I Made 20 Reed Diffusers by Hand, but the Fragrance Strength Varies Between Bottles — How Do I Improve Mixing and Filling Consistency?
Share
What can I check for free? Weigh all twenty, count reeds, note where each stood, and re-score blind from fresh air. See the audit.
How do I make them consistent? One master batch, stirred to the bottom and sampled at three levels, rested 24–48 h, filled to 40 g on the scale, 6 reeds each.
What does the test cost? ₹1,462.00 for six bottles and reeds with your own base and oil. See the kit.
Twenty bottles on your table, and they do not agree
Here is the batch as this page imagines it: twenty CSI 50 ml Gold bottles, each meant to hold 40 g of a 20% blend — 8 g of fragrance oil and 32 g of base — with a handful of black fibre reeds. They were made by hand over an afternoon or two. Some smell full and warm across the room; some you have to lean in to. The formula worked in your trial, so you are asking the right question: what in the making could make twenty bottles of one recipe disagree?
There are eight usual suspects, and the first four cost nothing to check. The order matters, because the cheap ones are also the common ones: it is far more likely that two bottles have different reed counts, or that you scored them in a different order on a different day, than that the oil separated. The diagram above sets them out; this section shows how to check the first four today, on your bench, without mixing anything.
Start with the audit sheet below. Line up all twenty bottles. For each, write down: the gross weight with reeds and cap off (on the same scale, one after another), the number of reeds, where it has been standing, the date it was filled, and whether it was mixed alone or poured from a shared batch. Weigh two empty bottles from the same pack so you know how much the glass itself varies; if two empty bottles differ by a few grams, gross weights alone cannot tell you the fill, but they can still show you outliers.
| Bottle | Gross weight with reeds (g) | Reeds | Where it stood | Filled on | Mixed alone or from a batch? | Blind score (0–5) |
|---|---|---|---|---|---|---|
| 1 | ___ | ___ | ___ | ___ | ___ | ___ |
| 2 | ___ | ___ | ___ | ___ | ___ | ___ |
| 3 | ___ | ___ | ___ | ___ | ___ | ___ |
| 4 | ___ | ___ | ___ | ___ | ___ | ___ |
| 5 | ___ | ___ | ___ | ___ | ___ | ___ |
| 6 | ___ | ___ | ___ | ___ | ___ | ___ |
| … 7 to 20 | ||||||
| Two empty bottles | ___ / ___ | — | — | — | — | — |
Then score them properly: one bottle at a time in a room that is not your workroom, the scorer coming in from fresh air, on a scale written down before the first sniff — for example 0 nothing, 1 at arm's length, 2 within a metre, 3 on entering, 4 throughout, 5 too strong. Code the bottles so whoever scores does not know which is which. If the twenty were all scored in one session on your bench, the ones scored first and the ones scored last were scored by two different noses: olfactory adaptation is commonly described, and coffee is a note that fills a room. Re-scoring blind sometimes makes half the "variation" disappear. The general mixing method is in the correct way to mix base and oil; the full method for one bottle is the 100 ml method pillar.
The eight suspects, and what each looks like in the audit
| # | Suspect | What it looks like | Free or test? | The check |
|---|---|---|---|---|
| 1 | Scoring conditions | Scores follow the order you sniffed, not the bottles | Free | Blind, coded, one bottle per room visit, fresh-air breaks |
| 2 | Reed count | Strong bottles have more reeds | Free | Count; re-reed all to one number from one pack |
| 3 | Fill weight | Heavy bottles score higher or last longer | Free | Gross weight minus an empty bottle's weight |
| 4 | Placement | Strong bottles stood near a window, fan, AC or heater | Free | Move all to one shelf for 3 days; re-score |
| 5 | Per-bottle mixing | Bottles mixed alone vary more than poured ones | Test | On a 1 g scale each was 18.75%–21.25%; remake from one batch |
| 6 | An unmixed batch | First-poured bottles differ from last-poured | Test | Three-level sample from a re-stirred batch |
| 7 | Rest time | Bottles bottled straight after mixing differ from rested ones | Test | Rest 24–48 h as the base pages ask, then fill |
| 8 | Oil lot | Bottles from a second oil bottle differ | Test | Same formula from each oil bottle, side by side |
Suspect 5 deserves arithmetic, because it is the one most hand-makers do not see. If each bottle was mixed on its own — 8 g of oil and 32 g of base weighed straight into the bottle — on a scale that reads in whole grams, each reading can be up to half a gram out. Half a gram on 8 g of oil is 6.25% of the oil. So one bottle might get 7.5 g of oil and the next 8.5 g: a true load anywhere from 18.75% to 21.25%, and up to 13.3% more oil in one bottle than another. Mix the same twenty bottles as one 800 g batch — 160 g of oil and 640 g of base — and the same scale holds the load between 19.938% and 20.062%. Every bottle filled from that batch carries the same load, because it is the same liquid. What scale accuracy you need for small trials and big batches works this through for other scale steps.
