Whats better for my application? - Page 2 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 1Likes
Reply
 
Thread Tools
post #31 of 56 Old 06-05-2014, 10:32 PM
AVS Forum Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 21,165
Mentioned: 836 Post(s)
Tagged: 1 Thread(s)
Quoted: 2920 Post(s)
Liked: 3626
"I'm flat out saying your wrong."

i agree.

the longer that the waves are relative to the room, the more the room behaves like the pressure model, even though technically all the wave information at the particle level is preserved. the predictions made by the pressure model work though (frequency independent spl below the critical frequency as bb points out and all the rest of it), which is why linkwitz, welti, and others can get away with calling it that. someone like geddes might say, "it's just waves, but i'll forgive you if you call it pressure." ;-)

then there is the other problem in that in the modal region, when waves start piling up on top of each other constructively, what do we call that? pressure modes. eek.gif

Listen. It's All Good.
LTD02 is offline  
Sponsored Links
Advertisement
 
post #32 of 56 Old 06-06-2014, 05:57 AM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5768
Quote:
Originally Posted by tuxedocivic View Post

what is room gain, how do you measure it discretely
The linked JBL document is a perfect example. The 'room' is too small for the observed gain to be caused by standing waves or room modes. The observed gain starts around 65Hz, where 1/2 wavelength is about 8.5 feet, which corresponds with the longest dimension in an average vehicle. If the size of the vessel didn't matter then one would have to conclude that all rooms would see a response rise starting at 65Hz.
This is another good source, assuming that you're an AES member:
http://www.aes.org/e-lib/browse.cfm?elib=16421
Bill Fitzmaurice is offline  
post #33 of 56 Old 06-06-2014, 06:03 AM
 
JWagstaff's Avatar
 
Join Date: May 2013
Location: Ontario, Canada
Posts: 853
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 28 Post(s)
Liked: 71
LTD, any idea where something like an open window, open door, crack under the door, air conditioning duct etc. come into play in the room gain formulas? What variable would they change? Any idea at what size something will affect if the room is "sealed" or not?
JWagstaff is offline  
Sponsored Links
Advertisement
 
post #34 of 56 Old 06-06-2014, 06:49 AM
AVS Forum Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 7,988
Mentioned: 227 Post(s)
Tagged: 2 Thread(s)
Quoted: 2300 Post(s)
Liked: 2225
Quote:
Originally Posted by Bill Fitzmaurice View Post

The linked JBL document is a perfect example. The 'room' is too small for the observed gain to be caused by standing waves or room modes. The observed gain starts around 65Hz, where 1/2 wavelength is about 8.5 feet, which corresponds with the longest dimension in an average vehicle. If the size of the vessel didn't matter then one would have to conclude that all rooms would see a response rise starting at 65Hz.
This is another good source, assuming that you're an AES member:
http://www.aes.org/e-lib/browse.cfm?elib=16421

The car examples make perfect sense. The smaller the "room" the earlier the reflections are constructive. I don't deny this. But its because of constructive reflections, the same as my measurement example posted earlier. How can we decide if there's a difference here?

Bill, are you suggesting room gain is a pressure type phenomenon? If so how do you explain Jwagstaff's question above? How about cars that have their windows open?

My youtube channel: Impulse Audio
tuxedocivic is offline  
post #35 of 56 Old 06-06-2014, 09:50 AM
AVS Forum Special Member
 
Ricci's Avatar
 
Join Date: Mar 2007
Location: Louisville KY
Posts: 5,478
Mentioned: 166 Post(s)
Tagged: 0 Thread(s)
Quoted: 373 Post(s)
Liked: 881
Over the past few years I've come to much the same viewpoint as Dave. Intuitively it just makes sense to me. As the wavelengths get longer the reflected energy from boundaries and objects has a larger and larger percentage that is constructive with each other.
tuxedocivic likes this.
Ricci is online now  
post #36 of 56 Old 06-06-2014, 10:29 AM
Advanced Member
 
laeriq's Avatar
 
Join Date: Aug 2006
Posts: 528
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 93
This forum is intense. I love it. Thanks for all the discussion.
laeriq is online now  
post #37 of 56 Old 06-06-2014, 11:02 AM
AVS Forum Special Member
 
NWCgrad's Avatar
 
Join Date: Feb 2010
Location: Frederick, MD
Posts: 3,625
Mentioned: 59 Post(s)
Tagged: 0 Thread(s)
Quoted: 1283 Post(s)
Liked: 1332
Should be relatively easy to emprirically determine the effect of room size using the scientific method.

______________________________________
Home Theater:
https://www.avsforum.com/forum/15-gen...sement-ht.html
NWCgrad is offline  
post #38 of 56 Old 06-06-2014, 11:08 AM
Member
 
sdaddy's Avatar
 
Join Date: Sep 2013
Location: KY
Posts: 73
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 16
Tux. If measurements were taken at various locations in the room, wouldn't modes for each location show up differently? Would room gain show the same? Maybe that's how you could measure RG.
sdaddy is offline  
post #39 of 56 Old 06-06-2014, 11:39 AM
AVS Forum Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 7,988
Mentioned: 227 Post(s)
Tagged: 2 Thread(s)
Quoted: 2300 Post(s)
Liked: 2225
sdaddy, you're right. When I move the sub (or mic) around the room, the response changes below 35hz, even below 17hz. Although the magnitude of the changes become less as frequency decreases because you simply cannot find a spot in the room where the reflections aren't piling on top of each other. This is also indicative that RG is a constructive reflection phenomenom.

