What each micron screen actually holds back
There is no chemistry in the bucket. Ice water makes the resin glands brittle, agitation knocks them loose, and a series of screens sorts them by one property only: size.
That is the whole process. What makes it interesting is not its complexity — it has none — but the fact that a number in microns has become a quality argument in the trade when all it measures is a diameter.
This page describes what actually happens at each step, what each fraction contains, and why the figure printed on a bag says nothing about the contents.
⚠️ Scope. This page describes a process and what can be verified from documents. It does not describe what a product does to a person, gives no amounts, and does not replace a healthcare professional.
Why ice water works
Cannabinoids do not dissolve in it
They are lipophilic. Water does not dissolve them, and that is precisely the condition that makes the process possible at all.
Cold makes the glands brittle
At low temperature the trichome wall hardens. Moderate agitation is then enough to break them off the plant.
The plant material stays whole
Wet leaves do not crumble. They remain in large pieces that the first screen holds back without difficulty.
What this implies
Water extracts nothing. It serves as a mechanical medium, and the sorting that follows is purely dimensional.
What each screen does
The work screen
Around 220 microns. It holds back the wet plant material and lets everything smaller pass straight through.
The middle meshes
Roughly 160 down to 90 microns. This is where the whole gland heads sit, and it is the most sought-after range.
The fine meshes
From 73 down to 45 microns. They hold smaller glands and fragments of broken heads.
The bottom screen
25 microns or less. It collects the finest debris and a large share of the impurities, and is usually sold on for further processing rather than as-is.
What the micron figure actually measures
A diameter, nothing else
The mesh lets through what is smaller and holds back what is larger. It is a geometric operation.
Not a potency
Two fractions of the same calibre can differ substantially in content depending on the starting plant.
Not a purity
An 80-micron fragment of leaf passes exactly like an 80-micron gland. The screen does not distinguish the nature of what it sorts.
Not a quality
A fine fraction from a poor harvest remains a poor fraction. Calibre corrects nothing.
The fractions and what they contain
The 160–120 fraction
Large whole heads, often the most abundant. It forms the core of a normal yield.
The 120–90 fraction
Considered the best in the trade. The heads are whole and the proportion of foreign matter is low.
The 90–73 fraction
Still mostly glands, with more fragments. The ratio of quality to quantity is often best here.
The fraction below 73
Plenty of debris and fine plant particles. It sells for less and is often destined for further processing.
Why “full melt” is not a label
What the term describes
Material that melts completely under heat without leaving residue. It is an observation, not a measurement.
What produces that behaviour
A high proportion of whole gland heads and a low proportion of plant matter.
Why it is not verifiable
No standard defines the test, no temperature is fixed, and no body oversees the use of the term.
What it is worth anyway
It describes an expectation you can hold the material against. Useful as description, worthless as proof.
What the process does not do
It does not remove heavy metals
They are in the plant and have no reason to leave it. Sorting by size does not concern them.
It does not remove pesticides
Some are fat-soluble and concentrate along with the resin. The result can be more loaded than the starting flower.
It does not separate cannabinoids from each other
CBD and THC travel together in the same glands. No screen tells them apart.
It does not improve a bad harvest
Like every separation process, it is capped by the raw material. It concentrates what already exists.
Why concentration raises a legal question
The enrichment factor
If ten parts of flower give one part of material, everything in the resin ends up multiplied by ten.
What that changes
A compliant flower can yield a concentrate that is not. Compliance does not travel with the material; content does.
The correct sum
Delta-9-THC plus THCA multiplied by 0.877. The factor comes from THCA losing a CO₂ molecule on heating and therefore becoming lighter.
Where to read the rules
On the page covering the legal position in the UK, with the corresponding official sources.
What the certificate must show
The four cannabinoid values
CBD and CBDA separately, THC and THCA separately. Only then can the correct sum be calculated.
The contaminants
Heavy metals and pesticides, with regulatory limit and measured value. On a concentrate, this is the most important line.
No solvent panel required
The process uses none. It is the one point where this production route genuinely simplifies verification.
The batch number
The link between document and goods. Without it the certificate describes some batch or other.
Drying, the step that decides everything
Why it is critical
The material comes out wet. Any residual moisture leads to mould, and that is the only real defect of this process.