Suspect 3 is the next most common. Pouring to a line by eye on a small bottle is commonly a few grams out either way; the example here is ±4 g, so fills between 36 g and 44 g. That does not change the load — each bottle still holds the same blend — but it changes how much liquid the reeds have to work with and how many days the bottle lasts. At an example 1.2 g per day, 36 g lasts 30.0 days and 44 g lasts 36.7. Customers comparing two bottles a month later will notice that. Measuring filling losses on the first ten bottles includes a filling-method test.
How to make the next twenty identical
None of this needs new equipment if you already own a scale, a dry vessel and a stirrer. The scale is the one tool worth checking: if it reads only in whole grams, the master batch is the fix, because it moves the weighing error from the oil of each bottle to the oil of the whole batch, where half a gram is a rounding error. If you are scaling beyond twenty — fifty or a hundred bottles — the same sequence applies with a bigger vessel and a bridge batch; the scale-up hub covers that, and the production SOP for 1,000 diffusers turns it into a written procedure.
The remake test: six against six
Once the free checks are done, one test settles whether the making was the problem. Take six of your original bottles — two strong, two middling, two weak by your blind score — and give them all 6 fresh reeds from one new pack, so reed count and reed condition are no longer variables; these six are the control arm. Then make six new bottles from one master batch of the same formula, same base, same oil bottle, weigh-filled to 40 g, with 6 reeds each from the same pack. Stand all twelve on one shelf, out of sun and draughts. The one difference between the arms is how the liquid was mixed and filled.
| Record | Original six (control) | Master-batch six | What you are looking for |
|---|---|---|---|
| Day 0 gross weight | 6 readings | 6 readings | Fill spread: originals vs ±0.5 g |
| g/day, days 0–7 | 6 rates | 6 rates | Spread of rates within each arm |
| Blind score day 3 and 7 | 6 scores | 6 scores | Spread of scores within each arm |
| Highest minus lowest score | ___ | ___ | The master batch should be tighter |
| Clarity at 7 days | per bottle | retained sample | Any haze points to rest or water |
Weigh every bottle with its reeds at the same time each day on days 0, 1, 3 and 7; grams per day is the loss between two weighings divided by the days between. Score blind on days 3 and 7. You are comparing the spread inside each arm, not the arms' averages. If the master-batch six are tighter than the originals on both grams per day and score, the making was the cause, and the method above is the fix. If the master-batch six are as spread out as the originals, the cause is something the two arms share — the room, the reeds' contact with the liquid, the scoring — and why only some reeds saturate or standardising testing conditions is the next read.
| If a bottle loses | Days to empty | Weeks |
|---|---|---|
| 0.8 g per day | 50.0 | 7.1 |
| 1.2 g per day | 33.3 | 4.8 |
| 1.6 g per day | 25.0 | 3.6 |
| 2.0 g per day | 20.0 | 2.9 |
The priced test kit — and what not to buy
The test uses your own base and oil, so the CSI kit is only the glass and reeds: six 50 ml bottles (one pack of ten) and 72 black reeds. If you do not have enough of your blend left for a 289 g master batch — 58 g of oil and 231 g of base — the second option prices the example base and oil.
