April 2015 Toole video on sound reproduction

[amirm]

And who knows what bass sound the producer and engineer of a particular recording is striving for.

That is true and is a fundamental flaw in audio reproduction. See http://www.madronadigital.com/Library/Room Equalization/Room Equalization.html

But that isn't a ticket to have any response we wind up having in our rooms. It is best to get a smooth one and then have adjustable target curves for different content. It is the best we can do given the cards that have been dealt to us.

 

[Fitzcaraldo215]

And who knows what bass sound the producer and engineer of a particular recording is striving for.

That is no doubt true, but whatever they were striving for is unlikely to be the typical roller coaster response with one room's own particular signature of peaks and troughs. As Amir says, flatten the response first, then add or subtract a shelf of bass energy to your own satisfaction. Personally, I am quite happy with the perceived realism of the stock Dirac Live target curve I use. I never fiddle with it, because to me it delivers awesome, non-boomy deep bass that plausibly replicates what I hear at live classical concerts.

As for me, I think DSP EQ one of the truly greatest breakthroughs ever in home audio reproduction. I can no longer imagine being without it.

 

[FrantzM]

That is no doubt true, but whatever they were striving for is unlikely to be the typical roller coaster response with one room's own particular signature of peaks and troughs. As Amir says, flatten the response first, then add or subtract a shelf of bass energy to your own satisfaction. Personally, I am quite happy with the perceived realism of the stock Dirac Live target curve I use. I never fiddle with it, because to me it delivers awesome, non-boomy deep bass that plausibly replicates what I hear at live classical concerts.

As for me, I think DSP EQ one of the truly greatest breakthroughs ever in home audio reproduction. I can no longer imagine being without it.

+1 but have not experienced Dirac but can attest to what can be achieved by judicious use of dSP in the bass...

 

[esldude]

Indeed. ..................................................

Here is the other thing electronic correction can do: match levels of channels accurately and include delays to make the arrival from both speakers the same. These two alone, with no EQ, can make a remarkable difference in imaging and overall perception of sound.

Yes, I once tried to get someone to get room correction just for those two reasons, telling them how effective it was. They didn't listen to me. I did send some pulses thru the system and record with a mic to use the result to carefully position the speakers physically. All things considered it is better than you can do with a measuring tape.

 

[Ronm1]

What is remarkable is that only part of the effect is in bass. A lot of it extends to higher frequencies! The ringing, continued playing of the loud bass clouds and muddles the higher frequencies. Voices, ambiance, etc. all become clearer without doing a thing to do them!

How true this is!! Brushes on cymbals, ability to distinguish diff drums. I was amazed at how clearer the upper freq's were, could tell the hall ambiance diff on good recordings near stage vs further back and dialog with movies improved. Not subtle at all.

 

[NorthStar]

Alright, there seems to be a common agreement; good, because I too believe that the most ultra hi-end system in the world in the best acoustically treated room can still benefit from judicious DSP correction/EQ in the bass. ...And from various readings here and there Dirac Live is a good one. And it can be manually mastered with different target curves.

It is in the digital domain with a combination of FIR and IIR digital filters, ...Audyssey is FIR only, and ARC (Anthem) is a combination of both but differently implemented as compared to Dirac Live. Music reproduction and movie reproduction are equally important.

Now, what to do with audiophiles who strictly have an analog rig (turntable and open-reel-tape deck)? Do they want to do A/D and D/A in their system?
We have members here who have some serious analog rigs, what are they doing, and what are they missing...sound reproduction wise?
Amir mentioned that aspect prior (analog systems), and I was expecting that reply from someone when I asked my question. ...It just happened to be him who referred to it.

Floyd E. Toole, is he familar with Dirac Live and what it can do to improve sound (music & movie) reproduction?
I read quite a bit about various EQ and room correction systems, by some high technical sound gurus (Michael is one of them right here), and Kal Rubinson another, and Nyal too, plus others, ...do we really have the complete picture/sound reproduction satisfaction?

 

[esldude]

Alright, there seems to be a common agreement; good, because I too believe that the most ultra hi-end system in the world in the best acoustically treated room can still benefit from judicious DSP correction/EQ in the bass. ...And from various readings here and there Dirac Live is a good one. And it can be manually mastered with different target curves.

It is in the digital domain with a combination of FIR and IIR digital filters, ...Audyssey is FIR only, and ARC (Anthem) is a combination of both but differently implemented as compared to Dirac Live. Music reproduction and movie reproduction are equally important.

Now, what to do with audiophiles who strictly have an analog rig (turntable and open-reel-tape deck)? Do they want to do A/D and D/A in their system?
We have members here who have some serious analog rigs, what are they doing, and what are they missing...sound reproduction wise?
Amir mentioned that aspect prior (analog systems), and I was expecting that reply from someone when I asked my question. ...It just happened to be him who referred to it.

Floyd E. Toole, is he familar with Dirac Live and what it can do to improve sound (music & movie) reproduction?
I read quite a bit about various EQ and room correction systems, by some high technical sound gurus (Michael is one of them right here), and Kal Rubinson another, and Neil too, plus others, ...do we have the complete picture?

It won't be agreed with by many, but you have little to fear in my experience of digitizing your analog signal. You would perhaps be surprised how little is lost doing that. You also have the option to replace the phono stage with a quality microphone preamp and then digitize the signal in order to then digitally perform the RIAA EQ though few do this. Such a practice has some advantages.

But let us say there was a 5% quality loss pushing your analog thru a digital system that follows. You will gain more than that 5% in the improvements wrought for the rest of the total system. In some cases far more.

 

[rbbert]

That is no doubt true, but whatever they were striving for is unlikely to be the typical roller coaster response with one room's own particular signature of peaks and troughs. As Amir says, flatten the response first, then add or subtract a shelf of bass energy to your own satisfaction. Personally, I am quite happy with the perceived realism of the stock Dirac Live target curve I use. I never fiddle with it, because to me it delivers awesome, non-boomy deep bass that plausibly replicates what I hear at live classical concerts..

My point was more that because many recordings are made with a particular "sound" in mind that it's hard to use what you remember about the sound of a bass as a reference for what any given recording's bass sounds like on your system. Nothing about DSP except as you say having good bass (flat, low distortion, good transient response, etc) in the room will let you hear better whatever is on the recording.

 

[dallasjustice]

Nothing about DSP except as you say having good bass (flat, low distortion, good transient response, etc) in the room will let you hear better whatever is on the recording.

Yes it will.

I understand the logic to the argument that the recording is the biggest factor in great music reproduction. This argument has that seductive post hoc ergo propter hoc logical fallacy feel.

Sometimes it's important to just look at the facts. The distortions inherent in recordings which include acoustical/room distortions are many times lower than the distortions present in your playback system at home or anyone else's playback system. THE reason for this is the playback speakers and the room. As Amir said; one only needs to use a mic to measure these distortions to fully appreciate how aweful they really are in almost all playback rooms.

 

[Mike Lavigne]

It won't be agreed with by many, but you have little to fear in my experience of digitizing your analog signal. You would perhaps be surprised how little is lost doing that. You also have the option to replace the phono stage with a quality microphone preamp and then digitize the signal in order to then digitally perform the RIAA EQ though few do this. Such a practice has some advantages.

But let us say there was a 5% quality loss pushing your analog thru a digital system that follows. You will gain more than that 5% in the improvements wrought for the rest of the total system. In some cases far more.

you are generalizing that all systems will get 'some' benefit from DSP in performance.

and you know that how?

not every acoustical problem is solvable by DSP.

and not every system needs any sort of acoustical solution.

but some do; maybe even most do.

and some that do have a problem solve it in the analog domain. so don't consider the DSP path.

I would agree that of the systems that need and get significant acoustical advantages from DSP, the more drastic degree of improvement, the easier to rationalize the compromise from digitizing (or re-digitizing) the signal.

and as far as what reduction in performance DSP extracts simply take an analog signal and run it thru 2 separate ADC<->DAC processes (using adc/dac's in the price range of your dsp product). compare the analog signal to the signal with 1 adc/dac and with 2.

so A, B, and C.

and do it for an extended enough time to get a real feel for it.

and then go out and find the analog solution to your acoustical problems like you should have to begin with.

