Acoustic panels, screens and rafts: where the absorption goes
Absorption is bought by area and placed by surface, and the ceiling carries most of the result. The published coverage figures, the two ISO 354 outputs that cannot be added together, the NCC group number that decides what may go overhead, and what each kind of absorber is actually for.
Acoustic treatment is bought by area and placed by surface, so the useful question is not how many panels but what share of which surface. Ecophon's published comparison of a room 4 m high, 10 m wide and 20 m long is the clearest number on the subject: untreated, with an exposed structure, that room runs a 3.0 second reverberation time and 21 per cent speech definition. Cover 30 per cent of the ceiling area with free-hanging units and it reaches 0.9 seconds and 54 per cent. Sixty per cent coverage reaches 0.6 seconds and 68 per cent. A full absorbent ceiling reaches 0.5 seconds and 75 per cent, an 8 dB reduction in room noise. Most of the result arrives in the first third.
That is the whole argument for treating the ceiling before anything else, and it is also why a schedule full of wall panels often underperforms the drawing it came from. This post is about the part of the decision that sits after the ratings: which kind of absorber goes on which surface, how it fixes, what the National Construction Code allows overhead, and what the manufacturers of the products actually publish. The ratings themselves, the αw and NRC and Rw figures and the absorption area calculation, are set out in Acoustic ratings on a plan: what NRC, Rw and dB actually tell you, and that post is the place to start if the question is what a number means rather than where the material goes.
The surface you treat decides most of the result
Ecophon's guide publishes the comparison as a set of four conditions in one room, at a mid-frequency average of the 500 and 1000 Hz octave bands. It is explicit that the figures assume a diffuse sound field, which most ordinary rooms do not have, so they size an order rather than settle a design. They are still the most honest published illustration of diminishing returns in commercial absorption.
| Condition | Units at 1200 by 1200 mm | Reverberation time | Speech definition | Noise reduction in room |
|---|---|---|---|---|
| No ceiling, exposed structure | none | 3.0 s | 21% | reference |
| Free-hanging units, 30% of ceiling area | 41 pieces | 0.9 s | 54% | 5 dB |
| Free-hanging units, 60% of ceiling area | 83 pieces | 0.6 s | 68% | 6 dB |
| Full wall-to-wall absorbent ceiling | continuous | 0.5 s | 75% | 8 dB |
Read the last two rows together. Doubling the coverage from 30 to 60 per cent buys 0.3 of a second and one more decibel. Going from 60 per cent to a continuous ceiling buys another 0.1 of a second and two more decibels. The same guide recommends not exceeding about 60 per cent coverage with free-hanging units, and gives a worked example where a meeting room target of absorption area equal to 1.1 times the floor area would need 76 per cent coverage, which it flags as beyond that limit. When the arithmetic asks for more than the ceiling plane can sensibly carry in discrete units, the answer is a continuous ceiling, not more baffles.
Why the ceiling, and not the walls, carries the load
Ecophon states it directly: a wall-to-wall ceiling is normally more beneficial than free-hanging units for acoustic efficiency, and is advantageous especially with regard to absorption at low frequencies. It recommends free-hanging units where a full ceiling is not possible, naming thermally active slabs and large glazed areas, and suggests supplementing them with wall panels or screens.
The second reason is arithmetic you can do on any floor plan. Take the same published room. The ceiling is 200 square metres in one uninterrupted plane. The walls total 240 square metres gross, which looks generous until the deductions start. The lowest metre of wall sits behind desks, storage and skirting, which removes 60 square metres. If the two 20 metre walls are glazed above that line, as a perimeter office floor usually is, another 120 square metres goes. What remains is 60 square metres of wall against 200 square metres of ceiling, and the wall area that is left is interrupted by doors, joinery, services and the things a tenant wants to hang on it.
That ratio is why a wall panel is a finishing move. It is also why the most effective wall treatment is usually a continuous lining of one wall rather than a scatter of panels across four, and why the walls that matter most are the hard parallel pair that faces each other across a narrow space.
A panel's absorption figure will not tell you how many you need
This is the mismatch that catches out an otherwise careful schedule, and it is written into the test method. ISO 354:2003 measures sound absorption in a reverberation room of at least 150 cubic metres, with at least 200 cubic metres strongly recommended for new test rooms. For a plane absorber, a lining material, clause 6.2.1.1 requires a test specimen with an area between 10 and 12 square metres, scaled up if the room is larger than 200 cubic metres. The result is a coefficient: a proportion, per square metre, of a large flat area of the stuff.
