SPLAYING WALLS, non-rectangular, averaging method?

Further thought... 1. Is it worth it to splay the side walls if I can't splay the ceiling at all? 2. What new measures do you think I should take into consideration for acquiring the appropriate Room Ratio Dimensions since it is NOT a perfect rectangle, NOR splayed walls AND ceiling, but instead ONLY splayed walls. 3. I've read something along the lines of having the rectangle 'best fit' inside the new room shape, but I was wondering if it is impractical/illogical to instead average the dimensions. like so: (just an example, not actual dimensions for my room....yet.) I will also note that in BOTH shapes the square footage remains EXACTLY the same. Yes, I can imagine it may not be relative to the whole Ratio's game because it's the sides' length that make the difference, I thought I'd just mention it in case it was. :o 4. Going back to number 1..... If I create a false ceiling containing DEEP absorption, then perhaps the effects of comb filtering and poor stereo image will be appropriately treated from JUST a splay of the side walls? 5. Now....how far an advantage is it to splay at all vs. a rectangular room? Believe me it's in my favor almost all for the extra space it creates in the back area of the room. However, seeing that the room is pretty small to start with, and I'd need great treatment along the side walls anyway, should I splay at all?

1. Yes. If your room is big enough to loose that space, do it. .- Generally 6 degrees per wall will give you a 12 degree splay. It will reduce flutter echo and 1st reflections. I dont want to be mean, but seriously, this has been covered a lot on the forum. Just take some time browsing it .

I have something to add to this....and it's not meant to derail the thread at all. When considering room ratios, how do you calculate those ratios for splaying walls. Do you first calculate the room ratio and THEN splay the walls 6 degrees outward? What about the ceiling if it is to be raked/splayed? Say I want a 12ft ceiling. Should it be 12ft at its lowest point with a 12 degree rake toward the back or should it be 12ft at the back of the rake?

Should it be 12ft at its lowest point with a 12 degree rake toward the back or should it be 12ft at the back of the rake?

Doesn't derail at all, in fact it reinforces my questions number 3, which is the question I need most help with. From what I gather, you are asking if you splay starting from the front to back, or vice versa, which can apply to both the side walls OR the ceiling. The only difference is that the ceiling requires 12 degrees, and the walls BOTH 6 degrees. What I'm trying to get across is if it is plausible/logical to splay from the MIDPOINTS of the wall dimensions, hence averaging the splay evenly, and maintaining the volume from the original shape. And what effects, better or worse or nothing, do you think would come of a room splayed in this manner VS splaying from the front wall, which is a method I have seen. Mathematically speaking, it seems wise to average measurements out when having to bend the rules (ratio of a non-rectangular room). However, this is acoustics and not just geometry, so in my lack of acoustics expertise I'm here posting my question. To clarify : (The dotted lines are ALWAYS the original shape. Some may overlap with solid, obviously.) any thoughts.....?

Keep in mind any figure you arrive at is an estimate. You'll still have to test the room and make final adjustments. With that in mind, calculate the volume of the room and average the LxWxD to match. Splaying the walls only affects reflections, not room modes.

there's more than one reason to splay the walls. Standing waves is only one of them. check out the reflection paths in this drawing