 

[JackD201]

It is the best we can do given the cards that have been dealt to us.

A point that is so often ignored.

 

[esldude]

you are generalizing that all systems will get 'some' benefit from DSP in performance.

and you know that how?

not every acoustical problem is solvable by DSP.

and not every system needs any sort of acoustical solution.

but some do; maybe even most do.

and some that do have a problem solve it in the analog domain. so don't consider the DSP path.

I would agree that of the systems that need and get significant acoustical advantages from DSP, the more drastic degree of improvement, the easier to rationalize the compromise from digitizing (or re-digitizing) the signal.

and as far as what reduction in performance DSP extracts simply take an analog signal and run it thru 2 separate ADC<->DAC processes (using adc/dac's in the price range of your dsp product). compare the analog signal to the signal with 1 adc/dac and with 2.

so A, B, and C.

and do it for an extended enough time to get a real feel for it.

and then go out and find the analog solution to your acoustical problems like you should have to begin with.

Never heard a system that didn't benefit. Haven't tried them all. Nor have you.

Never said every acoustical problem is solvable by DSP. I could name a dozen or more that are not.

I have done the AD/DA vs straight analog connection comparison. Which is why I think many people's imagination of what they lose doing that conversion is much higher than what will actually be lost. I have even done it with 2 such conversions though only for short periods of time.

Fixing your acoustical problems by other means when possible is always a good decision.

 

[amirm]

you are generalizing that all systems will get 'some' benefit from DSP in performance.

and you know that how?

Answering for him , it is physics. It is impossible to get flat bass response in small listening rooms. And there is no justification for having random peaks and valleys in the frequency response there. Or the time domain overhang they create.

not every acoustical problem is solvable by DSP.

Didn't think this was the topic. The topic was that the job is not done without DSP. If you have stopped short of that, there is very good chance that there is fidelity left on the table.

and not every system needs any sort of acoustical solution.

In bass frequencies, the *room* dominates. It is almost irrespective of the system, speakers or anything else. This is easily shown with measurements like this: http://www.madronadigital.com/Library/Computer Optimization of Acoustics.html

You are seeing multiple measurements in different parts of the room. In all of these cases, the "system" is obviously identical. All that is changed is the listening position. Change that and the response changes massively as shown below the transition frequencies. Therefore what is being seen/heared is the transformation the room is causing. How could we justify any of those low frequency graphs to be right???

Our listening ability to hear these response variations is excellent and similar to person to person. No way can we wave our hand and dismiss these. Until the distortion is taken away, you won't know what the problem is.

and some that do have a problem solve it in the analog domain. so don't consider the DSP path.

Analog domain? How?

 

[NorthStar]

Graphic EQ, Parametric EQ? ...But audiophiles don't use analog EQ. ...Or do they? ...No way, they don't even have Tone controls (Bass and Treble) on their preamps, ...and some don't even have a Balance control.

About Bass and Midrange and Treble control (tilt) knobs or switches in the back of their loudspeakers?
...For wall's proximity and other nasty first reflections or to simply adjust to taste.

Everybody loves full range loudspeakers (20Hz +/-3dB), ...about hi-end active monitors with dual subs that have their own DSP EQ?

 

[ddk]

Did I miss something? Toole only mentioned the room/speaker interaction and the issue with low frequency standing waves, something most of us are already very familiar with. According to the video M2's DSP and amp are part of the speaker's design to give it a smooth response, nothing to do with room EQ. He never did say how to deal with the standing waves. My own experiments with digital room EQ confirmed that its a waste of time and incapable of fixing the problem. The best digital equalizer I came across was Weiss EQ1. Fantastic and effective with digital but not for analog...

david

 

[NorthStar]

But DSP is used to give the loudspeaker the correct frequency response in relation to the room's acoustic properties @ the main listening position and with time domain compensation. /// Enter the world of Dirac Live. ...In particular from 10Hz to 250Hz. ...Above very little is needed, practically none.

A good loudspeaker with on and off-axis measurements, ...first and second reflections, from say 1kHz to 15kHz, is already EQued by its designer.
Positioning the speakers properly in the room where they perform is the first step, then the room itself, by judicious acoustical treatments...walls, ceiling, floor.The low bass though, that demands more care...bass traps, and DSP EQ.