Discrete absorbers are measured differently. Clause 6.2.2.1 puts rectangular unit absorber pads and baffles in a Type J mounting. Clause 6.2.2.2 says objects such as chairs or screens are tested in the same manner as they are typically installed in practice. Clause 6.2.2.3 requires, in general, at least three individual objects, chosen so the change they make to the room's equivalent sound absorption area is greater than 1 square metre but not more than 12 square metres, and if only one object is available it is tested in at least three locations at least 2 metres apart and averaged. The result is an equivalent sound absorption area, in square metres, per object.
Those two outputs are not interchangeable and they cannot be added. A coefficient for a lining tells you what a square metre of wall will do once it is covered. An equivalent absorption area tells you what one hanging baffle or one screen contributes wherever it is put. A schedule that lists a wall panel's class next to a pendant's class is comparing a rate with a quantity, and the specifier who asks the acoustic consultant for a target in square metres of absorption, then asks each supplier which of the two numbers they are quoting, gets a comparison that holds.
Specifier's note. Ask for the result, not the standard. The Australian PET panel makers name their test methods carefully and publish almost no measured outcomes. Klaro Design's published data for RAPT lists acoustics to EN ISO 354:2003, flame retardancy to EN 13501, fire testing to the NCC with AS ISO 9705 and AS 5637 classification, recycled PET staple fibre to AS/NZS 4266 and a weight range of 1300 to 3500 grams per square metre. It publishes no absorption coefficient, no class letter and no panel size. Offecct's published Soundwave data names EN ISO 354 and describes the panels as lightweight absorbers for 500 Hz and above, with the Basfill insert improving absorption between 150 and 500 Hz, without a coefficient on the product data either. None of that is misleading, and all of it is useless on a schedule. Request the test report number, the mounting it was tested in and the octave band figures before the panel goes on a drawing, because a naming of ISO 354 only tells you that somebody measured something.
Four ways to add absorption, and what each one is for
Geneva's acoustic solutions collection covers the four with products that publish enough to compare. The distinctions that matter are the surface each one uses, how it fixes, and the thing it cannot do.
| Type | Example | Surface it uses | Published construction | Fixing | What it will not do |
|---|---|---|---|---|---|
| Modular wall panel | Soundwave Swell, Ceramic, Village | Wall only | 585 by 585 mm, moulded polyester fibre, absorber for 500 Hz and above | Four adhesive Velcro attachments, stapled on uneven walls | Go on a ceiling, or reach low frequencies without the Basfill insert |
| Custom wall lining | RAPT | One wall, continuously | 9 mm and 12 mm PET boards combined, grooved to selected areas, pinnable | Set out to the wall, made to the measured dimension | Be reordered or extended later to the same cut pattern |
| Ceiling pendant or baffle | OM | Ceiling plane | Laser-cut 12 mm acoustic PET, set of three sizes, up to 60% recycled PET, no glue, five year warranty | Suspended, coordinated with services, optional LED | Work where the ceiling void is already full |
| Free-standing divider | Mr SCREEN | Floor, no fixing | Acoustic PET panels 24 mm thick, pinnable, two-colour, cut and assembled in Australia | None, portable | Stay where the drawing put it |
| Workstation screen | SCORE, Hedge | Worktop or floor | SCORE up to 80% post-consumer PET, NCC fire tested; Hedge 32 mm hoop pine plywood frame, acoustic foam, 12 mm PET at 75% recycled content | Worktop mounted, or steel feet with glides and castors on the single Hedge | Give speech privacy, at any height |
The last column is the one worth arguing about on site. Soundwave Swell is a wall panel in the strict sense: Offecct publishes that each panel fixes with four adhesive Velcro attachments, stapled on uneven walls, and that the panels are not designed for ceilings. RAPT solves the opposite problem, a wall that needs to be covered rather than decorated, by combining 9 mm and 12 mm boards into one made-to-measure design, which is also why it cannot be matched later. OM puts 12 mm PET in the only plane with area to spare. Mr SCREEN trades fixed position for the ability to zone a space this week and a different space next week.
What a screen is, and is not
A screen is an absorber that happens to stand in the way, and the standing in the way is most of its value. The published research on what that is worth, including the distraction distance and speech transmission index figures a consultant will use, is in Open plan office noise: what the research actually measured. The short version for a schedule is that a wrap-around screen interrupts more of the direct path between two seated people than a flat panel on a wall ever will, and that neither one encloses anything. Where the conversation has to be confidential, the answer is a room. Soft seating does something similar and is covered in Acoustic soft seating: what a high back actually does in an open plan.
