Measuring Amplifiers - Page 38

Thats a very good point, I just wish Velodyne tech and sales people would know this and answer my question properly. According to them if its set to GS (Generic Sub) then no filters exist and I specifically asked if there was a hard wired SFF inside, they said no. No huge deal for me in the end but I just hate hard wired SFF amps

When we look at penngray's first measurement on page 35 that has a 36db drop below tuning in less than an octave, nobody said a word until I mentioned it a page and a half later. That would lead me to believe most saw that and assumed it to be normal. In his measurements with a different amp, there is a much, much gentler rolloff bringing him down 6db at 10hz from the output around tuning. A gentle rolloff resulting in being 6db down at 10hz with a 16hz tune is not what I would consider an immediate death below tune.


Agreed. I would consider playing a 3-100hz sweep at progressively higher output levels with a >10hz tuned sub a not so intelligent idea though.


In my mind, the most stands to be gained (from an experimental point of view) with a flat anechoic response. That way, all non subwoofer created output becomes painfully obvious, AND, we more closely match a FR that our human limitations will allow us to appreciate. To do that though, at decent output levels, subwoofer drivers are not the right tool. Rather, how about a device with ~12.5 cubic feet of displacement (the equivalent of ~54 18" high excursion drivers) per stroke. At 1 meter ground plane, this will get you ~133db at 10hz, ~120db at 5hz, and ~88db at 1hz. A 100db sweep down to ~3hz while staying perfectly flat before room gain - now that's an experiment. That's what I'm working on, to be crossed over to a subwoofer at 20hz. Not nearly as costly or house-remodeling-necessary as a rotary woofer though.

Is less refined a good or bad thing??? Not sure which one your saying sounds better. Im thinking your saying that the Behringer sounds better, but I may be wrong.

I notice most every graph that's posted. I tried to suggest a wider BW at Sonny's BBQ, but it didn't play well. I think lots more could be learned about gain in-room, regardless of the sub measured, if input about the room and placement were included. Why offer REQW with capability to DC if you make people cut the graph off at 10Hz?

The problem in thinking that human hearing requires some sort of huge house curve is simple: If the soundtrack is encoded at -10dBSF, and your anechoic response is flat, but the BG/PVG is +20dB, you'll be experiencing the effect at a hugely distorted level relative to the rest of the BW.

I've always advocated flexibility in the anechoic response for that reason.

In any event, it certainly sounds interesting. How far are you from a test?

Bosso

what is the best way to boost a signal set from a receiver?? Ive been searching and thoroughly enjoying this thread and learning a lot...

the issue is i have a Mackie 2600 and an EP 2500...and right now the Ep is putting out more because of the volt sensativity. Im running the signal through a BFD...and they are running 2 eD 18's sealed.

how can i get more power out of the Mackie?? its a beast of an amp...i just know im not using all of it...any help would be great! thanks!

Ben

How do you know this?

Turn the Mackie's gain all the way up and turn the EP down until they match. Recalibrate the sub trim at the receiver.

Is there anyway to turn up the gain on the DSP1124 or FBQ2496? My DEQ2496 and DCX2496 can pump up the levels up to +15dB. So unless the CD mastering is screwed or what, usually there is enough voltage.

But if the output in the single digits - at the levels recorded on the media in question - aren't enough to really create a new and interesting experience, who cares? While a rising low end may be a distortion from a purist point of view, if it creates a more entertaining home theater experience without ruining other aspects of the presentation, the goal is achieved in my book.


Not too far.

EP2500 sounds much cleaner. I cant find more words to explain the difference, I would prefer the sound of EP2500.

Also, Tapco J-2500 now weighs 38 lbs, I measured it myself. Manual says the net weight is over 50 lbs...total BS.

are the eD 18" D2?

Yeah, sort of like the kids who drive by with 40Hz bumped 20dB.

'Purist' is a cheap shot. Imagining that a 20-30dB bump is necessary before you ever hear the soundtrack as intended is a bit much, really.

Starting at reference and playing around beyond that, for entertainment's sake, is one thing, but starting out at a grossly distorted point with no way to adjust to reference is completely another.

