Parts are a ratio, not a quantity
Written as 2:2:1:1, a recipe carries no size at all. It has to be multiplied by something before a scoop can be filled, and the multiplier is the batch divided by the sum of the parts. At the defaults on this page the parts sum to 6, and the loose total works out at 8.70 quarts, so one part is 1.45 quarts. The two-part ingredients get 2.90 quarts each and the one-part ingredients 1.45 each.
Those four figures add to exactly 8.70 because the page splits an integer number of hundredths using largest remainders rather than rounding each line on its own. Rounding independently is how a recipe ends up displaying four numbers that sum to 8.69 or 8.71, and on a printed sheet that discrepancy is the thing people notice and distrust.
Why the parts add up short
Pour a quart of coarse bark into a bucket and a quart of fine material on top of it, then mix. You do not get two quarts. The fine material goes into the void space between the bark pieces, and the total settles somewhere below the mark. That loss is real volume that will not come back, and it is the reason a mix planned to fill a specific number of pots comes up one pot short.
How much it loses depends entirely on the size spread of what you are combining. A blend of pieces that are all roughly the same size loses very little, because there is nothing small enough to fall into the gaps. A blend that pairs dust with chunks can lose 15 percent or more. The default of 8 percent on this page is a placeholder for a number you should measure once: fill a bucket with the loose parts, mark the level, mix thoroughly, and read the new level.
The correction runs the other way from the direction people expect. To finish with 8 quarts at an 8 percent loss you do not add 8 percent, you divide by 0.92, which is a multiplier of 1.087. The difference between adding 8 percent and dividing by 0.92 is small at 8 percent and not small at 25.
Anchoring on what you actually have
Batches rarely start from a target. More often one ingredient is the constraint: half a bag of something left over, one bucket of something else. The third batch mode anchors the whole recipe to ingredient 1, so entering 8 quarts scales everything to that being the two-part ingredient, and the loose total comes out at 24 quarts rather than 8.70.
Move the anchor to a different ingredient by reordering the list. The parts are what determine the scaling, so an ingredient with 1 part as the anchor produces a much larger batch than the same volume of an ingredient with 3 parts.
Scoops and the quarter-scoop rounding
Nobody measures 2.90 quarts. They measure three scoops of a quart tub, or six of a 2 cup jug. The scoop section takes whatever you actually measure with and reports the exact scoop count per ingredient, then a version rounded to the nearest quarter scoop with the total that version produces. At the defaults the rounded batch overshoots the target by a small margin, which is usually the right direction to be wrong in.
If the rounding drift bothers you, the fix is a smaller scoop, not more careful eyeballing. A 1 cup measure has four times the resolution of a quart tub, and the same recipe lands on whole scoops far more often.
Volume, not weight, and no opinions on content
Everything on this page is volume. Two ingredients at 1 part each fill the same space and can be an order of magnitude apart in weight, which matters when you lift the bucket and when you are translating a recipe that was written in pounds. There is no conversion between the two without a density for each material, and densities for the same nominal material vary with how wet and how compressed it is.
What goes in the mix is not something this page addresses. It scales the ratio you enter and stops there. Recipes are argued about endlessly and the right answer depends on the plant, the pot, the room and how you water, none of which is arithmetic. Nothing here covers fertilizer or any other chemical addition either.
Questions people ask
Do the parts have to be whole numbers?
No. The field takes any positive number, so 1.5:1:0.5 works exactly as well as 3:2:1, and the two produce identical proportions. What matters is the ratio between them, not their absolute size — a recipe entered as 20:20:10:10 gives the same volumes as one entered as 2:2:1:1. If you are transcribing a recipe that uses fractions, entering 0.25 for a quarter part is fine and the scoop rounding will still land on quarter scoops where the arithmetic allows it.
How do I measure the settle allowance for my own mix?
Fill a bucket with your ingredients in the right proportions, loose and unpacked, and mark the level with a pen on the inside. Mix it thoroughly, level the surface, and mark it again. The drop between the two marks divided by the original height is the allowance, near enough. Do it on a batch big enough that the marks are inches apart rather than a fraction of one. Wetting the mix will drop it further, so if you always wet before potting, measure it wet and use that figure instead.
Why does the finished volume matter more than the loose total?
Because the finished volume is what goes into pots, and that is what you were sizing the batch against in the first place. If you scale a recipe so the loose ingredients add to 8 quarts, you end up with about 7.4 finished at an 8 percent allowance, which is short of a shelf you had counted on filling. Setting the batch mode to finished mix makes the target the thing you care about and lets the page work backwards to the shopping list.
Can I use this for weights instead of volumes?
Not directly, and it would give the wrong answer if you tried, because the settle allowance and the scoop arithmetic are both about space rather than mass. A recipe by weight percentage is a different calculation with different failure modes, and the glaze batch calculator handles that shape: percentages of a batch weight, additions above 100 percent, and water to a target specific gravity. If you have a recipe in pounds and want it in quarts, you need a bulk density for each ingredient, which varies enough with moisture and compaction that measuring your own is the only reliable route.
What if I only have two ingredients?
Fill in two rows and leave the rest blank. Rows with no parts entered are skipped entirely, so a two-ingredient mix reports two lines and the percentages still sum to 100. The same applies in the other direction: all six rows work, and the largest-remainder split keeps the volumes adding to the loose total exactly no matter how many rows are in play or how awkward the ratio is.