The evidence I'm basing this on is as follows:

1. Open windows, doors, vents, stairwells, etc. do not significatly contribute to a loss of RG and in some cases have shown to help.
2. The location within the room changes the RG profile (RG based on pressurization would not cause this).
3. Loss of gain below the L/2 rule. ie. the 12hz dip forming in my posted measurement, as well as Ricci's results here: http://www.data-bass.com/data?page=content&id=80 this is a common location for a dip like this in larger rooms.
4. The lack of evidence shown by the pro-pressure advocates.
5. The abundance of evidence shown by the pro-constructive reflections advocates.
6. The similarity of RG profiles among various room sizes and shapes.
7. The predicted results are confirmed within the car audio world (I have personal experience here also and is what first caused me to abondon the idea of subs pressurizing anything, though pre-dating measurement abilities so I would not push on that very hard).
8. Physics.

My youtube channel: Impulse Audio
tuxedocivic is offline  
post #40 of 56 Old 06-06-2014, 11:55 AM
AVS Forum Special Member
 
bossobass's Avatar
 
Join Date: Mar 2003
Posts: 4,575
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 80 Post(s)
Liked: 690
Ahhh, the Harmon White Paper.

Geddes has said "It's just Harmon trying to sell you 4 expensive subwoofers". On this subject, this is the only thing Geddes has said that I completely agree with.

"Pressure Zone", as very briefly mentioned in the cited paper:
Quote:
The pressure zone response of the room. This is non-modal response of the room, i.e. the room acts as a 2nd order low-pass filter at low frequencies. The low pass characteristic can be seen below about 8 Hz in the above plot.

First, the referenced plot is simulated in MatLab. It was not verified in the actual room measurements, which were all cut off at 20 Hz (of course, since the Harmon boys were selling 20 Hz ported subs with this 'paper'). Note that the amplitude scale is 20dB/division, which will make the worst FR look flat.

I took the posted 'plot' and scaled it to a graph we're all more familiar with. The simulated room was 24' x 20' x 9'. That's a total of 4,320 cubes. The longest dimension = 32.5', so the L2 = 65', which translates to approximately 17.5 Hz in the mythological onset of pressure vessel gain formula.

Overlaid is my response in 3,500 cubes. The L2 in my room = 59.2' which translates to approximately 19 Hz, showing no such "pressure zone". Mine is an accurate measurement, the paper shows a simulation, not verified by an accurate measurement.

Also note that my result is with no post EQ and with the subwoofer in a single location.

70f7862a449ed69702562bce725870e8.jpg

There is no definition of "pressure zone response" and no formula from which they derived the 8 Hz number, nor any actual measurements to verify the model.

There is no mention of the construction of the room, therefore, no mention of transmission losses.

The seating area for 16 people is 6' x 6', or a bit more than 2 ft^2 per seat. My standard reclining seats require 18 ft^2 per seat with no room between rows when fully reclined. IOW, their simulated seating area would barely seat 2 people in my HT.

The paper mentions the skewing realized in actual measurements vs sim just by the addition of a projector and screen. There are no seats in the actual room, nor any other furnishings.

There were no simulations or actual measurements with subs stacked in a single location, as mine are and as I've found to be the best placement option in my room. Nor is there any experimentation with placement of a single sub at different heights off the floor. This was strictly a mono-plane exercise with one specific subwoofer.

The Harmon paper is so flawed as to be irrelevant. Toole-worship aside, I wish posters would simply use the search function of these forums, which contain all the empirical evidence and accurate data required. There is actually a simple formula for RG posted in this thread, actually, derived from the accumulation of data. RG begins in the 30-40 Hz decade and boosts the subwoofers GP response at an average rate of 7dB per octave as frequency decreases, the only variable being insertion of a transmission losses factor. The transmission losses factor is simple enough to arrive at because that data is available from various agencies and industries. I've just not had the gumption to wade through that part of it so that I could post that info here and be argued to death over it for the next 10 years.
bossobass is offline  
post #41 of 56 Old 06-06-2014, 01:33 PM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5768
Quote:
Originally Posted by tuxedocivic View Post

How about cars that have their windows open?
When you're in a room that contains a speaker for all intents and purposes you're sitting inside the speaker. Imagine it as a dual chamber box, with the enclosure that the driver is in being one chamber, and the room being the other chamber. If you make an alteration to the larger chamber you make an alteration to the entire system. In a home the size of a 'port' in a wall, ie, a door or window, is relatively small compared to the chamber, so it won't have a huge effect. In a car an open window makes a major difference in its behavior as a large port in a small room, so the system response shows a major change. Depending on the speaker inside the car it may be a positive change, it may be a negative change. The telling part is when you look at what guys in auto sound comps do to their cars to get crazy SPL readings. It includes not only sealing the windows tight, but also filling the body cavities with everything from foam to concrete, as well as sealing all the air vents, even welding the doors into their frames to prevent the escape of air.
Bill Fitzmaurice is offline  
post #42 of 56 Old 06-06-2014, 02:11 PM
AVS Forum Special Member
 
chalugadp's Avatar
 
Join Date: Jan 2013
Location: Vancouver, BC, Canada
Posts: 6,337
Mentioned: 58 Post(s)
Tagged: 1 Thread(s)
Quoted: 1683 Post(s)
Liked: 1572
Never thought of my room as a speaker box
chalugadp is offline  
post #43 of 56 Old 06-06-2014, 03:25 PM
AVS Forum Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 7,988
Mentioned: 227 Post(s)
Tagged: 2 Thread(s)
Quoted: 2300 Post(s)
Liked: 2225
Bill, that's been argued before and it's just not an adequate argument. We're going down the same path we've been down a hundred times before. I can't keep going on about this. Sorry for the thread jack OP. I'm responsible.