There is a huge misconception around here regarding the first octaves of the .1 channel. A 3Hz sound wave is in the neighborhood of 400 feet long. It can pressurize any home...entirely, before it leaves the subwoofer.

There is little chance you'll ever get to a listening level where the first octave is audible. That is irrelevant. It's the pressure that adds to the experience. It can be sensed at much lower levels than most here imagine.

Nousaine's comments on the TRW mentioned it generating 110dB in the first octave. Ssab's 4X18" system will easily do that, IMO, if he uses a wider BW power plant than the K2, which is down 20dB at 4Hz. It's a chuckle to me that Mr. Bass has never tested his IB (or any subwoofer) below 12Hz in his room, which would be the perfect environment and methodology for this subject.

Corner loaded in 7,000 cubes. Distance to the mic is irrelevant. Placement of the mic is irrelevant.

Still, I know you wouldn't be posting here about single digits if it was just blather, so I'm looking forward to your results. I hope you'll invest in a decent measurement rig and post some details. Debating the virtues of a listening experience based on guesses and opinions is pretty pointless. I'll be very surprised if you conclude that the first octaves are irrelevant once you've A/B'd with and without.

Bosso

My opinion is that the dip at 10Hz is a result of phase cancellation between the 2 subs vs the mic position, because with the single AV sub, there is no such dip and, at 10Hz, you'd need a room dimension of 56 feet to see a null.

One of the cool things about the sealed versions is that you could stack all 3 of them in a corner to see if the 10Hz dip disappears (or did you succeed in acquiring the 4th driver?).

It looks like a 10dB L/T or shelf boost will get you very good results, based on the driver/box model and graphs you've posted. In any case, having your detailed posts and graphs and my DIY hero Mr. Chas there will make for a great report and ensuing discussion, no matter what the results.

Bosso

I have a feeling it is a room / Sub cancellation as well. Hopefully spreading 4 subs around the room will solve this.

So far I have only 3 ava18's.....I wasn't able to acquire the 4th one. So I'm looking for a 4th as we speak. If I can't find one in the next week, I'm ordering a maelstrom X`18 from Kevin...it outta be as good!

The 1st cube is done...just needs some sanding and painting now. Chuck and I are hoping to power it up tomorrow or this week and play with it a little.

How would you know unless you've tried it?


No, four high excursion 18" drivers can't generate anywhere near 110db in the first octave in a home theater environment. Tremendous amounts of room gain *might* bring your near that level, but the subs themselves cannot generate anything approaching those levels. Using my device as an example again, even with ~12.5 cubic feet of displacement, only ~88db can be generated at 3hz at 1 meter ground plane. Plus, with the TRW or my theoretical device, that output would be largely distortion free, whereas room gain enhanced 110db output from quad Avalanche 18s at 3hz would very likely have well over 100% THD.


The main reason I am experimenting with such an idea is to bring real answers to this largely unknown area with a much more user friendly and inexpensive device than the TRW. You and I can postulate all we like, but until there is a device that can actually move enough air to produce sufficient output levels with minimal distortion in this frequency range, it's all guessing (sometimes educated, but guessing none the less).

Anybody got any info of the Intersonic's Sonic Boom Generator by Tom? Can't find much info, pix or price. Other than 132dB for an array of 6, 3Hz 2m. That's more than 100 TRWs!

Steve, 12.5 cubic feet of displacement? 88dB at 3Hz? You may want to check those numbers again.

4 Maelstrom-X 18s have only just over 1 cubic foot of displacement and will generate more output at 3Hz than 88dB @ 1M. Look at Ilk's LMS5400-18" test: 96dB at 10Hz. 84dB at 5Hz. 75dB at 3Hz. Add 12dB for 4 of them, 87dB at 3Hz.

25dB of BG/PVG and you have 112dB. Room gain being linear, there is a huge reduction in THD that's directly proportional to the gain, not that anyone would ever notice 100% THD at 3Hz as you'd have to get to H6 before it's an audible frequency).

