Yes, quite true, thanks. See how easy it is to confuse the two!!
It should read:
"Thermal modulation is also, for the most part, independent of the current voice coil temperature"
Absolutely, but consider the title of this thread. Why would anyone come here and use a "Pro" perspective? It's clearly not about "Pro"!
I get annoyed when people don't consider the context and criticize my comments from a perspective that is clearly NOT what this thread is about.
Actually you are incorrect. We mix for the highest attainable quality, and let the chips fall where they may in the field.
The producers of soundtracks are sound editors, sound designers, and re-recording mixers. These are quite different people than those who do the production of the film.
Understood, however I'm a pro engineer and was attracted by the thread title. I'm interested in the modularity of small subs for a few reasons: Scaleable arrays, multiple concurrent events, rentals etc. but also I find myself working in medium-small rooms where modes are very problematic. Using multiple subs at the nodes of the room modes has given me good results especially when the walls are pretty rigid.
That being said I do understand that thing on audio forums often get taken out of context.
-Matt
Is efficiency here the same as sensitivity?
If higher sensitivity is important, do we know where the upper limit is for an improvement? Or does it simply get better and better?
Though 95 dB is high compared to commercial speakers, it's a good distance to horn speakers with sensitivity of 100, 105 or 110 dB.
Yes, I used efficiency here as synonymous to sensitivity. "Better and better" is only relative to what is audible and we don't know what that is.
At first I doubted that this effect could be audible. But when I did some calculations and found that the difference between a 1" dome tweeter and a high sensitivity compression driver on a waveguide could be as high a three to four orders of magnitude I figured that audibility would surely lie in that range somewhere.
But I doubt that the situation could get "better and better" compared to a 105 dB per watt compression driver.
I have always been interested in why speaker sound quality falls apart as the SPL goes higher and why some speakers do so at much lower SPLs than others. I used to think that it was nonlinear distortion and so I studied this quite a bit (and I mean quite a bit!) As you know I, and almost everyone else who has looked at this, have concluded that nonlinear distortion is simply not the answer. So what is it? There is clearly a difference in the way speakers handle dynamics, but its not what we thought it was for a long time. It has to be something else. I now believe that it is diffraction and thermal modulation. Eliminate these two things and the speakers will play at insane levels without even the slightest loss of quality. I've done those tests.
Here's a funny story, to illustrate this:
An externally hosted image should be here but it was not working when we last tested it.
This week I was at CES, and had a chance to hear Dynaudio's new flagship. It's something like $95,000. This is your basic audiophile dream speaker, with imaging to die for, and a soundstage that's wide and deep.
While listening to it, I can honestly say that it's one of the best sounding speakers I've ever heard.
But someone asked Dynaudio to 'crank it up', and I noticed something...
It was starting to 'harden' at surprisingly low SPLs. I'd estimate around 100dB it was starting to compress.
I have a pair of Summas, and I've never even heard a *hint* of this, and I've listened at levels that exceeded what Dynaudio was playing by easily 20dB.
Also, this 'hardening' of the sound wasn't coming with big bass hits, or drums, or anything like that. It was just plain ol' vocals. If I had to hazard a guess, I'd bet that the low order crossover on the Dynaudio midranges was causing them to get close to their xmax.
That really made me think, though. IE, it *did* sound like one of the best speakers I've ever heard, but that was mostly because of it's enormous soundstage. And in my experience, most recordings don't have much of an image. For every well-recorded album there are ten more that are practically mono. And can you imagine spending ninety five thousand dollars on a set of speakers that image like a champ, but can't exceed 95dB without getting strident?
I have had a similar experience with a very highly regarded (within a small circle) two-way with a "waveguide on a 1" dome tweeter (not compression driver) and an 8" woofer known for low distortion.
At low levels it was gorgeous, but I could hear hardening starting at levels averaging 75dB, C-weighted. I sold them to someone who is still delighted with them after 2+ years but who, like most I guess, simply expect speakers to sound stressed when turned up. That is the common understanding of "loud" in the context of home hi-fi - the sound of stress.
I had always assumed it was the result of some form of non-linear behavior in one or more of the drivers - break-up - and am surprised to think that it may be a function of thermal modulation, but I am seeing this again and again. I had assumed that thermal modulation effects would occupy too long a time scale to affect short transients. I find it difficult to visualize, though, why diffraction effects would vary with loudness, if that I understand you correctly, Dr Geddes.
I remember an early ATC speaker being described as getting "bigger" sounding, more than "louder" as the volume control was advanced, and that description stayed with me.
