Your Grow Guide

At a glance

The one number that shapes this whole grow

Flowering cannabis yield rises linearly with light all the way to 1800 PPFD, measured at ambient CO₂ with no enrichment. Your fixture tops out near 1040. That is roughly 58% of the tested ceiling, and reaching 1800 in this footprint would take a 475–500W fixture.

That is not a criticism of the gear — it is the single most useful fact you own. It means when something goes wrong, light is the first thing to check and the last thing to reduce. It means CO₂ is pointless for you (you are not light-saturated, so there is nothing for extra CO₂ to unlock). It means defoliation to “improve penetration” costs you source leaves you cannot spare. And it means your realistic yield is 70–110 g for a first run, not the 1–2 g/W the internet quotes.

What coco actually is

Coco is soilless, not soil — effectively a simple hydroponic system you run like soil. It carries zero nutrition, so you feed every watering from the seedling stage onward. And it is an ion exchanger: it preferentially binds calcium and magnesium while releasing potassium and sodium, which is why cal-mag is not optional and why runoff pH tells you about the substrate rather than the root zone.

At a glance

↑ Open the VPD chart & calculator — pinned at the top of this page.

What VPD actually is

Tap any line to open it.

What your own tent settled — run 1, Aug 2026

Your readings, not somebody’s paper. Tap any line to open it.

What the Aug 2026 research pass settled

Four literature searches, ~370 sources.

The lights-off humidity spike (after the 12/12 flip — veg is 24/0)

At a glance

Why more light keeps paying

The definitive study is Rodriguez-Morrison, Llewellyn & Zheng (2021), University of Guelph, in Frontiers in Plant Science. They ran cannabis through flower at canopy PPFDs from 120 all the way to 1,800 and yield rose linearly the whole way — 116 to 519 g/m², 4.5× — with no plateau at the top. Two details matter:

First, this was at ambient CO₂ (437 ppm). No enrichment. The common advice that “1,000 PPFD is your ceiling without CO₂” is not what the research shows.

Second, THC percentage did not change. Potency was flat across every light level. But because yield quadrupled, total cannabinoids per square metre rose 4.5×. More light means more weed at the same strength — not stronger weed.

The subtlety people miss: individual leaves saturate somewhere around 500–900 PPFD, but the canopy doesn’t. Once the top leaves are maxed, extra photons punch deeper and light up the mid and lower bud sites that were sitting in shade. That is why whole-plant yield keeps climbing long after leaf-level photosynthesis has flatlined.

Your fixture vs the plant

You top out at roughly 55–60% of the plant’s demonstrated ceiling. Running click 10 in peak flower is not aggressive — it is simply everything the fixture has, and she would still take more. So the real question is never “is this too much light”; it is “what stops me running click 10”, and the answer is heat, then water, then CO₂.

How to actually measure it

The free Photone app turns your phone into a usable quantum meter — within about 10% of a real one, which is plenty.

Lay one sheet of plain white printer paper over the front camera as a diffuser. Skipping that is the single biggest source of bad readings. Set the source to LED. Put the phone screen-up at the top of the canopy, dead centre, then take a second reading at each corner. The corner-to-centre spread tells you whether to raise the fixture.

Re-measure weekly. Height is measured to the top leaf, so the gap closes as she grows and your PPFD drifts upward on its own.

Three ways to read this wrong

At a glance

Do this

• Pre-soak the coco.
• Squeeze-test it.
• Sow ½″ deep, pointy end down.
• Two seeds, both Lime Colada.
• Set the tent to 75–82°F and 70–85% RH.
• Light on click 1, 24/0, from day one.
• Set inflow pH to 6.1–6.3 and leave it there.
• Mark the sow date on the pot.
• At emergence: step to click 2.
• At first true leaves: hold 200–300 PPFD.

Don’t

If this happens

→ Full diagnostics: is it feeding, light, or environment?

What happens next

• 30–43 hShell splits, white taproot pushes out. Underground — you won’t see it.
• Day 3–5A pale loop of stem arches up through the surface. Click 1 → 2 now.
• +2 daysCotyledons — the two round seed leaves — open and go green.
• Day 5–10First true serrated leaves. Hold click 2 (24/0).
• Day 10Nothing up yet? Suspicious, not fatal. Hold.
• Day 14Still nothing — dead. Sow a replacement.
• → NextWeeks 1–2 — seedling. First feed at EC 0.4–0.5.