My youtube channel: Impulse Audio
tuxedocivic is offline  
post #44 of 56 Old 06-06-2014, 05:39 PM
AVS Forum Special Member
 
Mark Seaton's Avatar
 
Join Date: Jul 2000
Location: Chicago, IL, USA
Posts: 6,962
Mentioned: 210 Post(s)
Tagged: 0 Thread(s)
Quoted: 907 Post(s)
Liked: 2569
Quote:
Originally Posted by tuxedocivic View Post

Quote:
Originally Posted by Bill Fitzmaurice View Post

The linked JBL document is a perfect example. The 'room' is too small for the observed gain to be caused by standing waves or room modes. The observed gain starts around 65Hz, where 1/2 wavelength is about 8.5 feet, which corresponds with the longest dimension in an average vehicle. If the size of the vessel didn't matter then one would have to conclude that all rooms would see a response rise starting at 65Hz.
This is another good source, assuming that you're an AES member:
http://www.aes.org/e-lib/browse.cfm?elib=16421

The car examples make perfect sense. The smaller the "room" the earlier the reflections are constructive. I don't deny this. But its because of constructive reflections, the same as my measurement example posted earlier. How can we decide if there's a difference here?

Bill, are you suggesting room gain is a pressure type phenomenon? If so how do you explain Jwagstaff's question above? How about cars that have their windows open?

Quote:
Originally Posted by Ricci View Post

Over the past few years I've come to much the same viewpoint as Dave. Intuitively it just makes sense to me. As the wavelengths get longer the reflected energy from boundaries and objects has a larger and larger percentage that is constructive with each other.

Analyzing the issue as boundaries is a much more complete evaluation. The conceptualization of such situations as a pressurized vessel is an over-simplification, just as it is a gross oversimplification to state every line source falls off at -3dB/doubling of distance with uniform response over the length of it. You can find idealized situations/frequencies where this is almost true, but it is intended as a conceptual and thumb-to-the-wind understanding, not a mathematical definition. Many like to use a simplification of doubling of electrical power = +3dB. This too is a simplification. A precise statement is that 10x the power = +10dB, which is exactly from the math.

A much more appropriate example is that of the simplified T/S model of a ported box vs. a more complex model that would allow for cases of a transmission line / small horn. At extremes, the assumed conditions in the simplified model no longer hold true. I've modeled and built a few boxes that sit on the border of needing a more complex model, and there is not a specific point where you need the more complex model, but by understanding what defines those limits allows you to determine what model is appropriate for your use and interest.

We get into similar issues looking at how mid/HF speakers interact with boundaries. For many cases you can get plenty close for modeling and expectations by considering the boundary to be an acoustic mirror creating an identical source. In some cases you will want/need to more specifically model the detailed interaction with the boundary. The problems arise from those not realizing where the simplifications come from, or those stuck looking for simplified absolutes.
Quote:
Originally Posted by tuxedocivic View Post

Bill, that's been argued before and it's just not an adequate argument. We're going down the same path we've been down a hundred times before. I can't keep going on about this. Sorry for the thread jack OP. I'm responsible.

Indeed we have all posted much of this discussion dozens of times before. As is often the case, I take issue with the extremes of both views. It is much more useful and insightful to understand the boundary interactions, as this gives better insight as to why a big, high-ceiling cubic room will have a different gain profile than a room with a low ceiling and long depth. Analysis of measured data gets tricky as the gain from boundary confinement such as the ceiling height is entirely intermingled with the modal behavior as soon as one room dimension like the height is less than 1/2 any other dimension.

Mark Seaton
Seaton Sound, Inc.
"Make no little plans; they have no magic to stir men's blood..." Daniel H. Burnham
Mark Seaton is online now  
post #45 of 56 Old 06-07-2014, 12:15 PM
 
Mfusick's Avatar
 
Join Date: Aug 2002
Location: Western MA
Posts: 29,681
Mentioned: 23 Post(s)
Tagged: 0 Thread(s)
Quoted: 608 Post(s)
Liked: 2725
Quote:
Originally Posted by laeriq View Post

This forum is intense. I love it. Thanks for all the discussion.

+1
Mfusick is offline  
post #46 of 56 Old 06-11-2014, 12:05 AM
Member
 
cap'n's Avatar
 
Join Date: Jan 2009
Location: Los Angeles
Posts: 50
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 12
Gotta jump in here just a bit, to set a couple of things straight. The "How Many Subs..." paper was not intended as a treatise on room gain! If you are hung up on that, you completely missed the point. The point was about reducing seat to seat variation, and comparing different subwoofer configurations to each other (so things like room gain are constant and not of particular interest).
Quote:
Originally Posted by bossobass View Post

Ahhh, the Harmon White Paper.

