Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms

[Speedskater]

The first copies of the printed book are just now starting to ship. As of early last week, Dr. Toole had not yet received a printed production copy.

 

[TheWarrior]

The first copies of the printed book are just now starting to ship. As of early last week, Dr. Toole had not yet received a printed production copy.

He would probably not be pleased to know someone was working on Labor Day to deliver my hardcover...

 

[Pogre]

I got my copy yesterday. Haven't started it yet, but looking forward to learning some more.

 

[Johnny2Bad]

I've been reading Dr. Toole's writings for 30+ years. I have to say that his earlier stuff I didn't find very compelling, but he has continued to research and explore audio since those early missives, and his recent work is valuable to anyone seriously interested in audio and acoustics. I'll be putting it on my Amazon wish list immediately.

 

[andy_c]

I did a pre-order on Amazon a while back, and received the book this past Sunday, oddly. I started out on Chapter 8, which is the one about sound in the modal region. This chapter appears to be almost completely rewritten from the earlier edition and has lots of really good stuff. After that I started from the beginning of the book, and so far there seems to be extensive changes. So if anybody has the previous edition of the book and is wondering whether it's worthwhile to get the latest one, I'd say "definitely". So far, changes look to be quite extensive throughout.

 

[Dale Doback]

Just placed my order on Amazon.

 

[TheWarrior]

My favorite quote so far, pg 345 "When one sees a very smooth high-resolution steady-state room curve after equalization, there is a high possibility that something inappropriate has been done, and the quality of the sound may have been degraded."

 

[TheWarrior]

There are also e-books available for a few bucks less too!

https://www.routledge.com/Sound-Reproduction-The-Acoustics-and-Psychoacoustics-of-Loudspeakers-and/Toole/p/book/9781315686424

 

[TheWarrior]

Also, there is a companion website listed on the back of the book. It is not live yet, but will be, shortly.

I think someone mentioned Floyd dedicating the book to his father; Floyd is waiting on his hard cover copy to give to his Dad for his 105th. May that longevity be genetic!

 

[mtrycrafts]

... Floyd is waiting on his hard cover copy to give to his Dad for his 105th. May that longevity be genetic!

WOW, just wow.

 

[GregD]

What did you think? Get into Ch. 13, 'Making Bass Below Transition?

Hi, Sorry for the late reply. It is one of the chapters of interest for me and I found very educational and interesting. That said I still have a lot of questions. I was hoping I might be able to estimate how many sub-woofers I would need by making some measurements but it seems trial and error are a large part of figuring this out. This is difficult of course because without ordering the sub-woofers I can't determine experimentally what the optimal placement and number of sub-woofers required. Suffice to say I would like to get away with 2 but will 2 be sufficient or will there be a tangible benifet in 4. The issue is cost of course and having unused stock or not enough.

I assume you have the new edition, has this chapter been rewritten to make it clearer, perhaps with new information or is it basically the same?

Greg.

 

[TheWarrior]

Hi, Sorry for the late reply. It is one of the chapters of interest for me and I found very educational and interesting. That said I still have a lot of questions. I was hoping I might be able to estimate how many sub-woofers I would need by making some measurements but it seems trial and error are a large part of figuring this out. This is difficult of course because without ordering the sub-woofers I can't determine experimentally what the optimal placement and number of sub-woofers required. Suffice to say I would like to get away with 2 but will 2 be sufficient or will there be a tangible benifet in 4. The issue is cost of course and having unused stock or not enough.

I assume you have the new edition, has this chapter been rewritten to make it clearer, perhaps with new information or is it basically the same?

Greg.

Not at all!

Hopefully by now in your reading, you're coming to realization that everything matters.

You can predict what influences a room will have on your sound, but there's too many variables to predict a specific subwoofers performance.

"At resonance frequencies, the incoming sound is delayed by multiples of half wavelengths, summation occurs, and standing waves (room modes) are created - a pattern of higher and lower sound pressure levels within the room. Seat to seat variations in bass are an issue to be dealt with." - Ch.8 'Sound Reproduction' 3rd Edition

A subwoofer is a pressure source. The placement of that sub in relation to other subs, the room boundaries, and the listening positions, determines its ability to constructively or destructively drive those standing waves.

You'll need to start with predicting the room modes with the 'back of the envelope' calculations (p203). In the second edition, you'll see a diagram with suggested placement of two subwoofers on one wall, in Ch. 13 p221. That is an example of one way you can drive standing waves/room modes by using your pressure source (subwoofer) to drive or cancel out an offending frequency, without DSP.

Because loudspeakers are minimum phase devices, a steady state frequency response measurement will suffice for determining the phase of modal frequencies at the listening position. If a predicted mode (which is often a bit higher than measured, due to flex of boundaries by the inclusion of doors and windows) is measuring as a null, that indicates the mic location is out of phase at that frequency. You would then refer back to your predictions as to whether that mode is a height, width, or length mode and position the subs accordingly to drive it (null) or cancel it out(peak). DSP may be needed: Frequency = center of peak/null, Q= width in Hz of peak/null, Gain= dB that needs to added/subtracted to bring that frequency to the relative level.

After you've done what you can with placement and are ready to apply DSP, start with the lowest offending modes, and work your way up. You'll see with correct DSP filters, higher frequency peaks/nulls (that are unrelated to predicted modes) start to subside on their own. Some subs offer a few programmable DSP filters, like my SVS PC12's, but they only have two filters each. My open living room has a lot more modes than two/four filters can solve so I've had to incorporate a minidsp 2x4. But hopefully you've seen enough response graphs in this book that you understand that 'flat' is not the goal.

If you take the time to write all of that out, the trial and error will mostly subside. But it is a fair amount of work. Acousticians get paid a lot for a reason!

The 3rd Edition is thoroughly revised, and 'Ch.8 Below the Transition Frequency' is a bit easier to grasp than the former Ch.13 of the 2nd edition. I had to read the previous quite few times to get my head around it. But if you take the time to apply this knowledge, the process becomes more intuitive.

 

[Pogre]

@TheWarrior, how would I go about measuring my room to predict modes? It's not just a square or rectangular room. There's an open room to the right, vaulted ceilings and a hallway I'd need to account for.

 

[TheWarrior]

@TheWarrior, how would I go about measuring my room to predict modes? It's not just a square or rectangular room. There's an open room to the right, vaulted ceilings and a hallway I'd need to account for.

You will have to predict more modes! This includes any rooms your system is open to. I think I did 15 modes in my space, and still could measure more. 1131 fps / dimension in feet = modal frequency. Sound doesn't reflect once and dissipate, so reflections are logarithmic. Modal frequency x 2 = 2nd order x 3 = 3rd order and so on. I would predict the modes first, and after confirming the actual modal frequency with acoustic measurements, I would then calculate the higher order modes. Unless you really like crunching numbers.

Knowing the higher order resonances helps you to gauge the effectiveness of your placement/DSP solution, by seeing the cancellation of those higher order resonances stemming from the frequency to which you just applied a filter. If DSP is canceling out the modal frequency completely, the higher order resonances should dissipate. If not, either the filter is wrong or applied to the wrong sub(s), or placement/distance to boundaries needs adjusting. Inches matter!

 

[Pogre]

Yeah but, it would be almost astronomically hard to figure out the modes in my room(s), wouldn't it? I can't just figure out room modes for each individual room can I? Doesn't everything have to be factored as a whole? Like, I might show 1 room mode for the main area (35×15), but the room to the side (10x12) would completely change that, wouldn't it? Then the dog leg and vaulted hallway (?) would change that even more, no? How about the vaulted ceilings?

The entire interior is vaulted. I'm sure there's a formula that works, but wouldn't it be prohibitively difficult to get a true, accurate prediction in a layout like mine? Just taking the measurements alone is pretty daunting.

 

[TheWarrior]

Yeah but, it would be almost astronomically hard to figure out the modes in my room(s), wouldn't it? I can't just figure out room modes for each individual room can I? Doesn't everything have to be factored as a whole? Like, I might show 1 room mode for the main area (35×15), but the room to the side (10x12) would completely change that, wouldn't it? Then the dog leg and vaulted hallway (?) would change that even more, no? How about the vaulted ceilings?

