UJIAN LAMPU · THE LAMP TEST ·365 nm · 12 cm · f/8 · 1/60 s · iso 200

Jangan ambil kata kami. Pinjam lampu suluh. Don't take our word for it. Borrow a torch.

Produk ini lebih kecil daripada 10 nanometer. Anda takkan nampak. Tetapi ia berpendarfluor — dan itu boleh diuji di atas meja mesyuarat dengan lampu 365 nm yang anda pegang sendiri. The product is smaller than ten nanometres. Nobody will ever see it. But it is photoluminescent, and that can be performed on a meeting table with a 365 nm lamp you hold yourself, before any paperwork.

Beli mengikut hektarPINTU 01 · BUY BY THE HECTARE Kontrak satu penilaianPINTU 02 · CONTRACT AN EVALUATION

LAMPU MATI · LAMP OFF

A dull straw-yellow liquid in daylight.

SHOOT · matte black card ≥60 cm behind the subject · tripod · f/8 · manual WB from a grey patch · ColorChecker frame at the head of the session · a scale reference in frame · no props, no hands, no smoke, no reflectors, no bokeh, no gradient.

LAMPU HIDUP · LAMP ON

⟨emission maximum unmeasured⟩

THIS FRAME IS DELIBERATELY ACHROMATIC. Nothing we hold states QarboGrow's excitation or emission maxima, so no colour is rendered here. The correct response to an unknown accent is not a plausible cyan — it is a fluorimeter run, and until then the slot ships visibly empty. Half a day, one line of data.

365 nm · 12 cm · f/8 · 1/60 s · iso 200 · 2026-08-04 EXCITATION · LAMP DISTANCE · EXPOSURE · DATE

THE STAMP IS A REPRODUCTION RECIPE, NOT A RESULT. PUBLISHING THE CAPTURE CONDITIONS MEANS ANYBODY CAN REPEAT THE PHOTOGRAPH — YOU, A STATE AGRONOMIST, A RIVAL. IT ANSWERS THE ONLY QUESTION A TECHNICAL READER ACTUALLY ASKS ABOUT A GLOWING VIAL, AND IT ANSWERS IT BY INVITATION RATHER THAN BY ASSERTION. IT CONTAINS NO CLAIM, BECAUSE IT DESCRIBES CONDITIONS AND NEVER EFFECTS.

The lamp test proves that the product is what we say it is.
It proves nothing about yield.

And it is not a moat. Carbon dots are photoluminescent by definition, and Glaia Ltd — a University of Bristol spin-out — has sold a commercial carbon-dot foliar since 2024. We are a first mover on an image, never the only one, and any surface of ours that has said otherwise is being withdrawn. What the lamp shows you is identity, not performance. The number is below, on white, where numbers belong.

⟨emission — unmeasured⟩ THE SINGLE ACCENT OF THIS DIRECTION. IT IS THE SAMPLE'S OWN PHOTOLUMINESCENCE EMISSION MAXIMUM, AS A FLUORIMETER REPORTS IT. IT IS NOT A DESIGN DECISION AND WE ARE NOT MAKING IT.

60% FIELD · 30% PAPER · 10% EMISSION — AND TODAY THE SYSTEM RUNS 60 / 40 / 0, WITH THE ACCENT SLOT RESERVED AND LEGIBLY UNFILLED.

The accent is a client action, not a design decision.

UNTIL MEASUREDThe system runs achromatic on black. The 10% slot renders as paper or stays empty — and it is visibly empty.

THE CONSTRAINT IT MUST MEET≥ 4.5 : 1 on #000000 · relative luminance ≥ 18% (so it clears the 20–55% threshold band on the white side when the document mode inverts) · not within ΔE 15 of the consumer line's amber or violet, and not a category green.

IF IT FAILS THE CONSTRAINTThen the accent is paper, and this direction ships permanently achromatic on black. We publish whatever colour it actually is, even if it is an unglamorous dull blue — the alternative is colour-picking a brand from a lamp, which is the one thing this whole exercise refuses.

HOW TO CLOSE ITOne fluorimeter run at UPM ION2, or an afternoon on the client's own bench with the pen-lamp already in the supplied photograph. Half a day, one line of data.

