Group Delay in Passive Crossovers and Horn Loudspeakers

Measuring the group delay of a passive crossover and its audibility with horns concerns timing rather than a single headline specification. In a multi-way loudspeaker, the woofer and compression driver should deliver adjacent frequency bands with compatible phase and nearly equal arrival time. Small errors can blur transients, move the apparent centre of the stereo image, or cause tonal cancellation even when sensitivity and frequency response look excellent.

For Australian audiophiles, evaluation often begins in a specialist dealer room in Sydney, Melbourne, Brisbane or Adelaide, then continues in a home or apartment with different acoustics. Weekend auditions and evening listening at home are familiar buying habits, while open-plan rooms, nearby traffic and Australia’s heat and humidity make repeatable measurements especially valuable.

Why group delay belongs in crossover evaluation

Group delay describes how phase changes with frequency. In angular form, it is calculated as τg = −dφ/dω using unwrapped phase in radians. A network can have a modest average delay yet remain difficult to hear; what matters is whether delay changes rapidly across the crossover and how one band relates to its neighbour.

A passive crossover adds inductance, capacitance and component losses, so its phase response may differ from a textbook filter slope. The useful target is compatible phase at the handover, small off-axis level differences and stable imaging when both drivers operate, rather than the smallest possible group-delay figure.

Measuring the passive network electrically

Disconnect the drivers and terminate each crossover output with suitable non-inductive loads approximating the drivers’ impedance in the crossover region. A two-channel impedance analyser can reveal each branch’s phase and attenuation; a conventional terminal sweep mainly shows their combined effect.

Sweep widely enough to include the whole transition, export the raw phase and unwrap it carefully before differentiating. Repeat the sweep and check for cable delay, switching artefacts and numerical noise. Smoothing is valid only after phase-wrap errors have been corrected.

Component tolerances matter too. Film capacitors, air-core inductors and lead lengths can move each branch’s electrical centre. Comparing measured equivalents is more reliable than trusting catalogue values alone.

Measuring the complete loudspeaker acoustically

Place a calibrated, low-level microphone on axis at a fixed distance. Where access permits, record each driver separately and then together; aligning the traces by cross-correlation exposes relative arrival time, while the combined sweep shows whether magnitude and phase sum smoothly.

A single impulse response is a useful cross-check because pre-echo, a secondary arrival and cabinet or room reflections can be distinguished more easily than in a polished graph. Average several readings and move the microphone away from a wall before accepting a result.

For a horn-loaded system, the relevant reference is each driver’s acoustic centre, not the cabinet front. The compression driver’s throat, horn flare and apparent source location can place one band behind another even when both units appear flush.

Horn geometry, alignment and cabinet construction

Sunship’s systems use TAD-Pioneer compression drivers and woofers. A compression driver adds operating latency before energy reaches the horn throat, and a horn flare then spreads it through space. A bi-radial wooden horn provides controlled expansion and directivity, helping create one coherent acoustic origin rather than separate sources.

Sunship’s time-aligned passive crossover can compensate for fixed geometric offsets, but it cannot move a cone backwards. Alignment depends on the complete phase response, the horn’s effective source position and the woofer’s acoustic centre, so component values alone cannot prove correct timing.

Cabinet construction affects the measurement environment even when it does not set electrical group delay. Heavily braced birch plywood reduces panel resonance, while horn geometry and cabinet boundaries influence diffraction and source position. Sunship’s project gallery shows these elements as integrated systems.

When crossover timing becomes audible

Relative phase errors between adjacent bands can create comb filtering, level dips and partial cancellation. Unequal arrival can smear attacks or pull the sound away from the central axis. Listeners may describe this as reduced micro-dynamics, a thin centre or unstable placement, although the symptom depends on the recording.

Horn systems can make such faults easier to hear because high sensitivity and controlled directivity expose direct transients with fewer diffuse reflections. Good alignment can therefore sound precise and immediate. Directivity does not correct phase, however, and a low group-delay reading cannot guarantee an enjoyable loudspeaker.

There is no universal audibility threshold. Level, programme spectrum, reflections, listening position and whether delay is constant or frequency-dependent all shape perception. Controlled comparisons with familiar recordings are more informative than judging a scalar value in isolation.

Buying and using a custom system in Australia

Sunship’s Berlin demonstration room provides a controlled reference, whereas an Australian living room may place a horn near a wall, window or television. Small changes in cabinet angle or distance can alter the low-frequency boundary and image, so measurements should be repeated in the intended position.

Specialist dealers and custom installers can assess room suitability and coordinate a complete system before shipment, but regional buyers also need clear answers on freight, insurance, replacement parts, local warranty and installation responsibility. Horn mouth clearance and a stable base matter from the outset, since room placement cannot compensate for every system-design fault.

Goods supplied by an Australian dealer are covered by the Australian Consumer Law’s guarantees of merchantable quality and fitness for a disclosed purpose. Active amplifiers or processors may also fall within electrical safety and RCM requirements. For projects crossing into broadcast or professional work, this broadcast legal overview offers useful context for checking obligations beyond domestic playback.

A repeatable measurement and listening workflow

Before capturing data, confirm terminal arrangement and channel mapping with Sunship’s terminal bi-wiring guide. Keep microphone position, gain and processing consistent, and retain raw sweeps as well as processed graphs.

Measurements to prioritise

• Establish a repeatable baseline with fixed gain, averaging and no DSP.
• Capture wide sweeps and impulses for each driver and the combined output.
• Unwrap phase carefully and inspect the crossover region for sharp delay peaks.
• Repeat after repositioning the cabinet to separate design effects from room effects.

Listening checks for timing problems

• Use familiar recordings containing sharp transients, sustained bass and centred voices.
• Compare stereo placement and image stability, including mono playback.
• Move the listening position and note whether clarity or focus changes.
• Compare with a reference at equal level without altering system settings.

The strongest evidence is agreement between phase, impulse and listening results. Electrical delay alone is less informative than relative driver timing, crossover cancellation and behaviour in the final room.

Observation	Likely meaning	Useful comparison
Smooth phase and a clean main impulse	Drivers are combining coherently	Repeat at the final listening position
Rapid delay peak near the crossover	Component or branch sensitivity	Compare branches and measured parts
Secondary pulse before or after the main arrival	Driver offset, horn path or reflection	Cross-correlate traces and change microphone position
Repeating level notches in the combined sweep	Relative phase cancellation	Compare phase response and crossover slope
Strong changes after cabinet movement	Boundary or diffraction effects	Compare cabinet angles and positions