Why will this not tell me what colour I get?

Because the honest answer needs data nobody publishes, and the dishonest answer is worse than none. Mixing pigments is subtractive — each pigment removes part of the spectrum, and what comes back depends on how strongly each one absorbs and scatters light at every wavelength. Modelling that is Kubelka–Munk theory, and it needs absorption and scattering coefficients measured per pigment, which paint makers hold and do not publish. The shortcut a naive calculator takes is to average the two colours as if they were light, and that is not a rough version of the right answer, it is a different one: averaging blue and yellow as light gives grey, while mixing them as pigment gives green. On top of that, your screen is not calibrated, so even a correct prediction could not be shown to you faithfully. A confidently wrong colour on a site whose whole position is verified arithmetic would be the worst thing here, so the page calculates what it can calculate exactly — the volumes — and sends you to a board for the rest.

Why does the page reduce my ratio?

So that you can write it down and do it again. Seven hundred and fifty millilitres to two hundred and fifty is a perfectly good mix and a terrible record: it is tied to the batch size you happened to make, so reproducing it means remembering both numbers and dividing. Three to one is the same mix expressed in a form that works at any size, with any measure — three jugs to one jug, three cups to one cup, three tins to one tin — and it fits on a lid. Reducing also exposes when two mixes you thought were different are the same, which matters when you are working through several samples and trying to remember which was which. The one time not to reduce is when the parts are genuinely awkward, such as 7:5: that is already in lowest terms, and the right move is to pick a measure small enough that seven and five of them are a sensible quantity rather than to round it to 3:2 and change the colour.

How big should the test sample be?

Big enough to cover a board at full opacity, which is more than people expect. The commonest testing mistake is brushing out a thin smear: a thin film lets the board show through, so it reads lighter and less saturated than the finished wall will, and a mix judged that way comes out darker than intended when it goes on properly. Two coats on a piece of the actual substrate — lining paper is fine for a wall, a scrap of the real board for joinery — is the minimum that tells you anything. Then judge it dry, because almost every paint dries to a different colour from the one it goes on as, and judge it in the room, because a colour picked under a shop's fluorescent light is a different colour under a north window. Leave the board propped against the wall for a day and look at it morning and evening before committing to a batch you cannot unmake.

Can I mix any two paints together?

Only if they are genuinely the same product in different colours. Same brand, same line, same sheen, same base chemistry — then the mix behaves like the product and only the colour is uncertain. Change the sheen and you get a film that is neither: gloss level is set by the ratio of binder to pigment and filler, and a half-and-half mix of matt and satin has a sheen the manufacturer has never tested and cannot stand behind. Change the base chemistry and it is worse, because water-based and solvent-based products do not mix at all in any useful sense. Across brands, even two matt emulsions may use different binders, different extenders and different preservatives, and the result can look right and fail early. The trade exception worth knowing is that leftovers of the same product in similar colours are routinely combined for a first coat or for a garage wall, and that is fine — what it is not is a way to make a colour you intend to repeat.

Renovation & Construction

Paint Mix Ratio Calculator — Two Paints, Recordable Parts

The volumes of two paints for a chosen ratio, the reduced parts to write on the lid, and a sample batch at that ratio, so the mix can be made again.

Answers as you typeNo submit button
Every formula citedNamed published standards
Calculated in your browserNo account, nothing uploaded

SettingsSettings for this calculationUS

Market

Imperial · sales tax

Total mixed volume needed

How much of the mixed paint the job needs.

Mix the whole quantity in one go wherever it will fit. Two batches of the same ratio are closer to each other than two tins of one colour are, but they are not identical, and the difference lands wherever you happened to stop — the boxing calculator covers where that changeover has to fall.

Parts of the first paint

The first paint's share of the mix, as a whole number.

Whole numbers, because a ratio is only useful if it can be repeated with whatever measure is to hand. Three to one is a mix anybody can pour twice; 2.7 to 1.3 is a mix that gets rounded differently every time and stops matching.

Parts of the second paint

The second paint's share, as a whole number.

If you are working from a sample you already mixed, enter the measures you actually used rather than what you intended — the page reduces them for you, and the reduced form is what goes on the lid.

Sample batch to test first

A small quantity at the same ratio, for a test board.

Enough to cover a board properly at full opacity — a smear tells you nothing, because a thin film shows the substrate through it and reads lighter than the finished wall will. Set this to zero if you have already tested and are mixing the job.

Volume of the first paint

0.9375 gal

High confidence

3 : 1 by volume — write that on the lid before you start, because it is the only thing that makes this mix repeatable. Mix the sample first: 6.3 fl oz of the first paint to 2.1 fl oz of the second, on a board, at full opacity, and look at it in the light the wall will live in. The job itself is 0.94 gal to 0.31 gal. The volumes above are exact. The colour they make is not predicted here and is not shown: mixing pigments subtracts light rather than averaging it, and the arithmetic that would model it needs per-pigment coefficients nobody publishes. Mix the sample, judge it dry on a board in the room's own light, and mix the job only after that.

