DSP- one person's experience

marty

Well-Known Member
Apr 20, 2010
2,127
1,741
670
United States
DSP- One person’s experience

Among the very few premises to which all audiophiles would surely agree, is that good sound is often largely determined by the quality of the room in which the sound is being reproduced. Not far behind that premise is the general agreement that the nature of the specific speaker being used is also a large determinant to the resultant sound in a specific room. If it were easy to get great sound from any installation, there would be far fewer audiophiles reading columns such as this in an effort to maximize their listening experience. Acoustic suspension, ported box, transmission line, multifocal arrays, planar dipole, acoustic dipole, ribbons, and on and on are now widely offered speaker designs. Each offers strengths and limitations that need to be carefully optimized in the hope that the best sound for a given speaker in a given room can be obtained. There may be those who are fortunate and can achieve excellent placement of speakers in their particular room with wondrous sonic results. Whether done by “voweling” in, or obeying certain time-honored physical relationships or speaker placement to room dimensions, or intuition, or Quiji board, some folks are simply fortunate in that the maximum extraction of good sound from their speakers in their room can be achieved relatively easily. This does not mean that small changes of speaker location do not matter and that any old placement works just fine. That is rarely the case. Rather, these folks have been able to get satisfactory results by trial and error and careful acoustic observation. Sometimes these efforts are done in conjunction with room treatments of various sorts; some more costly and others not. Still, they are the lucky ones. For many of us, such good fortune is elusive, and thus we seek the use of other tools in an effort to obtain great sound from our gear.

One such tool that has become available at low cost is digital signal processing, or DSP. DSP can be employed as expensive outboard components, but is also readily available on many inexpensive receivers with multiple speaker surround installations. If you have a tendency to look down your nose at a 5 speaker, Blue- Ray, 1 box audio video system that can be optimized to any room with the push of a button for under $1000, you may want to think again as there is a reason that for most people, the extraordinary value that exists in those products is the very thing that is killing the high end, and not too subtly I would add.

Let’s start by asking the obvious question: who might obtain the greatest benefit from employing DSP in a high-end system, and who would not? For me, the answer starts with consideration of the most basic of all properties in any audio system- frequency response. After years of perspiration, inspiration and education, my first thought is that if you are fortunate to have a nearly ideal frequency response at the listening position, you may not benefit from DSP. There have been countless blog hours devoted to supporting and refuting the claims of what an ideal frequency response is, and thus I will share my view and state that in general, it approximates what B&K published a long time ago as typified in the curve shown here:


Screen shot..jpg

Two things are clear from this curve. First, it should be obvious that I am not a proponent of a flat frequency response in the listening position. For my taste, most audio systems are frankly too bright on top (some painfully so). Second, I learned a long time ago from Jason Bloom (of Apogee fame) that a mild enhancement of 2-4 dB in the 40-100 Hz region was euphonically pleasing and the B&K graphs reflects this, in part. The B&K graph is by no means carved in stone. My own system tends to be slightly flatter in the 300-10KHz range and slightly more rolled off from 10KHz-20KHz. But if you are in this general range, you may not need DSP to render a pleasing frequency response in your room.

However, if you are among the many that believe that the inclusion of any sort of tone control device is anathema to achieving good sound, then read no farther. I used to have such purist views, but they were soon cast aside by the results of careful listening. How many folks have speakers that have mid bass suck out from the floor effect in the 100-200Hz are? How many have searingly hot top ends? How may systems have the lovely “British sound” with slight reticence in the 2500Hz area? The answer is, a lot of them. And if that is your subjective preference, then so be it.

