Room Gain Complementary (RGC) tuning

[audiokinesis]

Lately I've been using a vented-box bass tuning approach that might be deserving of its own name and acronym. Sort of like EBS (extended bass shelf) tuning, which is where the bass rolls off gently at first starting fairly high up, then levels off (shelves) above the tuning frequency, and then of course plummets below the tuning frequency.

Let me say up front that I'm not actually doing something new (though it is a bit off the beaten path) - what's new is the descriptive name and accompanying acronym.

The concept behind Room Gain Complementary tuning is this: To have the speaker's natural rolloff rate be approximately the inverse of the anticipated boost from room gain. If we assume typical room gain to be 3 dB per ocatve below 100 Hz, then we'd want the woofer to be rolling off at 3 dB per octave below about 100Hz or so. RGC tuning isn't appropriate for every application - for example, I wouldn't use it for a stand-mounted monitor that's going to be placed well out into the room.

I think one of the main reasons many subwoofer systems sound slow and disconnected from the rest of the presentation is the result of room gain overboosting their output. If you have a sub that's flat to 25 Hz anechoic, after room gain it will be roughly +6 dB at 25 Hz, and that's too much for natural-sounding bass reproduction. In fact, the myth that smaller subs are "faster" may well arise from their typically rolling off higher up, so that they suffer less from undesirable deep bass overboost due to room gain.

In Earl Geddes' brief study of multisub systems published on his website, we can readily see the effects of room gain. This study probably assumes a more effectively-sealed room than most of us have, so the room gain here may be more than we can normally expect, but I think it's wortwhile to see how it plays out in the modelled response:

http://www.gedlee.com/downloads/sub%20study%20.pdf

Let me give an example of a commercial speaker with RGC tuning: The Guru Pro Audio QM10 loudspeaker. This is the little up-against-the-wall Swedish mini-monitor that supposedly goes down to 30 Hz. When used correctly (including correct stand height), I think the claim is plausible; at least, I think it could hit the mid-30's. Here's a link where you can see the anechoic frequency response that results in its RCG-style tuning:

http://www.soundstagenetwork.com/mea...proaudio_qm10/

The Aerial Acoustics model 10T looks to me like it has an RGC bass characteristic from 60 Hz down to 20 Hz:

http://www.stereophile.com/floorloud...66/index4.html

By way of contrast, here's the anechoic response of a speaker that does not use RGC tuning. This speaker's bass characteristic would be the better choice for a stand-mount placed well out into the room:

http://www.soundstagenetwork.com/mea...r_audio_pl100/

I suspect that the oft-cited preference for low-Q sealed subwoofers over typical vented subs arises in part from the low-Q sealed box's rolloff rate (approximately 6 dB per octave) being a reasonable first approximation of the inverse of room gain. If the room were removed and we were listening outdoors (with no room gain), a flat-tuned vented box would almost certainly be preferred over the low-Q sealed box. I think that, with an appropriately designed and tuned vented box, it's possible to improve on the room-friendliness of even a low-Q sealed box in a cost-effective manner.

Obviously the concept I'm using is nothing new, but perhaps by giving it a name and an acronym awareness will increase and it will be taken advantage of by more DIYers (and maybe more manufacturers).

I welcome comments and insights, especially regarding what typical real-world room gain characteristics actually are. I'm going off a couple of published sources, but there's doubtless more to be known about that side of the equation. In practice, I would expect room gain to vary considerably from one situation to another, but what I've described above is probably still a good starting point.

Duke

[sqlkev]

excellent post!!

from measurement point, I think it's not that hard to combat room gain by using a good active processor.
Keeping the bass sounding accurate and effortless from the sweet spot in a typical listening room is much harder (at least for me so far). I could never get my stereo subs to align right at the center of the room. (no amount of time alignment/phase correction could fix this)

Recently i've discovered your Swarm sub system and it piqued my curiosity with that multisub system concept.

[audiokinesis]

Thanks, sqlkev.

The best-of-all-worlds sub might be a large, high-displacement sealed box equalized to have an RGC response. I suspect that's what the JL Audio subs do when set up properly. Actually I'd want four of them, with Geddes-esque asymmetrical placement.

The approach I described above is the poor man's approach, as equalization isn't needed.

Duke

[gedlee]

Duke

I never suspected that anyone would use that report to support room gain - live and learn.

I am not a strong proponent of room gain because it requires a very well sealed room and I haven't seen any of those yet.

I recently did an extensive set of measurements in my listening room which is far far better sealed than most - its sound proof and any leaks are sound leaks. (I'm going to document the procedure for setting up multiple subs. ) Well there was no sign at all of any room gain - nada - nothing. Just two dominate modes at 20 Hz and 40 Hz that no amount of placements and adjustment could seem to get rid of - mind you because the room is so well damped these modes are only about +3 dB. But had I any EQ, I could have gotten a nearly ruller flat *in-room* LF response - multiple subs work. The response went down to 20 Hz and then dropped like a stone

Sorry - no room gain below that.

As to the room having +3 dB below 100 Hz, I guess that I'd ask relative to what?

[CLS]

May I ask a dumb question?

Do we have a definition of "room gain"? Do the supports by boundaries count? Say, 2pi radiation has the help from wall (or other infinite plane) while 4pi does not. Then we may say we get some gain on 2pi, etc, etc...

If so, then we don't have to seal the room to have some acousitc gain. All those boudaries are helping on restraining the window of radiation, and more so in lower frequencies. So we got more effective output in any ordinary rooms (related to anechoic chamber or true 4pi). Is that correct?

I'm somewhat confused here.


to Dr. Geddes,

In your previous post, you mentioned there's NO room gain in your well-damped listening room. I just think of an anechoic chamber. Is it possible that the walls/ceiling etc. absorb all frequecies too well, and effectively makes the acoustic environment tend to bahave like an anechoic room?

