 128F. The peak temperature happens at about 33 minutes after switching to the 1C bath and about 41 minutes after switching to the 30C bath, for a spread of about 7.6 minutes. So, this works for 30mm steaks. It works for 125mm roasts. I have checked this for thicknesses from 5mm to 150mm.

Yes there is a reason - you chill the food quicker. Chefs are typically taught that you "shock" the food to "stop the cooking". This is wrong - the degree of cooking is generally determined by the peak temperature and that is basically unaffected by a cold water shock. However, the food does chill faster in cold water. If you are doing cook-chill sous vide where you want to store the food cold then this helps with food safety. You want to chill as fast as you can in that case. If you look at the graphs in the post above you can see that the 1C cold bath gets colder faster than the 10C. In the case of the 30mm steak they each take about an hour to chill completely to the bath temperature.

I am still pressed for time so will be brief, but here are two graphs of a simulation that shows the peak phenomenon. The case shows here is that you take several steaks steak which are 30 mm thick. They start at a uniform 5C. First, the steaks go into a bath at T1 = 70C. At exactly 1100 seconds, the steaks are removed and put in a second bath at T2 = 1C, 10C, 20C, 30C. One is left in the 70C bath. The first graph shows what happens at the center of the steak. The blue line shows the steak that stays in the bath and is not removed. It finally equilibrates to 70C at about 4000 seconds (a bit over an hour). Each of the other lines shows what happens to steaks in the various second baths. They are also approaching equilbrium at the second bath temp, after about an hour. This graph shows a close up of what goes on in the transition between the two baths. At 1100 seconds we take the steak out of the first bath at T1 and put them in the second baths at T2. At that point the the core temperature of each steak is 50.08C. It takes a while for the center to even notice that anything has changed. Each of the lines representing a different bath has a SLIGTHLY different peak temperature (marked with dots the same color as the lines), at a slightly different times. However, for all practical purposes they are the same. The 30C bath steak peaks at 1246 seconds (146 seconds after moving to the second bath). The 1C bath peaks at 1216 seconds - or exactly 30 seconds sooner than the 30C bath, or about 2.5% of the total time. The others baths are in between the 30C and 1C cases. The peak temperatures are very close - 52.3 for the 1C second bath, and 52.67 for the 30C second bath, for a difference of only 0.37C, which is of zero practical importance. A difference of 29C between the baths produces a 0.37C difference in peak temperature. The difference in the peak is a factor of 78X smaller than the difference in the second bath temperatures. Somebody might object that in posts above I say that the peak would be the same, while these are different. First there is no practical difference for chefs. 0.37C is less than measurement error for all but the most expensive calibrated thermometers. It makes no difference you can taste in food. Second, the tiny difference that is shown here is primarily due to a mathematical approximation that I am using. I am assuming that I can switch the baths infinitely quickly, and bring the surface of the steaks to the new bath temperature very quickly. This is nearly true, but of course not quite. This approximation creates some unrealistically fast cooling right at the surface of the steaks, which contributes the 0.37C difference in peak temperature. Later on when I have more time I will do a better simulation that will account for this more realistically, but I don't have time to do that now.

Of course, but it takes time for the temperature difference to diffuse to the center. There is already a gradient of temperature difference in layers all the way from the surface to the core. Those interior layers will equilibrate the core before temperature stealing from the surface has time to get to the core.

Sorry, but this is the way it works. Your physical reasoning is wrong. You are certainly correct that heat and temperature are not the same thing. I clearly said in my earlier post that I was speaking loosely. You are also correct that this only applies to non-equilibrium cooking, where you use a bath temperature that is hotter than you want to achieve for the final. I totally agree that this is a bad idea that misses much of the point of SV. As to the physics I don't have time right now to give the full treatment, but trust me it really is the case. You have written a heat equation simulator, as I have - try it out and it will tell you exactly what I am saying (i.e. make the boundary conditions switch from T1 to T2 at time t, and see what happens for various T2 in relation to T1.)