But consider the following situation. A very famous maker of a $100K loudspeaker system offers zero tone controls; nada, zilch, none. The manufacturer has no idea of that speaker is going into a living room environment with plush drapes and carpet or stone walls and large glass windows. Why on earth would you buy speaker from a manufacturer who is so arrogant that he insists that its “his way or the highway” for you to get good sound with his speaker in your room, when having the flexibility to turn the treble down say 0.6 dB or 1 dB might make the difference between spectacular and enjoyable and “my ear hurts”? Everyone knows you can get a 2k pair of Vandersteens that have tone controls for the midrange and treble drivers that offer this flexibility. If the mega expensive speaker designer had even a modicum of tone adjustments in his loudspeaker to optimize the interface with the client’s room, perhaps you wouldn’t need a tool such as DSP? Oh, the expensive speaker manufacturer will blow smoke and tell you he can’t compromise his design, even though his driver is working through a resistor of known impedance that if changed ever so slightly, will mean little deviation in the design, except to help fine tune the tweeter driver to a particular listener’s environment. (This is the sort of arrogance that is also killing the high end. But I digress). If you can obtain a pleasing frequency response at the listening position without using any tone controls or DSP device, than by all means, please do so. If not, you might want to play with one to see if the benefits outweigh the liabilities of such a unit.

In my system, I do in fact rely on DSP to modulate and shape my frequency response. But it is only one of the two main reasons I employee DSP in my system. The second reason, equally important, is to help integrate subwoofers into my main midrange/tweeter line arrays by employing digital time delay to the mid/tweeters. If not for the second reason, I might consider a good professional grade analogue EQ device. Lord knows these have improved dramatically in quality over the years and I would not eschew them based on what you thought they sounded like 20 years ago. If you had any idea of the sorts of high quality digital and analogue EQ that can be found in some of today’s finest recordings, you might feel differently as well. But that is a topic best left for another time.

I should probably first comment that I believe I have tied to incorporate subs with the “mains” (i.e. midrange/tweeter) in my system every possible way without DSP. I have tried passive crossovers and active crossovers with every known filter and slope, etc. until I was blue in the face. But until I employed DSP, I never felt I could integrate them seamlessly. The reason is simple. Only with DSP can one impart a precise time delay to the mains such that the sound of the mains and the subs arrive at the ear at the same time. I won’t go into details as to why any system that simply uses a phase control to align subs can’t do that, but they can’t. In brief, using a phase control may allow you to get alignment of phase with sustained notes, but never, I repeat never, with transient notes. The sound of the main is always ahead of the subs if the subs are placed behind the mains, and there is simply no way to advance the sound of the sub to get the transient leading note to arrive at the listener’s ear at the same time as the transient signal from the mains. The only way to do this, is to digitally retard the signal from the mains. In my system, the sound of the L and R mains are retarded 7.45 and 7.50 msecs respectively. What is particularly nice is that I can validate the signal arrival time from the mains and the subs (JL Audio Gothams) to my ears by using a nifty little impulse program built into the TacT 2.2XP (Padilla modified), which is what I use for my DSP correction. (The “mains” are Pipedream Hemispheres18/36. There is a simple 6db/octave crossover built into in the Pipedreams that performs the crossover function that splits the midrange and tweeter drivers signal accordingly. The midrange and tweeter drivers are driven by the same amp, in this case, VTL Siegfrieds running in triode mode. With the JL Audio subs containing their own internal amp, the system is therefore a bi-amplified system.)

So there you have it. For me, the use of DSP has been extremely valuable in providing pleasing frequency response (at the listening position), but absolutely critical in enabling me to integrate he Gotham subs seamlessly with the Pipedream Towers. Do I wish I didn’t have to use DSP? Sure. I’ve been fortunate to hear a few great systems that have nearly perfect frequency response and driver integration without the use of DSP. But they are far and few between in my experience. The impression I would like to leave is that for many of us, DSP offers an engineering compromise that can often have far more benefits than liabilities. Don’t be shy in trying one because you previously thought that only a “purist” approach is the way to go. You may be surprised and very pleased, as I was, in the pleasures to be derived from using DSP thoughtfully in your system.