Yes, 4 high excursion 18" drivers can easily generate enough output to result in a clean 110dB at the LP.

No, there will not be 100% THD. There will be less THD than at 30Hz.

Remember, at this point it's you who is postulating. 8-15s will match or exceed 4-18s. I already have the graphs to back what I'm saying.

Attached is a single 2X15" in 2.25 cubes. It's only flat to 5Hz, but it can do 100dB at 5Hz in room, no sweat. The peak hold graph shows 100dB at 20Hz and 100dB at 5Hz. You can see the THD results.

Add 12dB for 4 (2X15) of them. Do the math. Calculate the THD.

Now...125dB in the first octave? How do I know if I don't try it? You may feel free to try it all you want and LMK how it works out.

Bosso

bosso,

"25dB of BG/PVG and you have 112dB... 4 high excursion 18" drivers can easily generate enough output to result in a clean 110dB at the LP."

Are you saying that because of level not decreasing so much with distance in the pressure region (a most interesting point BTW, thanks), SPL at LP is the same as at 1 m?

Have you verified that by measurement?

Yes, I'm saying that some of the gain realized at the LP is due to the fact that ULF travels much farther than higher frequencies without losing intensity.

Yes, I have measured the phenomenon. When I originally showed the graph, most said that it was an outside artifact, like an HVAC system, a passing train at the time of the measurement or some such other ambient noise.

I then took 2 measurements from the same mic position with the same sub and signal chain, etc. The only difference was the introduction of a 20Hz HP filter, which acted upon the measurement exactly as one would predict, eliminating the response to single digits.

Since this is a relatively new thought regarding in-room response, I am open to any opinions that may better describe what is happening in these graphs, but the evidence seems to point to the facts that:

1. All rooms produce gain below 10Hz because they are too small to set up modes due to the wave lengths involved.

2. Ultra low frequencies are largely immune to the inverse square law regarding loss of intensity over distance.

3. Boundary gain is only adversely affected by the construction of the corner used for placement. This can be shown by the increases observed by measurements of the same subwoofer placed in 3 different rooms of varying degrees of lossy construction.

This had led me to the conclusion that a portable corner of a certain design and build may possibly be used to retain some of the gain otherwise lost to poor room construction.

EDIT: Notice, in the first graph, the loss of exactly 12dB from doubling the distance twice from 1 meters to 4 meters, above 20Hz, as well as the reduction in loss of intensity as frequency decreases.

Notice the nearly perfect 2nd order roll off below 20Hz with the inclusion of the HP filter in the second graph, which would not have been the case if the single digit response were due to a constant in the ambient environment.

Bosso

Bosso, I would have to agree with you on the observation that room construction does make a HUGE difference in how much bass is reinforced, due to corner placement. A room of the same dimensions in a basement will have more gain in the ULF than a main floor, due to sturdier walls and concrete floor.

This discussion seems to be leading toward some resolution of long argued points of view and analysis and a possible useful solution to the question, which room size and shape is optimal for a home theater and how should it be built?
Is it possible there IS a "perfect room size and shape" that takes advantage of all possible factors of sound reproduction and transmission within a closed space using loudspeakers as a sound source?
With the popularity of home theaters becoming a part of new home construction, maybe your insight, experience, and discussion will impact that segment of the industry as well as supply us DIYers a better way to make it play.

Bosso,

#2 is most definitely not true. Low frequencies decrease in intensity at the same rate as mid frequencies in air. I think you need to consider what causes changes in sound intensity with distance.

Most speakers (6.5" woofer for example) generate spherical sound waves (below 1kHz in this example). Spherical sound waves increase in area by a factor of 4 as the distance from the sound source doubles. This is what causes the 6dB decrease in SPL every time the distance from the sound source doubles. If you have a very large driver area (in both the x and y dimensions) operating as a rigid piston, compared to the wavelengths its producing, you will get a plane wave. From a strictly power/area standpoint, the SPL from this source won't drop as your distance from it changes.

This is what frequently happens when you have a subwoofer in a closed room. The subwoofer causes one wall of the room to vibrate. Since this wall generates a plane wave, the SPL from it doesn't drop as you get farther away from it in the house.