Feed & water nowPre-soak the coco first — plain water at pH 5.5–5.8 until a little runs out the bottom. Coco arrives thirsty and will otherwise pull moisture straight back out of the seed. Damp, not soaked: a swamp at ½″ deep rots the seed before it can crack. Keep just the top inch evenly moist until it breaks ground. After that, Once it breaks ground: pump takes over — 9 sec hourly via halos — never the whole pot.

Sources: Geneve et al., Crops 2022 — cardinal germination temperatures · Šrajer Gajdošik et al. 2022 — photoperiod & lipid peroxidation · Moher et al. 2022 — light intensity & internode length · AC Infinity — germination stage · Coco For Cannabis — seedlings in coco · PNW Handbook — hemp damping-off · Purdue — DLI & transplant quality · Fluence — photobleaching · Zha et al. 2019 — continuous light is a DLI problem, not a clock problem · Temperature swing restores photosynthesis under continuous light · Sysoeva et al. 2010 — which species tolerate continuous light · Hermaphroditism in Cannabis sativa — documented inducers · Coco For Cannabis — pH by stage · Coco For Cannabis — fertigation frequency · UGA B1256 — pH and nutrient availability · Tyystjärvi & Aro 1996 — photodamage is proportional to photon dose · Elkins & van Iersel 2020 — photochemical efficiency falls as intensity rises · Elkins & van Iersel 2020 — equal daily light, longer photoperiod wins · Ikkonen et al. 2015 — a daily temperature drop prevents continuous-light injury · Avello et al. 2019 — a 7°F thermocycle resets the plant clock · Bok et al. 2023 — the only day/night temperature study on hemp · Rodriguez-Morrison et al. 2021 — cannabis yield linear to 1800 PPFD

At a glance

Do this

• Start fertigating 24 hours after the cotyledons open, at EC 400.
• Raise inflow EC 75–100 µS every single feed.
• Pump 9 sec hourly via halos.
• Light: click 4 @ 14″ = 550 PPFD.
• 2× S6 level 4 — one under the pots, one over, horizontal.
• pH the solution LAST, after every part is mixed in.
• Log runoff EC the moment you get runoff.
• Do nothing else.

Don’t

If this happens

→ Full diagnostics: is it feeding, light, or environment?

What happens next

• Day 5Cotyledons open 24 h. First feed: EC 400, pH 6.2–6.4. Click 1.
• Day 6–7First true leaf pair. EC ~500–600.
• Day 8–10Node 2 forming. EC ~700–850. Click 2. First real runoff usually appears.
• Day 10–12EC reaches ~1000 — full seedling strength.
• Day 12–14Node 3. Click 3. EC moving toward the 1100–1400 early-veg band.
• Advance when3 true nodes, consistent 10–20% runoff, roots reaching the pot wall.

Sources: Coco For Cannabis — seedlings in coco · Coco For Cannabis — EC targets · Coco For Cannabis — topping thresholds · Penn State Extension — damping-off · UC ANR — air movement and photosynthesis

At a glance

Running 24/0

24/0 spreads the dose, but the plan still climbs: 400–525 PPFD early veg → 630–735 late. Step one click or 2–3″ at a time, watch tops 48 h.

The stretch math — do this before you touch anything

Your tent is 72″. Subtract the fixture and hangers (~6″), the light-to-canopy gap you must maintain (18″), and the pot (~12″). That leaves 36″ of plant.

Sativa-leaning genetics multiply 2–3× in height after the flip. At 3×, 36 ÷ 3 = 12″ at flip. At 2.5×, 14″. At 2×, 18″. Lime Colada and Pineapple Lush are both sativa-leaning — plan for 2.5× and flip at 12–14 inches.

That number is your net height. Which is why the screen goes at 8–12″ above the pot: the net physically enforces the ceiling that the arithmetic demands.

Training — what is evidence and what is not

Topping has real data. In field hemp, topping above node 4 cut height 17–21% and raised dry biomass 25–29% against untopped controls. In medicinal cannabis, topping beat lollipopping and control on inflorescence dry weight — though the total cannabinoid yield difference was not statistically significant.

