“Let our rigorous testing and reviews be your guidelines to A/V equipment – not marketing slogans”
Facebook Youtube Twitter instagram pinterest

Legacy Audio Interview: Room Correction, DSP & EQ Processing of Loudspeaker Systems

by April 07, 2013
Legacy Whisper XD System

Legacy Whisper XD System

We had the good fortune this month of sitting down with Bill Dudleston, Chief Designer and Owner of Legacy Audio to discuss DSP and EQ processing for loudspeaker systems.

Audioholics:  What are your views on auto room correction systems? Do you have any particular favorites?  Why?

Bill Dudleston: Power response equalization is certainly necessary at minimum. I find that consumers and audiophiles alike are not aware that their speakers in real rooms are exhibiting response variations that are typically +6 dB to -15 dB, particularly below 500 Hz.

I have worked with Audyssey, Dirac, TACT, DSPeaker, DEQX and am presently looking at Bohmer. Thus far, my favorite is Dirac with TACT running second.  I personally believe Mathias Johansson at Dirac is going about it the right way.  However, all of them definitely seem to require voicing in the end, so the ability to specify the target curve is critical.  The optimal target curve is somewhat dependent of the radiation pattern of the speakers, the RT60 of the room and the length of the correction window.


Legacy doubleHelix Pro Speaker System2

Legacy Audio doubleHelix Pro Speaker System

AH: What do you feel are the common mistakes and/or limitations of most auto-eq systems?

Bill: Units with limited processing capability may not have the bit depth to accomplish resonance control without compromising dynamic range. Another common mistake is not allowing the user to create a target function (i.e. to voice the system). Auto-correction to a nominal flat curve can result in skewed overall power response with edgy upper mids and brittle lower treble. It is also important to understand that mutual coupling occurs at low frequencies between your left and right loudspeakers. This can result in a thickening that may not be desirable if the summation response is not considered in the equalization.


AH: What microphone placements do you find useful?

Bill: Multiple positions are important.  All multi-driver speaker systems experience some degree of lobing, and hopefully the designer has done his job well.  Overcompensation of high Q (narrow) nulls can result in strange sounding results. While an average of microphone positions can be useful, I prefer to heavily weigh the area near the primary listener position. It is also useful to make measurements 6 inches above and below the listener’s ear height, and also 2/3 the distance to the listener position. By averaging these positions, over compensations are avoided.


AH:  Do you think room correction is applicable for full bandwidth or just bass frequencies?  Why?

Bill: I prefer to use DSP correction full spectrum when possible. While time domain correction is very subtle at high frequencies if the speaker is already relatively linear, amplitude correction assures that the left and right loudspeakers appear nearly identical to the listener. Since imaging information is derived from what can be subtle left and right differences, small corrections can yield great benefits. Modal correction at low frequencies can greatly improve the weight of a system with eliminating ‘room boom’.

Wavelength Processor Algorithm

DSP Matrix for Legacy Whisper XD System

AH:  Do any of your products apply DSP processing and/or equalization?  If so, please list the products and provide an overview of the systems.

Bill: The 40-bit matrix processor that we use provides for high pass, low pass, gain, polarity, and mixing of any of the 4 inputs into any of the 8 outputs which allow gain control, parametric EQ, and time offset.

Presently we using DSP as the multichannel crossover, gain matching between channels, followed by time synchronization of each band, driver correction, system summation correction, and finally equalization to the room.  I think ‘room correction’ is a bit of a misnomer as it is obviously not the room that is being corrected, but the loudspeaker being adjusted to the room’s unique characteristics.  The Legacy Helix, Whisper XD and Aeris all utilize DSP as well as our Goliath subwoofer system. In our Professional Line, the doubleHelix, The Wall, Beast & Theater Towers all take advantage of this technology.

The 4 x8 matrix, for example, allows one to feed the internal ICEpower® amplifiers driving Whisper XD’s primary and secondary woofer arrays and synchronize them in the time domain with the upper frequency array. It additionally leaves a channel available to control an outboard subwoofer or a stereo pair of subwoofers.  Following are details of the Legacy Wavelaunch processor.  It is built by Xilica of Canada.  All algorithms are created here in house and are specific to the loudspeaker model and system it is used within.