I'm with everyone else here: Splaying is worth it for several reasons, including: first reflections, like John pointed, and flutter echo too. And it also just LOOKS cool! take a peek at some of John's designs if you don't believe me on that point. But getting on to the room ratio calculation: The dimensions for the room ration are ALWAYS measured from the hard finished surfaces of the inner leaf of the room, before treatment is applied. Theoretically, the only way to figure room modes accurately in a non-rectangular room is with finite element analysis software. However, my reasoning (and I may well be wrong, but none of the experts has said so yet... :)) is that the differences in the dimensions are so small relative to the sizes of the wavelengths, that I reckon approximation to a "bets fit" rectangle will give you a reasonable idea of the modes that you can expect. If you think about it, angling a wall by 6 degrees is only going to move one end a few dozen centimeters with respect to the other, but at the low end of the spectrum the waves are many METERS long, not just centimeters, so the small difference in distances just aren't going to make a lot of difference. They way I see it, up to about 200 Hz, using a best-fit rectangle to approximates the room dimensions is going to get you fairly close. The other is issue is that, since the walls are splayed, there should be much less tendency for axial standing waves to form anyway, since the surfaces aren't parallel, which leaves you with only tangential and oblique standing waves to worry about. The angles won't affect them much, other than to make the paths slightly longer or shorter, which might change the frequency by a couple of decimal points in the bottom octave, or maybe a couple of Hz. three octaves higher. So that's my reasoning for suspecting that a best-fit rectangle will give you a good approximation of how the modes will line up, especially the tangential and obliques. But once again, prediction is no substitute for measurement. Once the room is actually built, and the final inner leaf wall surfaces are finished, then you can actually measure the reality of the room, the real response and RT-60, and only THEN decide what to do for the treatment. Even in a perfectly rectangular room the actual response might not match the predicted response, and even more so with a non-rectangular room. So like Sandersd said: you still need to measure the actual response of the finished room in order to find out how it really responds. Mathematics and computer software will only give a rough approximation at best. My suggestion would be to splay your walls at 6 degrees (unless you also need to have your speaker axis intersection at 90 degrees instead of 60, in which case you should splay the walls 9 degrees, not 6), then use the best-fit rectangle for that shape to come up with a really good ratio, and build based on that. Then when you are finished building, measure it and decide how to treat it. - Stuart -

With an in-side out room using heavy 35-40kg insulation. Are the ratios measured from the "outer" plasterboard, the slats or the cloth finish? or a mix of both? :shock:

:D - I have no idea Deanne - I've never worked out a mode in my life :)

Good! I'm glad you said that. I think your attitude has rubbed off on song cave too :D Yanks and there formulas

ha ha - no it's Rules they love. :D

I'm still confused. Let's say I'm using the Sepmeyer #1 of 1: 1.14: 1.39 If I plug in 12 for the height, does that mean that the ceiling should be 12 at it's LOWEST point or HIGHEST point? Also, using that same height, it yields 13.68ft for the width. Again, is that at it's narrowest point, or widest point?

Myriad_Rocker wrote:

I'm still confused. Let's say I'm using the Sepmeyer #1 of 1: 1.14: 1.39 If I plug in 12 for the height, does that mean that the ceiling should be 12 at it's LOWEST point or HIGHEST point? Also, using that same height, it yields 13.68ft for the width. Again, is that at it's narrowest point, or widest point?

exactly :)

John Sayers wrote:

Myriad_Rocker wrote:

I'm still confused. Let's say I'm using the Sepmeyer #1 of 1: 1.14: 1.39 If I plug in 12 for the height, does that mean that the ceiling should be 12 at it's LOWEST point or HIGHEST point? Also, using that same height, it yields 13.68ft for the width. Again, is that at it's narrowest point, or widest point?

exactly :)

Ok..... :(

ya walked right into that one MR.

Sorry MR - I wasn't trying to put you down as you asked a very pertinent question and I'm sorry if I appeared to brush it off. Take Jamie's studio in Maryland - first studio on the current projects site http://johnlsayers.com/Pages/Projects.htm there is an obvious mode generated by the floor to ceiling waves and sure enough, Jamie picked it up when he fired up the speakers. We then added the ceiling cloud that sits over the console and is angled to break up that mode - it works - mode gone, well moved, so it doesn't interfere with the mix position.

Myriad_Rocker wrote:

I'm still confused. Let's say I'm using the Sepmeyer #1 of 1: 1.14: 1.39 If I plug in 12 for the height, does that mean that the ceiling should be 12 at it's LOWEST point or HIGHEST point? Also, using that same height, it yields 13.68ft for the width. Again, is that at it's narrowest point, or widest point?