How it is done
Spread thin, at low temperature, away from light. It takes days, not hours.
What happens when it is rushed
Excessive heat partly melts the glands and degrades the aroma. Haste costs more than it saves.
How to spot it
By smell. Material that smells musty or sour was sealed too early, and no appearance makes up for it.
How to read a commercial listing
The stated calibre
Useful if given, insufficient on its own. It describes a geometry.
The number of washes
A first extraction and a second do not give the same material. The second is more loaded with debris.
The production date
Resin hardens and loses aroma over months. This is often the most telling indication.
The certificate
The only document that speaks about contents. Everything else describes a container.
What distinguishes good execution
The water temperature
As low as possible without freezing. Every degree above detaches less cleanly.
The agitation time
Short and steady. Prolonged agitation eventually tears off plant material.
The number of screens used
More meshes allow finer sorting. It is an investment in time and equipment.
The cleanliness of the gear
Residues from a previous run mix into the next. A workshop detail with real consequences.
Why yield varies so much
The raw material decides first
A resin-rich flower gives a multiple of a poor one. No process compensates for that gap.
The harvest timing
Too early and the resin is missing; too late and it is partly degraded. The window is narrow.
The fineness of the sorting
Keeping only the best fractions mechanically reduces the weight obtained. That is a choice, not a failure.
The operator’s skill
Between an experienced operator and a beginner, with the same equipment, the gap is clear. It is the hardest variable to quantify.
What stays in the bucket
The spent plant material
It still holds residual resin. A second run recovers part of it, with a poorer result.
The wash water
It carries very fine particles. In careful workshops it is filtered one last time.
What that says about yield
Part of what the buyer pays for sits in these side streams. Yield is the exact reverse of it.
Why it concerns the buyer
Because the price of a gram reflects everything that was not recovered. That is arithmetic, not an opinion.
Comparison with dry sifting
What they share
Both sort by size and use no solvent. Neither adds anything whatsoever.
The decisive difference
Dry, the leaf shatters and its fragments reach the calibre of gland heads. Wet, it stays whole.
What follows from that
Wet separation reaches a higher purity. That is a mechanical consequence, not a matter of care.
What dry preserves better
Aroma, because there is neither water nor prolonged drying. Each route pays for something.
Where the process came from
A recent origin
Ice-water separation with mesh bags spread in the 1990s. It is a modern process, despite its handcrafted appearance.
What it replaced
Dry sifting, practised for centuries. The novelty was not the idea of sorting, but of using cold and water as aids.
Why it prevailed
Because it reaches a purity dry sifting cannot. The gap is structural, not a matter of care.
What has not changed
The principle is still dimensional sorting. All the sophistication concerns how glands are detached, never how they are told apart.
The equipment and what it costs
The mesh bags
A set of five to eight nested bags, from widest to finest mesh. This is the main purchase and it lasts years.
The bucket and the agitator
A simple vessel and something to stir with. At small scale, hand agitation is enough.
The ice
This is the real consumable, and in quantity. It has to keep the temperature down for the whole wash.
What that means
The equipment investment is low. The cost of bubble hash lies in yield and time, not in gear.
The exact role of the water
Its temperature
As close to zero as possible without freezing the mixture. It is the variable that decides how cleanly the separation runs.
Its quality
Water heavy in minerals leaves traces on drying. Careful workshops use filtered or demineralised water.
Its quantity
Enough for the material to move freely. Too dense a mixture agitates badly and detaches less.
What it never does
Dissolve anything useful. If water extracted cannabinoids, the entire process would collapse.
Dry flower or fresh frozen
The difference at the start
A plant frozen immediately after harvest has never dried. Its glands are more fragile and come off more easily.
What that changes in the result
An aroma profile closer to the living plant, because the volatile compounds had no time to evaporate.
What it costs
A cold chain from the field onward. That is a real logistical constraint and it is paid for.
What no analysis shows
The cold chain leaves no measurable trace. It is a claim resting entirely on the producer’s word.
Why whole glands are worth more
They protect their contents
An intact head keeps its volatile compounds enclosed. A broken head exposes them to air from the moment it cracks.
They age more slowly
Less exposed surface means slower oxidation. The same batch keeps better if it is made of whole heads.
They contain less foreign matter
A whole gland is almost purely resin. A fragment often comes with stalk debris attached.
What that justifies in price
A real difference, provided the fraction is named. A premium for pale colour with no stated fraction rests on nothing.
What the second wash produces
Extra yield
Already-processed material retains glands. A second run recovers some of them.
A different composition
What comes out second is more loaded with plant fragments. The most accessible glands are already gone.
A lower price
That fraction sells for less, and rightly so. The problem is not that it exists, but that it is sold as first-wash.
How to spot it
Under a microscope, by the proportion of debris. This is exactly the case where a pocket instrument pays for itself.
The traps of visual assessment
Colour
It depends on the cleanliness of the sort, the leaf content, heat and age. Four causes for a single observation.
Texture
It reports on pressure and ageing, never on potency. It comes last in the order of checks.
The product photo
Lighting and retouching change the rendering completely. Comparing two photos compares nothing.
What does work
Smell, then the certificate. In that order, and not the reverse.
The microscope, the check that settles it
What it shows
Translucent spheres are glands. Irregular green shards are plant matter. The difference is immediate and unambiguous.
Why it beats the naked eye
Because colour and texture each have several possible causes. The shape of the particles has only one.
What it costs
A pocket instrument for a few pounds. It is the best-value verification in the whole category.
What it does not show
Potency and contaminants. Like any observation, it stops where the laboratory begins.
How to store the material
Cold
The refrigerator slows oxidation. The freezer creates a condensation problem on opening.
Away from light
Light drives oxidation and turns the material pale, then brown. An opaque container is enough.
Airtight
The oxygen in the container is the reactant you want to limit. Less air is better than more.
Rarely opened
Every opening brings fresh air and a temperature jump. Several small containers beat one large one.
Why the vocabulary is English everywhere
The route of diffusion
The technique spread through English-language publications. The vocabulary followed the process.
What confusion that creates
Untranslated terms are perceived as technical and therefore reliable. “Full melt” sounds like a standard and is not one.
What is actually defined
The micron, because it is an SI unit. Everything else is commercial language.
What to take from that
A technical-sounding word adds no guarantee. Only the figure with its unit means something.
The most common mistakes
Agitating too long
Beyond a certain point, it is leaf fragments coming loose. Yield rises and purity falls.
Letting the water warm up
Water climbing in temperature detaches less cleanly. It is the most common error on large runs.
Overloading the bucket
An overfilled bucket circulates badly. Part of the load is never properly worked.
Drying too fast
This is the only irreversible mistake. The others cost yield; this one ruins the batch.
What to ask a seller
The fraction in microns
A range, not a single value. A fraction is defined by two consecutive meshes.
The number of washes
First or second. The difference is clear and the information is easy to give.
The production date
It decides what is left of the volatile compounds. The most telling of the four.
The full panel
Cannabinoids and contaminants, with batch number. A solvent panel would be beside the point here.
How to compare two offers
Reduce to the milligram
Price divided by total cannabinoid quantity. It is the only operation that makes two products comparable.
Compare equivalent fractions
A fine fraction and a wide one are not the same product. Comparing them by weight is meaningless.
Account for age
Recent material and last year’s are not equal at the same potency. Aroma is not in the certificate.
What actually settles it
The batch certificate, every time. The other three criteria orient; that one decides.
Why this process holds up well to verification
It uses no solvent
It is the only family of concentrates where the solvent panel is beside the point. One analysis line fewer, and one worry fewer.
Its steps are few
Chill, agitate, screen, dry. A short sequence leaves little room for grey areas.
Its technical vocabulary is numerical
The micron is an SI unit. That is rare in this category, where most terms are commercial.
What remains unverifiable regardless
The stated fraction, the number of washes and the cold chain. Three claims resting entirely on the producer’s word.
What the price per gram measures here
The fraction
Which screen the material came from. It is the figure usually sitting behind a pale colour, and it explains price better than appearance does.
The number of washes
First-wash material commands a premium over second. The difference is real and easy to state.
The freshness
Recent material sells higher because volatile compounds are still present. You are paying for the seller’s turnover speed.