Second paint: 0.31 gal
Reduced ratio, first paint: 3 parts
Reduced ratio, second paint: 1 part
Sample: first paint: 0.05 gal
Sample: second paint: 0.02 gal
Total mixed: 1.25 gal

Then change the inputs to see how far the answer moves.

Show calculation logic

How this was calculated

Formula source(s)

Volumes are the ratio applied to the batch: each paint's share is its own parts divided by the total parts, times the batch size. The parts are reduced by their greatest common divisor so the mix can be written down and repeated
A SAMPLE is mixed at the identical ratio rather than poured by eye, because a sample with no recorded ratio cannot be scaled up — the wall is the same two numbers multiplied, and nothing else reproduces it
No colour is predicted or displayed. Pigment mixing is subtractive and pigment-specific, modelling it requires per-pigment absorption and scattering coefficients that are not published, and an uncalibrated screen could not show the result faithfully in any case

Inputs used

Total mixed volume needed: 1.25 gal
Parts of the first paint: 3
Parts of the second paint: 1
Sample batch to test first: 0.07 gal

Intermediate steps

Second paint: 0.31 gal
Reduced ratio, first paint: 3 parts
Reduced ratio, second paint: 1 part
Sample: first paint: 0.05 gal
Sample: second paint: 0.02 gal
Total mixed: 1.25 gal

Final result0.94 gal

Confidence note: 3 : 1 by volume — write that on the lid before you start, because it is the only thing that makes this mix repeatable. Mix the sample first: 6.3 fl oz of the first paint to 2.1 fl oz of the second, on a board, at full opacity, and look at it in the light the wall will live in. The job itself is 0.94 gal to 0.31 gal. The volumes above are exact. The colour they make is not predicted here and is not shown: mixing pigments subtracts light rather than averaging it, and the arithmetic that would model it needs per-pigment coefficients nobody publishes. Mix the sample, judge it dry on a board in the room's own light, and mix the job only after that.

What this calculation does not cover

The volumes above are exact. The colour they make is not predicted here and is not shown: mixing pigments subtracts light rather than averaging it, and the arithmetic that would model it needs per-pigment coefficients nobody publishes. Mix the sample, judge it dry on a board in the room's own light, and mix the job only after that.
The two paints must be the same product, the same sheen and the same base type. A matt and a mid-sheen of the same brand are two different products, and mixing them gives a film with neither one's gloss, neither one's durability and no manufacturer's position on how it should behave. Mixing across brands adds an unknown compatibility to an unknown colour.
Measure by volume in the same vessel, or by weight on scales — but not one of each. The two paints may differ in density, so equal weights are not equal volumes, and a ratio poured as volume and topped up by weight is a third ratio that matches neither.
The sample has to be judged dry and at full opacity, on the substrate it will actually go on. Wet paint is a different colour from dry paint on almost every product, a thin test smear reads lighter because the board shows through it, and a colour judged under a shop light is not the colour of the room.
Nothing here is reversible. Paint added cannot be taken out, so mix the smaller quantity first and work up. If the job needs more later, the ratio on the lid is what saves you — mixing a second batch by eye to match the first is the failure this page exists to prevent.
Two batches at one ratio are close but not identical, and the difference lands wherever the first batch ran out. Mix the whole job at once where it will fit, and where it will not, put the changeover on a corner or a casing rather than halfway up a wall.

Estimated cost — your price

Price per gal ($)

This site holds no price list for this material — local prices vary too much to publish honestly. Enter your supplier's price and the result is costed with it.

Computed in your browser — nothing you enter is uploaded. Presented in US customary units and US trade terminology. Where a formula follows a published standard, that standard and its edition are cited beside it on this page; where none governs, the page says so. Local amendments override model codes — verify against the code in force where you build.

Sources checked 2026-09-15 · v1.0.0

Regulatory standards & verification citations3

Volumes are the ratio applied to the batch: each paint's share is its own parts divided by the total parts, times the batch size. The parts are reduced by their greatest common divisor so the mix can be written down and repeated
A SAMPLE is mixed at the identical ratio rather than poured by eye, because a sample with no recorded ratio cannot be scaled up — the wall is the same two numbers multiplied, and nothing else reproduces it
No colour is predicted or displayed. Pigment mixing is subtractive and pigment-specific, modelling it requires per-pigment absorption and scattering coefficients that are not published, and an uncalibrated screen could not show the result faithfully in any case

Cite this page

Your workspace

Most jobs need more than one number. Add the calculators you need next and they open right here, underneath this one — your figures stay on screen and nothing is lost to a page change.

Protection this work requires

Stop and check first. Paint under later coats in older housing can be lead-based, and sanding or heat-stripping it produces lead dust that is especially harmful to children and to anyone pregnant. Test before you disturb it — this is a job for a tested, controlled method, not a better mask.

Tools and safety for this job

To roll and cut in two coats and prepare a surface for paint. Generic types, no brands, no prices.

