Loudspeakers, because of their uncompensated mechanics, introduce the vast majority of errors and distortion that are heard in our audio systems, and as such, any major improvement we can make to their native (anechoic) performance is likely to be overwhelmingly more beneficial to the listener than whatever else can be improved in the audio chain.

By contrast, our ‘audiophile’ loudspeakers introduce about 1.5% distortion, even greater break-up and crossover distortion, and frequency-response errors of +/- 3dB, which equates to an extraordinary 4:1 power error. Timing or phase coherence is equally disturbed at different frequencies to the extent that mid-bass, for example, often lags the midrange by milliseconds (sound travels about one foot in one millisecond).

We’ve had no choice but to overlook these problems because designers have been constrained by a passive loudspeaker regime that has no provision to compensate for the inaccuracy problems of timing and frequency response, nor the ability to reduce distortion by properly using the ‘active’ speaker regime chosen by pro-audio for decades.

DEQX-HD introduces a new generation of DSP technology for loudspeakers that can deliver being there realism affordably using the new HDP-3 Pre-amp processor. DEQX-HD improves the audio experience by correcting the timing errors (group-delay and phase), as well as amplitude balance (frequency-response) without compromising sound quality. To significantly reduce crossover and break-up distortion, there is a need to ‘go active’ and the HDP3 can provide up to a 3-way active crossover to drive separate amplifiers for each type of driver, typically bass, midrange and tweeter. 

DEQX can be implemented at several levels, the ultimate being DEQX-HD, DEQX's High Definition ‘active’ implementation. This addresses the most serious problems at their source - the loudspeaker’s error-prone and distortion-prone drivers. The introduction of DEQX-HD’s active implementation for any speaker design, regardless of cost, can result in dramatic improvements in performance.

Where DEQX-HD’s intended requirement to ‘go active’ is not practical, significant improvements can still be realised with simpler implementations, such as correcting an existing passive speaker, perhaps integrating one or two subwoofers (which can also be corrected and time aligned with main speakers), then adding room correction and preference EQ. The HDP3’s remote control also allows 3-band tone controls, including one fully parametric band.

A problem with traditional DSP audio processing is its tendency to sound ‘digital’. In addition to its S/PDIF and AES3 digital inputs, the HDP-3 Preamp-processor contains balanced and unbalanced analogue inputs utilizing highly transparent Analogue to Digital Converters (ADCs) running at 24-bit 96kHz HD resolution. Two SHARC 32-bit floating-point DSPs provide internal processing resolution in the order of 160dB, followed by audiophile 24/96 Digital to Analogue (DAC) converters and six 256-step resistor ladder analogue volume control chips to maintain the full dynamic range at low listening levels. Nine separate power-supply regulation stages, including four to provide the main analogue rails, provide very low measured distortion and fatigue-free listening transparency. Almost half a million lines of software refined by engineers over twelve years distinguish DEQX-HD from other digital processors and crossovers.

Using the HDP-3‘s active features, DEQX-HD’s ‘linear-phase’, rather than ‘minimum-phase’, filters virtually eliminate crossover distortion while dramatically improving natural dispersion. The HDP3 is one of very few units available delivering steep linear-phase crossovers. In traditional high-end passive speakers, the only way to achieve close to a liner-phase crossover filter is to limit the filter to only one pole (6dB/octave slope).

Although single pole crossovers can provide near linear phase crossovers that audiophiles appreciate, they provide almost no protection to tweeters and midrange drives from the excessive excursion that causes beak-up distortion, compression and reduced resolution. They also promote ‘beaming’ of higher frequencies from the bass and midrange drivers which cause uneven dispersion that sounds unnatural in most rooms.

Compared to the significant phase distortion caused by steeper than single-pole filters (6dB/octave), or the significant different problems caused by shallow single-pole filters (albeit more linear-phase), the HDP-3 provides typically 8-pole filters (48dB/octave) while maintaining linear-phase and quarantining undesirable driver excursion. This retains each driver’s ‘best case’ resolution while providing some driver overlap. The benefits are significant reductions in dynamic, crossover, break-up distortion and compression, with corresponding improvement in natural dispersion, resolution, three-dimensional imaging, and volume capacity.

Other DSP active crossovers typically simulate traditional analogue filters and some also provide parametric EQ and ‘timing correction’. Normally these filters are not linear-phase, and ‘timing correction’ simply allows one driver to be delayed (time-aligned) with another. While driver timing alignment is essential, it cannot correct the numerous ‘group-delay’ errors (frequency-related phase) introduced by the drivers themselves. Here it is necessary to adjust phase and amplitude to different degrees for numerous different frequency ‘groups’.

The HDP-3 achieves this automatically by analyzing the pseudo-anechoic (native) measurements of each speaker driver in active designs, or the overall speaker response in passive designs, and by then creating corresponding ‘correction filters’. These ‘filters’ are FIR (Finite Impulse Response) filters which are quite unlike traditional EQ filters because, in their attempt to correct frequency-response, the FIR filters actually maintain linear phase, rather than, like traditional EQ filters, introduce non-linearity. This returns the actual measured acoustic timing coherence to correspond to the original audio source, while frequency-response is corrected in similar detail. In effect, the measurement and analysis procedure detects which frequency groups are being comparatively delayed and DEQX software then appropriately delays the faster groups till all groups are again coherent.  

It is also necessary to correct frequency-response errors, so that the volume of every pitch needs to be reproduced equally. Speakers are unique in their inability to provide an accurate frequency-response, even when listening directly in front of them, let alone ‘off-axis’.  In contrast, natural sound sources, such as musical instruments and voice, typically distribute frequencies more omnidirectionally. DEQX-HD provides a more balanced distribution of frequencies, and therefore a closer approximation to natural sound dispersion in the listening room, which further enhances the being-there experience. 

A DEQX ‘speaker correction’ measurement ignores room reflections in determining accurate native (anechoic) speaker performance in the time domain (phase and group-delay) as well as frequency domain (frequency-response). After a speaker’s native  frequency-response and dispersion have been corrected, there is less requirement, except in the bass region, for ‘room correction’, which compromises phase accuracy when applied to midrange and high frequencies. Once a speaker is corrected, DEQX measures its behavior in the room from the listening area/s and the results are displayed graphically using a PC for set-up. Typically only three to six bands of DEQX’s parametric-EQ are required to resolve bass in room issues, which can be set automatically or manually to preference.

Beyond DEQXs timing and frequency-response correction, the HDP-3 allows seamless integration of mono or stereo subwoofers, whereby the crossover from the main speaker and subwoofer/s can be aligned in time using measured or real-time adjustment by ear.

Up to four different combinations of speaker and room corrections can be stored in the HDP-3 for instant recall. These can include an ‘uncorrected’ (bypass) mode to remind you (and demonstrate to friends) what your system used to sound like before the HDP-3. The HDP-3 remote control also provides a three-band tone control, so without a PC connected, you can store 100 different EQ settings for different recordings or simple room EQ. You can use the bass, mid and high default settings, or for more detailed control, you can move a band’s centre frequency across 10 octaves, with semitone resolution, while its bandwidth can be adjusted from just one semitone to four octaves, providing cut/boost in one dB increments.

The remote control is also used to select inputs: two analogue and two digital.