 

[ddk]

But DSP is used to give the loudspeaker the correct frequency response in relation to the room's acoustic properties @ the main listening position and with time domain compensation. /// Enter the world of Dirac Live. ...In particular from 10Hz to 250Hz. ...Above very little is needed, practically none.

Unless I missed it that's not what Dr. Toole talked about in this video. His point was that with help of a digital crossover they've come up with an active speaker with minimal coloration, it had nothing to do with room acoustics. I'm familiar with Dirac, not for me but if I only had a digital system I'd use the Weiss Eq1.

david

 

[Rodney Gold]

DIRAC is free - well the trial is .. all you require is a Umik1 mic ($70 from miniDSP) and that in itself is a bargain purchase as you can measure the room with a calibrated mic..
Just try it , the cost is pennies compared to what some of us spend on a cable...

Lowbass DSP is almost a given for any system and room , unless you happen to luck on a speaker/sitting position that is serendipity and often the best place for bass is not the best place to listen to the rest.

Get the bottom end under 200hz right , you can passively treat for the rest.. passive treatment for a +10db hump at , lets say , 60hz is almost impossible to do.
If you have taken the time and money to spend on a megabuck system , get it right...

I actually redid my whole listening room with extensive treatments and what motivated me to do so was my speakers (g1's)

If you have a dedicated room , your attack is multipronged.. passive treatment , active treatment , bass distribution , DSP , eq etc... you will be amazed at the difference.
with a dual purpose room , you are somewhat limited and if you cannot get treatments or optimal positioning in there.. go DSP

Ignore arguments based on wide area listening position (HT) premises , for most of us with 2 chan audio , only the sweet spot counts.
there is no universal truth or reference for ANY recorded work .. no one barring the guy at the mastering desk knows what the thing actually sounded like .. reproduced music is nowhere near to anything live.. all it has to do is give you the illusion of "being there" and using eq/tone control in the chain is perfectly ok if it makes that illusion seem more real .. crossovers are equalisers , RIAA stages are equalisers ... speakers are equalisers ..mastering used equalisers .
At the end of it all , DSP makes a much bigger change than any box swapping does.. just try it.

As to target curves , well they are all taste based .. no one corrects for flat at listening position (unless you like no bass and really hot treble)

 

[amirm]

Did I miss something? Toole only mentioned the room/speaker interaction and the issue with low frequency standing waves, something most of us are already very familiar with.

That doesn't match my experience. Acoustic science remains a highly complex field, with so many differing opinions that most people do not have a clear idea of what it all means. Dr. Toole makes these presentations precisely because there is so little awareness of the critical aspects of this field. Did you know for example that our hearing sensitivity to broad resonances in low frequency is as low as 0.5 db? Yet we have rooms with excursions of 20 dB up and down! Here is my own room:

Unless everyone has such measurements and has played with optimizing the room, the awareness I am afraid if it is there, is superficial and of little value.

According to the video M2's DSP and amp are part of the speaker's design to give it a smooth response, nothing to do with room EQ.

Actually it does both. The DSP plays dual role in creating the crossover and room EQ. From the M3 brochure: http://www.jblpro.com/ProductAttachments/M2_Brochure_Jan2013.pdf

"Tuning and Room Integration
Room acoustics can play a big part in what you hear
at the mix position, particularly in the room-dependent
low frequency bands, where resonance caused by room
modes can give a false impression of bass in the mix.
While offering exceptional accuracy “out of the box,”
the power of the M2 is fully realized with its intelligent
in-room tuning and integration capabilities, ensuring
optimum performance in your listening environment.
Room optimization is achieved through the use of
floating-point digital signal processing integrated into
Crown iTech HD power amplifiers and BSS Soundweb
London processors. HARMAN System Architect™
Software is included to provide external control of system
EQ and tuning capabilities. Used in conjunction with
external measurement hardware and software, the
complete M2 tuning and room integration system
addresses non-linearity in the room. "

That is indeed how it is working in our theater at work. The SDEC-4500 (consumer version of BSS above) is performing the optimization post measurements using ARCOS.