Geddes has said "It's just Harmon trying to sell you 4 expensive subwoofers". On this subject, this is the only thing Geddes has said that I completely agree with.

Really? So we're trying to sell one more than Earl does using his approach? To put it in perspective, Floyd originally asked me to investigate the approach used by a company called C.A.T (California Acoustics Technology or somesuch) who were advocating 12 or 16 subs in a room, which sounded ridiculous. We ended up suggesting that 4 worked well (a lot LESS than 16). Also note that I have noted in that paper and others that "Two subwoofers at the wall midpoints is very nearly as good and has very good low frequency support as well. "
Quote:
Originally Posted by bossobass View Post

"Pressure Zone", as very briefly mentioned in the cited paper:
First, the referenced plot is simulated in MatLab. It was not verified in the actual room measurements, which were all cut off at 20 Hz (of course, since the Harmon boys were selling 20 Hz ported subs with this 'paper'). Note that the amplitude scale is 20dB/division, which will make the worst FR look flat.

The 20 dB per division was for the purpose of getting the additional modal diagrams (stems with circles showing eigen-frequencies) above the FR plots, and keep things condensed. Also, flatness of the FR's was not really relevant to the paper.
Quote:
Originally Posted by bossobass View Post

There is no definition of "pressure zone response" and no formula from which they derived the 8 Hz number, nor any actual measurements to verify the model.

There is no mention of the construction of the room, therefore, no mention of transmission losses.

Not sure what you mean by the 8 Hz number. THere is a dip there where the falling response of the subwoofer ("The computer model includes a realistic power response for subwoofers, based on near field measurements of a 10” Infinity Entre subwoofer. ") meets the room gain. I probably should have just plotted above 20 Hz. Transmission losses are irrelevant. Absorption was specified at 0.05 on all surfaces. Again, it is a simplified model, with low absorption.
Quote:
Originally Posted by bossobass View Post

The seating area for 16 people is 6' x 6', or a bit more than 2 ft^2 per seat. My standard reclining seats require 18 ft^2 per seat with no room between rows when fully reclined. IOW, their simulated seating area would barely seat 2 people in my HT.

The seating area is more of a grid, rather than distinct seats. you are over-analyzing this, I think.
Quote:
Originally Posted by bossobass View Post

There were no simulations or actual measurements with subs stacked in a single location, as mine are and as I've found to be the best placement option in my room. Nor is there any experimentation with placement of a single sub at different heights off the floor. This was strictly a mono-plane exercise with one specific subwoofer.

Correct, but how often do people stack subwoofers? Not saying you can't, but probably only best if you're going for max output. Subs generally go on the floor, next to one or more walls.
cap'n is offline  
post #47 of 56 Old 06-11-2014, 08:41 PM
AVS Forum Special Member
 
bossobass's Avatar
 
Join Date: Mar 2003
Posts: 4,575
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 80 Post(s)
Liked: 690
Quote:
Originally Posted by cap'n View Post
Gotta jump in here just a bit, to set a couple of things straight. The "How Many Subs..." paper was not intended as a treatise on room gain! If you are hung up on that, you completely missed the point. The point was about reducing seat to seat variation, and comparing different subwoofer configurations to each other (so things like room gain are constant and not of particular interest).
Quote: Originally Posted by bossobass

Ahhh, the Harmon White Paper.

Geddes has said "It's just Harmon trying to sell you 4 expensive subwoofers". On this subject, this is the only thing Geddes has said that I completely agree with.


Really? So we're trying to sell one more than Earl does using his approach? To put it in perspective, Floyd originally asked me to investigate the approach used by a company called C.A.T (California Acoustics Technology or somesuch) who were advocating 12 or 16 subs in a room, which sounded ridiculous. We ended up suggesting that 4 worked well (a lot LESS than 16). Also note that I have noted in that paper and others that "Two subwoofers at the wall midpoints is very nearly as good and has very good low frequency support as well. "
Quote: Originally Posted by bossobass

"Pressure Zone", as very briefly mentioned in the cited paper:
First, the referenced plot is simulated in MatLab. It was not verified in the actual room measurements, which were all cut off at 20 Hz (of course, since the Harmon boys were selling 20 Hz ported subs with this 'paper'). Note that the amplitude scale is 20dB/division, which will make the worst FR look flat.


The 20 dB per division was for the purpose of getting the additional modal diagrams (stems with circles showing eigen-frequencies) above the FR plots, and keep things condensed. Also, flatness of the FR's was not really relevant to the paper.
Quote: Originally Posted by bossobass

There is no definition of "pressure zone response" and no formula from which they derived the 8 Hz number, nor any actual measurements to verify the model.

There is no mention of the construction of the room, therefore, no mention of transmission losses.


Not sure what you mean by the 8 Hz number. THere is a dip there where the falling response of the subwoofer ("The computer model includes a realistic power response for subwoofers, based on near field measurements of a 10” Infinity Entre subwoofer. ") meets the room gain. I probably should have just plotted above 20 Hz. Transmission losses are irrelevant. Absorption was specified at 0.05 on all surfaces. Again, it is a simplified model, with low absorption.
Quote: Originally Posted by bossobass

The seating area for 16 people is 6' x 6', or a bit more than 2 ft^2 per seat. My standard reclining seats require 18 ft^2 per seat with no room between rows when fully reclined. IOW, their simulated seating area would barely seat 2 people in my HT.


