Every ceiling calculator I could find does the same sum: total area, divided by the area of one board, rounded up. It is tidy and it is short every single time, because a ceiling is not one big surface, it is a set of separate rooms, and each room needs its own part-width last row and its own cut end board. Those offcuts do not walk down the passage into the next bedroom.
Three 3m x 3m bedrooms make the point cleanly. 27m2 divided by the 3.6m2 of a 1200 x 3000mm board says 8 boards. Laying them out properly needs 9, because each room takes 3 rows on its own. Scale that up to a whole house and you are the person driving back to the merchant on a Sunday afternoon.
So this one lays every room out individually, and then does the part the others skip entirely, the structure underneath. Brandering at Gyproc’s specified 400mm centres, running at right angles to the boards, with an extra pair wherever two boards meet end to end. Timber, cornice and joint strip all come in fixed lengths and get cut up, so it solves that as a real cutting list rather than dividing metres by stock length, and it picks the stock length that wastes least per room width. Screws come off total brandering metres at 150mm centres, nails off the tie beam crossings.
A couple of things fell out of the research that I did not expect:
Gyproc’s own estimating guide publishes the exact 900/1200 board combination for each room width (3000mm = 2 x 900 + 1 x 1200, 3300mm = 1 x 900 + 2 x 1200, and so on). The Best fit option reproduces that table exactly without being told the answer, which was a nice check that the maths is right.
Their rule of thumb of 0.75 linear metres of M-strip per m2 lands exactly on what the layout maths gives for a 4m x 4m room. Independent confirmation that joint strip belongs on the long edges only.
DIY saves a lot less on a ceiling than you would think. A contractor buys the same board and brandering at trade prices, usually 20% to 30% under retail, so a good part of their labour is already funded by a discount you cannot get. The tool says so rather than pretending otherwise.
It covers 6.4mm gypsum with M-strip (the standard nailed-up ceiling), 9mm taped and skimmed, and 4mm Nutec for bathrooms and patios, plus cornice, adhesive, trapdoors and optional insulation. All prices are current South African retail and all editable.
Let me know if the quantities look off against a job you have actually done, that is the part I most want checked.
Small update to this one. Someone made the point that the board count would land a lot better if you could actually see the layout it came from, so every room now gets a scale drawing underneath the room-by-room table.
The first version of the drawing shaded the boards that get cut, and it turned out to be useless. The reason is worth knowing. Most rooms in a South African house are narrower than a board is long, so there is one board per row and every single board gets cut to length, which meant the whole ceiling shaded in and said nothing at all.
So the boards are now drawn at their full bought size instead. The ones on the end of a row and the ones in the last row hang out over the wall, and that overhang is hatched off as waste. It reads the way the job actually goes: you buy this many whole boards, you cut this hatched strip off and throw it away, and that is exactly where the offcut percentage comes from. The brandering is dashed in underneath at right angles to the boards, the doubled pair shows up wherever two boards meet end to end, and each sketch is captioned with the real rip, something like “the last row is ripped down to 200mm wide”.
One thing it made obvious that I have not fixed. Put a small bathroom in with 4200mm boards and the drawing comes out mostly hatching, because the tool counts one board per row and never lets you use a long end offcut on the next row of the same room. A real installer gets two 1.8m rows out of one board. The auto length picker mostly hides it by picking a shorter board, but the sketch shows it up honestly, and I would rather leave that visible than quietly fudge the picture to flatter the maths.
So a question for anyone who has actually boarded a small room. How much of that end offcut do you genuinely get to reuse before it turns into more measuring and fiddling than it is worth?
Honestly, showing the overhangs hatched out is a much better way to visualize the waste. Good call on leaving the drawing honest rather than fudging the math to make the picture look prettier, it’s always better for a calculator to estimate a little high than to leave someone short on site.
To answer your question about small rooms: an installer will generally only reuse an offcut if it covers the entire span of the next row without needing a joint.
Gypsum ceiling boards are relatively cheap compared to the labor of finishing them. If someone cuts a 1.8m length out of a 4.2m board, they are absolutely using that remaining 2.4m piece for the next row. It’s a clean cut and drops right in.
What they won’t do is piece together smaller scraps to make up a row. As soon as reusing an offcut means adding an extra butt-joint in the middle of the ceiling, which means more taping, skimming, and sanding, or adding an ugly H-profile strip in a small bathroom, the piece goes straight to the skip. The extra measuring and finishing just isn’t worth the few bucks saved on material, and moisture in bathrooms makes extra joints a bad idea anyway.
If you ever wanted to tweak the logic down the line, you could just have the tool carry an offcut over to the next row only if its length is greater than the room’s width. But even without that, it sounds like a solid update to the tool.
You were right and the tool was wrong. The count follows exactly the rule you described: an offcut goes back up only if it covers the whole of the next row on its own, and anything shorter goes in the skip rather than being pieced in. So one board yields as many full rows as fit into its length, and none of that applies once the room is longer than a board, because then every row is being joined anyway and the leftovers would have to be staggered.
The passage was the worst case. 6m long and 1.2m wide in 4200mm board was costing 5 boards. It now says 2, which is what you would actually walk out of the merchant with. The bathroom that started this, 1.8m span out of a 4.2m board, went from 3 boards to 2.
What did not move is the part I was most worried about. The worked example on the page is still 9 boards and R7 112, and the three 3m x 3m bedrooms still come to 9 against the shortcut’s 8. That whole argument was untouched, because the auto board length was already picking sizes where the reuse rule never kicks in.
One thing I checked properly before shipping it. The premise of the entire tool is that dividing your area by the area of one board is never enough, and a change that lowers board counts could quietly break that. So it is now proved and pinned as a test: a count of whole boards can never come out below area divided by board area, whatever the room shape. Six room shapes by four board lengths, in the test suite, so nobody can undo it by accident later.
The drawings had to change with it or they would have contradicted the numbers. Waste is now hatched only at the end of each run of rows cut off the same board rather than on every row, so the passage draws as two boards with a small tail after the third row and a bigger one after the fifth. The caption spells it out: one 4.2m board cuts into 3 full rows at 1.2m.
The cost article is updated too. The seven-room house went from 26 boards to 24, so materials excluding cornice came down from R14 633 to R13 319, and the DIY saving against an installed quote moved from about 26% to about 32%.