He never did say how to deal with the standing waves.

The presentation is not meant to be all inclusive. Suggest reading Dr. Toole's excellent book,Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms http://www.amazon.com/Sound-Reprodu...=1432795572&sr=8-1&keywords=toole+loudspeaker

Some quick quotes:

Getting back to the listening room problem, although moving the chair
eliminated the objectionable resonance—but only for the single listener—there
was a problem. The chair could not be left permanently in the middle of the
living room floor, and moving it was leaving tracks in the carpet. What next?
Equalize, of course. Put the chair back where sensible room décor suggests
it should be, closer to the wall, and attenuate the resonance with a single parametric
filter tuned to 42 Hz, the appropriate Q and attenuation required to
create a frequency response that looked like the one measured at the previously
preferred listening location.

[...]

Equalization had the huge advantage of allowing the listener to sit in a decorative
location. Acoustically, there were advantages, too. With up to a 14 dB
amplitude reduction around 42 Hz, the woofers no longer had to work so hard,
distortion was lower, and they could play louder. There was also much less
energy everywhere in the room at 42 Hz. This was noticeable as improved sound
quality at other listening locations. This was a good solution to a personal
problem, as well as a learning experience: The right kind of equalization sounds
just fi ne, and electronics can provide an option equivalent to natural acoustical
manipulations. Of course, it works best for a single seat.

[...]

The ultimate form of intervention in controlling the sound field in a room
is the type of process exemplified by Sound Field Management (Section 13.3.5).
It is an optimization algorithm that begins with measurements in the room and
ends with a description for processing of the signals fed to each of multiple
subwoofers. Here we are actually manipulating the room resonances, the standing
waves, to minimize variations among some number of designated listening
positions. High-resolution measurements and parametric equalization at low
frequencies are essential ingredients of any of these systems.

One subwoofer used with high-resolution measurements (at least 1/10
octave) and parametric equalization can deliver good bass to one
listener. All other listeners in the room will take their chances because
of standing waves. Seat-to-seat variations are large.

Because low-frequency room resonances behave as minimum-phase systems,
we need high-resolution measurements (1/10- to 1/20-octave) and parametric
equalizers to work with them.

A parametric equalizer can attenuate the amplitude and reduce the audible
ringing from gross resonances (see Figure 13.24). This is a great benefit, but it
only works at the location of the measurement microphone. This should be
done for the prime listening location (yet another bonus!). In general it is recommended
to adjust the parametric equalizer to match the shape of and to
reduce the amplitude of any upward thrusting peaks in the frequency response.
Narrow dips should be left alone, but broad depressions may be boosted if the
amount of boost is not more than about 6 dB."

SFM mentioned above goes way beyond simple filtering by utilizing DSP and multiple subwoofers to create smooth response across multiple seats. Something that cannot be done by hand due to sheer number of combinations. Again, see this article I wrote drilling down into this feature with a number of real examples: http://www.madronadigital.com/Library/Computer Optimization of Acoustics.html

My own experiments with digital room EQ confirmed that its a waste of time and incapable of fixing the problem.

That is shocking to hear. One of the easiest places to apply EQ is in low frequencies. Find a peak, pull it down, and you make an immediate improvement. It is not hard at all once you have the measurement and have some familiarity with how to program the parametric EQ. If you have nulls, those can't be fixed but the peaks are readily fixable.

The strategy here is to get rid of the nulls with placement, and multiple subs and let the peaks become whatever they become. Then use EQ to pull down the peaks.

 

[TheMadMilkman]

Amir, can you post a graph of your room after EQ? I'm curious to see.

EQ is pretty high on my "to try" list. My room has bass issues. Since the system is in our main living space, most room treatments really aren't an option. My current listening chain is Mac Mini --> Schiit DAC --> NAD integrated amp --> Revel M105 speakers. Given that I intend to remain digital-only, is EQ best applied through software (Amarra w/ Dirac or similar) or through a separate box between the Mini and the DAC? Or does it really matter, since both are working with the digital files before analog conversion?

I also intend to eventually add a Revel sub. The sub has built-in EQ. Would the proper setup be to integrate the sub through its own software, then rerun and reapply the room EQ?