Making air visible
A beam of light is invisible. You see the lamp, and you see what it lands on, and in clean air there is nothing in between. Every shaft of light the industry designs around is not light at all — it is a few grams of suspended particles, each one scattering a little of the beam sideways into somebody's eye. Almost everything else on this page follows from how big those particles are.
The one number
Particle size decides all of it
Light scatters off a particle in a way that depends on how the particle’s size compares with the wavelength of the light. Visible light runs roughly 0.4 to 0.7 µm, and the particles we put in the air to catch it run from about 0.1 to 10 µm — which places atmospheric effects squarely in the Mie scattering regime, the awkward middle where the maths has no shortcut and the behaviour changes as the droplets grow.
That is not trivia. It is the whole design space. Small particles scatter short wavelengths more strongly than long ones, which is why a well-hazed room has a faint cool cast and why haze reveals a beam without greying out the scenery behind it. Large droplets attenuate every visible wavelength about equally, which is why fog reads neutral white and why you stop being able to see through it.
So haze and fog are not two settings of the same idea. They are two different optical regimes, and choosing between them is choosing whether the audience looks at the air or through it.
Haze
under ~1 µmBeams become visible; the room does not. Hangs for hours and spreads evenly, which is why it is laid in before the house opens rather than cued.
Fog
a few µmA visible cloud with a shape and an edge. Reads as an event, obscures what is behind it, and dissipates on a timescale you can cue.
Smoke
combustion particulateSolid products of burning rather than a condensed fluid. Behaves differently, stains, smells, and is treated as a different hazard class.
What is actually in it
Glycol, glycerin and mineral oil
Most of what the industry calls fog is a water-based fluid: a glycol or glycerin dissolved in water, pushed through a heat exchanger, flashed to vapour and condensed into droplets as it hits room air. Change the glycol, the ratio or the block temperature and you change the droplet size and how long it hangs — which is why fluids are not interchangeable between machines and why using the wrong one is both an effect problem and a warranty problem.
White mineral oil hazers work differently: the oil is atomised mechanically rather than heated, so the particles are very small and very consistent, and the output hangs for a long time. It is the classic film and concert haze, and it is a different substance with a different exposure profile — which is exactly why the standard names it separately.
Cracked oil is the older name for the same mechanical principle. It earned a poor reputation from machines that spat rather than atomised, and left a film on everything downstage of them.
| Effect | How it is made | Hangs for | What it is for |
|---|---|---|---|
| Haze | Glycol/water heated, or mineral oil atomised | hours | Beam visibility across a whole space |
| Fog | Glycol/water, higher output, larger droplets | minutes | A cloud the audience notices as an event |
| Low fog | Fog chilled below the temperature of the room | 1–5 min | A layer that stays on the deck |
| Dry ice | Solid CO2 into hot water | 1–3 min | Dense white low fog, no machine in sight |
| Cryo jet | Liquid CO2 released through a nozzle | seconds | A sharp white plume, usually on a beat |
| Coloured smoke | A pyrotechnic composition subliming a dye | varies | A coloured plume outdoors; it is pyro, not fog |
Why it stays on the floor
Low fog is a temperature trick, not a different fluid
Fog hangs at eye level because its droplets are close to the density of the air around them. Cool that same fog and two things happen at once: the air carrying it contracts and becomes denser than the room, and the droplets stop evaporating as fast. The result sinks, spreads and sits.
Every low-fog method is some version of that. A chiller passes ordinary fog over ice or a refrigerated coil. Dry ice drops solid CO2 at −78 °C into hot water, which does the chilling and the fog-making in one gesture. Liquid nitrogen does it harder and colder.
And all of them end the same way: the layer warms to room temperature, becomes neutrally buoyant again, and rises as ordinary fog. Low fog does not disperse where it lies. It gets up and walks into the beams twenty seconds later, which is a lighting problem nobody cues for the first time they use it.
CO2 displaces air, and it pools. Dry ice and cryo effects put carbon dioxide into the room, and it collects in exactly the places the effect is designed to fill: low, still, enclosed. Orchestra pits, traps, under-stage voids and basement corridors are the hazard, and a person lying down in it — a performer playing dead, somebody who has fainted — is in the worst part of it.
This is a ventilation calculation and a monitoring decision, not a thing to judge by eye. Carbon dioxide is colourless and the fog you can see is not the gas you cannot.
The detector problem
Why the fire panel is part of the design
The same property that makes atmospherics work — suspended particles scattering light — is the property a smoke detector is built to notice. A photoelectric detector looks for light scattered into a chamber by particles, which is precisely what haze does. An ionisation detector senses particles disturbing a small current, and reacts to very fine particles particularly well. A beam detector across a large space measures obscuration, which is what fog produces by definition.
So this is never a question of whether the system might trigger. It is a question of which zones are isolated, by whom, under what written permission, and who is watching the space while the detection that normally watches it is switched off. That last part is the actual safety measure: isolation removes a protection, and something has to replace it.
Get it wrong and the cost is not embarrassment. It is an evacuation mid-show, a brigade attendance, a charge for the call-out, and a venue that will not let you haze again.
The health question
What the standards actually say, and what they do not
Two ESTA standards carry this. ANSI E1.23, now titled Design, Execution, and Maintenance of Atmospheric Effects, covers glycol, glycerin and white mineral oil effects, and requires that exposure be monitored rather than estimated. ANSI E1.5 sets out acceptable components and the exposure limits themselves for glycol and glycerin fogs.
The limit worth knowing the shape of: for glycerin, E1.5 has carried a figure of 10 mg/m³ as a time-weighted average and 50 mg/m³ as a peak. Figures for the other fluids are in the standard and are not reproduced here, because the values circulating secondhand for glycol did not reconcile against a primary document and a wrong number in a risk assessment is worse than no number.
The part almost nobody quotes is the scope. E1.5 describes fogs not likely to be harmful to otherwise healthy performers, technicians or audience members aged 18 to 64. It makes no statement about anybody else. Child performers, a cast member with asthma, an audience that includes both — the standard is silent, and silence is not permission. A 2017 review commissioned for the US film industry said so explicitly.
Monitoring means measuring. E1.23 asks for assurance that recognised limits are not exceeded, and a handheld meter does that only once it is calibrated for the particular machine and fluid in use. The same reading means different concentrations for different fluids.
The practical version: a rig that has been hazing into a sealed room with the HVAC off all afternoon is a different exposure from the same rig in the same room with air moving, and nothing about the look tells you which one you are in.
Pyrotechnics
The vocabulary, and where this page stops
Pyro is a different discipline with a different legal basis, and the competence to fire it is licensed rather than read. What follows is a naming reference, so that a designer can ask for the right thing and a technician can understand a cue sheet. It is not a method and there is deliberately nothing here about composition, quantity or firing.
The governing documents in the US are NFPA 1126 for pyrotechnics before a proximate audience and NFPA 160 for flame effects; other jurisdictions have their own, and the local authority having jurisdiction has the final say regardless.
Coloured smoke is pyrotechnic, not atmospheric. It works by burning a composition that sublimes a dye, so it involves combustion, heat and solid residue, and it falls under pyrotechnic rules rather than under the fog standards. The name invites the confusion and the two are regulated differently.
It is also, in practice, an outdoor effect. Indoors the residue coats surfaces, the smell persists, and the particulate is a genuine respiratory exposure rather than the engineered one a fog standard describes.
- Hazing after the house opens. Haze needs time to spread evenly and the machine is audible and visible while it works. It goes in early, with the air handling set the way it will be during the show — because changing the HVAC later changes the look.
- Assuming the HVAC is a constant. It is usually on a schedule, and the schedule often changes at the exact moment the house opens. A room hazed to look right at 18:00 can be clear by 19:30 for no reason anybody on headset can see.
- Low fog that comes back. It warms, rises, and arrives in the beams a minute later looking like a hazer fault.
- Fluid swapping. A fluid formulated for another machine changes droplet size, output and residue, and can block a heat exchanger. The exposure profile changes too, so the monitoring you did no longer describes what is in the air.
- Treating the permit as admin. The isolation is a removed protection. Somebody has to be the replacement, and that person needs to know they are it.
Where this sits
The rest of the chain
Atmospherics are a lighting tool before they are anything else — the beam you are revealing was shaped by optics, and the colour you see in it is being filtered by the air. The outdoor page covers the wind that makes all of this unpredictable, and the room covers the air handling that decides whether your haze stays where you put it.