WHAT WE WILL NOT DRAWThe mechanism. Not a spectral-conversion arrow, not an emission curve into the chlorophyll peaks, not a wavelength-in / wavelength-out diagram. This direction may photograph the emission. It may not diagram the mechanism, because no sourceable statement of that mechanism exists to draw from.

TIER D · DARK PLATE

brightness(1) contrast(1) saturate(1)
No grading — and the identity is declared in the code so its absence would be visible too. The emission colour is the datum. Grading it means choosing the brand colour in Lightroom. This tier is specified as a capture, not a filter.

TIER L · DAYLIGHT RECORD

saturate(.80) contrast(1.05) brightness(1.02) + the solid black foot band
20% desaturation stops a phone's auto-vivid field green fighting the black around it. The foot band is what lets a daylight photograph sit on the same page as a dark plate without either being graded toward the other.

TIER X · ARCHIVE

grayscale(1) contrast(1.12) brightness(1.04) + a SOURCE: line
Same ethic as the other two directions: monochrome means not ours. A reader can tell at a glance which images we made and which we cited, without reading a word.

TIER NIL · INSTRUMENT PLATE

COMMISSIONNEVER GENERATE

HRTEM / FESEM of the actual particle.

UPM ION2 runs FESEM at 1 nm and HRTEM as a paid service, and it is the founder's own institute. Ask first whether they already exist. A genuine micrograph is a booking, not a fabrication.

filter: none !important · no foot band on a data record
Deliberately empty. And one boundary specific to this direction: an instrument plate placed on black is not inverted, brightened or knocked out to match the page. If a white-ground spectrum looks wrong on a black spread, the spread changes, not the plate.

THE HAZARD UNIQUE TO THIS DIRECTION, AND IT IS SEVERE. A PHOTOLUMINESCENCE PHOTOGRAPH LOOKS LIKE A FALSE-COLOUR INSTRUMENT IMAGE. IF IT EVER TAKES INSTRUMENT GRAMMAR — A SCALE BAR, AN N =, A PANEL LETTER, AN FFT INSET — IT BECOMES A FABRICATED INSTRUMENT READING, WHICH IS THE ONE THING THIS BRAND CANNOT SURVIVE. A VIAL PHOTOGRAPH IS A PHOTOGRAPH. IT GETS A PHOTOGRAPH'S CAPTION: SUBJECT · CONDITION · DATE. IT GETS INSTRUMENT GRAMMAR ONLY WHEN AN ACTUAL FLUORIMETER TRACE IS PUBLISHED BESIDE IT, AND THEN THE GRAMMAR BELONGS TO THE TRACE, NOT TO THE PHOTOGRAPH.

THE TARGET IN TIERS D, L AND X IS A SYNTHETIC PATCH SET, NOT A PHOTOGRAPH. IT IS IDENTICAL IN ALL THREE SO THE FILTERS ARE VISIBLE RATHER THAN DESCRIBED. THE INSTRUMENT FRAME CARRIES NO IMAGE AT ALL.

SINI HALAMAN BERTUKAR PUTIH · THE PAGE TURNS WHITE HERE — because the numbers cannot live on black. dim at 25.4% and mist at 45.6% both sit inside the 20–55% photocopier dead band, so a scanned annexe resolves them either way, unpredictably, page to page. Below this line the site borrows Direction 1's document spine whole: ink / graphite / axis / rule / wash / paper. That is an architecture, not a tweak, and a client who picks this direction is buying two systems.

1.0 — NOMBOR YANG JUJUR / THE HONEST NUMBER

Purata, julat, n, tanaman, tarikh. Mean, range, n, crop, date.

Every performance number here is drawn as a range with its mean marked. No single value, no “up to”, and no counting animation — a number that performs is a number selling itself.

11.57% PURATA · MEAN
n = 300+ PETANI / FARMERS
500 ha · MALAYSIA + INDONESIA · 2024
PUNCAK / PEAK 45%

Average +11.57% across 300+ farmers on 500 ha in Malaysia and Indonesia, 2024 — Qarbotech's own pilot data; peak result 45%. Results vary by crop, variety and conditions.