| Item | Need | Buy | Price |
|---|---|---|---|
| Gold Reed Diffuser Bottle 50 ml | 6 bottles | 1 × pack of 10 | ₹826.00 |
| Fiber Diffuser Reeds (Black) | 72 reeds | 1 × pack of 50 + 1 × pack of 25 | ₹636.00 |
| Kit with your own base and oil | ₹1,462.00 | ||
| Depth Base (only if needed) | 231 g | 1 × 500 ml | ₹1,099.00 |
| Roasted Coffee (only if needed) | 58 g | 1 × 50 g + 1 × 15 g | ₹323.22 |
| Kit with base and oil | ₹2,884.22 |
The six new bottles are sellable if they pass, and the other four in the pack start your next batch. Black reeds are ₹9.44 each in the 25-pack and ₹8.00 from 50; for 72 reeds the cheapest listed combination is 1 × pack of 50 + 1 × pack of 25. If you use Depth Base and Roasted Coffee, note that Depth Base's page names coffee among the rich gourmands it is for — a statement to test, not a result. Roasted Coffee is ₹4.60 per gram at 50 g and ₹3.90 at 100 g; on CSI's listed catalogue the one reliable per-gram step is from 15 g to 50 g or 100 g.
What a consistent bottle costs
| Line | Quantity | Cost |
|---|---|---|
| Roasted Coffee | 8 g | ₹31.20 |
| Depth Base | 32 g | ₹70.34 |
| Gold bottle 50 ml | 1 | ₹82.60 |
| Black fibre reeds | 6 | ₹48.00 |
| CSI materials per bottle | ₹232.14 |
In a 50 ml diffuser the liquid is ₹101.54 of the ₹232.14; the bottle and reeds are the rest. That makes inconsistency expensive in a quiet way: a bottle that disappoints wastes the glass and reeds as well as the liquid, and a customer who compares two of your gifts remembers the weak one. For the next twenty, the batch is 866 g including a 40 g retained sample and a 3% labelled loss — 173 g of oil (2 × 100 g) and 693 g of base (2 × 500 ml at the light density). The full cost ledger shows how to price the bottle, and organising five fragrances on one bench keeps the next batches apart.
The decision rule
| If this is true | It means | Do this |
|---|---|---|
| Blind re-scoring removes most of the spread | It was the scoring | Buy nothing. Score blind from now on |
| Strong bottles have more reeds | Reed count | Buy nothing beyond reeds; one count, one pack |
| Heavy bottles score higher or last longer | Fill weight | Fill by weight from now on |
| Strong bottles stood somewhere warm or draughty | Placement | Standardise the shelf; re-score after 3 days |
| Master-batch six are tighter than the originals | The making | Adopt the method; kit ₹1,462.00 was the only spend |
| Both arms equally spread | Shared cause | Room, reeds or scoring — 8272 |
| The master batch hazes when the trial did not | Process or water | Stop and diagnose — 8288 |
Five of the seven lines end in no purchase at all, or in reeds and bottles. That is the honest shape of this problem: uneven bottles are almost always a process problem, and process problems are fixed on paper. When the twenty agree, the formula can finally be judged — and if it is then too weak or too strong for your customers, testing throw without simply raising the load is where to go.
The CSI products in this test
| Variant | Price | Per bottle | Stock (15 Oct 2026) |
|---|---|---|---|
| 50 ml · pack of 10 | ₹826.00 | ₹82.60 | In stock |
| 120 ml · pack of 10 | ₹1,062.00 | ₹106.20 | In stock |
| Pack | Price | Per reed | Bottles covered |
|---|---|---|---|
| Pack of 25 | ₹236.00 | ₹9.44 | 4 bottles at 6 reeds |
| Pack of 50 | ₹400.00 | ₹8.00 | 8 bottles at 6 reeds |
| Pack of 100 | ₹800.00 | ₹8.00 | 16 bottles at 6 reeds |
| Pack of 500 | ₹4,000.00 | ₹8.00 | 83 bottles at 6 reeds |
| Size | Price | Per unit | Stock (15 Oct 2026) |
|---|---|---|---|
| 500 ml | ₹1,099.00 | ₹2.20 per ml · ₹2.20 per g at an example 1.00 g/ml | In stock |
| CSI HyperDiffusion Base™ (different product) | ₹1,499.00 | — | Out of stock — not offered |
| Pack | Price | Per gram | Bottles at 8.0 g oil |
|---|---|---|---|
| 15 g | ₹93.22 | ₹6.21 | 1.9 |
| 50 g | ₹230.00 | ₹4.60 | 6.2 |
| 100 g | ₹390.00 | ₹3.90 | 12.5 |
| 500 g | ₹1,890.00 | ₹3.78 | 62.5 |
| 1 kg | ₹3,738.00 | ₹3.74 | 125.0 |
| Item | Price | Note |
|---|---|---|
| Weighing Scale | ₹354.00 | Resolution not published — check the display steps |
| Silicon Measuring Beaker 250 ml | ₹600.00 | For batches above about 100 g |
| CSI Pro Mixer (cordless handheld) | ₹349.00 | Optional; a clean glass rod or spoon also works |
Where to go from here
This page belongs to the scaling-up section. If you are about to make fifty or a hundred bottles, read fifty 120 ml bottles by weight and base for a hundred after losses next; if the next batch is several fragrances, avoiding cross-contamination covers the bench layout. The section capstone pulls the method into a commercial product.