MID-point! It's the best-fit you are looking for, so if your ceiling slopes evenly from one end of the room to the other, then the mid-point of the ceiling will be the average, or best-fit, AFAICS. So in your case, the ceiling would be 12 feet high above the middle of the room, several inches lower at the front, and the same number of inches higher at the back. Ditto for side walls. Once again, this is just my own line of reasoning here, and I haven't found much literature to support it. But so far none of the acoustic experts have shot the idea down in flames either, so I'm hopeful that I'm not too far wrong.... :) - Stuart -

Room ratios only relate to modes. The whole point to room ratios is to evenly distribute the mode peaks across the spectrum. Modes and reflections are two completely different items. I think your confusion is due to a less than thorough understanding of how modes are propagated. The calculations are based on a rectangular room. If you splay the walls, you must average the dimensions to equal the volume of a rectangular room in order to compare. Splayed walls have a nominal effect on modes. I want to reiterate the result is only an estimate. The further away from rectangular you get the less meaningful the results will be. You must test the room after construction and make final adjustments via room treatments. That's all there is to it. I assume you already have copies of Everest's (Master Handbook of Acoustics) and Newell's (Recording Studio Design) books...

Ah YES! The answers come forth! :D

MID-point! It's the best-fit you are looking for...

The midpoint! I don't know if there already was information supporting splaying from the midpoint, but I just couldn't find it. Thank you Soundman for clearing that up. I understand there's no support for this method "written in stone", but that makes 2 of us!

If you splay the walls, you must average the dimensions to equal the volume of a rectangular room in order to compare. Splayed walls have a nominal effect on modes.

YES YES! I totally comprehend that there is a complete difference between modes and reflections. And no matter what, TESTING the actual room once built is certainly the only definite answer. Unfortunately it is incredibly easy to become obsessed with finding the 'perfect ratio' that simply can never exist. Thanks for making that point Sanders.

there's more than one reason to splay the walls. Standing waves is only one of them.

And yes John, among ratio obsession I began to lose the primary idea of creating a RFZ. Thanks for the friendly quips! :D And now to the drawing board....

The midpoint! I don't know if there already was information supporting splaying from the midpoint, but I just couldn't find it.

I guess I didn't explain what I meant very well! (Either that, or I'm not understanding what you said very well... Too many late nights catching up on me, maybe...) The idea is not to splay from the midpoint, but rather just to place the edges of your imaginary rectangle on the midpoints of the splayed walls. In other words, you are basically averaging out the splaying. And if you have a section of the wall that straightens out again to parallel, or even splays back in again (which occurs in some designs), then average those out too. And once again, like everyone said: it wont be accurate, there are more reasons than one to splay, and you still need to measure. I totally get where you are coming from with obsessing about dimensions and ratios! Been there, done that... :) - Stuart -

the idea is to splay to a point that is past the listening position.

The idea is not to splay from the midpoint, but rather just to place the edges of your imaginary rectangle on the midpoints of the splayed walls. In other words, you are basically averaging out the splaying.

Hmmmm....well I must say that when I splay from the midpoint it inherently places the edges of the imaginary rectangle on the midpoints of the splayed walls. And in either case, the volume remains the same, as per what Sanders suggested. Both fer both! I think I can see how this may not always work out, but in my basic situation it sure does. :D

What John is saying, Danny, is that the splaying does not need to continue all teh way to the back wall. Once you get past the mix position, you can straighten out those walls again, so that they are parallel once more, if you want to. That helps a lot in the case of a very small space, since it allows you to have the room a bit wider overall. If you look at some of John's designs on this forum (and other designs too) you'll see what he means. Once the wall gets past the mix position, the rear section of the room can go back to being rectangular. - Stuart -

:oops: I SEE. Okay...now i get it. Man, simple and clear as things can be, I have a bad habit of making a big mathematical equation out of it and/or semi-ignoring it. That one sure flew passed my head. Thanks for ALL the replies! ! !