Joint-compound dust is very fine and travels: a P1/FFP1 mask is the minimum, and a damp-sponge finish avoids most of it.
A dust mask does nothing about solvent vapour: ventilate through, and use an organic-vapour (A1) cartridge respirator in an enclosed room.
Most fatal falls on small jobs are from under three metres: use a tower or a ladder tied at the top, and never work off the top two rungs.

Show the 11 tools this job needs

Essential

Brushes
Dust sheets and masking
Filling knife
Roller frame, sleeves and tray
Scraper

Recommended

Caulking gun
Extension pole
Pole sander
Random orbital sander
Step ladder

Optional

Dust extractor or vacuum with a fine filter

Also needed as materials: abrasive discs, caulk and sealant cartridges, masking tape, plastic sheeting, roller sleeves, sanding screens or paper.

what each painting tool is for, and the spec that decides which to buy where one does.

The data behind it: Production rates by operation

How to calculate paint mix ratio — two paints, recordable parts in 5 steps

Total mixed volume neededHow much of the mixed paint the job needs.
Parts of the first paintThe first paint's share of the mix, as a whole number.
Parts of the second paintThe second paint's share, as a whole number.
Sample batch to test firstA small quantity at the same ratio, for a test board.
Volume of the first paintThe tool computes the volume of the first paint from those figures and shows the formula, its sources, and a confidence rating alongside it.

Volume of the first paint by total mixed volume needed

Page defaults, not your figures above.

Total mixed volume needed	Volume of the first paint (gal)
1 gal	0.75
1.5 gal	1.13
2 gal	1.5
2.5 gal	1.88

Frequently asked questions

Why will this not tell me what colour I get?: Because the honest answer needs data nobody publishes, and the dishonest answer is worse than none. Mixing pigments is subtractive — each pigment removes part of the spectrum, and what comes back depends on how strongly each one absorbs and scatters light at every wavelength. Modelling that is Kubelka–Munk theory, and it needs absorption and scattering coefficients measured per pigment, which paint makers hold and do not publish. The shortcut a naive calculator takes is to average the two colours as if they were light, and that is not a rough version of the right answer, it is a different one: averaging blue and yellow as light gives grey, while mixing them as pigment gives green. On top of that, your screen is not calibrated, so even a correct prediction could not be shown to you faithfully. A confidently wrong colour on a site whose whole position is verified arithmetic would be the worst thing here, so the page calculates what it can calculate exactly — the volumes — and sends you to a board for the rest.
Why does the page reduce my ratio?: So that you can write it down and do it again. Seven hundred and fifty millilitres to two hundred and fifty is a perfectly good mix and a terrible record: it is tied to the batch size you happened to make, so reproducing it means remembering both numbers and dividing. Three to one is the same mix expressed in a form that works at any size, with any measure — three jugs to one jug, three cups to one cup, three tins to one tin — and it fits on a lid. Reducing also exposes when two mixes you thought were different are the same, which matters when you are working through several samples and trying to remember which was which. The one time not to reduce is when the parts are genuinely awkward, such as 7:5: that is already in lowest terms, and the right move is to pick a measure small enough that seven and five of them are a sensible quantity rather than to round it to 3:2 and change the colour.
How big should the test sample be?: Big enough to cover a board at full opacity, which is more than people expect. The commonest testing mistake is brushing out a thin smear: a thin film lets the board show through, so it reads lighter and less saturated than the finished wall will, and a mix judged that way comes out darker than intended when it goes on properly. Two coats on a piece of the actual substrate — lining paper is fine for a wall, a scrap of the real board for joinery — is the minimum that tells you anything. Then judge it dry, because almost every paint dries to a different colour from the one it goes on as, and judge it in the room, because a colour picked under a shop's fluorescent light is a different colour under a north window. Leave the board propped against the wall for a day and look at it morning and evening before committing to a batch you cannot unmake.
Can I mix any two paints together?: Only if they are genuinely the same product in different colours. Same brand, same line, same sheen, same base chemistry — then the mix behaves like the product and only the colour is uncertain. Change the sheen and you get a film that is neither: gloss level is set by the ratio of binder to pigment and filler, and a half-and-half mix of matt and satin has a sheen the manufacturer has never tested and cannot stand behind. Change the base chemistry and it is worse, because water-based and solvent-based products do not mix at all in any useful sense. Across brands, even two matt emulsions may use different binders, different extenders and different preservatives, and the result can look right and fail early. The trade exception worth knowing is that leftovers of the same product in similar colours are routinely combined for a first coat or for a garage wall, and that is fine — what it is not is a way to make a colour you intend to repeat.

Preliminary estimate, not certified engineering. This tool produces an indicative quantity calculation for planning purposes only — it is not a certified structural analysis, a guaranteed material takeoff, or a substitute for building department approval. Always verify measurements on-site and have a licensed contractor or structural engineer review any load-bearing, code-sensitive, or safety-critical work before purchasing materials or starting construction. Spotted an arithmetic or standards error? Report it to contact@craftquantities.com with your inputs — a confirmed fix gets a permanent check of its own, so the same mistake cannot come back.