Group number: the test that decides what can go overhead
A panel that satisfies the acoustic consultant can still be refused by the building surveyor, and the refusal usually concerns the ceiling. Specification 7 of NCC 2022 Volume One sets the fire hazard properties of wall and ceiling lining systems. Clause S7C4(2) requires the group number, and the smoke growth rate index or average specific extinction area, to be determined in accordance with AS 5637.1. Clause S7C4(1) requires a lining in a building not fitted with a sprinkler system complying with Specification 17 to have a smoke growth rate index not more than 100, or an average specific extinction area less than 250 square metres per kilogram.
Table S7C4 then sets the permitted material groups. The row that covers a commercial office is Class 5, 6, 7, 8 or 9b schools.
| Area | Unsprinklered, walls | Unsprinklered, ceilings | Sprinklered, walls | Sprinklered, ceilings |
|---|---|---|---|---|
| Fire-isolated exits and fire control rooms | 1 | 1 | 1 | 1 |
| Public corridors | 1, 2 | 1, 2 | 1, 2, 3 | 1, 2, 3 |
| Specific areas | 1, 2, 3 | 1, 2 | 1, 2, 3 | 1, 2, 3 |
| Other areas | 1, 2, 3 | 1, 2, 3 | 1, 2, 3 | 1, 2, 3 |
Three consequences for a furniture and finishes schedule. First, Group 1 is the best performing of the four groups and Group 4 the worst, and Table S7C4 never permits Group 4 anywhere. Second, the unsprinklered specific areas row is the only place in that block where the ceiling is held to a tighter set than the wall, Group 1 or 2 against Group 1, 2 or 3, which is precisely the situation where a wall panel gets proposed as a ceiling baffle. Third, sprinkler protection opens public corridors and specific areas to Group 3, so the same product can be compliant in one building and refused in the next for reasons that have nothing to do with the product.
Loose furniture is outside this. A free-standing divider or a worktop-mounted screen is not a wall or ceiling lining system, which is why the relevant published evidence for those products is different, and why the fabric and foam side of the question is handled in Commercial upholstery ratings: which fabric numbers belong on a schedule. The moment a PET product is fixed to a wall or suspended from the slab as a lining, Specification 7 applies to it.
The fixing detail that decides whether it can be built
Most of the trouble on site comes from the back of the panel rather than the front.
Adhesive fixing and the substrate
Offecct's published method for the Soundwave panels is four adhesive Velcro attachments per panel, stapled where the wall is uneven. That works on plasterboard and on timber, and it is what makes a panel array adjustable on the day. It also means the panel is fixed to the finish rather than the structure, so the paint system and the plasterboard's condition become part of the specification, and a wall that has been repainted several times is a different substrate from a freshly set one.
Suspension and the ceiling void
A pendant absorber has to be hung, and the slab it hangs from is already carrying sprinklers, lighting, detection, speakers and air. The OM kit is published as a set of three pendants in one colour with an optional LED light, which makes the lighting coordination easier than it looks, but the suspension points still have to be agreed with the services contractor before the ceiling is set out. This is the item most often value-managed out of a project, because by the time anybody draws it the void is full.
Made to measure, and what that costs you later
RAPT is published as custom designed to the wall, combining 9 mm and 12 mm boards with grooves applied to selected areas. The benefit is a wall that is covered rather than decorated. The consequence is that a churn event which moves a door or a glazed screen leaves a lining that cannot be extended to match. Where the brief includes future reconfiguration, a modular panel array or a free-standing divider survives the change and a bespoke lining does not.
Nothing fixed at all
Mr SCREEN is published as free-standing and portable in 24 mm acoustic PET, cut and assembled in Australia. Hedge is published with steel feet and glides, with castors available on the single screen. A movable absorber is the only kind that can be repositioned after the floor is occupied and somebody discovers where the noise actually is, which is a real advantage and the reason it belongs in the budget rather than as a substitute for the ceiling.
Putting it on a drawing so it survives the set
Six things settle an absorption package before it reaches tender.
- A target, in square metres of absorption, from the acoustic consultant, not a coverage percentage copied from a brochure. Coverage follows from the target and the room.
- The surface allocation, decided first. Ceiling, then one continuous wall, then furniture. Reversing that order is what produces a schedule of panels that does not reach the target.