No offense, just a plain question. Because I don't know the construction of your room.


This reminds me one thing. 10-some years ago, I worked in a local hi-fi store. There was one of the demo room constructed as a major show room (with lots of acoustic "treatments"). It was a dedicated listening room with a special door which can effectively seal the whole space. Double walls and double window panes made it sound proof. Inclined multi-angle ceiling and front walls with various treatments made it look quite like something...

However, so interestingly, almost NO speakers sounded good in that room. Or I should say, all speakers we were selling could not work properly in the room, including all those big-name/big-money ones at that time. ( some 4-peice monuments, and the famous spiral shell.... ) All speakers in our store sound too lean and dry in that room. It seemed room treatments ate up the bass (maybe also other frequencies...). And none of those comercial speaker products were designed for such space. I surely believe that one can design a system with such dead room AND make it sound good. It just could not be done at the situation we were in back then

So I think it's very important to have a system with highly adjustable subs to solve this room-related problem.

[simon5]

To help this discussion, I'll show you a old Adire Audio Shiva paper made by Dan Wiggins with a room gain curve.

It shows 2 dB to 3 dB of room gain per octave.

Look at page 5.

http://web.archive.org/web/200408130...plications.PDF

[audiokinesis]

Earl, my understanding is that "room gain" is the gain one can expect from placement of the speaker near the room boundaries, like on the floor and/or against a wall. Maybe I'm misusing the term - perhaps "boundary gain" would have been a better wording.

Here's a link to an alleged "room gain curve". I read somewhere that the source was Martin Colloms, but I can't verify that:

http://www.speakerbuilding.com/content/1020/rge.gif

My other source was the Dan Wiggins paper that simon5 linked to. See page 5.

Duke

[BHTX]

Originally posted by gedlee
Well there was no sign at all of any room gain - nada - nothing.

Strange. On the other hand though, imagine sitting in your listening position, and all the walls in your room fell over, flat to the ground beside them (if your room was all by itself away from other connecting rooms). What would the low frequencies sound like then? Not much I'm sure. No room gain? Riiight.

You've probably measured your subwoofers outside though, right?
That's why I find your resulting in-room measurements a little odd.

[gedlee]

I'm not sure that I can clear up all these questions in a simple post.

There is "room gain" in a weak sense, it comes from the modes, but the simple boundary gain analogy is misleading. But "gain" requires a reference to something as normal and there-in lies an issue. A boundary adds gain, yes, but a room has several boundaries and they interfeer to creat gain and no-gain, i.e. there are modes or "standing waves" (Yuck I dislike that term!!). So, to me, there is only a room with modes. The more modes the better.

Room gain historically came from the static mode that pressurizes the room below the first mode - to me this is what the term means. This mode CAN be quite pronounced in a car, but a van does not seem to have it as the volume gets too great. But a listening room in a home, as I said, does NOT have a static mode of any relavence. There is substantial "gain" in a small vehicle at very low frequencies "relative" to what would be there in a larger room from this same source.

I think the question about a room where the walls go flat is a good one. Lets call this the reference. Now as I add one wall, I get some gain from the reflected sound - actually gain and loss as the reflected wave adds and substracts from the direct one. As I add more walls there is gain, but also cancellation from the different walls. The gain occurs at a mode, but between the modes there is no gain as the response will go back down to what it was with no walls. If one wants to define "room gain" as the mean level increase from the modes with frequency, then the room is in fact a HP filter as the modal density will decrease as frequency goes down so the mean level has to decrease. So to be accurate the room gain occurs at HFs NOT LFs.

Now if we dampen the room then the peaks of the modes come down and the valleys fill but the mean level will tend to drop further accentuating the HP characteristic (if I just damped the LFs).

As to the Dan Wiggens discussion, that was all theoretical and not founded in a good model IMO. Its not based on actual measurements.

Thinking about LF sources independent of the room is IMO not a very useful thing to do. The two things are one system, they act together and modify each others characteristics. A room has modes and we must deal with these modes - embrace them if you will. Thinking of a room as a corner or a wall is not accurate - there are eight corners and six walls and we have to consider ALL of them.

CLS

As to a "sound treated room" making the speakers sound bad - yes this will occur. Thats because most people treat a room wrong. If you read my work you will see that I use almost NO HF absorption at all - my rooms are very live, but I do use a lot of LF absorption. This is difficult to do and I suspect that the room that you were talking about was done wrong.

[simon5]

Originally posted by gedlee
[...] As to the Dan Wiggens discussion, that was all theoretical and not founded in a good model IMO. Its not based on actual measurements. [...]

Originally posted by Dan Wiggins
[...] Our models are based upon application of accepted physical principles, as well as confirmation with empirical measurements in several different rooms. [...]

I don't want to contradict you since you are one of the legendary and mythical people around here but let dissect Dan Wiggins argumentation please.

The typical room he seems to be using is a 50 m^3 room with dimensions 4m x 5.5m by 2.3m if you refer at page 4.

Pressure vesel gain would start when the frequency is longer than the 2.3m walls, then more gain when frequency is longer than the 4m walls, then more gain again when frequency is longer than 5.5m.

2.3m = 130 Hz
4m = 75 Hz
5.5m = 55 Hz

Since the subwoofer is corner loaded, those cumulative gains will not show at exactly those frequencies, but it will be a shallow gain curve because the actual distance between the walls and the subwoofer is not constant.

According to his curve, his room seems to fit the data. The first fast +6 dB gain is caused by corner loading. Then, the shallow gain of 2 dB per octave starts at around 160 Hz, a bit higher than expected, but not impossible.

What do you think about that ? Thank you very much !