At low levels it was gorgeous, but I could hear hardening starting at levels averaging 75dB, C-weighted. I sold them to someone who is still delighted with them after 2+ years but who, like most I guess, simply expect speakers to sound stressed when turned up. That is the common understanding of "loud" in the context of home hi-fi - the sound of stress.
I had always assumed it was the result of some form of non-linear behavior in one or more of the drivers - break-up - and am surprised to think that it may be a function of thermal modulation, but I am seeing this again and again. I had assumed that thermal modulation effects would occupy too long a time scale to affect short transients. I find it difficult to visualize, though, why diffraction effects would vary with loudness, if that I understand you correctly, Dr Geddes.
I remember an early ATC speaker being described as getting "bigger" sounding, more than "louder" as the volume control was advanced, and that description stayed with me.
I find it strange that they are set up to fire straight down the room. I cannot think of any circumstance where this would be the preferred configuration - even if the speakers need to be heard off axis due to nasty beaming of mids or highs. This only works well when the listener is precisely centered; with any lateral movement the image moves to the near speaker, doesn't it?
Crossing axes in front of the listener reduces this, widens the listening area and also reduces the first side wall reflections, yet I see this all the time.
The only rationale for it can be the visual aspect (some kind of symmetry) or the speakers are way too hot in the HF and need to be listened to 30º off axis to be tolerable. I guess.
Last edited:
They're not, look at the baseplate and the feet of the enclosures, a little bit of too in.
A bit off-topic, but how does the JBL 2242H compare to the B&C 18SW115? The 2242H had higher sensitivity and require less power. But also needs a bigger cabinet to reach same lows. Anything else?
The JBL 2242H has only 7.87mm Xmax compared to the BC18SW115's 15mm Xmax, and 800 watt power handling compared to 1500.
That means you would need almost two 2242H to equal the output of one BC18SW115.
Of course, if you subscribe to the multiple location sub approach, and don't need 122-128 dB output per sub, both these drivers are overkill.
It's definitely true that the specs favor ,b&c, and I can personally say that the 12" and the 21" in the TBX line are beasts sounding great even at high spl so I'm sure the 18 neo version is too. However, I have worked with the 2241H(G) for about 9 years and I love it(read: slightly biased). It has been the most reliable subwoofer driver for me. I've only messed around with the 2242 but it has the same motor structure but with 9mm of xmax compared to the 7.6 of the 2241(though Art may be citing his direct measurement not the spec sheet.). My experience, and the numerous papers I like that JBL has put out, has led me to believe that JBL often understates its it's power handling.
Less than a year ago I drove some sr4719xs with dual 2241s(reconed) with over 2000w each to the point where I heard the vc former touching(euphemism) the back plate. Naturally practically toppled the entire crowd of people on my way to adjust, but held it at just a bit inside what we can call xlim. This was probably right around 2000w. Amazed at the new found power handling I kept this up for 3-4 nights a week for 3 months. The only failure I had was one broken lead wire that was easily repaired.
The unique venting that the 224x series has deserves a closer look I think.
@Bateman thanks for sharing that tidbit, it makes me feel all warm inside.
It would seem high spl testing is important, I guess us pro sound guy can be of some use after all
Send me your drivers and I will break them!
And just to tie us back to the topic: driving multiple subs at eigenmode mode node points is good, but how does one deal with problems in the height of a room?
-Matt
Matt,
Usually one deal with problems in the height of a room with a building contractor, but as this is a DIY forum, go ahead and get out the chainsaw and wrecking bar.
I have not measured the 2241 or 2242, just going from the JBL TS spec sheet, which shows 7.62mm Xmax for the 2241, 7.87mm for the 2242.
I have measured the BC18SW115-4, it definitely has 15mm Xmax with distortion under 10%. It also will not "bottom out" with 120 volts (about 3600 watts) of sine/square wave, as I found out three times in a row when I made a testing mistake
.Art
Attachments
Dr. Geddes, a few questions about your statements this ~week:
Interesting. Now I feel less like a slacker and more like someone who was on the cutting edge for having never gotten around to building 4th order BP subs to compare with the closed boxes I've been using (set up by your methods, in different rooms) for years.
But - leaving aside the lower distortion of the 4th order BP (I never thought that would be a major concern if one uses basically good drive-units to start with), you're not longer that impressed with the efficiency gains of the bandpass? That is to say, with enough cones pumping and sufficient power, you're no longer convinced that the efficiency bump a 4th order BP provides in the bottom half (or so) of its passband is that useful.
