Sous Vide

Credit: Useful Info gathered from eGullet Forum

Posted 20 August 2004 - 10:29 PM
FG and I have discussed making a tutorial on eGullet but so far various things have gotten in the way. Here is a very simplified story in the mean time.

I use bags that are designed to be boiled, and I use a commerical vacuum packer machine (ARY Vacmaster). I have the SVP-15, but the SVP-10 has less maintanance. These are expensive, commercial machines that are $1200 or more, but they work great. There are many other brands at various price ranges. The cheapest are FoodSaver and Rival.

Here are the specific bags

The FoodSaver web site says that their bags can be boiled, which is the key thing.

I use pint, quart and gallon bags. Pint and quart for single servings, gallon sized for large things.

By the way, the bags are very nice for use in storage - for example when I freeze chicken stock I put it in a bag, seal it under vacuum, and freeze it. This is not sous vide, but it is very convenient.

The other thing you need for Sous Vide is a way to cook the bags. Typically people use one of two things - a computer controlled convection steam oven such as the Rational CombiTherm. These are fantastic (and not just for SV), but expensive and large and not at all for home use unless you have a completely pro kitchen at home.

The alternative is to use laboratory water baths. Personally, I have and use both Rational ovens and lab water baths depending on what I am cooking.

The easiest for home use would be the laboratory water baths - you can get them on Ebay for $100 to $500, and they are reasonably small (size of a slow cooker, or a bit bigger). Some people on the list have suggested trying to use slow cookers, or crock pots, but they generally do not go low enough in temperature, and are not accurate enough. You can't really cook this stuff on top of the stove manually - you need to control the temperature to within 1 degree, and you just can't do that normally.

The main point of Sous Vide cooking is to cook very gently without oxygen. There are several reasons different motivations for this:

- Can cook at lower temperatures than normal without the same food safety concerns, because the lack of oxygen supresses growth of bacteria.

- Can cook for long periods of time. In part this is because the cooking method (water bath or steam oven) is very accurate. In part it is because you can use low temperatures and food saftey as above.

- The food is sealed in the bag, so it does not get diluted through contact with a poaching liquid.

- Can keep the food longer - becaues the vacuum packed bag is almost like a can used for canning. This was the original reason for Sous Vide - cooking in a central kitchen, for distirbution as catering or to remote locations where it would just be reheated. This is sometimes called cook and hold.

- The food is convenient to store and reheat in a sealed bag.

- Oxygen discolors some food (artichokes, fennel bulbs).

Personally, I do not do the cook and hold approach - I cook and serve immediately. SV gives you more control, and more delicate cooking, than you could do otherwise.

Most cooking is about using a high temperature medium to cook food for a fairly short period of time. For example, you might roast a chicken in an oven where the air is 350F, and roast until the core temperature of the flesh is at 160F (at least that is the standard poultry temp). You have to time things carefully so that you stop cooking in time - otherwise you overcook.

I do chicken breasts Sous Vide in a water bath at 141F for 30 min to an hour, until the core temperature is 141F. This is typical of SV - you cook with the external temperature at the same temperature as you want the core. That way you can leave it in as long as you want (within reason) and not worry about overcooking. Some SV authorities use sligthly higher external temp than core (by say 5 -10 degrees). This is faster, but then they have to worry about overcooking.

Normally, you would worry about cooking anythnig at less than 140F because that is the general food safety limit. You would not even roast a chicken at 140F beacuse the internal parts of the chicken would take a long time to come up to temperature, and would stay in the "danger zone" below 120F for a long time. SV removes oxygen so it is much more tolerant of low temps. Also, you generally have a smaller piece of food in a single serving bag, so that helps the time issue.

Note that US food safety guidelines are generally very conservative. Note also that health inspectors in the US have little or no experience with SV and tend to mistrust it becaues it seems to be cooking within the danger zone. 20 years experience in Europe says that it is OK (with proper precautions), but you may have issues with US based health inspectors. Some people on the list have reported problems along these lines. On the other hand, most of the really top chefs in the country are using SV. At any rate do this all at your own risk.

Risks are greater if you are are doing SV to store and serve later with reheating (i.e. cook and hold) rather than immediate service.

Timing is done by experience, or the accurate way is to get a very good laboratory style digital thermometer to check the core temperature. This can even be done through a vacuum bag.

Fish is done SV to 113F for most fish, but as low as 104F for salmon. Meat served rare to medium rare is done between 120F and 130F. Chicken, duck and many other things to 140F.

Most vegetables are done at just below boiling - 190F to 200F.

Most SV cooking is like conventional cooking in that you bring the food to the desired core temperature, then serve.

however, there are some products that need to sit a long time at the proper temperature to effect a physical or chemical change in the food. For example, you can SV cook beef short ribs at 150F for 36 hours. The texture is much better than normal braising, but because the temperature is much lower than normal braising it takes longer to cook.

SV is a good candidate for most foods that you would ordinarily poach, steam or braise. Because the temp is low and the food is in a bag, it will not brown. So you either need to serve it that way (as you would a poached item), or you quickly glaze it, or brown it in hot saute pan or under a broiler.

Hope that helps - review the other posts in this thread for more ideas...

==

Posted 04 August 2005 - 10:58 PM
I started this thread nearly a year and a half ago asking about information on Sous Vide, particularly recipes which would give me cooking times and temperatures. Ironically, I am now going to largely answer my own question.

In the intervening period of time I have bought and read every book on SV in English, French and Spanish, talked to chefs, and done countless experiments of my own. Here are my conclusions on the basic issues for time and temperature.

There are several factors that go into determining the cooking temperature and time:

1. Is SV for immediate service, or to cook and hold? SV got its first big push in cooking meals on an industrial scale for later reheating – for example for airline catering, or servicing low to mid range restaurants from a central large scale kitchen. That is the cook and hold method – the food is stored up to three weeks after the SV cooking. More recently SV has been used by high end chefs for immediate service – i.e. just like roasting, frying or other techniques.

2. What is the product? Obviously, fish and meat are cooked to different temperatures, and even within something like beef there is a wide range of temperatures depending on the cut and on personal preference. The main split is pretty simple however. One set of cooking processes occur rapidly, so the goal is to bring the food to a temperature. Another set of cooking processes – such as tenderizing tough meat – require a lot of time. Collagen breaks down into gelatin at or above 130F / 54.4C, but the rate is very slow until you get a higher temperature.

3. Do you need to sterilize / pasteurize? This is a food safety question, but it is not as simple as you make think because most of what chefs are told about food safety is actually wrong. Sterilization means heating the food to a high enough temperature, and leaving it there for a long enough period of time that it kills microorganisms that could be harmful. Sterilization is required by law for some food products but there is no legal requirement for sterilization of many foods. Otherwise we couldn’t have sushi, carpaccio, or even a rare steak. Cook and hold SV requires sterilization no matter what the product is, but that does not necessarily mean you have to cook at high temperatures (see below).

4. Do you want to cook at the final temperature, or above? Conventional cooking is almost always about using a heat source (oven, steam, poaching water) that is much higher temperature than the food. As a result, timing is important. In SV you have the option to cook at, or only slightly above the final temperature. The cooking is very slow, but as a result timing is not critical you can hold the product for a long time without risk of overcooking. Or you can cook at a higher temperature. This has the benefit that the cooking time is shorter (see below) but you can’t help but overcook part of the product, and you must be careful about timing.

The food safety issue is particularly interesting. Most people think this means cooking to 140F / 60C – that is not correct, it is about both time and temperature. 140F/60C is NOT ENOUGH if you do it for a short period of time – typically you need to be at that temperature for at least 12 minutes. Conversely, red meats can be cooked at 130F / 54.4C and be just as sterilized if they are held for 112 min. Poultry is frequently incinerated in a well meaning attempt to meet food safety guidelines, but in fact the US FDA says it is perfectly acceptable to sterilize lean turkey or chicken at 136F/ 57.8C for 64 minutes. These times come from the following FDA document http://www.fsis.usda.gov/OPPDE/rdad/FSISNo...ltry_Tables.pdf

Many people who have the phobia that below 140F/60C is unsafe do not realize this and will say things like “it is unsafe to do long time duration cooking at 130F”. In actual fact it is SAFER than bringing food to 140F, unless you keep it there for 12 minutes. As long as you follow

In addition, it is not necessary to cook all foods to the sterilization limit – the majority of our food is not cooked to that limit even in conventional cuisine. So, most things that are served medium rare or below, including fish, beef, lamb, duck breasts and so forth are not cooked to sterilization normally speaking, and it is not clear why we would have to do so with SV (for immediate service).

Here is what I do.

I almost always cook for immediate service, not cook and hold. Nothing against it, but that isn’t what I do.

I use laboratory water baths for most of my SV cooking, or a Rational combi-oven in steam mode if there is a large quantity.

I almost always use a cooking temperature that is just marginally above the final temperature. This makes cooking times longer, but it gives a better, more uniform result. If necessary I cut the product thinner to achieve reasonable times. Note that this approach is very different than many SV practitioners, such as Joan Roca. His book recommends cooking most food at a temperature substantially above the final internal temperature, although this is not consistent - for some things he recommends cooking at the final temperature.

I cook most fish (including prawns) to 45C / 113F. That is below the sterilization limit. That is perfectly OK within food safety guidelines so long as the cooking time is not too excessive (i.e. so long as it is less that a couple hours). I would NOT cook thick pieces of fish this way because the cooking time would be so long that some spoilage could result.

Beef and other red meat I cook to 130F / 54.4C. This is what I would term “medium rare” but there is a lot of disagreement on mapping subjective terms like that to temperatures. Depending on the situation I might cook it to sterilization, or I might not.

Tough red meat – such as flat iron steak, short ribs etc I will cook at either 130F/54.4C or 136F/58.8C, from 24 to 72 hours depending on the cut and how tough it is. This gives time for the collagen to break down and make a tender result.

Tender cuts of pork, such as tenderloin I generally cook to 140F/60C. Trichinosis is killed at 137F, and anyway has been eliminated from the food supply in the US and many other places. I would cook wild meat a bit more, and/or pre-freeze it.

Chicken and turkey also gets cooked to 140F, and I make sure that I follow the sterilization times.

Duck breast goes to 130F/54.4C. Duck confit is 180F/82.2C for 8-12 hours – pork or lamb confit is the same.

OK, so given all that the only remaining question is how long to cook a given piece of food. The answer is that it depends on one thing – the thickness of what you are cooking. Heat slowly diffuses through a piece of meat or other food product. The time it takes depends on the thickness and is very nonlinear – i.e. doubling the thickness more than doubles the cooking time.

In a flat slab, such as a steak or cutlet, the relevant thing is the thickness. In a more irregular shaped thing, such a chicken breast or a whole chicken, the cooking time is going to be constrained by the thickest part.

This is true for any kind of cooking actually. Most of what we are taught about cooking is actually wrong. For example, anytime somebody tells you it is “10 minutes per pound” they are saying something that cannot possibly be accurate, because this would imply that cooking time is proportional to the weight. If you take something like a whole bird and scale it up you will find that cooking time is actually proportional to something like (weight)^(2/3) – weight to the 2/3 power. This is because increasing the weight scales up the thickness by the cube root. Since most people are not accustomed to taking things to fractional powers, people substitute a linear relationship. That might work out OK in practice over a small range, but it can’t be accurate over a large range. For example, if you double the weight, the linear relationship would tell you that you double the cooking time. The 2/3 power would tell you to increase it by 59% - that is a pretty big difference.

As a general rule of thumb heat diffusion times go as the square of the thickness increase. So, doubling the thickness results in FOUR TIMES the cooking time. That is a rough general rule of thumb, which is not perfect but it illustrates the nonlinearity of the system. I don’t know why this isn’t taught to chefs more often because it is a fairly easy thing to grasp and use.

A more accurate way is to compile a table of times and thicknesses. I have done lots of experiments, and I have also written some software to solve the partial differential equations governing heat flow. The theoretical results agree with the experiments to extraordinary precision – a tiny fraction of a degree, and/or a tiny fraction of a second. The result of all of this is that I have produced the following set of tables.

Instead of using a table you can use a thermometer probe, inserted though closed cell foam tape. However, it is still useful to have the cooking times so you can plan ahead and know roughly how long things will take.

In each case I list different options of thickness, and different cooking temperatures. I use the low temperatures, but I have included the result for using higher temperatures, for example as recommended by Joan Roca in his book. I have also listed the time difference required to reach one degree less, or one degree more – i.e. it is the time range between undercooking or overcooking by 1 degree C (1.8 degrees F). This gives you the margin of error that you have in the timing. If the cooking temperature is the same as the final temperature, you can’t overcook it, so that is infinite.

Note that the thermal diffusivity of various forms of animal flesh (chicken, meat, fish) is very similar so within the accuracy displayed here, the product type does not affect the cooking times.

I will start with the table for a final core temperature of 54.4C / 130F. I’ll put other tables in other posts.

The first column is the cooking temperature (water bath or steam oven) in degrees C, the second is degrees F. The next column is the thickness of the slab in mm, followed by the thickness in inches. Then comes the cooking time, and the resting time. The resting time as I use it here means the time until the core temperature stops rising and starts to fall. The next two columns are the core temperature at the time you stop cooking, in C and F. Even if you use a digital thermometer probe, you can’t just wait until the food is at the final temperature, because if you do it will overshoot. The overshoot is minimal or non-existent if the bath temperature is close to the final temperature. However, the higher you go, the more the risk of overshoot.

Finally I have two columns showing the cooking time to go one degree C lower, and one degree C higher. This shows the amount of time latitude you have. In some cases it is quite surprising – a multi-hour cook time might only have 5 minutes plus or minus leeway.

It turns out that it is easier to get the HTML formatting correct if I put the table in a post by itself, so it comes next.

This post has been edited by nathanm: 05 August 2005 - 09:24 AM
Nathan

==

Here is a little bit more about the big tables for cooking times and temperatures.

As explained in a post above, I generally cook at the lowest temperature setting, or perhaps just one above. So, I will cook at 131F / 55C for a 54.4C /130F final core temp. Joan Roca recommends 65C/149F. I don't know why he prefers the higher temperature - it speeds things up, but not enough that it is worth the various negatives.

Consider a 1 inch / 25mm thick piece of meat cooked to 54.4C/131F. At 55C it takes 41 minutes. At 65C it takes 17 minutes. Unfortunately if you really want the meat to be within +- 1 degrees C (or 1.8 F) then you must time this to plus or minus 50 seconds otherwise you overcook it. This is not the end of the world, but it is easy to mess up this way. And, if your meat is not exactly 25mm, well that makes a variation too. In the 55C case, you just leave it in a little longer and it can't overcook. Not so with the higher temperature bath.

If you want to sterilize it for food safety reasons (which is generally not required for beef steak, at least in the US) then you must use a second bath at 55C becaues it must sit there for an additional 112 minutes (see FDA chart I linked to in a post above) after the core reaches 54.4C/130F. You can't do that with your 65C bath. So, why not cook it at 55C to start with? I will probably write to Roca and ask.

Note that the cooking times depends very strongly on temperature and on thickness. Moving the temperature down 10C (65C to 55C) increases the cooking time by almost 3 fold.

Double the thickness to 50 mm and the 55C time goes to 2.5 hours, 65C time goes to just over an hour. That is a BIG change.

Note that the exact times here are for SV cooking, but the same principle holds for ANY kind of cooking - increasing the thickness, or the temperature differential, radically changes cooking times.

In SV we have an option that does not exist in most other kinds of cooking - we can cut the food to portion size prior to cooking. It is easy to do small vacuum bags that have an individual portion in them. In many approaches to meat cookery thick pieces are used primarily to prevent the center from being dried out or overcooked. That is not an issue with SV.

One big lesson from the chart is that cutting the food thinner radically reduces cooking time. So, rather than increase heat, I tend to cut the food size.

The only reason not to do thin pieces is if the post-SV cooking would be problematic. Many people like to brown or crisp the outside after SV, with a salamander, blow torch, hot pan or other method. A very thin piece of meat is easy to ruin that way. There are also aesthetic and other reasons to have big chunks of food. But the lesson from the tables is that thickness comes a big price.

This post has been edited by nathanm: 05 August 2005 - 08:32 PM

==

Posted 11 July 2008 - 12:39 AM
For what it's worth. I have done briskets at 135F and 147F (several times at each temperature) and would say that to my taste, 135F was vastly preferable. The meat had better texture and was moister at 135 (for 40 to 48 hours) than at 147.

No matter what you do, a lot of moisture will leave the meat. Also, you need a well-trimmed brisket (especially at 135F) that also has reasonable marbling. The flat of some briskets does not have enough marbling and will yield something that turns out "dry" no matter what temp.

At 135F for 48 hrs, you end up with meat that is fork tender, deliciously pink and has its full visual integrity (it doesn't look stringy or fibrous at all).

Posted Image


The picture doesn't do the meat justice (it was a lot pinker than it looks in the pic).

---

I think that the difference is easy to explain. At 135, you have medium rare brisket, while at 147, it is medium to medium well done. With the connective tissue reduced to gelatin in both cases, the tenderness is a function of internal temperature. If you think about it, we are used to brisket being well done.
---