The seating area is more of a grid, rather than distinct seats. you are over-analyzing this, I think.
Quote: Originally Posted by bossobass

There were no simulations or actual measurements with subs stacked in a single location, as mine are and as I've found to be the best placement option in my room. Nor is there any experimentation with placement of a single sub at different heights off the floor. This was strictly a mono-plane exercise with one specific subwoofer.


Correct, but how often do people stack subwoofers? Not saying you can't, but probably only best if you're going for max output. Subs generally go on the floor, next to one or more walls.


Not hung up on anything. It seems someone alerted you to my post because if you'd read any of this thread you'd know Room Gain IS the context.

What's of little significance is frequency response variation from 20-80 Hz across a 2 seat "grid" of 36 ft^2.

Bossobass, Carp, Beastaudio, MKTheater, nube, notnyt, BassThatHertz, N8 and many more members... all stack subs as will pretty much anyone who's serious about accurate reproduction of modern soundtracks.

In most cases, it's about displacement, not max output.

Have another look at the posted graph. With a single placement the result is +/-3dB and no post setup smoothing EQ was used, not to mention a full 2 octaves more bandwidth. So, flatness does indeed have everything to do with the paper. If the response across your 2 seat grid was flat, the exercise would be moot.
bossobass is offline  
post #48 of 56 Old 06-12-2014, 01:54 AM
AVS Forum Addicted Member
 
BassThatHz's Avatar
 
Join Date: Apr 2008
Location: Northern Okan range (NW Cascades region)
Posts: 10,646
Mentioned: 222 Post(s)
Tagged: 0 Thread(s)
Quoted: 3682 Post(s)
Liked: 4180
8 of my 16 subs are stacked and within 3.5ft of the front corners.
88% of them are semi-colocated along the front wall.

I had to stack them to save on floor space.
My subs take up 71.5 cubic feet of air space and unstacked would be 40sqft of floor space, it's 33sqft stacked.

In order for me to add any more subs to my room I'll have to start stacking dual-opposed boxes. 32 18's in ~38sqft is doable LOL

In my current room every time I tried to place subs behind me it degraded the sound;
not sure why, I haven't really looked into that much mostly because it currently doesn't apply with the success I'm having without. One day I'll revisit it (maybe).

Now that I have efficient SEOS tweeters, most people cry for mommy well before my mains run out of excursion or power handling or watts clipping (ignoring ULFtard tracks); in fact your ears themselves begin to distort it seems

That's what 4400watts times 3 into a tripplet of hot-rodded Cheap Thrills will do. (and if that wasn't enough, another 4 when movie-mode is activated.)

Only bassheads and drunken fools can tolerate my system at full blast for any extended amount of time (say 30 seconds to 5 minutes); and even then hearing damaged is almost a given in such a case.

The longest I lasted was 4 hours at full blast. Yes I'm stupid, I know...

Up until 2013 prior to owning quad LMS's, my system was actually power limited and then thermally limited within 5 to 20 minutes of high output.

Now that I have 16 subs I'm still power limited, that's fixable though, but I finally haven't noticed any further thermal issues at high volume anymore.

I still haven't reached the "bass limit" of my ears yet where I cry uncle (i.e. Max Bass SPL. I still want more subs), with the exception of that one time I played a 60hz sinewave @ 130db @ the LP on my TermLab.

But even at moderate playback levels the excursion rarely exceeds 4mm p-p; most of the time the subs don't even seem to visibly move at all. A lot of that has to do with displacement / efficiency. Only with infrasonics at heavy playback levels does the excursion increase noticeably.

Above >40hz my system sharply rises to 100db/watt efficiency,
below <40hz it drops to about ~90db/watt (or something like that),
and then at <25hz the room-gain comes into effect.

Last edited by BassThatHz; 06-12-2014 at 02:48 AM.
BassThatHz is offline  
post #49 of 56 Old 06-12-2014, 03:49 PM
Member
 
cap'n's Avatar
 
Join Date: Jan 2009
Location: Los Angeles
Posts: 50
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 12
Quote:
Originally Posted by bossobass View Post
Not hung up on anything. It seems someone alerted you to my post because if you'd read any of this thread you'd know Room Gain IS the context.

What's of little significance is frequency response variation from 20-80 Hz across a 2 seat "grid" of 36 ft^2.

Bossobass, Carp, Beastaudio, MKTheater, nube, notnyt, BassThatHertz, N8 and many more members... all stack subs as will pretty much anyone who's serious about accurate reproduction of modern soundtracks.

In most cases, it's about displacement, not max output.

Have another look at the posted graph. With a single placement the result is +/-3dB and no post setup smoothing EQ was used, not to mention a full 2 octaves more bandwidth. So, flatness does indeed have everything to do with the paper. If the response across your 2 seat grid was flat, the exercise would be moot.
... and YOU need to read my post again. I said that room gain was not the subject of MY paper, which is being referenced in this thread. The variation across the seating area that I am interested in in my paper is seat to seat variation, and it is of primary importance if you HAVE a seating area (such as in home theaters), as opposed to one seat.