Purata +11.57% merentas 300+ petani di 500 hektar, Malaysia dan Indonesia, 2024 — data perintis Qarbotech sendiri; keputusan puncak 45%. Keputusan berbeza mengikut tanaman, varieti dan keadaan.

MEASURED · QARBOTECH 2025 MIT SOLVE APPLICATION, 2024 RICE PILOT.
OUR OWN SURFACES HAVE CARRIED “UP TO 60%”. THAT IS THE PEAK OF A RANGE WITH THE RANGE DISCARDED — 5.19× THIS AVERAGE — AND IT IS BEING WITHDRAWN.

In our chilli trial, measured on a LI-COR LI-6800: CO₂ assimilation +21%, stomatal conductance +33.9%, intrinsic water-use efficiency +14.7% against a named control plot (Braintree Technologies, Selangor, Jul 2024 – Mar 2026). The farm also ran JADAM practices concurrently; the study does not isolate the two.

2.0 — DUA PINTU / THE TWO DOORS

Beli, atau kontrak satu penilaian dahulu. Buy, or contract an evaluation first.

The torch is a demonstration, not a transaction. These are the two transactions, and most buyers should take the second one first.

Beli mengikut hektarBuy by the hectare

Sold by the hectare into a programme, not by the bottle into a shop. What exists on the public record is a pass schedule and a per-hectare product price. What does not exist is set in angle brackets, at the size of the content it is holding, so the hole is the right shape.

UNIT OF SALEhektar / hectare

PASSES4 · d40 / d50 / d60 / d70

PRODUCT PRICERM320/ha · on record

PACK SIZE⟨pack size pending⟩

PER-LITRE PRICE⟨price pending⟩

CONCENTRATION⟨dose pending⟩

SPRAY VOLUME⟨volume pending⟩

PHI⟨PHI pending⟩

SABAHnot listed, not shipped

Kontrak satu penilaianContract an evaluation

A funded trial in your own operating environment: an untreated control strip inside the same block, a named agronomist, a written protocol, and the harvest date declared in advance, before anybody knows the answer. The data is yours whichever way it goes.

We hold MySTI certificate STI202300224. It is an innovation-directory certificate — not a product registration, and not an approval — and we write it that way every time.

OUTPUTa dossier

CONTROL STRIPadjacent, in-block

AGRONOMISTnamed

HARVEST DATEdeclared in advance

DATA OWNERSHIPyours

MySTISTI202300224 · held

INCORPORATION1291227-W · held

ePEROLEHAN⟨registration pending⟩

SDS⟨SDS pending⟩

⟨ ⟩ ANGLE BRACKETS ARE THIS DIRECTION'S MARKED-PLACEHOLDER MARK, AND IT OWNS THEM HARDEST BECAUSE ITS OWN BRAND ACCENT IS ONE. PACK SIZE, PRICE, CONCENTRATION, SPRAY VOLUME AND PRE-HARVEST INTERVAL ARE ALL UNPUBLISHED. THEY APPEAR AS VISIBLE, SIZED HOLES — NEVER AS INVENTED NUMBERS, NEVER AS SILENCE. NO TANK-MIX TRIAL EXISTS ANYWHERE, SO NOTHING IS STATED ABOUT COMPATIBILITY WITH ANY FERTILISER OR PESTICIDE PROGRAMME, IN ANY WORDING.

3.0 — REKOD PROGRAM / THE PROGRAMME RECORD

Ini fakta penggunaan. Ia bukan keputusan. This is a fact of use. It is not a result.

A state department ran a programme with us, at scale, and it is on the public record. That is evidence it was used. It is not evidence that it worked, and this page does not let the two blur.

2,000hektar · hectares
Seberang Perai U/T/SMEASURED

1,319pesawah · paddy farmersMEASURED

RM320sehektar, harga produk
per hectare, product priceMEASURED

RM1.622juta · million
peruntukan projek · project allocationMEASURED

4semburan · passes
hari 40/50/60/70MEASURED

SUMBER / SOURCE

Jabatan Pertanian Pulau Pinang, in kerjasama strategik with Qarbotech Sdn Bhd. Launched 1 October 2025. Reported by BERNAMA in Bahasa Malaysia and in English. MySTI STI202300224.