Plant density has real data, and it says something useful: 10 plants/m² at 28 days veg yielded 1009 g/m²; 18 plants/m² at only 10 days veg yielded 1091 g/m². Area yield tracks canopy fill, not per-plant size. With two plants in 8 ft² and a legal cap you can’t exceed, filling the canopy is the entire job.

SCROG versus LST versus untrained has no peer-reviewed comparison at all. Neither does tucking cadence, net height, percentage fill at flip, or “stop training N days before flip.” All of it is grower convention. It is sensible convention, and the density data supports the underlying logic, but nobody has measured it.

Verdict for this tent: top once at 4–5 nodes, then SCROG. The screen controls height better than LST because it is a hard physical ceiling, and height is your binding constraint. Net 8–12″ above the pot, flip when it is 70–80% full.

Do this

• Top once, above node 4–5.
• Install the trellis net around week 2 of veg.
• Feed to 10–20% runoff every single time.
• Measure runoff EC at every feed.
• Start Athena at 1.4–1.6 EC and climb toward 2.1.
• Hold VPD at 0.8–1.0 kPa.
• Stop all training 7–10 days before the flip.

Don’t

If this happens

→ Full diagnostics: is it feeding, light, or environment?

What happens next

• Wk 3 d1–2Top above node 4–5.
• Wk 3 d3–7Recover. Two new apical shoots. Begin bending the mains outward.
• Wk 4Net goes in 8–12″ above the pot. Start weaving. PPFD to ~500.
• Wk 5Net filling. ~600 PPFD. EC at the top of range.
• Wk 6Stop all training. Let wounds close for 7–10 days.
• Flip whenNet 70–80% full, canopy even, height at or under 12–14″, no open wounds.

Sources: Hemp topping trial — height and biomass · Plant density and area yield in cannabis · Llewellyn & Zheng — light intensity and yield · Coco For Cannabis — EC targets · Coco For Cannabis — watering coco · Athena — veg stage

At a glance

Three photoperiod myths, checked

“Step the light down gradually.” No evidence. Every Guelph photoperiod trial flips abruptly and gets flowering initiation in 8–11 days. Go straight from 24/0 to 12/12.

“A light leak will turn them hermie.” A 2024 study of 403 indoor plants used distance-to-door as a proxy for dark-period light and found it explained 1.6% of the variance in male flower formation — the authors advised growers not to spend resources sealing door leaks. Bugbee’s threshold is that cannabis cannot even perceive light below about five times full moonlight. The real risk from a leak is delayed flowering, not hermaphroditism — measured at roughly 6 days delay at 62 µmol of near-infrared, 12 days at 121.

“12/12 is optimal.” It is convention. Ten cultivars tested across 12 to 15 hours all flowered up to 14 h, and three initiated at 15 h. Peak floral biomass landed at 12.6–13.0 h depending on cultivar, and a follow-up found 13 hours yielded 35–50% more than 12 in two cultivars with THC unchanged or slightly higher. There is no universal critical daylength. If you want to test 13/11, run it on one plant, not both.

The stretch

Starts within 2–4 days of the flip, peaks days 7–14, essentially finished by day 21. What actually controls it, in order of evidence: light intensity (low PPFD produces etiolated stretch — measured, plants under low light were shorter and wider with thinner stems, high-light plants taller with denser flower), far-red (increases height in every genotype tested; your EVO3 has three 730nm diodes, always on, not separately switchable), temperature differential (flattening day/night restrains elongation), and training.

Honest gap: no peer-reviewed stretch multiplier by cultivar type exists. The 2–3× sativa figure is entirely breeder and vendor convention. The Guelph papers report “growth index,” not stretch ratios.

Sex identification — day 8 onward

Pistils (female): two white or cream hairs emerging from a teardrop-shaped calyx, at the nodes and apex. Pollen sacs (male): smooth round-to-oval balls in grape-like clusters, no hairs, on a short stalk. Nanners: elongated yellow or lime hooks protruding from inside a bud, often in bunches — these shed pollen without bursting, which makes them more urgent than an obvious sac.

Feminized seed keeps your risk low but not zero: a study of about 1000 plants across three cultivars found hermaphroditic anthers in 5–10%, appearing in weeks 4–7 of flower. The progeny were 100% female. Your real trigger is cold — exposure near 55°F raised male-flower incidence by roughly 26 percentage points. Hold the dark period at 68–72°F.

Do this

• Flip at 12″ or under if you can.
• Net is already in from veg — tuck every 2–3 days.
• Hold EC 2.1 through the flip; ramp toward 2.4 from day 8.
• Measure PPFD at canopy weekly and raise the light as the plants rise.
• Keep the dark period genuinely dark.
• Hold cal-mag at full veg dose through day 21.
• Log height daily for 21 days.

Don’t

If this happens

→ Full diagnostics: is it feeding, light, or environment?

What happens next

• Day 1First 12 h dark. ON 8 PM → OFF 8 AM. EC holds 2.1. PPFD 600.
• Day 2–4Stretch begins. Visible vertical growth. No defoliation.
• Day 5–7Fastest elongation. Tuck under the net. PPFD to 700.
• Day 8–10Pre-flowers appear. Sex-check every plant under a loupe. EC to 2.2.
• Day 11–14Peak height gain — 60–70% of total stretch is banked here. PPFD 800. Supercrop outliers now if headroom is tight.
• Day 15–18Decelerating. Calyxes stacking. Light interior defoliation now acceptable.
• Day 19–21Stretch done. Measure final height, confirm 10″ of clearance. EC 2.4–2.5.

Sources: Ahrens, Llewellyn & Zheng 2023 — there is no single critical daylength · Oliver 2024 — dark-period light and sex expression · Punja & Holmes — hermaphroditism rates and progeny · Rodriguez-Morrison et al. — light intensity and plant architecture · Coco For Cannabis — why coco needs cal-mag

At a glance

You are light-limited. This changes everything.

The definitive study ran flowering cannabis from 120 to 1800 PPFD at ambient CO₂ — 437 ppm, no enrichment. Dry inflorescence yield rose linearly the whole way, 116 to 519 g/m², a 4.5× increase with no plateau. The authors: even at ambient CO₂, photon availability was still limiting whole-canopy photosynthesis. A second trial confirmed it: +4.6 g per plant for every extra 100 PPFD.

This kills the common claim that you need CO₂ above 800–900 PPFD. You do not. And it means every photon you leave on the table below your fixture’s maximum is forfeited yield.

Where that puts you: at ~1040 PPFD you are at 58% of the tested ceiling. Reaching 1800 in a 2×4 would need roughly a 475–500W fixture. Your 280W physically cannot get there. That is fine — it just means light, not nutrition or genetics, is your binding constraint. Run click 10 every day of weeks 3 through 7 and stop looking for the problem elsewhere.

One thing more light does not buy you: potency. Measured across the full 120–1800 range, there were no intensity effects on cannabinoid concentration at all. Terpenes are the exception — total terpenes rose 1.4×, myrcene doubled, limonene up 1.5×. So more light means more grams and richer aroma at the same percentage THC.

Defoliation — the honest answer

The only controlled cannabis study tested eight architecture manipulations, including systematic fan-leaf removal and lollipopping, against untouched controls on two cultivars. Result: no reliable increase in total yield. The measurable benefit was cannabinoid standardisation — more uniform concentration top to bottom, from better light penetration.

Schwazzing has no published controlled data whatsoever. It is a commercial protocol promoted through a book and never validated in a trial. Every defoliation study in comparable crops — tomato, soybean, strawberry, sugar beet — runs neutral to negative, because removing leaves removes the source tissue that fills the flower.

For you specifically: you are photon-limited, not penetration-limited the way a 1000W grower is. Stripping leaves removes source capacity you cannot afford. Tuck rather than cut, and remove only leaves physically buried inside colas — and do that as a bud-rot intervention, which is a defensible reason, not as a yield play, which is not.

Do this

• Run click 10 every day, weeks 3 through 7.
• Buy the HYDRONE 5 ($139) at the flip.
• Step RH down 5 points per week from week 4.
• Keep the bottom S6 aimed under the canopy.
• Inspect the six densest colas daily from week 5.
• Keep cal-mag in every flower feed (Athena CaMg, low dose — your tap is Ca-rich).
• Verify the tent is at negative pressure.
• Lollipop the bottom 25% BEFORE week 3.