As for stacking maybe in your world you do (one of the OP's mentioned stacking to save floor space, which I understand). For the 12" subs that i was using, I doubt there would be too much difference between stacking, side by side, or just a 6 dB boost in gain. In any case it was simply not feasible to investigate every possible way a person can put subs in a room.

So. you're saying i have to stack my subs to get good sound? Not buying it.

I think you know what I meant about max output, are we going to quible about "max output" vs "max diplacement" vs "max volume velocity"?

Your plot only shows one seat so again you show that you have completely missed the entire point of my paper. I suggest a refresher read.

...and finally, what exactly do you mean by "accurate reproduction of modern soundtracks."? Accurate compared to what?
cap'n is offline  
post #50 of 56 Old 06-12-2014, 04:55 PM
Member
 
cap'n's Avatar
 
Join Date: Jan 2009
Location: Los Angeles
Posts: 50
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 12
Quote:
Originally Posted by bossobass View Post


The seating area for 16 people is 6' x 6', or a bit more than 2 ft^2 per seat. My standard reclining seats require 18 ft^2 per seat with no room between rows when fully reclined. IOW, their simulated seating area would barely seat 2 people in my HT.
Your math is off. Draw a square grid of 16 points separated by 2 feet. Draw edge to edge squares around each seat. you will see that the area is 4 sq ft, not 2. The total seating area represented by the grid centerpoints is 8 x 8 feet = 64 sq ft, not 36 as you stated elsewhere.

Quote:
Originally Posted by bossobass View Post
The paper mentions the skewing realized in actual measurements vs sim just by the addition of a projector and screen.
Huh? Where did I say that?!? how would adding a projector affect the measurements?
cap'n is offline  
post #51 of 56 Old 06-12-2014, 07:09 PM
AVS Forum Special Member
 
bossobass's Avatar
 
Join Date: Mar 2003
Posts: 4,575
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 80 Post(s)
Liked: 690
Your drawing is off. I plan and draw residential structures for a living. You might have hired someone else to do the graphic. The "grid" is clearly marked, from end-to-end, 6' x 6', and the note to the right refers to the 6' x 6' grid of seats. My math is fine.

In either case, nice try with spacial averaging over 16 seats. Here is your drawing to scale with theater seats also to scale:



16 seats... not happening. More like 3 seats, 6 seats if you allow for some more room than you did but then what happens to the other 10 guests? Saying you tested for 16 seated persons in a 20' x 24' virtual room is where the paper lost my interest.

You say your paper is about spacial averaging and said earlier that transmission losses are of no consequence but the paper mentions the "absorption" of the room and the fact that there is sheetrock and "a door" or a flimsy door or some such, that both influenced the liveness of the room, etc. So, what's the difference? Were you saying the sheetrock and a flimsy door result in longer decay time at <100 Hz frequencies?

It's irrelevant to me how you or anyone else gets good sound in a room. I responded to the flippant 'who stacks subs anyway?' remark. You simulated 5,000 subwoofers in the virtual room that hasn't enough space to hold 5,000 subs, but couldn't possibly have investigated the effect on the ceiling/floor relationship, etc., from stacking the subs? Not buying it.

And, yes, we'll have to quibble about max output vs max displacement if you insist that they refer to the same thing. Volume velocity is irrelevant.
bossobass is offline  
post #52 of 56 Old 06-12-2014, 09:58 PM
Member
 
cap'n's Avatar
 
Join Date: Jan 2009
Location: Los Angeles
Posts: 50
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 12
Quote:
Originally Posted by bossobass View Post
Your drawing is off. I plan and draw residential structures for a living. You might have hired someone else to do the graphic. The "grid" is clearly marked, from end-to-end, 6' x 6', and the note to the right refers to the 6' x 6' grid of seats. My math is fine.

In either case, nice try with spacial averaging over 16 seats. Here is your drawing to scale with theater seats also to scale:



16 seats... not happening. More like 3 seats, 6 seats if you allow for some more room than you did but then what happens to the other 10 guests? Saying you tested for 16 seated persons in a 20' x 24' virtual room is where the paper lost my interest.

You say your paper is about spacial averaging and said earlier that transmission losses are of no consequence but the paper mentions the "absorption" of the room and the fact that there is sheetrock and "a door" or a flimsy door or some such, that both influenced the liveness of the room, etc. So, what's the difference? Were you saying the sheetrock and a flimsy door result in longer decay time at <100 Hz frequencies?

It's irrelevant to me how you or anyone else gets good sound in a room. I responded to the flippant 'who stacks subs anyway?' remark. You simulated 5,000 subwoofers in the virtual room that hasn't enough space to hold 5,000 subs, but couldn't possibly have investigated the effect on the ceiling/floor relationship, etc., from stacking the subs? Not buying it.

And, yes, we'll have to quibble about max output vs max displacement if you insist that they refer to the same thing. Volume velocity is irrelevant.
As I already said, the grid does not represent a specific seating layout. I wanted to have the grid points reasonably close together so it could represent a seating area, not discrete seats like you are suggesting. Looking back at the paper, i see that i did use little "heads" to represent the grid points - that may be partly to blame for your confusion. Each point represents an area around it where we expect the response to be reasonably close to the reference point. It's not an unreasonable assumption, the same one you make whenever you measure at one seat and assume that when you move your head a foot one way or the other the response will still be reasonably the same. With that in mind, the seats on the outside of the grid also represent an area including one foot outside the outer grid points. If you draw the picture like I suggested, you will see this. and you are also incorrect about the square footage that each point represents. Unless 2 x 2 somehow does not equal 4, you math is off.

