In addition, mDSP enabled me to flatten the peaks below 40Hz for better management of the dominate room modes. We have leveled the playing field. Also notice the 80Hz dip in the front row was significantly reduced with the 5th sub added. Without mDSP, I wouldn't have been able to do this. The independent adjustment of delay and crossover slope in the mDSP device afforded me this luxury. We now have a bass response from 15Hz to 100Hz +/-8dB across all six seats and this is done with NO added passive bass trapping!
For comparative purposes I wanted to show the difference at the primary listening position between running just my front speakers bi-amped vs having the other subwoofers and SFM engaged.

Umik 1 Calibration File

Front Speakers (Red trace) vs All Subs + mDSP (Blue Trace) @ Primary Listening Position

Subterranean Bass

I thought it prudent to note that these frequency graphs extend down to 10Hz. Most in-room measurement graphs we've seen published in journals and manufacturer websites stop at 20Hz with the bass already down 3dB or more. While some folks de-emphasize the importance of sub frequency bass below 20Hz, I personally feel its importance is underestimated. The reality is, it takes multiple high excursion subs and lots of power to dig this deep. But, if your subwoofer system is capable, you will reap the rewards in LFE-content-driven movies and some multi-channel music too. The scene in Dark Knight Rises, you can actually feel your spinal column shake when the Batwing takes off in our reference room. It's eerie but quite intoxicating and in my opinion the true hallmark of an extreme Bassaholic.
How was I able to achieve this?
  • I added a 5th sub to deal with mother-in-law seat (necessary in my asymmetric room). For ideal room shapes, typically 2 or 4 subs are sufficient.
  • * I have each sub play the same signal (LFE + summed mono bass for speakers set small).
  • Using Superposition – I added one sub at a time and made sure they integrated, then I got the best summed response of all subs blended with the main speakers.
  • After the subs were integrated with each other to yield the best possible summed output, I worked on flattening their combined response using multi-band PEQ via mDSP.
Abbreviated Equipment List (related to bass section)
  • Processor: Denon AVP-A1HDCI Preamp/Processor
  • Amplifiers: Denon POA-A1HDCI 10CH amp / 2x Emotiva XPR-1 mono-blocks, RBH SA-500 sub amp
  • Speakers: Status 8T and 8C (front LCR) bi-amped bass modules connected to sub L/R channels, RBH 66-SE/R (side surrounds), RBH SI-61/R (back surrounds)
  • Subwoofers: 2x Velodyne DD-15+, 1x RBH SI-1010 in-ceiling subwoofer
  • Bass Management: mDSP 2x4 balanced ($135 with power supply)
*Editorial Note about my Bass Management Configuration:
In myparticular case, my main speakers are bi-amped with their bass portionsused as separate powered subs configured as stereo(L/F + summed bassfrom 'small' channels and LFE). I run the subs for my front speakersout to 250Hz for optimal integration with the mid/tweet satellitesection while the other powered subs in my setup get crossed over at thetypical 80Hz 24dB/octave. THX recommended setting. This worksextremely well for my situation but we typically recommend sending monosummed bass to all subs crossed over at 80Hz.

Time Domain Analysis and Waterfall Graphs

There are some that insist that frequency response analysis alone is NOT a good indicator of great bass response. Unfortunately for them, the science is NOT on their side.
flattening the bass in the frequency domain also improves its corresponding time domain behavior too.
As the folks at Harman have proven in the past via various white papers and AES contributions regarding the benefits of multi-sub, loudspeaker transducers, including woofers and their enclosures, are minimum-phase devices/systems, and therefore respond to accurately matched parametric equalization based on anechoic data. In small rooms, the individual resonances behave as minimum-phase systems to the extent that one can attenuate prominent peaks in a high-resolution room curve and thereby both attenuate the peak and dampen the ringing. Again matched parametric EQ is needed. Above the room transition frequency (roughly 300Hz), EQ becomes less predictable since it’s difficult to separate the speaker and room contributions. At low frequencies (< 300Hz), the room dominates the response which means if you fix the frequency response, then the time domain response is also corrected since they are related. Smooth bass response means there will also NOT be any excessive ringing within the bandwidth of concern. In other words, decay times are directly correlated to frequency response in the subwoofer frequency bands.
Note: Transition frequency – the frequency at which standing waves dominate the loudspeaker response in small room acoustics.
Time domain behavior is often represented in waterfall graphs, but such graphs are easily manipulated as can be seen in CH 13 of 'Sound Reproduction: Loudspeakers and Rooms' by Dr. Floyd Toole. Floyd discusses how waterfall graphs can be manipulated to appear bad by jacking up the frequency response resolution at the expense of the time domain resolution, while the latter is what we are actually most concerned about in such graphs.
To illustrate ringing in a waterfall graph, here is a measurement at my primary listening seat with just one subwoofer playing with NO EQ.

Waterfall Graph: Primary Seat, 1 sub, no EQ

Here is the corresponding frequency response measurement.

Frequency Response Graph: Primary Seat, 1 sub, no EQ

Notice the prominent peak in the 45-65Hz region. There is some ringing going on here. But let's see what happens at the exact same seating position with multi-sub + DSP using the exact same settings in my REW measurement program to generate the corresponding waterfall plot below.

Waterfall Graph: Primary Seat, 5 sub, EQ

Frequency Response Graph: Primary Seat, 5 sub, EQ

Notice how the response is much smoother now. The ringing has been damped which corresponds to the much flatter frequency response behavior. The science works!
But what about the other seats? How do we know the time domain behavior is good there too? Remember that we fixed the frequency response for ALL seats so it's good at those seats too but the proof is in the pudding so let's show a couple of examples.

Waterfall Graph: Front Right Seat, 5 sub, EQ


Waterfall Graph: Back Center Seat, 5 sub, EQ

Once again we see well-damped time domain behavior for the various seating locations, all of which correspond to their smooth frequency response measurements.

The Bottom Line

Claims that there is only one way to skin a cat, or in this case, fix your bass, are unfounded and frankly intellectually dishonest or naive. Bass trapping is one method, multi-sub + DSP is another, and a combo of both can also be an effective solution. Be leery of anyone who tells you otherwise. The science has proven the benefits of multi-sub with SFM (or in our case manual delay and EQ compensation via mDSP) as a viable and often preferred solution to improving bass response for reproduction systems in small rooms. When fine-tuning bass in your system, careful examination and improvement of frequency response behavior across all of your seats is a key metric of achieving great bass.

The End Result

Bass nirvana was finally achieved in the Audioholics Showcase theater room system through multi-sub plus equalization through the mDSP device (review forthcoming). I now have smooth, even, tight, extended bass for EVERY SEAT, not just a couple of preferred seats. Bass level and tonality sounds similar no matter what seat you chose and this is easily demonstrated using the Diana Krall - Love Scenes DTS CD, which is a very bass-intensive recording. The mother-in-law seat I often talk about in my room has been terminated once and for all. Using EQ, I could still flatten the 20-30Hz region a bit more, but quite frankly, I like the sound of that region slightly elevated since it adds to the tactile feel in scenes like when the Batwing takes off in Dark Knight Rises. Another take away from this is if you flatten out the bass and fix your seat-to-seat consistency, research shows that a boost of up to 5dB in the bass region is perfectly acceptable and often preferred by home theater enthusiasts.
As good as the bass is in my system, it doesn't mean I have completed my journey. Anyone who knows me appreciates that I am not one to rest on my laurels. I still plan on investigating the benefits of employing modest amounts of low frequency corner bass trapping with some Gik Acoustics products to see if that further improves the bass in my system. With eight 10' and two 15' bass drivers, and over 7 kwatts of power behind them, I certainly have enough power and piston area for the subwoofer channel to sacrifice a little efficiency if I can further fine tune the bass sound in my room. Would you expect anything less from a true Bassaholic?

Acknowledgements:

Many thanks to Dr. Floyd Toole, Kevin Voecks and Todd Welti of Harman for their contributions and peer review of this article.