Don’t

If this happens

→ Full diagnostics: is it feeding, light, or environment?

What happens next

• Week 3Stretch ending, bracts stacking. EC 2.5, RH 55–60%, 80°F. Lock fixture height, go to click 10. Last chance to lollipop. Set up the under-canopy fan.
• Week 4Bud sites fatten and connect. Smell arrives — verify filter seal and negative pressure. EC 2.6 peak. RH to 50–55%. Start the daily cola inspection habit.
• Week 5Rapid bulking. EC 2.4. RH 50%, 78°F. Colas are now dense enough to hold their own humidity — bud rot risk begins in earnest here.
• Week 6Peak density. EC 2.3. Day temp down to 74–78°F, RH 45–50%. Lower fans yellowing — intended. Judge each plant separately; your two will not finish together.
• Week 7Ripening. EC to 1.7. RH 45%. Pistils receding. Start trichome checks with a 60× loupe. Expect 8–9 weeks total, so this is ripening, not harvest.

Sources: Rodriguez-Morrison, Llewellyn & Zheng 2021 — yield linear to 1800 PPFD · Llewellyn & Zheng 2022 — 600/800/1000 PPFD, no CO₂ · Danziger & Bernstein 2021 — the only controlled defoliation study · Ohio State Extension — Botrytis conditions · Punja 2023 — Botrytis in cannabis inflorescences · Athena Blended feed schedules

At a glance

The flush question — there is actual data, and it says no

A controlled trial ran 0, 7, 10 and 14 day flush periods, 12 plants per treatment, four replications, and measured yield, THC, terpenes, tissue minerals, extraction efficiency and a blind expert sensory panel on flavour, harshness and ash colour.

No significant differences in anything. Yield 97.3 g/plant average, THC 21.9% average, terpenes flat. No difference in flavour, smoothness or ash colour. Ash was black or grey across every treatment. And directionally, the 0-day flush scored best — 36% rated smooth — while the 14-day flush was rated harshest at 41.7%. Nitrogen in tissue was only 6.7% lower after fourteen days of plain water.

The physiology never made sense anyway. The claim assumes starving the plant makes it export minerals out of the buds. Remobilisation is real but it runs leaf to sink — and the buds are the sink. Nothing exits the plant. Calcium, boron and manganese barely move at all. That 6.7% nitrogen delta is the measured proof.

So what actually causes harsh smoke? Post-harvest handling. Drying too fast leaves green-leaf volatiles and coumarin in the flower — that is the hay smell, and it is a drying failure, not a nutrient one. Drying too slow or too wet gives you ammonia and mould. The smoke quality you are chasing is decided after the cut, not before it.

Reading trichomes properly

Clear → cloudy → amber. But the browning is oxidation of phenolic compounds in the gland head, accompanied by roughly eight-fold cuticle thickening — it is a maturation and oxidation marker, not literal THC converting to CBN. That common explanation is a simplification the trichome literature does not support.

Check mid-upper bracts, never sugar leaves. Sugar leaves amber first and will lie to you by a week. Maturity is asynchronous across the plant, with more stalked trichomes on the lower bract surfaces.

Pistils are more reliable than growers give them credit for. Across 25 genotypes over 119 days, total cannabinoid concentration peaked at mostly or fully amber stigmas in 22 of 25. Use pistils as the coarse signal and trichomes as the fine one.

“Amber equals couch-lock” is folklore. No controlled human-effects trial supports it, and the CBN mechanism people cite is far too slow — CBN accumulates on a multi-year timescale, not over a fortnight.

Every other sign, ranked by how much to trust it

What the “two weeks out” look changes in the bucket. Mostly cloudy, no amber yet, stigmas about half darkened: that is the batch to drop Balance and CaMg. The chart drops CaMg at week 8 — let the plant pick the week, not the chart. Everything else follows the chart as printed — Bloom steps down, PK steps up, Cleanse holds. Staged harvest: tops come off first; the lowers keep the normal mix until their heads say so.

