Replacing IEC 60268-5: what comes next and why

Geoff Hill — chair of the AES SC-04-03 working group on loudspeaker modelling, member of IEC TC 100/TA 20/PT 60268-22, CTO of Hill Acoustics — on why the standard behind every T/S datasheet in this wiki was withdrawn, and what replaced it. Published in audioXpress; T1, because it rests on the author's own committee work and on the standards' contents rather than on reporting about them.

This is the spoke's first T1 source and the first that describes the measurement regime rather than a measurement.

The standard the datasheets came from

IEC 60268-5, first issued 1972, treated a loudspeaker as "entirely passive." By its 2003 form it specified the Thiele/Small parameters, compliance, the anechoic-chamber and IEC-baffle setups, test boxes, harmonic distortion and noise signals — four stimulus types (sine, broadband noise, narrowband noise, impulsive) across free-field and diffuse-field conditions.

That is the document thiele-small-parameters and every datasheet this wiki reads implicitly stand on.

Why it was retired

Hill's summary is that "loudspeaker theory and practice in 2021 is much more advanced compared to the 2000 — let alone the 1980s." The specific failures:

It assumes a passive device. Active and digital loudspeakers — amplifier, DSP and driver sold as one unit — cannot be characterised by a standard written for a passive element. That is precisely the device camilladsp and the DSP-correction thread describe.
The measurement techniques are dated. No chirp or multi-tone stimuli, no 3D sound-field assessment, no defect detection.
Manufacturing QC and R&D used incompatible procedures, so a production line and a design lab could not compare numbers.

What replaced it

Three standards, not one:

IEC 60268-21:2018 — output-based acoustical measurements.
IEC 60268-22 — electrical and mechanical measurements on transducers (the part Hill sits on).
IEC 63034 — micro-speakers.

The new set adds chirp and multi-tone stimuli, 3D sound-field assessment, and a Rub & Buzz test for detecting mechanical defects — a class of fault the old standard had no way to name.

What this changes here

The spoke has treated T/S parameters as the design language and asked how they are measured (ts-parameters-wikipedia answered that with the impedance-sweep method). This source adds the layer above: the measurement method belongs to a standard, that standard was written for a passive device in 1972, and it no longer exists. Two consequences worth carrying:

A datasheet's provenance now has a date attached. Numbers produced under 60268-5 and numbers produced under 60268-21/-22 are not guaranteed comparable, and nothing in this corpus says how a builder tells which they are holding.
The active/DSP loudspeaker, which this spoke reached from the hobbyist side, is the same device the standards body reorganised itself around. The DSP bridge is not a niche — it is why the standard broke.

thiele-small-parameters · loudspeaker-driver · ts-parameters-wikipedia · camilladsp · geoff-hill · audio-engineering-society · iec · synthesis

Why this page is the closest this wiki will get

Added 2026-08-30. The obvious next step from here is to read IEC 60268-21, -22 and 63034 themselves. This wiki will not, and the reason is worth recording rather than leaving as a silent gap.

The standards are sold, not published. Three routes were tried across two hunts. IEC's text is behind a paywall, and iTeh, BSI, the ANSI webstore and GlobalSpec all return the scope paragraph and none of the procedure. AES sells its own standards the same way — every document routes through a store, and the standards committee's temporary PDF links expire. What is freely readable is vendor material: instrument-maker slide decks about the IEC project, an application note, a consultancy's guide to reading a chart. Useful, and not a specification.

So the position is structural, not a matter of effort: a wiki built from openly readable sources cannot hold the documents that define how its own numbers are produced. Every T/S parameter, frequency response and distortion figure in this spoke traces back to a procedure none of us can read. geoff-hill describing the standards he helped write, in a magazine, is the substitute — a participant's account standing in for the text. Worth remembering when a measurement here is compared against another: two labs can both claim conformance to a document neither reader has seen.

The edge asking for the text was dropped on 2026-08-30 at the curator's decision, after the third failed route.