- Which of the two ISO 354 outputs each supplier is quoting, a coefficient for a lining or an equivalent absorption area per object, with the test report number and the mounting.
- The group number to AS 5637.1 for anything fixed as a lining, and the smoke growth rate index or average specific extinction area if the building is not sprinklered, checked against the area of the floor the lining is in.
- The substrate and the suspension, agreed with the builder and the services contractor rather than noted as by others.
- What moves. At least part of the package should be repositionable, because the first month of occupation always finds a noise the model did not.
The last point is the one that separates a package that works from one that measured well on paper. Absorption is the part of a fitout doing an invisible job, and the temptation is to make it announce itself. A continuous lining on the one wall that needs it, a considered pendant array and a few movable screens will usually read better and measure better than forty panels distributed for visual balance.
Frequently asked
How many acoustic panels does an office actually need?
Enough to cover a stated share of one surface, which is why the answer is given as a coverage percentage rather than a panel count. Ecophon's published comparison of a room 4 m high, 10 m wide and 20 m long puts an untreated room with an exposed structure at a 3.0 second reverberation time and 21 per cent speech definition. Thirty per cent coverage of the ceiling area with free-hanging units, which it gives as 41 pieces at 1200 by 1200 mm, takes that to 0.9 seconds and 54 per cent. Sixty per cent coverage, 83 pieces, reaches 0.6 seconds and 68 per cent. A full wall-to-wall absorbent ceiling reaches 0.5 seconds and 75 per cent. The figures are a mid-frequency average of the 500 and 1000 Hz bands and assume a diffuse sound field, which most real rooms do not have, so they size the order rather than settle it.
Is it better to put acoustic treatment on the ceiling or on the walls?
The ceiling, first, for two reasons. Ecophon states plainly that a wall-to-wall ceiling is normally more beneficial than free-hanging units for acoustic efficiency and is advantageous especially with regard to absorption at low frequencies. The second reason is arithmetic. In that same published room the ceiling offers 200 square metres in one uninterrupted plane. The walls offer 240 square metres gross, but the lowest metre sits behind furniture, which removes 60 square metres, and if the two long walls are glazed above that line another 120 square metres goes with them. The ceiling is usually the only surface with enough continuous area left to do the heavy work, and panels on walls are the top-up that gets the last part of the way.
Can an acoustic wall panel be used on a ceiling?
Not unless the manufacturer says so, and two separate gates apply. Offecct states for Soundwave Swell, Ceramic and Village that each panel fixes to the wall with four adhesive Velcro attachments, stapled on uneven walls, and that they are not designed for ceilings. That is a fixing limitation and it does not have an exception. The second gate is the National Construction Code. Table S7C4 of Specification 7 in NCC 2022 Volume One permits a narrower set of material groups on ceilings than on walls in some cases: in an unsprinklered Class 5 office, specific areas allow Group 1, 2 or 3 on walls but only Group 1 or 2 on ceilings. A lining that is compliant on the wall is not automatically compliant overhead.
What fire rating does an acoustic panel need in an Australian office?
A group number determined in accordance with AS 5637.1, and the permitted group depends on the class of building, whether it is sprinklered, and which part of the floor the lining is in. For a Class 5 office that is sprinklered, Table S7C4 permits Group 1, 2 or 3 on walls and ceilings in public corridors, specific areas and other areas, with Group 1 only in fire-isolated exits and fire control rooms. Unsprinklered, public corridors tighten to Group 1 or 2 and specific area ceilings tighten to Group 1 or 2. Clause S7C4(1) adds that a lining in an unsprinklered building must also have a smoke growth rate index not more than 100, or an average specific extinction area less than 250 square metres per kilogram.
Will acoustic panels stop people overhearing each other?
No. Absorption reduces the sound that arrives by reflection, so it shortens reverberation and lowers the general level in a room. It does not interrupt the direct path between two people, and speech privacy depends on that path being broken or masked. A free-standing divider such as Mr SCREEN, or a wrap-around workstation screen such as SCORE, interrupts more of the direct path than a flat panel on a far wall, and still does not enclose it. If the brief is confidentiality rather than comfort, the answer is a room or a booth, and the panels are what stops that room sounding like a bathroom.
Where to start
If you have an acoustic target and a plan but no allocation yet, send both: Geneva will set the absorption against the ceiling, wall and furniture surfaces you actually have, and come back with the published construction and fire evidence for each item rather than a panel count. Start at our contact page, or look through the acoustic solutions range and the Klaro Design and Offecct brand pages first.