The entire interior is vaulted. I'm sure there's a formula that works, but wouldn't it be prohibitively difficult to get a true, accurate prediction in a layout like mine? Just taking the measurements alone is pretty daunting.

I wouldn't be concerned with ceilings in other rooms, focus on the walls that are parallel to your listening space.

As I said above, Acousticians get paid a lot for a reason. It's work. But doing that work saves the hassle of endless guessing and tweaking.

Hopefully 1131/35 = 32 and 1131/15= 75 will help get you started

 

[Pogre]

I wouldn't be concerned with ceilings in other rooms, focus on the walls that are parallel to your listening space.

As I said above, Acousticians get paid a lot for a reason. It's work. But doing that work saves the hassle of endless guessing and tweaking.

Hopefully 1131/35 = 32 and 1131/15= 75 will help get you started

I don't have parallel walls. To my right there's a short boundary that turns into a large archway opening into a 10' x 12' room. To my left there is a room boundary with a sliding glass door. The hallway is part of the space too and it doglegs. The vaulted ceilings run through the whole interior of the house so there's no real height measurement I I know how to do that I could plug into a formula.

It seems like A LOT of effort, like I said, just to even do the measurements with any degree of accuracy. I have rough mathed it to come up with about 5900-6000^3' of space I'm pressuring and my room has odd bumps, spaces, hallway, furniture and vaulted ceilings that make it really difficult, don't you think?

*Edit: Forgot to mention kitchen behind me. It's not a straight wall with the pantry and fridge jutting out a few feet. Cabinets, countertops... how would I account for all of that? Seems like I should if it's to be accurate.

 

[KEW]

Pogre,
For example, if I have a open door to my kitchen and my kitchen is 10 feet deep. I would take the length of my room I already measured and add the 10 feet extra depth to identify an additional mode for that opening.
This assumes that the back wall of the kitchen is parallel to the wall opposite of it in my listening room!

 

[TheWarrior]

@Pogre
Width:
1. (glass door to short boundary before archway)
2. (glass door to parallel wall through archway)
3. (whatever is parallel in the kitchen behind you, that runs side to side of your seating position)
4. (sounds like he short boundary before archway would be parallel to the other end of the hallway)

Length:
1: (back of kitchen wall, ignoring cabinets, to wall your display is on)
2: (front and back of room on right)

Height:
1: (high point of vault)
2: (low point of vault)
3: (repeat if the kitchen behind has different heights)

Try filling those in, determine the modal frequency, and refer back to your baseline measurement without DSP or room correction. See how those predicted modes line up with the acoustic measurement. Peaks and dips should correspond at or below the predicted frequency. The center of the peak or dip is the modal frequency. Start from 20 hz and work your way up to the LFE crossover. As you find predicted modes that correspond with measurements, take the actual measured resonant frequency and calculate the higher order modes.

 

[GregD]

Thanks for the ideas and experience. I downloaded the spread sheet from JBL that is also used in the book. Unfortunately it assumes a square room. I am installing a system for my patients, its rectangular but has a shot hallway and opening to another large room as well as a lot of glass on one side and a bit on the back. Listening to tones, as well as some measured frequency sweeps, already suggest to me that acoustically it's far from a perfect rectangle. I read Gene DellaSala's article about the miniDSP he used with his sub-woofers. I also read that the RCA version has a level drop that needs to be compensated for and can reduce headroom. Not sure if the XLR version also suffers from that; if not perhaps it would be better to drive the XLR version in unbalanced mode to bet better headroom? I am a little sketchy on the details because I only noted it in passing.

In reading Pogre's comments, and thinking about my situation too, I am wondering how much I can learn about the room by making some measurements with one sub-woofer driving the a corner. In theory it should stimulate all modes of the room (to some degree) and hence the resonant modes become obvious through measurement (assuming the sub-woofer is not too close to the microphone)?

Greg.