For the numbers and the look-up, the glossary carries the vocabulary bilingually, and the governing standards are indexed with access links on the standards pages.
5 questions on the mechanism rather than the trivia. Nothing is scored and you can answer again — the explanation under a wrong answer is the point of it.
-
1A designer asks for "more haze" because the beams are not reading. The hazer is already at full output and the room looks slightly milky. What is most likely wrong?
-
Possible, and it changes droplet size — but a milky room says particles are present in quantity, so this is not the first thing to check.
-
Right. A milky room is a room with plenty of particles in it. If the beams still do not read, the usual culprit is air handling: haze needs to sit and spread evenly, and a room with air moving through it never builds the even suspension that makes a beam look solid. Turning the machine up adds output to a space that is carrying it away, and the result is a grey room with weak beams — the worst of both.
-
Beam visibility is about what is in the air, not about output. A brighter lamp in moving air gives you a brighter lamp and the same weak beam.
-
That is how you make low fog, which is a different effect. Chilled haze would sink to the floor and leave the beams emptier still.
-
-
2Your low fog looks perfect for the number. A minute later the lighting designer is complaining that the beams have gone soft. What happened?
-
Right, and it is the thing that catches everybody once. Low fog is ordinary fog chilled below room temperature; that is the entire trick. As it warms it stops sinking, and it does not disperse where it lay — it rises into the beams as ordinary fog. The effect has a tail, and the tail belongs to the lighting department.
-
Contamination changes output and residue, not the buoyancy of a layer over the course of a minute.
-
Air handling certainly moves it, but the layer would thin and scatter rather than arrive in the beams on that timescale.
-
Worth ruling out, but the timing points elsewhere — a minute is exactly how long the real cause takes.
-
-
3A venue agrees to isolate the smoke detection in the auditorium for your haze. What has actually been agreed?
-
Isolation is a venue and system decision. It is not an approval of your effect and it does not transfer responsibility.
-
They cannot distinguish. A photoelectric detector looks for light scattered by particles, which is exactly what haze is for.
-
Isolation says nothing about the haze. It says the thing that would have noticed a fire has been switched off.
-
Right, and this is the part that gets treated as paperwork. Detection is a safety system; isolating a zone removes it for the duration. The permit is the record of what stands in its place — usually a named person watching that space who knows they are the replacement. If nobody can say who that is, the isolation has not been arranged, it has just been requested.
-
-
4Coloured smoke is requested for an indoor finale. What is the first thing to say?
-
An aesthetic point, and secondary. The first answer is that this is a different discipline with a different legal basis.
-
There is no fluid for it. Coloured smoke is not made by a fog machine at all.
-
Right. Coloured smoke burns a composition that sublimes a dye, so it involves combustion, heat and solid residue, and it falls under pyrotechnic rules rather than the fog standards. The name invites exactly this confusion. Indoors it is also a poor idea on its own terms: the residue coats surfaces, the smell persists, and the particulate is a real respiratory exposure rather than the engineered one a fog standard describes.
-
It will, but that is true of the atmospheric effects too. The distinction that matters here is the legal and hazard class.
-
-
5ANSI E1.5 describes fogs "not likely to be harmful to otherwise healthy performers, technicians or audience members aged 18 to 64". Your cast includes two children. What does the standard tell you?
-
The standard does not say that. Reading a margin into silence is inventing a figure the document declines to give.
-
Right, and this is the most-skipped sentence in the whole subject. The scope is explicit about the population it covers, and children are outside it. A review commissioned for the US film industry made the same point about child actors and about audiences generally. Silence is not permission: it means the standard has not answered your question and you need a different basis for the decision.
-
The standard does not say that either. It declines to make a statement, which is not the same as a prohibition — but it does mean you cannot cite E1.5 as your justification.
-
That number is not in the standard, and a figure somebody invented is worse in a risk assessment than an acknowledged gap.
-
Answers are kept in this browser on this device. Nothing is sent anywhere.
- Haze or fog?
- Why did my low fog rise?
- Will it set the alarm off?
Where to go next
These pages are a chain rather than a library — the next one along usually answers the question this one raised.
Everything on this page, in the index
Each of these has its own entry with sources, ports, gotchas and an open gap where a field is missing.