What it does not measure
Potency. Only dividing price by total milligrams makes two offers comparable, and for that you need the certificate.
A short glossary
Trichome
The gland at the tip of a stalk where cannabinoids are produced. It is the head the process aims to recover intact, because whole it keeps its volatile compounds enclosed and ages more slowly.
Micron
One thousandth of a millimetre. The unit expressing mesh diameter, and the only numerical, standardised quantity in the entire vocabulary of this category.
Fraction
What is recovered between two consecutive meshes. Each fraction has its own composition and price, and is always defined by a range — never a single value.
Full melt
A commercial term for material that melts entirely without residue. No temperature is fixed, no test is defined, and no body oversees its use.
Cold chain
Keeping temperature low from harvest to processing. It justifies a price premium and leaves no measurable trace in the finished product.
Batch
One production run. The certificate applies to the batch, never to the product name or a stated calibre, and two batches from the same workshop can differ noticeably.
The five checks that apply to any product
The batch certificate
Request it with the number printed on the packaging. A general document for the product line does not cover this batch, and those differences are exactly what it should record.
Total THC
Delta-9-THC plus THCA multiplied by 0.877, because THCA becomes THC under heat and the limit applies to the sum, not to either value alone.
Price per milligram
A division, not an opinion. The only figure that makes two products comparable, whatever the format, process and pack size.
The claims
Without an authorisation number in the EU Register these are assertions, not data — and for cannabidiol that register contains no authorised entry to this day.
The seller’s details
Company name, address and contact. Without them there is also no counterparty to turn to if something turns out to be wrong.
How this connects to the rest
With dry sifting
Both processes sort by calibre. The difference lies in the state of the leaf, not in the principle.
With pressing
Pressing comes after separation. It cannot improve what the screen did not deliver.
With colour
It partly depends on how cleanly the sort was done, so indirectly on the process. It says nothing about potency.
With the certificate
Everything ends there: batch number, date, cannabinoids, contaminants. Calibre appears on no analysis report.
What this is for in practice
Facing a listing
Read the calibre as a geometric indication and ask for the date, the number of washes and the certificate.
Facing a price gap
A fine fraction costs more because less of it comes out. That is justified, provided the fraction is named.
If you simply want to understand
That sorting happens by size, and that size says nothing about the nature of what is sorted.
What this page publishes
Processes, verifiable documents and arithmetic. No promises of effect — for that, the EU Register contains no authorised entry.
Frequently asked questions
Does a lower micron figure mean better quality? No. It indicates a mesh diameter. A gland and a leaf fragment of the same size pass in exactly the same way, and the finest fraction contains mostly debris.
What does “full melt” mean? That the material melts entirely under heat without leaving residue. It is an observation, not a standard: no temperature is fixed and no body oversees the use of the term.
Does the process remove contaminants? No, it concentrates them. Heavy metals remain and some pesticides, being fat-soluble, follow the resin. The contaminant panel is therefore essential.
Why does drying matter so much? Because the material comes out wet, and any residual moisture leads to mould. It is the only real defect of the process, and smell reveals it immediately.
Is a solvent panel needed? No, no solvent is involved. It is the one point where this production route genuinely simplifies verification — contaminants, however, remain mandatory.
How do I recognise a second wash? Under a microscope, by the proportion of debris. What comes out second is heavier in fragments, the most accessible glands having already gone. The lower price is legitimate; selling it as first-wash is not.
Can I use product photos to compare colour? No. Lighting, retouching and screen all change the rendering. Colour in a photograph is a design decision, not information about the goods.
Fresh frozen or dried: what is the difference? A plant frozen right after harvest keeps an aroma profile closer to the living plant. But the cold chain leaves no measurable trace — it is an unverifiable claim.
What we check and what we do not
We check what can be checked from documents: official registers, declared contents, certificates of analysis, divisions. We do not test products, do not produce rankings and do not claim any effect.
On this subject the boundary is sharp: a screen calibre appears on no analysis report, and the term “full melt” corresponds to no defined test. What remains verifiable is the batch certificate with its four cannabinoid values and its contaminant panel.
That is precisely why this page explains what the figure measures — a diameter — and what it does not. A screen sorts by size, and size says nothing about the nature of what is sorted.