I Want to Make a Minimalist White Luxury Candle — Which Wax and Fragrance Combinations Are Less Likely to Create Obvious Discoloration?

What CSI actually publishes about discolouration

Discolouration is the best-documented appearance fact in the CSI fragrance range, and it is still only documented on a majority, not on all of it. Across the 107 fragrance-oil product pages, 62 mention vanillin or discolouration in some form and 45 mention neither. Of the 62, seventeen put a number on it — fourteen at 0%, one at 0.2%, two at 2.5% — and six describe vanillin only as "Present". The rest describe the consequence in words instead of figures, and seven of those use the same blunt sentence: the oil will cause browning in white or light-coloured wax.

That distribution is the whole strategy for a white collection. You are not looking for the oil that smells most like a luxury candle. You are looking for the intersection of an oil you want to sell and a page that commits to a number. Everything else is an unpriced risk you are carrying on behalf of your customer, and the customer notices it four weeks after the sale, not four days.

The fourteen 0% vanillin oils, priced per test jar

Here is the entire shortlist with the two numbers that matter — the vanillin figure and the stated soy ceiling — plus what one 70 g test jar of each costs in oil at a 10% load on wax weight. Thirteen of the fourteen state up to 10% for soy wax. The fourteenth, Citronella, states the vanillin figure but no candle load at all, which is a useful reminder that a page can be precise about one thing and silent about another.

Two practical readings. First, the spread from the cheapest to the dearest of the fourteen is ₹183.50 on a 50 g pack, which inside one 70 g test jar is only ₹23.34. At screening stage, price is not the variable worth optimising — the vanillin figure and the smell are. Second, notice that the shortlist is heavily weighted towards florals, citrus, herbals and woods, and contains no gourmands. That is not an accident of stocking; it is what the chemistry of vanillin-bearing notes does, and it is why vanilla oils turn wax yellow is a separate question with a separate answer.

The wax half of the question has no published answer

This is the part most guides skip. You can look up a vanillin figure for fourteen oils, but you cannot look up a colour for a single CSI wax. Across the 15 waxes in the wax collection — soy, coconut-soy, paraffin, beeswax, gel, microcrystalline, soy butter — not one product page states a colour, a whiteness grade, a bleach level or a discolouration position. The only colour words in the entire wax range appear on the Beeswax Pellets page, which mentions white and yellow while only the yellow variant is listed as a Shopify option.

So the honest specification is: pick the wax on the specifications it does publish — the stated load ceiling, the stated temperatures, the stated use — and settle appearance with your own side-by-side pour. That is a two-hour job, not a research project, and it is the only evidence that will survive a customer question. Below are the container waxes worth putting in the comparison, with what each page actually commits to.

For a white brief, Luxury Soy Wax Chunks is the strongest starting point in the range, and not because of anything to do with colour. It is because its page publishes the most numbers: a stated range rather than a single ceiling, a stated add temperature, a stated pour temperature and a stated cure instruction. When you are trying to isolate one variable — does this oil change the colour? — every other variable needs to be nailed to a published figure so that it cannot move underneath you. A wax with no stated temperatures forces you to invent two, and then your screen has three unknowns instead of one.

Two ceilings, and which one governs

Every pairing in this post has two published ceilings: the wax page's load range and the oil page's stated candle rate. Use the lower one. Luxury Soy Wax Chunks state 8–13% of total wax weight; Jasmine states up to 10% for soy wax. The pairing's ceiling is therefore 10%, and a 10% pour puts you at the top of the oil's stated range while sitting comfortably inside the wax's. If you want headroom below both numbers on a first test — which is the right instinct — pour at 8%, the bottom of the Chunks range, and log it.

All arithmetic on this page is percentage of total wax weight, the convention CSI's own wax pages use. Some makers calculate load as a share of the finished candle instead, so for clarity: 10% of wax weight is 9.09% of the finished candle, and 8% of wax weight is 7.41%. Pick one convention, write it on your batch sheet, and never mix them inside a single screen.

The costed twelve-jar screening basket

Here is the whole shopping list for a screen that answers four questions at once. Four oils from the 0% shortlist, three jars each, two wick sizes distributed across them. Twelve White Matte Votive Jars at a stated ~70 g wax fill — a real gram figure, which is why these are the screening vessel rather than a jar that states millilitres. Three spare jars are in the basket because a cracked or badly wicked jar mid-screen otherwise costs you the whole comparison.

Running the 21-day colour screen

A colour screen fails for one reason above all others: the maker looks at the jars on pour day, likes them, and ships. Colour drift in wax is a slow process and the only way to catch it is to build the delay into the method and to keep an unburned control you can hold next to the burned ones. Twelve jars, three per oil, one photograph protocol.

Scaling to the finished white collection

Say two oils pass. You now build a ten-candle launch run in the Clear Frosted Glass Jar, five of each scent. The first thing to do is the thing nobody does: weigh one filled jar. The page states a fill weight of 200 ml, and millilitres are not grams — soy, coconut-soy and paraffin all fill a 200 ml vessel to a different weight, and your own pour line will differ from ours by several grams either way. Fill one jar with melted wax to the height you like, let it set, weigh it, subtract the empty jar weight, and call that number F.

Then the run arithmetic is three lines. Wax per candle = F ÷ 1.10. Oil per candle = F − (F ÷ 1.10). Batch wax = 10 × that, plus a labelled 4% wastage allowance for what stays in the pouring jug, the top-up pour and spills. If your weighed F comes out at 170 g, that is 154.55 g of wax and 15.45 g of oil per candle; ten candles need 1,545.50 g of wax plus 4% = 1,607.32 g. That is the rounding trap in this series in one number: it is more than a 1 kg bag plus a 500 g bag (1,500 g), so the smallest sufficient purchase is two 1 kg bags at ₹1,198, leaving 392.68 g in hand — or a 5 kg bag at ₹2,995 if a second run is already planned inside the month. The leftover is your next test batch, not waste, but the rounding is real and you should see it before you price the collection.

The last row of that table is the quiet advantage of a white brief. A minimalist white candle has no colourant in it, which removes the single most common cause of a wick that behaved during testing and misbehaved in production. Colourant load competes with the wick for the flame's attention; a white candle hands the wick nothing extra to burn through. If you later decide the range needs a coloured sibling, that is a new formula with a new wick test, covered in choosing a candle dye without hurting wick performance and liquid dye versus dye chips for a deep burgundy.

Frequently asked questions