One final note. I employee both phono and CD sources. While the TacT 2.2XP can be used as a preamplifier, I choose to bypass its preamplifier function and use it only for its DSP and digital crossover features. The majority of my system gain comes from using an ARC Reference 5 preamplifier which then drives the TacT. For the digital sect out there, this seems to offer proof positive that I have surely lost my mind. But at the end of the day, the only thing that matters is the sound and what works for you.
 
Last edited:

Ron Party

WBF Founding Member
Apr 30, 2010
2,457
5
0
Oakland, CA
Marty, I whole-heartedly agreed that the use of DSP is indispensible for proper subwoofer integration and, IMO, optimal bass reproduction. I am curious what are your thoughts on the Toole/Olive approach about positioning the subs at the midpoints of the side walls, which appears to be the second best implementation, or 4 subs, one in each corner of the room, which is the best approach. There is strong empirical, objective evidence that, even before DSP, correct sub positioning is the only effective way of dealing with nulls. And, there is strong objective evidence that, tested blindly, most people's subjective preferences align perfectly with the empirical evidence.
 

JackD201

WBF Founding Member
Apr 21, 2010
11,848
882
700
Manila, Philippines
Hi Marty, nice to meet a fellow member of "a very small minority". Great write up. I heartily agree. It is the end result and not the philosophy that counts. :)
 

audioguy

WBF Founding Member
Apr 21, 2010
2,778
59
545
Near Atlanta, GA but not too near!
You are the man!

As I have stated in another thread, way too many two channel listeners are of the less is more mentality. And I have visited hundreds of listening rooms to prove Marty's point and our solution did NOT have the ability to integrate a physically separate sub appropriately. I, too, once spent countless hours trying to integrate Dunlavy Subs with Dunlavy speakers and it never did work because, of course, the subs and mains could not occupy the same space at the same time, so they were not time aligned.

I have recently made the decision to purchase speakers that only go to about 40Hz and will be integrating my subs with them also with the Tact 2.2XP. (In house but in the box at the moment).

Stay tuned.
 

Mike Lavigne

Member Sponsor & WBF Founding Member
Apr 25, 2010
9,627
4,223
1,070
Marty,

i wholeheartedly agree with your fundamental point....that even though there are quite a few valid approaches to the end result, one must solve the speaker--room interface with respect to frequency response for superior performance. one can get lucky, do room mods, build the 'perfect' room, use DSP or in-speaker adjustments.....but the problem must be solved.

i also agree with your point about speakers presuming to be state-of-the-art need a way to to optimized for specific rooms and placement. otherwise performance cannot help being compromised.

my path started as 'build the perfect room from scratch'. but i've learned that only gets you part of the way. the perfect room only exists in theory. even with speakers which are highly adjustable i needed to do significant mods to my room before i was able to end up with a good frequency response. without an adjustable speaker i'd still be looking for 'other' solutions.

would those include DSP? i cannot say really without actually listening to what current SOTA DSP does to my analog signal path. you have done the work of going thru multiple possible solutions and arrived at your current direction. it's likely i might come to the same conclusions too. if you had been a primarily analog listener i wonder would you have used DSP? whatever our answers to these questions DSP is a credible solution to the problem.

my Evolution Acoustics MM3's do have built-in powered subwoofers which allow considerable adjustments as well as high frequency attenuation. i don't have the phase issues separate subwoofer enclosures have.

i wish time would have allowed a visit to your place last month when i was in DFW. i would really like to hear your room. i'd also enjoy it if you could return for another visit to my room; it is fundamnetally different that the last time you were here.
 

FrantzM

Member Sponsor & WBF Founding Member
Apr 20, 2010
6,464
11
505
You are the man!

As I have stated in another thread, way too many two channel listeners are of the less is more mentality. And I have visited hundreds of listening rooms to prove Marty's point and our solution did NOT have the ability to integrate a physically separate sub appropriately. I, too, once spent countless hours trying to integrate Dunlavy Subs with Dunlavy speakers and it never did work because, of course, the subs and mains could not occupy the same space at the same time, so they were not time aligned.