Does that change (assuming a suitable PR is available - I had no idea PE discontinued their "shallow" subs and PRs) if one takes a "2-way" sub approach, i.e. in your terminology separate "BroadBand" fairly high-tuned boxes to play in the modal region, and a single closed box for the ULF?
When mentioning efficiency as relevant, do you mean primarily for headroom in the modal region, or overall?
After all, isn't the "efficiency" of a subwoofer at the bottom of its range determined substantially by the size of the box, regardless of the driver's native efficiency?
Just to be clear, that means for subwoofers one needs to look at thermal modulation. That is to say, diffraction isn't so much an issue for bass drivers, though it's of high importance for mains.
Interesting. Now I feel less like a slacker and more like someone who was on the cutting edge for having never gotten around to building 4th order BP subs to compare with the closed boxes I've been using (set up by your methods, in different rooms) for years.
But - leaving aside the lower distortion of the 4th order BP (I never thought that would be a major concern if one uses basically good drive-units to start with), you're not longer that impressed with the efficiency gains of the bandpass? That is to say, with enough cones pumping and sufficient power, you're no longer convinced that the efficiency bump a 4th order BP provides in the bottom half (or so) of its passband is that useful.
Does that change (assuming a suitable PR is available - I had no idea PE discontinued their "shallow" subs and PRs) if one takes a "2-way" sub approach, i.e. in your terminology separate "BroadBand" fairly high-tuned boxes to play in the modal region, and a single closed box for the ULF?
When mentioning efficiency as relevant, do you mean primarily for headroom in the modal region, or overall?
After all, isn't the "efficiency" of a subwoofer at the bottom of its range determined substantially by the size of the box, regardless of the driver's native efficiency?
Just to be clear, that means for subwoofers one needs to look at thermal modulation. That is to say, diffraction isn't so much an issue for bass drivers, though it's of high importance for mains.
Just to be clear, Earl wrote In post 1510:
"As expected people are confusing thermal compression with thermal modulation. they are related but distinctly different effects.
"Thermal compression" is well understood, documented and sometimes even measured. It is a long term effect that happens after a long exposure to a signal and the entire motor structure heats and eventually reaches a thermal equilibrium. In this effect the heat dissipation is critical as it determines the long term temperature."
And in 1541:
"Thermal modulation is also, for the most part, independent of the current voice coil temperature."
The thermal modulation Earl believes happens in ultralight voice coil dome tweeters would not happen in the large, heavy voice coils used in a good subwoofer, and thermal compression is not a problem unless high average levels are maintained.
As he wrote in post #1517:
"The time constants tend to be too long to be a factor for low frequencies, but not for tweeters. A burst of HF material and the tweeter heats and a burst about 1 ms later will be lower in level with a change in frequency response.
This is dominately a HF effect, which is why its kind of out of place here."
Art
Last edited:
Cabinet design affects a drivers output more than sensitivity, but I dont think there is a direct relationship between sensitivity resultant sound quality. You can avoid power compression by using a high power handling driver in a cabinet that results in it only using half that power, and there are driver out there like that. You just need enough headroom in your system so your not pushing it too hard and you will avoid things like power compression whether your using one big sub or multiple subs that are large or small.
The efficiency gain and bass extension of the bandpass is just not a good "a value", not that it doesn't exist. I cease to see any situation in audio where "value" is not a real criteria (except arguments on forums like this). There is always a cost involved and it is virtually always a factor. A closed box is simply the best value when one is going to use DSP and multiple subs. Adding a bandpass port increase the size substantially and a little to the cost (people tend to not like either of those) and using a PR adds a lot o the cost and a little to the size (same issues). Closed box with ample power handling is cost effective and as small as practicable.
Efficiency is simply a means of looking at headroom. When I did tests of "powered subs" in almost every case the subs "crashed" from amp clipping even when it appeared that sufficient amp power was available. the higher the efficiency the lower likelihood of this happening. Yes the native efficiency and the "bottom end" efficiency are different, but they are completely correlated, so if one is "high" then so it the other" Box design has an effect, but its not huge, more of a secondary thing.
Subs have no issue with "thermal modulation", its a mid to high frequency problem. And diffraction at sub frequencies is not an issue either. For subs its all about how much SPL can be achieved without massive overloading, i.e. distortion in the 10's and 20% as occurs when something is being used beyond its design limitation. As I said, I find that this is usually an amp that is too low power, but sometimes a driver that goes beyond its magnetic gap limits. The magnetic gap thing however can be very low order and hence inaudible, while the amp clipping is very high order and highly audible. So, to me, X-max is less of an issue in a good sub driver than power handling and efficiency.