Thanks for the info. I would tend to agree with you were it not for the information I've gathered over recent years on the subject, which maybe we could discuss:






(If the inverse square law applies, what would the dBSPL be attenuated to from 90dB over 4,000 meters? down by 74dB to 'up to' 16dB?)





Bosso

which would explain why the low rumble of thunder is heard and not the high pitched crackles when a mile or so away.

Bosso,

Your quotes are discussing several different things, which I believe are all accurate, but which don't apply to soundfields or propogation in small environments (houses).

The intensity drop with distance from a sound source has primarily to do with the fact that the area the waveform occupies increases. This means that the SPL must drop due to less energy per unit area. When considering this behavior, there is no energy loss in the system.

Sound energy is "lost" as it travels through the air. By lost, I mean that heat is generated. This does happen more at high frequencies, due to moisture, than at low frequencies. However, this decrease in sound energy is very small compared to the loss in SPL due to the inverse square law.

Some of the other quotes (elephants) are dealing with sound transmission through solid matter, such as the ground. In acoustics, this isn't considered sound because it's not vibrations transmitted through air. I wasn't including it in my original comment because of this.

Acoustics, by Berenak has probably the best overall explanation of the physics involved.

Room gain will amplify the distortion just as it will the clean output. Yes, it very likely will be 100%+ at 3hz.

Unfortunately your measurements don't mean much to me, as the starting point (super high distortion) is not what I am after. That aside, room gain won't be more than 12db/octave in the absolute most ideal situation, so I'm not quite understanding how you have a rising low end unless you are EQing flat to single digits anechoically, in which case you would be contradicting what you advised against doing in previous posts? That, or your measurement gear isn't calibrated correctly.

"1. All rooms produce gain below 10Hz because they are too small to set up modes due to the wave lengths involved."

I guess that's correct but it seems like an obtuse way of describing it.

A more clarifying way to think about it is like waves in a bathtub.

If you beat the water quickly with your hand shorter waves will be sent out with many peaks and troughs along the length of the tub.

If you submerge your hand slowly enough, the water level rises in the tub everywhere at once.

"2. Ultra low frequencies are largely immune to the inverse square law regarding loss of intensity over distance."

"#2 is most definitely not true. Low frequencies decrease in intensity at the same rate as mid frequencies in air. I think you need to consider what causes changes in sound intensity with distance."

It's not true outdoors, but I believe the discussion is directed at enclosed spaces.

However #2 is only strictly true at 0 Hz, where the sound pressure is truly at the same phase everywhere in the space, and that above 0 Hz the situation will be something in between, which was why I asked about measurements.

"Room gain will amplify the distortion just as it will the clean output."

The distortion is higher in freq and not boosted by the room gain.

Another way to think about it is that the woofer doesn't have to be driven as hard for the same output, so the distortion is less.

Thanks Noah, I didn't even know where to start to pick through the wild posturing and misattribution of real facts...


Very nice analogy, as it also holds if the tub then becomes a swimming pool or a small cup of water.



It doesn't matter the condition. From an extreme mathmatic perspective, a sound source ALWAYS drops at a rate of 20*Log(D/Di). Any surface that is large relative to the wavelength involved can be viewed and considered as a miror of the source. If the complex situation of a room is viewed as an array of virtual sources instead of a ray tracing problem, we can get a different perspective on the behavior we can expect that makes it much easier to conceptualize. Once you have an array of spaced sources, we can evaluate the problem based on wavelength vs. spacing. Mathmatically the result is the same for any given point, but it makes for a lot of messy math to observe what is reasonably obvious to anyone who has modeled and tested the interactions of multiple sources such as pro audio loudspeaker arrays.

Getting back to Bosso's point, it is a valid observation with the wrong attributed cause. It is true that the lowest frequencies will have a more uniform sound level throughout the room as compared to higher frequencies, but this is due to the reflective qualities of the walls. Low frequencies disperse the same as higher frequencies, and the walls are not acting as planar subwoofers. From the perspective of virtual sources, the listener, and the entire room in fact, is within a dense field of infinite sources. If the reflections have loss, the further sources will attenuate in level, but below some frequency the sources will be very closely spaced WRT frequency, so you will observe a mostly uniform field moving through them.