Inputs and Outputs

  • Maximum Level: +20dBu
  • Type: Electronically balanced XLR, output relays

Audio Performance

  • Freq Response: +/- 0.1dB (20 to 30kHz)
  • Dynamic Range: 115 dB
  • Crosstalk: < -100dB
  • Distortion: 0.002% (1kHz @ +4dBu)

Digital Audio Performance

  • Processor (DSP): 40-Bit Floating Point
  • Sampling Rate: 96kHz
  • Analog Converters: Super Performance 24-bit
  • Propagation Delay: 1.5ms

Audio Control Parameters

  • Gain: -40 to +15dB in 0.25 dB steps
  • Polarity: +/-
  • Delay: Up to 650 ms per I/O

Equalization

  • Parametric EQ:  8 on each input and output
  • Gain: -30 to +15dB in 0.25dB steps
  • Graphic EQ:  31 Band 1/3 Oct steps on each input

Dynamic Control

  • 1 Limiter per output channel
  • Threshold: -20 to +20dBu
  • Attack: 0.3 to 100ms
  • Release: 2 to 32x the attack time

Crossover

  • 2 Individual filters per I/O
  • Type: Butterworth, Linkwitz Riley, Bessel
  • Slopes: 6 to 48dB per Octave   

 

AH:  Do you believe full-range loudspeakers and dedicated powered subwoofers can co-exist in a multi-channel system?

Bill: An overwhelming YES. Since distortion is proportional to displacement and displacement requirements increase by a factor of four each time the frequency falls an octave lower.  I’ll take all the piston area I can get to maximize dynamic range, minimize distortion and improve transient behavior.  I think people are often surprised when they hear four subwoofers properly configured in the same room with a pair of full-range speakers. All woofers are traveling a much shorter distance and overshoot is greatly reduced as pressurization/rarefaction occurs at the lowest frequencies. The mutual coupling and boundary reinforcement enhances the radiation impedance effectively raising output and further reducing distortion. The key to success is proper level & phase adjustment. Four or eight bass drivers can play just as softly as two, they just do it cleaner.


AH:  How would you manage the bass in such systems?  (ie. run the main channels fullrange, or run the bass portions of the main channels as part of the mono summed Bass + LFe channel that connects to the other powered subwoofers)

Bill: Since the majority of Legacy speakers are full-range, I like to mix the LFE output into the main channels on our larger speakers such as Helix, Whisper XD, Aeris, & sometimes Focus SE. When using side subwoofers, I love to phase them in quadrature; leading 45 degrees on the left and lagging 45 degrees on the right. This reduces the intensity of axial resonance measurably.


AH: Do you prefer to individually EQ the bass of each subwoofer or do a global correction to all subs in a system?

please see:  Setting up Multi-Subwoofer System for reference)

Bill: I prefer to EQ each subwoofer individually first relative to its placement and then adjust the final response relative to the summation. Otherwise, things can get a bit chunky & thick resulting in a loss of attack and decay.  I have set up large theaters with 3 of more side subwoofers firing sequentially.   The bass propagation is quite natural this way for a wide range of listener positions.

 

AH:  Where can fellow Audioholics experience Legacy demos of such systems?

Bill: You can hear the full line of Legacy speakers at our factory in Springfield, IL or our Los Angeles showroom. We also have about a dozen additional domestic showrooms and a growing number of international distributors. One extraordinary venue for our commercial line is the state-of-the-art Giant Screen Theater at Peoria Riverfront Museum. This theater utilizes 5 of our doubleHelix loudspeakers across the front and includes 40 DSP channels. More information on this particular install is available here.

 

Many thanks to Bill Dudleston of Legacy Audio for participating in this enlightening interview on DSP and EQ loudspeaker room correction.

 

About the author:
author portrait

Gene manages this organization, establishes relations with manufacturers and keeps Audioholics a well oiled machine. His goal is to educate about home theater and develop more standards in the industry to eliminate consumer confusion clouded by industry snake oil.

View full profile