CANDLEMAKINGSUPPLIESINDIA supplies raw materials, not finished diffusers. The easy answer is "switch to a stronger base". This page says the variation is almost always in the making, prices a remake test at a pack of bottles and reeds using the base and oil you already own, and says plainly that we publish no throw, evaporation or longevity data for any base, oil or reed.
Diffuser bases, fibre reeds, Gold bottles, beakers, scale and mixer are CSI live pricing, re-checked 15 October 2026. Fragrance oils are CSI live pricing, pulled 24 September 2026, stock re-checked 14 October 2026. Every gram split, ml-to-g conversion, per-bottle cost and kit total on this page is computed and shown to two decimals so you can check it; densities are labelled examples, never CSI specifications.
Bottle counts per arm, clarity checkpoints, weighing schedules, room-scoring scales and loss allowances are labelled working practices, not industry standards — none exists for reed diffuser performance. The 40 g fill, the 1 g scale step, the ±4 g pour-by-eye spread, the 1.2 g/day rate, the 3% loss and the 40 g retained sample are labelled examples you replace. For help building a test kit, bulk quotes, GST invoicing, MSDS or IFRA documents, message us on WhatsApp at +91 7397976926.
Frequently asked questions
- I Successfully Made a 100 g Reed Diffuser Trial Batch — How Do I Scale the Formula to 5 kg Without Changing Fragrance Strength or Clarity?
- I Want to Make 50 Reed Diffusers of 120 ml Each — How Do I Calculate Base and Fragrance Quantities by Weight When My Bottles Are Labelled by Volume?
- I Am Making 100 Reed Diffusers with 20% Fragrance Oil — How Much Base Should I Buy After Accounting for Filling Losses and Density?
- I Am Making Five Reed Diffuser Fragrances with One Base — How Do I Organise Production to Avoid Fragrance Cross-Contamination?
- My 500 g Trial Formula Was Clear, but the 10 kg Batch Became Cloudy — Could Mixing Conditions or Raw-Material Differences Explain the Problem?
- I Want to Make Reed Diffusers Using a Digital Weighing Scale — What Accuracy Do I Need for 100 g Trials Compared with 10 kg Batches?
- I Am Ready to Manufacture 1,000 Reed Diffusers — How Do I Create a Production SOP for Weighing, Mixing, Resting, Filling, Sealing and Batch Tracking?
- I Want to Turn My Successful Homemade Reed Diffuser Formula into a Commercial Product — How Do I Finalise the Base, Fragrance Percentage, Fibre Reeds, Stability Testing, Costing and Wholesale Raw-Material Purchase?
Customer reviews: The six reviews at the top of this page are genuine, published Judge.me reviews left by CSI customers on the product pages named on each card — Eco Candle Wicks, Premium Candle Wicking Needle and Bamboo Wick Holder. They are general product reviews, not assessments of the guidance set out here. Names, star ratings, dates and products are as recorded by Judge.me and wording is unedited. "Verified buyer" appears only on the three reviews Judge.me recorded as verified — Monika Deep, Subhankar Roy and Sneha Tamang; the reviews from Ashwini, LJ and Lalitha Jagan are published but not recorded as verified, and carry no badge.