RM320/ha is the price of the product. RM1.622 million is the total project allocation — everything, not product. We do not multiply RM320 by 2,000 and call it a contract value; we do not divide RM1.622 million by 2,000 and call it a product price; and we do not add the 1,319 Penang farmers to the roughly 1,200 reported across three states, because those two figures are not additive.

4.0 — SKOP / SCOPE

Apa yang kami tidak tuntut. What we are not claiming.

Every competitor prints sesuai untuk semua jenis tanaman. We do not. The blank cell below is a decision, and it renders before anything else on the page.

JADUAL 4.1 — SKOP TERISYTIHAR / DECLARED SCOPE · QB-DOC-0021 REV B · 2026-08-04
KumpulanGroup	KedudukanPosition	AsasBasis	PenentuanDetermination
C3	KuatStrong	Rubisco is substrate-limited at ambient CO₂, so extra activity has headroom to convert into fixed carbon. This is where our commercial data sits.	det. ⟨name⟩ 2026-08-04 MEASURED
C4	TIADA TUNTUTAN WE ARE NOT MAKING A CLAIM	Our own position is unresolved. We will publish it the day it is settled and we will date it. The cell renders blank with the page, at full opacity, first, and it never animates — a cell that fades in reads as loading, and this one reads as decided.	det. pending — UNRESOLVED
Pokok berkanopi besar Large-canopy tree crops	Belum terbuktiUnproven	Delivery to a tall canopy is solved in the published literature. The blockers are ours: a multi-year lag between a source-side intervention and a harvestable sink, and no residue or pre-harvest-interval statement published anywhere.	det. ⟨name⟩ 2026-08-04 DERIVED
Akuatik Aquatic	TidakNo	Submerged. There is no leaf in air to spray. A physical kill, and nothing reverses it short of different hardware.	det. ⟨name⟩ 2026-08-04 MEASURED

CAVEAT · WHAT THIS PRODUCT IS NOT

Ini bukan pengganti baja. Kau tetap membaja macam biasa. QarboGrow is not a replacement for fertiliser — use it alongside the programme you already run. It carries no nitrogen, no phosphorus and no potassium. It is not a pesticide, and we make no disease, pest or virus claim of any kind, in any language, on any surface.

Mobile — and the surface where this direction is weakest

The two-mode page holds up on a phone: black above the fold, white the moment a number appears. The lamp still crossfades on tap at 250 ms, both states from the same frame, same crop, same exposure.

The forward test is where it breaks, and we say so. A dark photograph re-encoded through a WhatsApp forward becomes a black square with a smear across it. So the forwardable object is not the photograph — it is a white card carrying the condition line and the attestation block, which survives compression because it is line and type. That fallback is shown below the device, and it is part of the system rather than an apology for it.

FORWARD FALLBACK · WHITE CARD, LINE AND TYPE ONLY QARBOGROW · PENGGALAK FOTOSINTESIS
365 nm · 12 cm · f/8 · 1/60 s · iso 200
PURATA / MEAN 11.57% · PUNCAK / PEAK 45%
n = 300+ · 500 ha · MY+ID · 2024
QARBOTECH SDN BHD · 1291227-W

UJIAN LAMPU / LAMP TEST372 × 690

QARBOTECH · 365 nm · 12 cm · f/8

Jangan ambil kata kami. Pinjam lampu suluh.

Don't take our word for it. Borrow a torch.

LAMPU MATI · LAMP OFF

Tap to switch state

LAMPU HIDUP · LAMP ON

⟨emission maximum
unmeasured⟩

CROSSFADE 250 ms · BOTH STATES FROM THE SAME FRAME, SAME CROP, SAME EXPOSURE. NO GLOW, NO BLOOM, NO SPEED CHANGE ON THE TRANSITION.

Beli mengikut hektar PINTU 01 · BUY BY THE HECTARE Kontrak satu penilaian PINTU 02 · CONTRACT AN EVALUATION

HALAMAN BERTUKAR PUTIH · THE PAGE TURNS WHITE

11.57% PURATA / MEAN · n = 300+ · 500 ha · 2024 · PUNCAK / PEAK 45%

C3	Kuat · Strong
C4	TIADA TUNTUTAN · NO CLAIM
Pokok berkanopi besar	Belum terbukti
Akuatik	Tidak · No