I said the TL is irrelevant because it is only the absorption that matters for the modal calculations. TL and absorption are not the same thing. A well designed wall has high TL but could either have higher or lower absorption coefficient.

You say "It's irrelevant to me how you or anyone else gets good sound in a room." Then why are you on this forum?

As for the sub stacking, yes there are a lot of things I could have included in that paper. There just simply wasn't time to do them all. With my (newer) modeling software, I can fairly quickly model two subs side by side versus stacked for a number of different seating configurations, and look at the effect. Would you like me to?
cap'n is offline  
post #53 of 56 Old 06-13-2014, 02:25 PM
Member
 
cap'n's Avatar
 
Join Date: Jan 2009
Location: Los Angeles
Posts: 50
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 12
Bossobass, I looked over you website and now understand what you mean when you talk about stacking subwoofers. I didn't realize you were talking about stacking that high. I see that you show measurements, and that is commendable. It does bring up a couple of questions:

Why do you accuse me of writing papers just sell more subwoofers, when you yourself are selling systems with anywhere form 4 to 16 drivers in them?

Picking out one mode, such as the floor-ceiling mode, and focusing on it to the exclusion of other modes can get you into trouble. The probelm is that there are many modes and they tend to interact. That is why I wrote my modeling software. It usually more complex than you think. Looking through my modelling results i don't see a consistant notch like you show, for the floor-ceiling mode (1st axial). Looking at the measured data from my original subwoofer study (which did include real rooms), I can plot measured responses for 16 seats and an average.



Here's the plot for one sub in the corner, the dotted line shows the calcuated floor ceiling frequency for this room:
Click image for larger version

Name:	infRoom_16mics_cornersub.jpg
Views:	4777
Size:	69.0 KB
ID:	114481

Here's one for one sub in each corner:

Click image for larger version

Name:	infRoom_16mics_4corners.jpg
Views:	4827
Size:	51.8 KB
ID:	114489

It seems overall that there is not a consistant notch like you show in your measurements. Using the 4 corner setup eliminated any hint of it. and these are measured data.
cap'n is offline  
post #54 of 56 Old 06-15-2014, 11:02 AM
AVS Forum Special Member
 
bossobass's Avatar
 
Join Date: Mar 2003
Posts: 4,575
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 80 Post(s)
Liked: 690
Here is why I come to this forum...

I am and have been an advocate for accurate, full bandwidth playback of modern soundtracks and music in HT. For years the experts, accredited or self-appointed making no difference in a forum where new-comers are just looking for options, howled and jeered over the very idea of playback of anything <20 Hz. The arguers were as grossly incorrect as they were prolific. "The content <20 Hz is unintended noise", "You can't hear those frequencies anyway", "The equal loudness curves dictate that you would need 140dB <'x' Hz", "No subwoofer can play that low with enough authority cleanly", "You'd need a nuclear power plant for mains and you'll have to rob a bank", "Room gain is dependent upon how sealed the room is", "Room gain begins at a frequency whose wavelength is 2* the rooms longest dimension", "Rooms >2000 ft^3 will have little, if any room gain", etc., etc.

Most of this has been dispelled over the past decade but some of it persists, like the discussion in this thread re; room gain. What it is, how it works is still largely a mystery to the average enthusiast because of the mysterious explanations by people who have never experienced <20 Hz reproduction.

Here is what brought your paper to the thread, from a post by LTD:

Quote:
Originally Posted by LTD02 View Post


todd welti also refers to the zone below the first mode as the "pressure zone".* again though, same result, so use whichever term you like.
I quoted the pressure zone mentioned in the paper:

Quote:
The pressure zone response of the room. This is non-modal response of the room, i.e. the room acts as a 2nd order low-pass filter at low frequencies. The low pass characteristic can be seen below about 8 Hz in the above plot.
(That's ^^^ where the 8 Hz came from)

LTD claims, among other strange things on the subject, that the "pressure zone" acts as a 2nd order shelf boost <L2, but the paper seems to contradict that, so not sure in what context he cited your paper.

Regarding the other comments:

The bottom line with the grid drawing of little heads in a 6' x 6' square space, it's just incorrectly presented. The reader should not have to re-read and re-draw to get the correct context. I appreciate the actual procedure now, but it took you to explain it. Illustrations are used in lieu of or to support explanations. If drawn correctly, a discussion as to what the drawing represents would not be necessary.

Like so, for example:



For the record, we both know that no HT will have folding chairs for seating, squeezing listeners at 2' O.C., but, I do appreciate the time you took to lay it out for me. Over the past 12 years, this is the first time in dozens of posts that referenced the paper that anyone took the time to explain the grid details so that they made sense.

Stacking:

I agree that concentration on any single mode is not wise. I haven't said otherwise. I just feel that the floor-to-ceiling mode, which exists in every room, is as important as any other parallel surfaces, yet is completely ignored by the professionals. The ceiling is also potentially an additional boundary of purely constructive reflections through the typical 80 Hz crossover region when the point source is the correct distance from it. This is not trivial.