Figures re-checked 15 October 2026 and pulled 24 September 2026: Diffuser bases, reeds, bottles and tools — CSI live pricing, re-checked 15 October 2026: Pure Performance Base ₹325.00 / 100 g, ₹1,099.00 / 500 ml; Longevity Base ₹1,296.00 / 500 ml; Lift Base ₹1,099.00 / 500 ml; Depth Base ₹1,099.00 / 500 ml; Softness Base ₹1,099.00 / 500 ml; CSI HyperDiffusion Base™ ₹1,499.00 (out of stock); Fiber Diffuser Reeds (Black) ₹236.00 / 25, ₹400.00 / 50, ₹800.00 / 100, ₹4,000.00 / 500; Fiber Diffuser Reeds (White) ₹265.00 / 25, ₹450.00 / 50, ₹900.00 / 100, ₹4,500.00 / 500 (both 20.3 cm long); Gold Reed Diffuser Bottle 50 ml ₹826.00 / pack of 10, 120 ml ₹1,062.00 / pack of 10; Weighing Scale ₹354.00; Silicon Measuring Beaker 100 ml ₹300.00, 250 ml ₹600.00; CSI Pro Mixer ₹349.00. Fragrance oils — CSI live pricing, pulled 24 September 2026, stock re-checked 14 October 2026: Roasted Coffee Fragrance Oil ₹93.22 / 15 g, ₹230.00 / 50 g, ₹390.00 / 100 g, ₹1,890.00 / 500 g, ₹3,738.00 / 1 kg.
Non-price figures and assumptions: Fragrance load is a share of the total blend by weight (oil g = blend g × L; base g = blend g × (1 − L)). No density is published for any CSI base or oil: every ml-to-g conversion uses a labelled example density and shows ±5%; weigh 100 ml of your own blend to get yours. Fill 40 g per 50 ml bottle at 20% (8 g oil + 32 g base). Example scale step 1 g: per-bottle load band 18.75%–21.25%; 800 g master batch 19.938%–20.062%. Example pour-by-eye spread ±4 g; example rate 1.2 g/day. Remake test 6 + 6 bottles, 6 fresh reeds each, weighed on days 0, 1, 3 and 7. Allowances 40 g retained sample and 3% loss. Per-bottle cost uses Roasted Coffee at the 100 g rate and Depth 500 ml at the example density. What CSI's base pages state (quoted as claims to test, not CSI test results): every base is a "Professional … Reed & Car Diffuser Base", crystal clear per its page, used at 10–30% fragrance of the total blend with a 24–48 hour rest before bottling. Pure Performance — neutral, "no added fixatives, no complex enhancements", for all families, "start at 10% … up to a maximum of 30%", "20% recommended", a "typical working blend 75% base + 25% fragrance", the example "for a 100 ml finished diffuser at 25%: 75 ml base + 25 ml fragrance oil", and "flip reeds after the first hour". Longevity — oriental/amber, woody, resinous/balsamic, spicy and gourmand; "avoid light florals, fresh citrus, aquatics — use Lift Base"; slower "cold diffusion"; "if clouding occurs, reduce fragrance percentage and retest"; "allow 2–3 hours before the first flip". Lift — floral, citrus, green and fruity. Depth — chypre, leather, smoky and rich gourmand; "stabilises tricky sweet and spicy fragrance oils". Softness — soft florals, powdery/musky, light gourmand. Reed page — 20.3 cm long; "for a standard 50–100 ml diffuser bottle, 6–8 reeds gives a good balance". Bottle page — variants 50 ml and 120 ml (its text says "50 and 100ml"); neck diameter, fill line and any reducer are not stated. CSI publishes no composition, density, viscosity, evaporation rate, flash point or shelf life for any diffuser base; no throw, evaporation or longevity result for any base, oil or reed; no reed diameter; no bottle neck size or fill line; no IFRA percentage for any oil; and no bulk base price. The five bases are told apart only by what their pages say they are for — never by a measured result. IFRA: reed diffusers are listed under Category 10A in current IFRA amendments as commonly summarised in suppliers' guides; read the Category 10A limit on the certificate for your oil, which CSI provides on WhatsApp request. Material costs exclude labour, packaging, freight and GST. Diwali 2026 falls on Sunday 8 November, 24 days after this page was written on 15 October 2026.