How Loud Are Firecrackers?

125 dB(AI), or 145 dB(C) peak, measured 4 metres from the point of bursting — that's the number every legally sold Indian firecracker has to sit under. It was set by the Supreme Court's order of 18 July 2005 in In Re: Noise Pollution (V), Writ Petition (Civil) No. 72 of 1998, and it hasn't moved since. The 2018 green-crackers judgment, the 2020 NGT orders, and every state notification that followed all work within this same figure rather than resetting it; none of them created the 125 dB standard, they enforce it.

PESO, the Petroleum and Explosives Safety Organisation, is the body that tests a formulation against this ceiling before a manufacturer can legally produce it. A cracker that hasn't cleared that test has no legal basis to be sold, whatever the box claims about "low sound."

Decibels climb on a logarithmic scale, not a linear one: a 10 dB rise isn't "ten percent louder," it's roughly a 10x jump in sound energy and close to a doubling in what the human ear perceives as loudness. The jump from 85 dB (a NIOSH-safe workday) to 125 dB (the cracker ceiling) is 40 dB, which works out to roughly 10,000 times more acoustic energy, not 40 units more.

There's a second distinction that matters here and gets flattened in most comparisons. Traffic and conversation are measured as dB(A), a continuous reading averaged over time. A firecracker is measured as dB(AI), the impulse-weighted figure for a single sharp bang lasting a fraction of a second. Two numbers that look close on paper aren't hitting the ear the same way: a sustained 120 dB concert and a 125 dB cracker blast are different kinds of loud, and the impulse kind is the one CDC/NIOSH flags as capable of causing immediate, permanent damage above roughly 140 dB peak, with no exposure-time cushion the way a continuous sound gets one.

Nine years later, the same testing partnership between MPCB and the Awaaz Foundation ran 16 widely-sold firecracker varieties through the RCF ground in Chembur, Mumbai. Every sample stayed under 125 dB, the first time since testing began there in 2008 that a full random batch came back fully compliant. The loudest reading, 114 dB, came from a bound string of 10,000 ladis, still 11 dB under the flat single-cracker ceiling despite being a joined product that should legally answer to a much lower reduced number.

That gap between "passed the flat ceiling" and "passed its actual reduced ceiling" is worth sitting with. A 10,000-cracker string has an N large enough that its legally reduced ceiling would sit well below 114 dB; whether the tested sample was measured and rated against its correct series ceiling, or just checked against the standalone 125 dB figure, isn't settled by the reporting available. It's the exact ambiguity the next section explains.

You can't hold a decibel meter to a listing photo, so the practical check is different: look for PESO test marks on the packaging rather than an unverified "low noise" claim, and for joined or series products, ask specifically whether the certification covers the series' reduced ceiling rather than the flat single-cracker figure. If keeping noise down for infants, elderly family, pets or an apartment building is the goal from the start, our low-noise cracker finder filters stock by decibel threshold instead of leaving that check to guesswork after the packs arrive.

What is the legal noise limit for firecrackers in India?

125 dB(AI) or 145 dB(C) pk, measured 4 metres from the point of bursting, fixed by the Supreme Court's 18 July 2005 order and unchanged since.

How loud is a single legal firecracker compared to a jet engine at takeoff?

About 15 dB quieter on paper (125 dB vs. roughly 140 dB), but because decibels are logarithmic, that gap represents a large real difference in acoustic energy, not a small one.

Do garland or series crackers follow the same 125 dB limit as a single cracker?

No. Series and joined crackers get a reduced ceiling of 5 log₁₀(N) dB below 125 dB(AI), where N is the number of crackers in the string — 115 dB(AI) for a 100-wala string, 110 dB(AI) for a 1000-wala string.

Did Indian test data ever show firecrackers exceeding the legal limit?

Yes. Maharashtra Pollution Control Board's 2013 tests found single crackers up to 132.4 dB and series crackers up to 118 dB against their applicable ceilings, both over the legal limit at the time.

Are firecracker decibel readings measured the same way as traffic or conversation noise?

No. Traffic and conversation are measured as a continuous dB(A) average; a firecracker is measured as dB(AI), an impulse-weighted reading for a single sharp bang, which the ear and safety agencies like CDC/NIOSH treat as more damaging than a steady sound at a similar number.

Compare specific product noise levels against your own threshold with the low-noise cracker finder, or browse the rest of the data library for more sourced Diwali datasets.