The floor-to-ceiling mode is also far more universal than other modes because most rooms have a ceiling height of 8' (70 Hz) or 9' (62 Hz) and are both in the crossover region, vs other boundaries which are infinitely different from room to room and outside of the typical crossover region. If you ignore stacking, it will depend on where you sit in the room, as your measurements show. You're actually off by a couple of inches in your stated ceiling height, IMO. Move your marker up a couple of Hz. I can't move my primary seat and planar placement options have no effect.

Finally, stacking offers the most output in the least area of floor space, which is a checklist item that's usually at or near the top of the list. Most people here build overkill systems for headroom at the extreme low end and to handle whatever software the industry throws at them. When it comes to <20 Hz, displacement is king and that means multiple drivers. It's an unavoidable requirement of the recipe. In most rooms, placing 30 or so liters of displacement in proper sized enclosures around the room, on the floor, is not an option.

So, I advocate multiple drivers for bandwidth, not smoothing response or any other reason. Because most people don't or conventional wisdom ignores, these are not reasons to not explore the possibilities and present the data.

I see that Kreisel's new stuff is made to stack, the alignment is sealed, the drivers are dual opposed, etc., so some industry folk are either paying attention or just finally getting around to dealing with the same reality we've evolved to of how to separately place the required 8 or more drivers.

TL:

To my knowledge, sheetrock and doors do not absorb anything <40 Hz, the BW of interest. The pressure waves are either reflected off a rooms boundary back into the room or they're lost through the boundary to outside the room or somewhere in between those two. Drapes, traps, sofas, etc., have no absorption authority over low frequency soundwaves.

If your test room was constructed of sheetrock and a flimsy door, where it concerns subwoofer frequencies, it was not a live room. Live or the opposite, where low frequencies are concerned, has everything to do with TL and little if anything to do with absorption.

This is the essence of room gain. It's the result of wave interference that is progressively more constructive vs destructive as wavelength increases, minus boundary TL and including modal influences <30-40 Hz with the final arbiter being signal chain roll off at the extreme low end.

I don't expect most people to wade through what I write or have written, much less give a hoot. But, some do and these are the basic facts of the matter. When people post that multiple sub locations = smoother response, they need to do a lot of qualifying for that statement to be true. Because, of course, every room is different and placement options vary in each room. When people post that sound waves turn into liquid below L2 as in an infinitely rigid cylinder of infinite length, or whatever the Star Trek parallel universe theory of room gain is on paper, they should be ready to present actual data and a reasonable explanation of what the he!! they're talking about.

The projector:

I was going from (ancient) memory, but I looked up the actual quote in the paper:

Quote:
Due to several factors, including a sizable rear projection television located at the front center of the room, and a limited number of available subwoofers, the configurations are slightly different than those modeled in Investigation 4.
It's no big deal, it just illustrates the point I tried to make; that real rooms will skew the results of a simulated empty room of rounded-off dimensions.

I appreciate the time you've spent here, and… can you please tell us more about the software program you wrote?
OKGeek likes this.
bossobass is offline  
post #55 of 56 Old 06-15-2014, 01:42 PM
AVS Forum Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 21,165
Mentioned: 836 Post(s)
Tagged: 1 Thread(s)
Quoted: 2920 Post(s)
Liked: 3626
"This is the essence of room gain. It's the result of wave interference that is progressively more constructive vs destructive as wavelength increases"


"room gain" is relative to something like 2pi space where there is no destructive interference. so yes, below about the 1/2 wavelength point, the waves begin to progressively become more and more constructive (both positive and negative amplitudes increase) and the spl in the room begins to approach the pressure model. no disagreement.

Listen. It's All Good.
LTD02 is offline  
post #56 of 56 Old 06-15-2014, 10:44 PM
Member
 
cap'n's Avatar
 
Join Date: Jan 2009
Location: Los Angeles
Posts: 50
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 12
Bossobass, i won't quote your post, as it is a long one There are some valid points made. Some I could quible about. I won't get into the room gain debate, but only say that it is entirely possible that there might be more than one way to think about it that might be equally valid.

My focus has always been on getting consistent seat to seat response, mostly because that is a challenging problem, and hard to solve without multiple subs. It doesn't mean I think other issues (low bass, high SPL, minimizing footprint) are not important too.

Since seats are normally not stacked, but are distributed in the x and y (length/width) dimensions of the room, I believe the length and width modes are slightly more important, and that locating the subs around the room should be preferred if possible. However, adding stacking to the mix only makes it better. And you have inspired me to look into that more. BTW my latest AES paper uses a grid spacing of around 8", with random perturbation of the locations. There can easily be 2000 receivers in the "seating area"! Results can be seen at http://www.harman.com/en-us/ourcompa...itepapers.aspx

Software program uses the same basic solution as before, I just made it much much faster and it calculates ranges of room dimensions now, not just one room. Also very easy to set up and run models now. Is whatever metric I am interested in, plotted over a range of possible room dimensions, for a given sub and seating configuration.
Attached Thumbnails
Click image for larger version

Name:	RMRC.jpg
Views:	60
Size:	28.6 KB
ID:	120130  
cap'n is offline  
Sponsored Links
Advertisement
 
Reply DIY Speakers and Subs

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off