Since the math is still the same, this gives us a clear depiction of why room gain occurs... The closer the walls, the higher density of virtual sources. The more reflective the boundaries, the higher in level all of those virtual sources will be. Look at the problem closely enough and you might make a similar observation I did, where the hardest acoustic problem to overcome in a small room is the rear wall.



Absolutely correct. Distortion is not affected if gain is constant with frequency, as is observed in ideal ground plane conditions. A good example of this form of distortion reduction on the other end of the spectrum is with constant directivity horns. A 1" dome which has a flat response on a baffle, will, in general terms, have a response that rises somewhere below 10kHz and then increasing to some lower frequency as dictated by the size of the horn. When the dome is driven identically as on a baffle, all factors which create distortion are the same, but the fundamental frequency can be raised by the confinement of the horn, which still aligns with the virtual source perspective, but the boundary is only significant to higher frequencies. The similar case can occur at lower frequencies, and yes, a large peak in the upper bass response will amplify distortion products which co-incide with such a peak.

Please note that this effect only happens when the variable gain WRT frequency occurs AFTER the production of the distortion. In other words, EQ in front of a subwoofer has no effect on the distortion per given SPL per given frequency.

Were we talking about amplifiers at some point here?

The discussion regarding THD in the first 2 octaves is an absolute non-issue. Noah and Mark are correct, but I have always said that, in either case, the harmonics of these frequencies are inaudible. They are inaudible. They have zero negative effect.

There is also the measurements that I've taken at 4Hz, 5Hz, 6Hz, all at the same dBSPL at the LP and all with extremely low HD showing up on the graph. If there was 100% THD, the harmonics peaks would be there on the graph.

You may as well argue the audibility of group delay at 5Hz.


I understand the theory involved, but I haven't seen any measurements that bear it out. The evidence suggests that ULF does indeed travel much farther at an intensity that contradicts the law. If the explanation is that higher frequencies are attenuated by refraction and/or absorption, the formula for the law is incomplete and has no application in HT discussions.

GP tests such as Ilkka's sweep of the LMS5400-18" L/T, expanded to 1Hz, could easily nail this one. A measurement at 2M and at 16M should show a perfectly uniform drop of 18dB across the entire BW. End of story.

To date, there is no such measurement comparison that I can find anywhere.


I suspect that this is the correct explanation. The room creates virtual sources and those virtual sources vary in intensity according to their refractive properties and are frequency dependent.

Of course, again, the evidence shows that this theory is also suspect. ULF appears to be less affected by walls and in fact travels better through the earths crust than it does through air. This makes it hard to understand how a 4" wall in a typically constructed home offers any sort of refraction as frequency decreases, much less offering more refraction.

In the end, I am less interested in the precise mathematical formulae involved and much more interested in the results. I'm not a physicist. I'm a HT enthusiast who has a library of source that has content that's intended to be reproduced as part of the A/V presentation.

I continue to be taken to task over this by the weakest arguments imaginable. "It's impossible with conventional drivers. Flat isn't loud enough. There's too much THD. The source is mistaken, it's just artifacts. It adds nothing to the experience. There isn't enough juice in a 20A circuit. The signal chain rolls off. GP measurements to below 10Hz won't offer any useful information. There was a bus driving by. The HVAC skewed the measurements. It's one of the walls. The box is too small. Ported is better..."

Mark is right, this is just a silly thread jack. I've gotten Mark so befuddled that he doesn't know where to start. I apologize, Chas. This is a great thread that should have all of these first octaves posts deleted.

My original intent was to lead to the suggestion that a FR graph for each amplifier be included because reproduction of the LFE channel, especially DD TrueHD and DTS Master Audio, includes information that a SW amp should be capable of amplifying. It seems that watts at 20Hz vs rated watts is all that matters, which is cool. It's a great thread.

Back to amplifier tests...

Bosso