I have recently made the decision to purchase speakers that only go to about 40Hz and will be integrating my subs with them also with the Tact 2.2XP. (In house but in the box at the moment).

Stay tuned.

Hi

I am presently at a customer site, I've been there for the week-end .. Cut-over from an Analog PBX to an IP one... Banging my head on non-sense documentation written by a person who want to appear geeky :( I am almost done though, the customer will have his system tomorrow, hopefully with few bugs ...
I am interested in this thread, very much so but will not be able to write much on it today. Two things:

Toole and Olive got it right in the sense that more sub provide better smoother low frequency reproduction. I am not totally in agreement with the placement method, as asymmetry seems to work very well in the low frequencies... There again DSP can correct the phase and level of these subwoofers in ways that analog method can't ...

Second concerning speakers limited in the bass, not necessarily a good thing. more LF source increase thelikrlihood of smoother Low response curves ...

Aside from that I am quietly contemplating a DSP future, the Tact starts around $3700 MSRP .. so ...

later

Frantz
 

marty

Well-Known Member
Apr 20, 2010
2,127
1,741
670
United States
Ron,
I am not familiar with the Toole/Oiive approach. But I have attached an earlier picture of my room, shot prior to replacing the Pipedream subs with the Gothams. As you can see, placement of the subs is not feasible at the midpoint of the sidewalls, unless I want to roast a subwoofer in the fireplace!
Marty

WaxMedia1..jpg
 

Ron Party

WBF Founding Member
Apr 30, 2010
2,457
5
0
Oakland, CA
Marty, that is an absolutely beautiful room - one of the best I've seen. Consider me invited over if I ever make it to Dallas.


Toole, Todd Welti, Sean Olive, and Allan Devantier have undertaken a lot of research into the behavior of deep bass in rectangular, square and odd-shaped rooms. They undertook not only room measurements, but also blind listening tests to determine preference.

The conclusions apply mainly to rectangular rooms:

•Using one subwoofer, only one listener can truly enjoy smooth and extended deep bass response in a single seating location; listeners in other locations will hear different and irregular deep bass.

•Multiple subwoofers, optimally placed in rectangular rooms, will greatly reduce the seat-to-seat variations normally experienced using a single subwoofer, resulting in listeners in different seats hearing much smoother and more consistent deep bass.

•Even numbers of additional subwoofers work best.

•Two subwoofers or four subs deliver the greatest benefits in smoothing out irregular bass for multiple listening seats (four subwoofers are more effective than two). They concluded that 2 subwoofers give you about 90% of the performance that is possible, and that 4 take you about as far as you can reasonably expect to go."

•There seem to be no advantages to using more than four subwoofers in most rectangular rooms.

The most effective arrangement of two subwoofers in a rectangular room includes one at the front wall in the middle of the shorter dimension, and one sub at the back of the room in the same relative position. Alternatively, one sub can be placed along each sidewall (the longer dimension) at roughly the middle with the second sub against the opposite sidewall in the same relative location.

Four subwoofers were found to be most effective when two subs are placed at the middle location of the short wall at front and back and two subs at the middle location of each sidewall, opposite each other.

Placing one subwoofer in each of the room's four corners was found to be similarly effective. The third beneficial arrangement of four subs located two across the front wall (shorter dimension) and two across the back wall, opposite the two subs at the front wall.
 

RUR

WBF Founding Member
Apr 21, 2010
647
0
0
SoCal
Marty, do you happen to have that B&K curve in TacT file format? I can recreate it using the GUI, but if you've already done the heavy lifting..... It's considerably different from other "preferred" curves e.g. Harman's (slide 23) in that it's shallower (Harman's drops ~9dB! from 20-20kHz), pushes 20-~200Hz and the crossover @ 0dB is 2kHz vs 1kHz.