Sources: Plants 2025 — stigma colour vs cannabinoid peak, 25 genotypes · J Cannabis Res 2023 — Punja, Sutton & Kim, trichome maturation · J Nat Prod 2016 — cannabinoid & terpene evolution by chemotype · Athena — when to harvest

Do this

• Taper to EC 1.5 for the final seven days.
• Drop RH to 45–50%.
• Scope trichomes daily from day −10.
• Prep the freeze dryer BEFORE you cut.
• Harvest the two strains separately.
• Keep Cleanse running to the end.
• Drop Balance and CaMg the batch after the “two weeks out” look.

Don’t

If this happens

→ Full diagnostics: is it feeding, light, or environment?

What happens next

• D−14 to −12Last full-strength feed. Final airflow defoliation. Begin daily trichome scoping.
• D−11 to −8Hold EC. RH toward 50%. Night temp 65–68°F. Watch for Botrytis daily.
• D−7Switch to the fade — EC 1.5. Fan leaves will start yellowing. That is the point.
• D−6 to −4RH 45–50%. Assess pistil stage. Freeze dryer ready.
• D−3Final irrigation at EC 1.5. Confirm: mostly cloudy, 5–15% amber.
• D−2 to −1Optional dark period. No feed.
• D−0Harvest. Cut just before 8 PM lights-on.

Sources: RX Green Technologies flushing trial — full report · Maillard et al. — nutrient remobilisation in plants · Livingston & Samuels — trichome development and gland chemistry · Plants 2025 — stigma stage vs cannabinoid content, 25 genotypes · Plants 2025 — drying method, decarboxylation and microbial load

At a glance

Wet trim vs dry trim — settled by the freeze dryer

The only real argument for dry trimming is that the leaf envelope buffers moisture loss and slows the dry. A freeze dryer replaces the slow-drying window entirely, so that argument evaporates. Wet-trimmed buds also pack more densely on trays, and leaving sugar leaves on complicates separation after the cycle.

Worth knowing: there is no measured terpene data comparing wet and dry trim in the peer-reviewed literature. None. Every claim on both sides is preference plus mechanism reasoning. Treat “dry trim preserves terpenes” as a hypothesis.

Freeze drying — and the finding that contradicts the marketing

Freeze drying works in three stages: freeze solid, then primary drying where vacuum sublimes ice directly to vapour, then secondary drying which pulls off bound water. Roughly three times faster than hanging.

Here is the part vendors do not tell you. A 2024 peer-reviewed volatilome study compared aggressive freeze drying against gentle tray drying and found freeze drying worse for terpenes — 45 compounds negatively affected, 28 of them terpenes, including myrcene, pinene, limonene and terpinolene. It concluded freeze drying “results in a loss of the predominant cannabis key, multi-property and aromatic compounds.” It did better preserve acid cannabinoids by avoiding decarboxylation.

The likely reconciliation: deep-vacuum, high-heat lyophilisation strips monoterpenes. Low shelf heat may not. That reconciliation is untested — but it is why you run Pharma manual control at 45–55°F shelf, not the 125°F food default. This single setting is the most consequential decision in your whole post-harvest.

Does freeze-dried flower still need a cure? Only if you run it to the beep. A Harvest Right cycle ends around 2–7% moisture, which is below 0.45 water activity — and 0.45 was measured to smoke more harshly with less terpene delivery than 0.65. Pull at ~12% instead — the jar settles at 58–62% RH, no pack needed. Boveda 62% = rescue only if you overshoot.

Don’t trust the machine’s “done” beep. Consumer units terminate on generic time, not on a moisture endpoint. Weigh the trays in and out — you want roughly 75–80% weight loss — then verify with a hygrometer in a sealed jar reading 58–62% after 12 hours.

Do this

• Cut at the end of the dark cycle.
• Work branch by branch, not whole plant.
• Wet trim straight onto the trays.
• Pre-freeze solid at −30°F before starting the cycle.
• Single layer on trays, below the rim.
• Weigh trays before and after.
• Keep the tent exhaust and carbon filter running through the entire trim.
• Save everything trichome-bearing.

Don’t

If this happens

Sources: Freeze drying vs tray drying — volatilome study · ASTM D8197 — water activity standard for cannabis · Water activity and smoking quality — sensory panel · Plants 2024 — controlled atmosphere vs open-air drying · Maryland cannabis home-grow rules

At a glance

Water activity — why the industry abandoned “moisture content”

Moisture content is total water by mass. Water activity is the free water available to microbes, on a 0–1 scale, equal to equilibrium RH divided by 100. Bound water counts toward moisture content but cannot feed mould — so moisture content does not predict spoilage and water activity does. That is why ASTM standardised it: 0.65 is the safety ceiling, 0.55 the quality floor.

