The recipe is the destination, not the shopping list
Earthen mix recipes are almost always written as the proportions of the finished mix: so much sand to so much clay. That is a useful way to describe a mix and a useless way to order material, because the subsoil you are digging is already a blend of the two. If your pit reads 45 percent sand and you want 70, you are not adding 70 units of sand for every 30 of clay. You are closing a 25 point gap, and the arithmetic for that gap is short.
Take the mineral you will handle as one unit. Some fraction of it is your subsoil and the rest is clean sand. The sand in the result is the subsoil fraction times its own sand content, plus all of the added sand. Set that equal to your target and there is exactly one split that satisfies it. At 45 percent subsoil and a 70 percent target it comes out at 54.5 percent subsoil and 45.5 percent added sand, which is 0.83 of sand to one of subsoil and reads nothing like the recipe it satisfies.
The direction of the error is always the same. People read a sand-heavy recipe, shovel sand in that proportion on top of soil that was already sandy, and end up with a mix that will not hold together. The fix costs nothing: a settling jar, a day, and a reading.
Loose in, placed out
The other number that does not behave is volume. Loose subsoil and loose sand heaped on a tarp occupy more space than the same material wetted, trodden and pressed into a wall, because mixing fills the voids and treading closes what is left. That ratio is a property of your material and your process and there is no book figure for it. Measure it once: count the barrows into a test batch, build the batch into a known length of wall, and divide.
Everything downstream runs on the loose figure. Barrow loads, tarp batches, the size of the sand delivery, how much of your own hole you need to dig — all of it is the loose number. Only the wall itself is the placed number. Quoting one and buying the other is how a job runs out of material two thirds of the way up.
Straw is dosed by weight
A parts recipe that includes straw by volume is describing something that cannot be measured usefully. Loose straw is mostly air, it compresses differently depending on how it was baled and how wet the mix is, and a heaped barrow of it is not a repeatable quantity. Weight is repeatable. Work out on a test batch how many pounds of straw go into a known volume of finished wall, and then the whole job scales off that single figure.
What this page refuses to do
It supplies no target. The sand share you are aiming for is an input because it depends on your clay, your straw, what the wall is doing and what has worked in test bricks made from your own soil. There is no universal cob ratio, and any page that gives you one is telling you about somebody else's dirt.
It also supplies no strength, no thickness, no height limit. Cob and other monolithic earthen walls sit outside the prescriptive path in most adopted codes; getting one permitted normally means an alternative materials review and often an engineer. Ask the building department early. Making test bricks is cheap; discovering at inspection that the wall needs a stamp is not.
Questions people ask
How do I find the sand content of my subsoil?
The usual method is a settling jar: soil and water in a straight-sided jar, shaken and left to settle, and the bands read off as a share of the mineral column with sand at the bottom. It is rough and it is enough for this purpose. What matters is that you read your own pit rather than assuming, because the whole point of the blend arithmetic is that your soil is not the soil the recipe was written for.
My subsoil is sandier than my target. Do I add clay?
That is what the arithmetic says, and the calculator switches to it automatically. Adding sand to a mix that is already too sandy moves it further from the target, so the correction is clay or a clay-rich soil. Sourcing clay is a different job from ordering sand, and any clay soil you find is itself a blend, so test it too before you count on it.
Why does the calculator ask how much loose material equals a unit of wall?
Because loose material and placed wall are different volumes and everything you buy and carry is measured loose. Mixing fills the voids between grains and treading closes what is left, so a barrow on the tarp is less than a barrow in the wall. The factor is a property of your soil and your process. Measure it once on a test batch and it holds for the job.
What sand to clay ratio should cob be?
This page does not say and will not. The right proportion depends on the clay you have, the straw, what the wall is carrying and how it is protected, and the way people settle it is by making test bricks from their own soil at several ratios and seeing which survives. A ratio quoted without the soil it belonged to is not information you can use.
Does this cover the plaster and the foundation?
No. It works out the mineral blend, the straw and the batch count for the cob itself. The plinth or stem wall the cob stands on, the roof overhang that keeps rain off it, and the plaster that finishes it are all separate and all more important to the wall lasting than the mix is. For a bag-yield render system there is a stucco calculator on the site; for the earthen plasters people usually use over cob, the mix is the same kind of blending problem as this page solves.