Example of the Effects of Room Placement & Baffle Step

[---k---]

I'm making my way through Floyd Toole's book, "Sound Reproduction". So far, it has me some what shaking my head in disagreement, but that is another thread (and whom am I to argue with Floyd Toole?). One thing I just came across that I thought would be very helpful, because it comes up all the time here, if measurements of the effect room placement on a speaker. I know that most of us already know this, and Jon has some of his own graphs that he posts, but, I like these graphs. They are nice and simple. I thought that this would be a good info for our FAQ section.

Here are the measurements of a typical speaker (Infinity Primus 160) taken out in space away from all boundaries, placed in wall, placed on-wall, and placed in a wall with openings around it such as would happen if it were placed inside an entertainment center (I like this graph the best, since I've never seen one like it before). You can easily see how the placement affects the speaker, especially the bass. We call this Baffle Step Compensation.

[augerpro]

Great book. Interesting that the ideal 4pi and 2pi match the actual response so well. The onwall mounting dip is something I'm modeling (V16 SE) right now for the BoseBuster2. I'm trying to keep it is the enclosure depth as shallow as possible so the dip is *hopefully* just past the crossover frequency.

[fbov]

Ryan,
Two comments.

Toole's opening sections are psychoacoustic in focus but scientifically sound, if counterintuitive in result. Remember that he's trying to find out how auditory perception works, before we recognize it as "sound." It made more sense on a second read. There are plenty of "optical illusions" that illustrate visual perception, but very few that address auditory "illusions" in a manner that reveals the illusion. That's one reason I find audio so fascinating!

He provides copious amounts of data, with references to original research, so you can argue with his conclusions, but not the data. The man is a scientist and he's giving you a synopsis of his life's work in a manner that's directly applicable to us and of no direct benefit to him. I find that refreshing.

And I've wanted to post charts, too...

Have fun,
Frank

[dlneubec]

I would like to read that book. One thing of note is that what you see in those graphs tells only part of the story. You have at least a coule things at work here, I think. One is with regard to baffle step compenstation, having to do with the baffle size itself and when the driver changes from 4pi to 2pi space, resulting in a 6db increase in on axis response and a similar decrease in off axis response. The other is with regard to boundary reinforcement from surrounding surfaces, which as you move closer to them, can effectively swamp out any baffle related issues, since the driver pretty much stays in 2pi space when flush with the wall.

The other issue and one I see here in some of these graphs, is the position of the driver in terms of proximity to additional boundary surfaces. For example, take the graph with the big dip at 200hz, where the speaker is on the wall. Chances are if you make that a tower and move the woofer down near the floor, so it is in close proximity to two boundaries, the floor and the front wall, you can might fill in that null. This is essentially what is in Roy Allisons research on boundary reinforcement. That is why you see much of Allisons designs having the woofer very close to at least one, usually two boundaries and the speakers are designed to be placed tight against the wall. There is some intesting stuff in his research. I have a little program that is based on Roy's research that can simulate the low end response by inputting the drivers location relative to the floor and the front and side walls.

[augerpro]

Dan SE V16 has included a boundary diffraction calculator based on Allison's and some other guy's research. It combines both the baffle diffraction model with the room boundary model.

[dlneubec]

Cool, I didn't know that had been added. I guess I really need to upgrade from V13 soon! ops:

[---k---]

Yeah. I mainly posted the graphs for those people who come here and ask the question, "I want to build a pair of XXX speakers, but I have to mount them inside my entertainment center (or whatever). Is that OK?" or similar questions.

What I didn't mention above though is one of Toole's conclusion for the in-wall is:

p.190
The overall conclusion is that mounting this excellent little loudspeaker in a wall has left its overall performance substantially intact, but the bass output has been greatly increased, making it sound fat, thick, and tubby. ... The solution in this case is to turn down the bass. Any competent equalizer can do it, or the old-fashion bass control may just be optimum if the 'hinge' frequency is around 500 hz.

[BOBinGA]

Ryan, does Toole also provide a measurement with the speaker placed about three feet out from the wall about ear height? This is the most common living room placement and would be the most interesting data for most of us.

-Bob

[---k---]

Maybe in a later chapter.

[---k---]

Thomas,
I keep reposting this. Can you move this thread to the FAQ section to make it easier to find?
Thanks.

[joeybutts]

Very interesting. I'm an intermediate audioist (new word....kind of like it since I wouldn't consider myself an audiophile...) and just experimented with this at my Grandmother's house. She had purchased some Klipsch Bookshelfs (can't believe it) and was complaining about their lack of low end. She couldn't really hear any bass. True to point I went and listened and found that she had the speakers SITTING ON THE FLOOR AGAINST THE WALL, flush up against the side of an entertainment center. The room is rather large for these (about 20 x 14). These bookshelves have a ported back, it looks like a 5" woofer and a Piezo tweeter (I think that is what it is called, noobishness rears it's ugly head).

So in listening I played with their position and moved them around placed them on a chair stacked with books and whatever I could find. She came back in and was amazed at the difference in their sound. I was amazed at the change in sound. So I told her I would build her some stands and we would position them where they need to go.

Kind of bloggish and in no way scientific, but just thought I would share that I learned the importance of placement.

[---k---]

We recently heard a significant improvement when we pulled Mike's Dayton Classic WMTMW out from the wall another 12 inches. (Front baffle moved from ~18" to ~30" away the wall) It very obviously cleaned the muddy sound up.

[fjhuerta]

Thanks a lot for this thread, Ryan. I think I need to buy that book! It's very interesting.

[HokieJoe]

We recently heard a significant improvement when we pulled Mike's Dayton Classic WMTMW out from the wall another 12 inches. (Front baffle moved from ~18" to ~30" away the wall) It very obviously cleaned the muddy sound up.

Would the addition of some insulation (of sufficient thickness and density) on the wall behind the speakers accomplish the same thing?

[---k---]

Maybe some. A lot depends on the thickness. Problem with bass is the insulation needs to be pretty thick.

[HokieJoe]

What frequency's cause the most issue for close wall placement- 50-100Hz*?

*For floor and stand-mounted speakers

[---k---]

Well, that depends on the speaker. As you can see from the graphs above, for a speaker that was designed to be 3' from the wall, if you push it too the wall you get a big hump in the bass from 0hz all the way up to 200hz. But, if the speaker was designed to be closer to the wall, you won't get this as much. The opposite is also true. For example, the Khanspire's were designed to be close to the wall. Pulling them 3' out into the room results in very lean bass.

The other problem with onwall mounting is the null caused by the reflection off the back wall merging with the sound radiating from the driver. In this example, that is ~200hz. It will vary based on the depth of the speaker. But, it will be around 200hz. The frequency it occurs at is equal to the wavelength/2.