And there is measured sensory data on where inside that band to sit. A study with 315 consumers and 38 trained judges compared 0.45, 0.65 and 0.85 aw. 0.65 delivered significantly more terpenes in the smoke than 0.45, was rated less irritating, produced lighter ash, and scored higher overall (54.2 versus 45.2). Sit at 0.60–0.62 — high side, under the 0.65 ceiling. Weight difference alone was 18 g per pound.

The hay smell is not chlorophyll

Chlorophyll is a large non-volatile molecule. You cannot smell it, and it does not leave the plant during cure. The actual chemistry is two things: green leaf volatiles — six-carbon aldehydes and alcohols like hexanal and hexenal, produced enzymatically from linolenic acid when tissue is wounded and senescing — and coumarin, the sweet dried-hay note released as cut tissue dries. Both are volatile and dissipate over time. That is what curing actually accomplishes on this axis. The chlorophyll story is folklore, repeated everywhere.

What curing does and does not do — measured

Terpenes only go down from harvest onward. Measured: after one week of dry and cure, myrcene dropped 55%, with pinene and limonene also falling. Meanwhile humulene appeared to double and germacrene rose 154% — but much of that “increase” is arithmetic. Monoterpenes are light and volatile and leave first; sesquiterpenes are heavy and stay, so their percentage rises because the denominator collapsed. Nothing is being created.

Cannabinoids are essentially flat through cure. Under 10% change across drying methods, no significant cure-driven shift. Curing does not increase potency. Any “stronger after cure” reading is water weight concentrating the percentage.

So what does cure genuinely do? Three things, all real: moisture equilibration across bud and stem to one uniform water activity; volatilisation of the green-leaf volatiles and coumarin, which is the hay smell leaving; and microbial stabilisation — measured yeast and mould counts held or fell through cure. What it does not do is “develop” terpenes or build cannabinoids.

CBN, for the record: the CBN-to-THC ratio reaches 2.5% at one year and 9.4% at three. It forms on a multi-year timescale. Nothing you do in a four-week cure moves it.

Aspergillus matters more than Botrytis

Botrytis ruins the bud. Aspergillus can hurt you. Several species are opportunistic human pathogens, the spores survive combustion, there are over 24 documented cases of cannabis-acquired aspergillosis including the only documented cannabis-related deaths, and a CDC analysis found cannabis users 3.5× more likely to acquire a fungal infection. Oregon flower failed Aspergillus testing at 8% in 2023.

It is frequently invisible. A clean-looking bud can be contaminated. This is the real reason to respect the 0.65 ceiling rather than eyeballing it — and the reason a jar with mould gets discarded, not re-dried, washed or extracted. Mycotoxins and spores survive all three.

Do this

• Buy a water activity meter, or at minimum a calibrated hygrometer in the jar.
• Target 0.58–0.62, not 0.55.
• Pull the freeze dryer short of its own “done” call.
• 62% pack = rescue only.
• Cure in glass, dark, filled about 75%.
• Store long-term in the fridge, vacuum-sealed or Mylar.

Don’t

If this happens

What happens next

• Freeze dry — D0Cut, wet trim, single layer on trays, buds not touching. Pre-freeze 1 h at −20 to −30°F.
• D0–D1Run at 45–55°F shelf, Pharma manual. Pull at ~12% moisture, short of the done call.
• D1–D10Jar. Settles 58–62% RH — 62% pack only if it reads low. Burp every 2–3 days.
• D10–D14Ready. No extended cure needed chemically — the jar time was equilibration.
• Machine down?Hang whole branches 60–68°F / 60→55% RH, 10–14 d, then jar.

Sources: ASTM D8197 — water activity standard · Water activity and smoking quality — sensory data · Plants 2025 — drying method, decarboxylation, microbial load · Plants 2024 — controlled atmosphere vs open air, cultivar variance · Green leaf volatiles — the real cut-grass chemistry · Medicinal Genomics — Aspergillus risk in cannabis · UNODC — cannabinoid degradation in storage