Time is of the Essence

The step response of a loudspeaker is an expression of how well it can reproduce a sudden and sharp transient in amplitude across the audible frequency range. As a note is first struck, an initial short complex burst of mixed frequency energy occurs making time coherence essential for perceiving an instrument as authentic.

The thin steep vertical brown line represents the TCA-M step response of a 10Hz – 24KHz frequency sweep. There is just 0.05ms between each vertical gridline and thus the TCA-M step response transition occurs in less than 0.1 milliseconds.

The measured step response provides a useful view of how the complete loudspeaker integrates its different frequency ranges in time. The TCA-M exhibits one dominant, clean transition in less than 0.1 milliseconds, reflecting the close acoustic time alignment of its bass, midrange and treble sections.

Psychoacoustics of transients and identifying sounds

It turns out that frequency response does not sufficiently describe the qualities that enables us to identify instruments or other sounds we easily recognise.

"We now know that we normally identify what instrument is being played, almost entirely by starting transients, the complex patterns of rapidly changing sounds at the beginning of notes. That surprises some people; it is contrary to popular — and Helmholtzian — belief. But that is how we identify all natural sounds, noises and words, by transient pattern changes.

In general, one cannot get any orthodox instrument to produce sound without giving it a 'kick’; even the most gentle start to a sound is a transient. Pickering (1989) showed that it can take a fifth of a second or more for a violin string to settle down to steady vibration, and much depends on how the note is started (Beament, 1997). And the change from one pitch to another is a transient, however legato the music is played. Percussive sound, including that of pianos, is entirely transient.

A recording of a piano or guitar played backwards is unrecognisable. If we make magnetic tape recordings of people playing normal steady-pitched notes, and then remove the starting transients of the notes, the first fifth of a second or less, it is difficult to identify what instrument was responsible for the sound. We are not used to identifying instruments by the steady-pitched sound which follows the transient, and if we splice the starting transient of one instrument onto the continuing sound of another instrument we can be confused, identify the instrument by the transient and think it has rather odd tone.

If an instrumentalist plays more than half a dozen notes per second, the sound consists entirely of starting transients. We can identify the instrument from these transients, and we can discern pitches, because there is sufficient steadily repeating vibration in the transients. But transients have no regular pattern of sizes of their harmonics, and random noise is also present^[1]".

[1] How We Hear Music: The Relationship Between Music and the Hearing Mechanism, By James Beament

Conventional loudspeaker design step and transient response

A conventional audio system consisting of a multi-way loudspeaker with a passive crossover presenting a variable load to an external power amplifier, is unlikely to be phase linear. Consequently an ideal step response is not possible, as seen in the example above, and results in some amount of variable group delay increasing at low frequencies and a related relatively poor transient response.

Frequency-dependent phase rotation and delay can alter the relative timing of the fundamental, harmonics and transient components that together define the attack and character of sounds such as a snare drum or piano note.

The TCA-M combines close acoustic time alignment, near-constant group delay and controlled frequency response through integrated active DSP.

The measured TCA-M group delay remains nearly constant from approximately 30 Hz to 24 kHz, with an increase of less than approximately 2 milliseconds towards 20 Hz.

Group delay and phase-linearity

Group delay describes how different frequency components of a signal can emerge from a loudspeaker at different times, even though they subsequently travel the same distance to the listener.

A time-coherent loudspeaker maintains group delay as nearly constant as practicable across its operating bandwidth, thereby preserving the temporal relationships within the recorded signal. Achieving this is technically demanding because most loudspeakers divide the audible spectrum between multiple, physically separate drivers, each optimised for a particular frequency range. The drivers, their acoustic alignment, crossover filters and enclosure behaviour can all introduce frequency-dependent phase rotation and delay.

In conventional passive loudspeakers, the electrical crossover, the acoustic responses of the drivers and their physical separation interact to produce a frequency-dependent phase response. Designers must therefore balance frequency response, directivity, driver integration and phase behaviour. Consequently, even well-designed passive loudspeakers will generally not achieve constant group delay or full time coherence across the entire audible range.

Research demonstrates that group-delay variation can become audible under suitable conditions, with sensitivity depending strongly on frequency, signal characteristics and the magnitude of the delay. TCA therefore treats low and frequency-dependent group delay as a parameter to control deliberately.

To learn more, please download our “Room Acoustics, Psychoacoustics and TCA-M Design Concepts” white paper and the Features and Technical Specification.