Ask most shop owners how loud their building is and you get one number, usually estimated while standing next to the bridge saw. That number is close to useless. Noise in a stone shop is not a single value; it is a field that changes with every step you take, with which machines happen to be cutting, and with how much hard reflective surface surrounds you. The saw operator, the polisher three bays away and the office door live in different acoustic environments.
Mapping is what turns guesswork into a plan. A grid of measured points across the floor, taken under known conditions and written down, shows which sources dominate where, which walkways are quietly exposing people who never touch a machine, and which controls will actually move the number. It also produces the evidence you need if a regulator, an insurer or an employee asks what you did about it. The method is straightforward and the instruments are inexpensive relative to the exposure they document.
Pressure, Power and the A-Weighted Number
Two quantities get used interchangeably in trade conversation and they are not the same thing. Sound pressure is what a microphone at a given location senses; it depends on distance, on the room, and on everything else running at that moment. Sound power is a property of the source itself, independent of where you stand and independent of the building. A machine has one sound power level and an unlimited number of sound pressure levels around it.
That distinction matters the moment you compare two machines. Manufacturers publish declared emission figures determined under standardised conditions — ISO 3744 is the common engineering-grade method, in which sound pressure is measured over a surface enveloping the source in an approximately free field over a reflecting plane, then converted into a sound power level. A power figure derived that way can be compared fairly between machines. A pressure reading someone took on a shop floor cannot, because the building is baked into it.
On the floor you measure pressure, and that is correct, because pressure at the ear is what damages hearing. The point is to know which quantity you are holding. Never put a supplier's declared power figure into exposure calculations, and never judge a competing machine by a pressure reading taken in someone else's building.
A-weighting is the filter that makes a raw pressure reading meaningful for hearing risk. It shapes the meter's frequency response to approximate the sensitivity of the human ear, discounting very low and very high frequencies the way the ear does. Occupational limits are written in A-weighted decibels, so every reading you take for compliance purposes is set to A-weighting with slow time response. Unweighted or C-weighted readings have their uses in diagnosis, but they are not the compliance number.
Distance behaves predictably in the open and unpredictably indoors. For a point source radiating into a free field, sound pressure falls about 6 dB for every doubling of distance. A stone shop is close to the opposite of a free field: concrete floor, block walls, steel roof, glass, and slabs standing on racks, all reflecting energy back. Beyond a certain radius from a machine, the reverberant field dominates and moving further away stops helping.
Sources also stack in a way that is unintuitive to anyone used to arithmetic. Two equal sources running together produce roughly 3 dB more than one, not double the reading. The practical consequence is that silencing the second-loudest machine in a bay barely changes the total while the loudest one still runs. Ranking sources by contribution, and attacking them in order, is the only sequence that pays.
Building the Grid Map
Instruments and Settings
A sound level meter meeting ANSI S1.4 Type 2 is the minimum instrument for compliance work, and Type 2 units are the general-purpose class used for industrial hygiene surveys. Set it to A-weighting with slow response. For personal exposure you need a dosimeter meeting ANSI S1.25, and for OSHA-style dose the standard configuration is A-weighting, slow response, a 5 dB exchange rate, a 90 dBA criterion level and an 80 dBA integration threshold.
Calibrate at the start and end of every session with an acoustic calibrator, typically producing a known level near 94 dB at 1 kHz, and record both checks. If the instrument has drifted more than about 1 dB between them, the session data is questionable and should be repeated. A survey with no calibration record cannot be defended later, however careful the measurements were.
Laying Out the Grid
Start with a scale drawing of the floor with machines, racks and walkways marked. Overlay a grid; three metres between points suits most fabrication shops, tightened to one metre around machines and loosened in storage areas. Number every point on the drawing and use those numbers on the sheet. A map with unlabelled readings is impossible to reproduce six months later when you want to prove a control worked.
Take readings at ear height for a standing worker, roughly 1.5 metres, and keep the microphone at least a metre clear of walls and slab faces so you are not measuring a local reflection. Point the microphone as the instrument's manual specifies, since Type 2 meters have a defined reference direction. Add extra points wherever people actually stand for long periods: the saw console, the polishing bench, the layout table, the office doorway.
Running the Survey
Do the whole grid twice on the same day. The first pass runs during normal production with everything that would normally be operating; the second runs with the machines idle and only the constant services, typically the compressor and the dust or slurry system, still live. The second map is your background floor, and it is often the more interesting of the two.
Write down what was running at each point, not just the number. A given level means one thing when the saw and two grinders are cutting and something else entirely when the shop is between jobs. Note the time, the machines active, whether doors were open, and anything unusual. Those notes let you interpret the map instead of merely admiring it.
Spot readings and dosimetry answer different questions and you need both. The grid tells you where the noise is and which source dominates each area, which is what you need for engineering controls. A dosimeter worn for a full shift by a representative worker in each role tells you what that person actually accumulated, including the time they spent away from their machine. Neither substitutes for the other.
| Shop zone | Dominant source | Control approach | What to re-measure |
|---|---|---|---|
| Bridge saw bay | Blade contact with the slab, plus the water pump | Partial enclosure, damped blade core, operator station offset from the cut line | Operator ear position and the adjacent walkway |
| CNC cell | Spindle, tool changes and vacuum pumps | Full enclosure with interlocked doors; relocate pumps outside the cell | Just outside the cell door line and the programming station |
| Hand polishing and edge work | Angle grinders and line polishers | Task barriers between benches, tool selection, operator rotation | Each hand station and the bench immediately next to it |
| Compressor and pump room | Rotary screw or piston compressor | Move it outside the production envelope; isolate the mounts | The doorway and the nearest permanent work station |
| Slurry and dust handling | Blower fan, ducting runs and cyclone | Lag the ducts, fit a fan silencer, hang runs on resilient supports | Directly under the duct run and beside the fan housing |
| Loading and slab handling | Stone contacting steel racks and A-frames | Rubber pads on contact points, revised handling sequence | The handling area during an actual load-in, not a quiet hour |
Once the grid is plotted, contour it. Group the points into bands and shade them on the drawing. Patterns emerge quickly: a corridor of elevated readings following a duct run, a quiet-looking bench that sits in the reflected field of the saw, a storage aisle nobody thought about that is louder than the machine bay. Those patterns are the argument for spending money in a particular place, and they are far more persuasive to an owner than a list of figures.
Pro Tip: Run the grid twice on the same day — once at full production and once with only the compressor and dust system live. Subtracting the two maps separates the background floor from the process peaks, and it usually shows that the constant services, not the saws, are what keeps every point in the building elevated all shift.
What the Occupational Standard Actually Requires
Occupational noise exposure in general industry is governed by 29 CFR 1910.95. The permissible exposure limit is 90 dBA as an 8-hour time-weighted average. Exposures at or above that level oblige the employer to use feasible engineering and administrative controls, and to fall back on personal protective equipment only where those controls do not bring exposure within the limits of the standard's table.
The action level sits lower, at an 8-hour time-weighted average of 85 dBA, equivalently a 50 percent dose. Employees at or above the action level must be brought into a hearing conservation programme, with monitoring, audiometric testing, protector availability and training. A shop can be fully below the permissible limit and still owe every one of those obligations, which is the point most operators miss.
The exchange rate is 5 dB. Each 5 dB increase in level halves the permitted duration: eight hours at 90 dBA, four hours at 95 dBA, two hours at 100 dBA, and so on down. The standard's table also caps continuous exposure at 115 dBA, allowed for no more than fifteen minutes. A footnote adds that exposure to impulsive or impact noise should not exceed a 140 dB peak; because it says should rather than shall, it functions as guidance rather than as an enforceable limit on its own, though the agency treats exceeding it as a serious hazard.
It is worth knowing that the health-based recommendation is stricter. The recommended exposure limit published by the national occupational health institute is 85 dBA as an 8-hour average with a 3 dB exchange rate, on the reasoning that equal energy, not equal decibels, drives hearing damage. Under a 3 dB rule the permitted time at 100 dBA collapses to a small fraction of the 5 dB allowance. Shops planning capital equipment for a twenty-year life are better served designing to the stricter figure.
Published measurement work on masonry and stone saw cutting consistently puts the operator position in three figures. Studies characterising cutting of concrete, block and stone report operator-ear levels above 100 dBA, and reviews of construction power tools generally place that class of equipment in a comparable band. Exact values depend on the material, the blade, the machine and the room, so treat the published work as an indication of magnitude and measure your own equipment.
Controls, Protectors and Re-Measurement
Enclosure is the strongest control available and the least popular, because it interferes with loading. Partial enclosures around a saw, or a full interlocked enclosure on a CNC, work well when the panels are sealed rather than merely present. A gap in an enclosure leaks sound far out of proportion to its area, so the discipline is in the detailing: sealed access doors, brush strips at conveyor openings, and no permanently propped panels.
Isolation and damping deal with the structure-borne half of the problem. Machines bolted rigidly to a slab drive the floor, and the floor radiates. Resilient mounts under compressors, pumps and blowers break that path. On the cutting side, blade cores with damping layers or slotted designs are made specifically to suppress the ringing of the steel disc, which is a large part of what a running saw radiates between cuts.
Relocating the compressor is usually the highest return per dollar in a stone shop. It runs all day, it contributes to every point on the background map, and it does not need to be inside the production envelope. Moving it to an outbuilding or a sealed lean-to lowers the floor everywhere at once, which is worth more than shaving a few decibels off a machine that runs intermittently.
Absorptive treatment is where a wet stone shop differs from a dry factory. Conventional porous absorbers foul with slurry and dust and lose their performance, so the usual approach is barriers and enclosures rather than blanket ceiling absorption. Where absorption is used it needs a washable, sealed facing, and it belongs high on walls and above the roof line of machines rather than anywhere slurry can reach it.
Hearing protection is the last layer, not the first, and its rating needs derating before it means anything. The regulator's own appendix method subtracts 7 dB from the labelled noise reduction rating and applies the remainder to the A-weighted average, with a further halving commonly recommended to reflect real-world fit. The health institute takes a different route, derating by 25 percent for earmuffs, 50 percent for formable earplugs and 70 percent for other plugs.
Whichever derating you use, comfort and fit beat the number on the box. A high-rated muff that an operator lifts every ten minutes protects less than a modest plug worn continuously, and protectors that cut a worker off from machine sounds they rely on create their own risk. Fit testing is worth more than any label comparison.
Re-measure after every change, at the same grid points and under the same recorded conditions, and keep both maps. That is the only way to know whether an enclosure earned its cost or whether the improvement came from a machine that happened to be idle. Repeat the full survey annually and whenever equipment moves, since a shop layout drifts more than anyone expects across a couple of busy years.
Noise mapping sits alongside the other exposure work a fabrication shop has to document, and the same survey discipline applies to airborne hazards; our silica dust safety guide covers that side in detail. If you are planning a building or a move rather than retrofitting one, the notes on shop layout, equipment and costs and on dust and slurry extraction systems will help you place the noisy services where they belong from the start.
Equip the shop, then quiet it down
From blades and core plates to dust and slurry handling, the right hardware changes what your noise map looks like.
Shop the full catalogue →Free Tool
Free Guides & Tools — a working library of shop references covering layout, equipment selection and safety programme basics, useful when you turn a noise map into a control plan and a budget.
Open the guides library →