Is it just me, or are there an inordinate number of TacT users on this forum?:p
 

Steve Williams

Site Founder, Site Owner, Administrator
Marty, do you happen to have that B&K curve in TacT file format? I can recreate it using the GUI, but if you've already done the heavy lifting..... It's considerably different from other "preferred" curves e.g. Harman's (slide 23) in that it's shallower (Harman's drops ~9dB! from 20-20kHz), pushes 20-~200Hz and the crossover @ 0dB is 2kHz vs 1kHz.

Is it just me, or are there an inordinate number of TacT users on this forum?:p

It's not you. That is why I created the forum.
 

RUR

WBF Founding Member
Apr 21, 2010
647
0
0
SoCal
It's not you. That is why I created the forum.
Steve, you're still using a 2.2XP in combo with your Lamm analog pre, no?

I think I mentioned that I followed Marty's directions and tried my 2.2XP with my analog pre and it didn't do happy things for me in my system. :-( At this point, I'm all-digital with the 2.2XP+S2150M amp. Both fully Padilla-ized.
 

Steve Williams

Site Founder, Site Owner, Administrator
Steve, you're still using a 2.2XP in combo with your Lamm analog pre, no?

I think I mentioned that I followed Marty's directions and tried my 2.2XP with my analog pre and it didn't do happy things for me in my system. :-( At this point, I'm all-digital with the 2.2XP+S2150M amp. Both fully Padilla-ized.

Ken I am one of the members who doesn't use TacT. I have talked with Marty about it. He tells me not to. Last week Amir spent a few hours with me. I had a long discussion with him about integrating my Gotham subs as well as DRC. Amir felt I would notice improvements BUT at the expense of sacrificing much of the magic that the Lamm produces in the midrange and highs
Suffice it to say I am not prepared to do this. I would love to get a 2.2 into my system to A/B. Amir on his next trip down can hopefully bring his
 

audioguy

WBF Founding Member
Apr 21, 2010
2,778
59
545
Near Atlanta, GA but not too near!
Steve, you're still using a 2.2XP in combo with your Lamm analog pre, no?

I think I mentioned that I followed Marty's directions and tried my 2.2XP with my analog pre and it didn't do happy things for me in my system. :-( At this point, I'm all-digital with the 2.2XP+S2150M amp. Both fully Padilla-ized.

How did you have the system configured with your analog pre and the Tact?
 

RUR

WBF Founding Member
Apr 21, 2010
647
0
0
SoCal
How did you have the system configured with your analog pre and the Tact?
Marty was kind enough to walk me thru the setup, but the simple answer is that the analog preamp feeds the 2.2XP, which is set to "enable" digital amplification. In this configuration, gain control is accomplished via the analog preamp, with the TacT handling crossover, RC and speaker delays.
 

audioguy

WBF Founding Member
Apr 21, 2010
2,778
59
545
Near Atlanta, GA but not too near!
Marty was kind enough to walk me thru the setup, but the simple answer is that the analog preamp feeds the 2.2XP, which is set to "enable" digital amplification. In this configuration, gain control is accomplished via the analog preamp, with the TacT handling crossover, RC and speaker delays.

Does that mean that your digital inputs get converted to analog, fed to the preamp, fed to the Tact and then re-digitized?
 

RUR

WBF Founding Member
Apr 21, 2010
647
0
0
SoCal
Does that mean that your digital inputs get converted to analog, fed to the preamp, fed to the Tact and then re-digitized?
Yes, D/A in CDP, A/D/A in TacT - or that last conversion in an external D/A. In my case, I didn't feel that the analog pre added anything and I've gone to great lengths to avoid redundant D/A in my current setup: PS Audio PWT>2.2XP>S2150 = single D/A at the very end of amplification. 2.2XP analog output to the single f112 sub.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing