The total weight of the down fill in an item is the most often used method for determining the warmth rating of that item. However, despite it's widespread use, total fill weight is a number that is subject to so many variables that it is far too unreliable to be used accurately to determine warmth. The total fill of an item is spread across it's total area. Problem is, total area can vary quite a bit while still being classified as similar items. Here are a couple examples.
I could build two jackets. Fill each chamber with the exact same calculations. Both with the same labeled size, same shoulder/chest width, same arm length, same torso length.....for all intents and purposes, similar items. However, if one jacket has a tapered torso and the other has a straight cut, the area of the item could change by 20%, meaning the total fill weight could change by 20%. The jacket with the straight cut has 20% more fill weight in it, and it is considered a warmer jacket when looking at fill weight. But it is not warmer. Pick any given section of the two jackets and they are the exact same R-value. In real life, the straight cut jacket might actually be colder. That larger, straight cut could just leave a bunch of extra space that your body needs to warm up. Since each item has the same R-value in any given spot, the one that conforms to your body (the tapered fit) is actually the warmer since there isn't a bunch of empty space. This scenario often applies to many different variables. One with longer arms being compared to one with shorter arms. One with insulated pockets vs one without. And even, to my dismay, jackets with hoods being compared to jackets with collars. A hood might contain 10% of the total fill weight in a jacket, which is often, using total fill weight as a comparison method, considered as a jacket that is 10% warmer than a jacket that has a collar. Obviously, attaching head insulation to a jacket does not increase it's warmth over a jacket in which you bring separate head insulation for. It's actually a colder jacket with a hat sewn onto it.
I could build two quilts/bags. Each with a top dimension of 52 and a foot dimension of 38. By changing the way I taper between these two dimensions I can vary the area and total fill weight by at least 20%. Again, two items, same R-value but 20% difference in total fill weight.....and again, on paper, the one with the large taper looks 20% warmer, when in fact, it's the same R-value in any given area and overall a colder item because of all the extra space your body needs to burn calories to heat up.
So why is this the industry standard? I don't know, but here is my assumption. Loft is really the first metric we should be using to indicate warmth. The amount of space filled with down between your body and the outside cold, is, by far the most influential metric for determining R-value. Problem is, there are more variables. I could fill a space with double the amount of down and it is warmer than the same space filled with half that amount. Same loft, but the one with more down is warmer. However, it is not as warm as if I took that "double amount" and let it loft to 100%. So, down, like most fibrous loose fill insulations, is most efficient per weight at it's full loft. But if we were to build an item with only enough down to fill the space, it would be the most efficient for the weight in a lab setting, but in real life there would not be enough density to keep the down from shifting around, leaving gaps. Therefor, the goal for a builder is to calculate the loft needed to achieve a certain temp rating and add to it just the amount of density needed to keep the down from shifting around. As a consumer, in lieu of an actual tested R-value measurement, what you really want to know are the numbers that the manufacturer uses to build and fill each chamber of an item. Basically, it's three numbers for a box chamber item.
- Calculated loft . When a builder fills a chamber first they find the area. Then they take that area of a chamber and multiply it by the desired loft. So a chamber 10" x 15" has an area of 150cu. If I'm building a 20f degree system I would take that area and multiply it by 2.5" and get a volume of 375cu. That, divided by the fill power of the down and I end up with the exact amount of down to fill that volume.
- Overstuff. We have calculated how much down is needed to JUST fill the volume of a chamber. However, as mentioned above, in the real world that density would shift around and leave big open gaps. So we need some density added on top of this to keep things from moving. We call this "overstuff" and it's typically expressed in a percentage. So for 30% overstuff I would add 30% to the base amount needed to just fill the space.
- Baffle height. So we have calculated the amount to just fill a space and an amount of overstuff to increase the density. However, 30% overstuff is actually not that much. It's still a pretty low density chamber, especially if there is a differential cut used to allow room for loft. So I could add more overstuff but then I am exceeding the target temp rating. So how do I increase the density without overshooting the temp rating? By reducing the baffle height. In my systems I calculate the desired loft, add 30% overstuff, and the reduce the baffle height by 50%. This gives me an 80% density and an average loft near where the target temp rating. By "average loft" I mean at the baffle it's half the desired loft. By mid chamber it might puff out to twice the desired loft, ending up with an average desired loft appropriate for a certain temp rating.
For a sewn through item, first see Sewn through baffle construction and it's effect on warmth. After reading through that, these are the numbers you need for determining warmth.
- Calculated loft. A sewn through item technically has a volume of zero since it is two dimensional. The calculated loft actually pushes the shells apart to make a three dimensional space. So you're actually calculating loft and overstuff together in this scenario, but it works in conjunction with.....
- Baffle spacing. Wider sewn through chambers allow for more loft and make fewer cold spots. Narrower spacing keeps the down compressed more and makes a bunch of spots with zero loft or fill.
- Measured loft. The end result of the above metrics is the actual measured loft in the middle of the chamber. So you'll know the loft and density which you can subtract the number of sewn through cold spots from. Essentially you're looking for higher loft and wider spacing but there is a point where fill control can become an issue. I feel that calc loft of .75 can do well in a 4" wide chamber, but anything wider doesn't have enough control. Similarly, I could potentially reduce the calc loft if I reduce the chamber width.
By using these metrics a consumer can easily compare dissimilar items. You could compare a wide/long quilt to a narrow/short quilt. You could compare a small hooded jacket to an XL jacket with collar. An item with extreme tapers to a boxy fit.....and it will always give you the numbers to compare the R-value of any section, regardless of size, fit, dimensions, etc. Every builder/manufacturer uses these numbers to fill the item so it merely a matter of just handing it over, whereas total fill weight is something abstract that doesn't apply to the actual build process. It's something we have to intentionally tally up for this purpose.
Of course there is more to an items warmth than just R-value. The metrics described above will get you a much, much more accurate picture of warmth than total fill weight, but there are still more metrics that should be considered to the discerning eye.
- Chamber size. A very large chamber will have less control and more shifting than a smaller chamber. I can put 80% density into a typical horizontal chamber and it results in a very secure fill that doesn't move much. However, if you put 80% density into a really wide or really long chamber that has no interruptions then the mass of the fill is so high that it has enough force to compress itself. Very small chambers could potentially get away with 30% overstuff just fine, but in full length vertical chambers (or worse) it will never be enough to keep it from shifting around.
- General design. To use the house analogy again.......obviously you can have all the R-value you want, but if you leave the door open it doesn't matter. Similarly with a sleep system you could have nice, lofty, dense chambers but if it doesn't seal well around it's openings, then the cold will get in. You can have all the R-value that you want but if you have a 4000 square foot house then it will be hard to heat. Similarly you could have nice, lofty, dense chambers but if the item is sized large your body heat is dissipated in the empty space.
I could build two jackets. Fill each chamber with the exact same calculations. Both with the same labeled size, same shoulder/chest width, same arm length, same torso length.....for all intents and purposes, similar items. However, if one jacket has a tapered torso and the other has a straight cut, the area of the item could change by 20%, meaning the total fill weight could change by 20%. The jacket with the straight cut has 20% more fill weight in it, and it is considered a warmer jacket when looking at fill weight. But it is not warmer. Pick any given section of the two jackets and they are the exact same R-value. In real life, the straight cut jacket might actually be colder. That larger, straight cut could just leave a bunch of extra space that your body needs to warm up. Since each item has the same R-value in any given spot, the one that conforms to your body (the tapered fit) is actually the warmer since there isn't a bunch of empty space. This scenario often applies to many different variables. One with longer arms being compared to one with shorter arms. One with insulated pockets vs one without. And even, to my dismay, jackets with hoods being compared to jackets with collars. A hood might contain 10% of the total fill weight in a jacket, which is often, using total fill weight as a comparison method, considered as a jacket that is 10% warmer than a jacket that has a collar. Obviously, attaching head insulation to a jacket does not increase it's warmth over a jacket in which you bring separate head insulation for. It's actually a colder jacket with a hat sewn onto it.
I could build two quilts/bags. Each with a top dimension of 52 and a foot dimension of 38. By changing the way I taper between these two dimensions I can vary the area and total fill weight by at least 20%. Again, two items, same R-value but 20% difference in total fill weight.....and again, on paper, the one with the large taper looks 20% warmer, when in fact, it's the same R-value in any given area and overall a colder item because of all the extra space your body needs to burn calories to heat up.
So why is this the industry standard? I don't know, but here is my assumption. Loft is really the first metric we should be using to indicate warmth. The amount of space filled with down between your body and the outside cold, is, by far the most influential metric for determining R-value. Problem is, there are more variables. I could fill a space with double the amount of down and it is warmer than the same space filled with half that amount. Same loft, but the one with more down is warmer. However, it is not as warm as if I took that "double amount" and let it loft to 100%. So, down, like most fibrous loose fill insulations, is most efficient per weight at it's full loft. But if we were to build an item with only enough down to fill the space, it would be the most efficient for the weight in a lab setting, but in real life there would not be enough density to keep the down from shifting around, leaving gaps. Therefor, the goal for a builder is to calculate the loft needed to achieve a certain temp rating and add to it just the amount of density needed to keep the down from shifting around. As a consumer, in lieu of an actual tested R-value measurement, what you really want to know are the numbers that the manufacturer uses to build and fill each chamber of an item. Basically, it's three numbers for a box chamber item.
- Calculated loft . When a builder fills a chamber first they find the area. Then they take that area of a chamber and multiply it by the desired loft. So a chamber 10" x 15" has an area of 150cu. If I'm building a 20f degree system I would take that area and multiply it by 2.5" and get a volume of 375cu. That, divided by the fill power of the down and I end up with the exact amount of down to fill that volume.
- Overstuff. We have calculated how much down is needed to JUST fill the volume of a chamber. However, as mentioned above, in the real world that density would shift around and leave big open gaps. So we need some density added on top of this to keep things from moving. We call this "overstuff" and it's typically expressed in a percentage. So for 30% overstuff I would add 30% to the base amount needed to just fill the space.
- Baffle height. So we have calculated the amount to just fill a space and an amount of overstuff to increase the density. However, 30% overstuff is actually not that much. It's still a pretty low density chamber, especially if there is a differential cut used to allow room for loft. So I could add more overstuff but then I am exceeding the target temp rating. So how do I increase the density without overshooting the temp rating? By reducing the baffle height. In my systems I calculate the desired loft, add 30% overstuff, and the reduce the baffle height by 50%. This gives me an 80% density and an average loft near where the target temp rating. By "average loft" I mean at the baffle it's half the desired loft. By mid chamber it might puff out to twice the desired loft, ending up with an average desired loft appropriate for a certain temp rating.
For a sewn through item, first see Sewn through baffle construction and it's effect on warmth. After reading through that, these are the numbers you need for determining warmth.
- Calculated loft. A sewn through item technically has a volume of zero since it is two dimensional. The calculated loft actually pushes the shells apart to make a three dimensional space. So you're actually calculating loft and overstuff together in this scenario, but it works in conjunction with.....
- Baffle spacing. Wider sewn through chambers allow for more loft and make fewer cold spots. Narrower spacing keeps the down compressed more and makes a bunch of spots with zero loft or fill.
- Measured loft. The end result of the above metrics is the actual measured loft in the middle of the chamber. So you'll know the loft and density which you can subtract the number of sewn through cold spots from. Essentially you're looking for higher loft and wider spacing but there is a point where fill control can become an issue. I feel that calc loft of .75 can do well in a 4" wide chamber, but anything wider doesn't have enough control. Similarly, I could potentially reduce the calc loft if I reduce the chamber width.
By using these metrics a consumer can easily compare dissimilar items. You could compare a wide/long quilt to a narrow/short quilt. You could compare a small hooded jacket to an XL jacket with collar. An item with extreme tapers to a boxy fit.....and it will always give you the numbers to compare the R-value of any section, regardless of size, fit, dimensions, etc. Every builder/manufacturer uses these numbers to fill the item so it merely a matter of just handing it over, whereas total fill weight is something abstract that doesn't apply to the actual build process. It's something we have to intentionally tally up for this purpose.
Of course there is more to an items warmth than just R-value. The metrics described above will get you a much, much more accurate picture of warmth than total fill weight, but there are still more metrics that should be considered to the discerning eye.
- Chamber size. A very large chamber will have less control and more shifting than a smaller chamber. I can put 80% density into a typical horizontal chamber and it results in a very secure fill that doesn't move much. However, if you put 80% density into a really wide or really long chamber that has no interruptions then the mass of the fill is so high that it has enough force to compress itself. Very small chambers could potentially get away with 30% overstuff just fine, but in full length vertical chambers (or worse) it will never be enough to keep it from shifting around.
- General design. To use the house analogy again.......obviously you can have all the R-value you want, but if you leave the door open it doesn't matter. Similarly with a sleep system you could have nice, lofty, dense chambers but if it doesn't seal well around it's openings, then the cold will get in. You can have all the R-value that you want but if you have a 4000 square foot house then it will be hard to heat. Similarly you could have nice, lofty, dense chambers but if the item is sized large your body heat is dissipated in the empty space.