Explain the ratios?

[Coytee]

I copied this from Artoo who posted it back in 2003 or so. (hope that's acceptable)

"There is no such thing as an absolute optimum room size for the

Klipschorn. Obviously, the larger the room, the fewer acoustical

problems you will have, especially at lower frequencies.

More

important than room size, are the rooms proportions. As mentioned

above, there is a mathematical ratio commonly called the Golden Mean.

Also known as the Golden Section or Fibonacci Sequence. As applied to

room acoustics, it can predict the Eigenton distribution of a room (the

rooms modes where resonant frequencies will occur, and/or overlap).

The smaller the room, the more modal problems occur, primarily in the

lower frequencies. Also the smaller the room gets, the more (higher)

frequencies are affected.

The Golden Mean ratios for proper room

proportions is: 1:1.26:1.59 (1 to 1.3 to 1.6) (height to width to

length). There are a wide range of ratios that are suitable. And

certainly good sounding rooms can be found outside of these

proportions. Obviously the worst case condition would be cubical where

all 3 room dimensions are the same as that of a particular wavelength

of sound, causing a substantial room resonance at that frequency.

A

range from 1 to 1.26 to 1.59, to 1 to 1.6 to 2.5 can be considered

good. The old studio 116 at Klipsch in Hope, AK. was 10x16x25, which

falls outside of the ideal ratio but was still considered good.

Just

as important (if not more) are the rooms acoustics. It will do you

little good to get the rooms proportions ideal, and then ignore

proper BROADBAND diffusion & absorption, and reverberation time

relative to frequency.

See arttos klipschorn room in the

architectural topic area for a more in depth description of

construction methods & acoustic techniques."

Ok, can someone put this into a bit of

english? If I draw a room that has the front wall 10 feet wide,

then is the depth 10x1.26 to 1.59 multiplied or is it 10x1.6 to 2.5 and

how do you factor in the ceiling height?

I'd like to make a spreadsheet so I can do some room figuring and need

to understand the ratios and exactly what is multiplied by what to get

what.

??

[Champagne taste beer budget]

Hopefully I've gained a small understanding of this as I've perused this forum the last few years, I'll try and get you headed the right direction.

The ratios are H x W x L of the room, height, width and length. If you're working with an 8 ft ceiling, that would be your 1, so your 1.26 would be 10.08, (8 x 1.26) for the width and 1.59 would be 12.72 feet. As I recall from Artto's threads, you can use the "Half Room" principle, since the above measurements would make for a small room, and not many people have a 16' ceiling!. That would bring you up to 20 wide and 25 long still maintaining the 8' ceiling. The "1" is always the ceiling height, the other dimensions are based on that. Somewhere on the forum is a link to a "ModeCalc" program, a little thing that will do some quick math and give you the ratios of any rooms dimension, will also show you the modes, or resonant frequencies, that will develop based on the dimensions. I believe you want as many of them as possible, spaced as evenly as possible. Stay away from the 19' length, as that is the wavelength of a 60hz signal, and with your system running on AC with 60 hz, anything that gets through to the speakers will be amplified based on the rooms dimensions.

I'll see if I can find the program and link to it for you, I don't know how to get the program from my puter

onto the board so it can be downloaded.

Good luck!!

[Coytee]

* If you're working with an 8 ft ceiling, that would be your 1*

THAT was the secret ingredient I was missing, thanks!!

[formica]

There are actually quite a few of good ratios, like mentioned in that

quote. Obviously the height is usually one of the tougher

dimensions... as 20' front wall would require either a 12'6" or 16'

ceiling to work according to that particular ratio.

Here are a couple of Ratios collected from reliable sources by Eric Desart...

1- L. W. Sepmeyer: 1965 - 1:1.14:1.39

2- L. W. Sepmeyer: 1965 - 1:1.28:1.54

3- L. W. Sepmeyer: 1965 - 1:1.6:2.33

4- M. M. Louden: 1971: Best ratio as per Louden - 1:1.4:1.9

5- M. M. Louden: 1971: 2nd best ratio - 1:1.3:1.9

6- M. M. Louden: 1971: 3rd best ratio - 1:1.5:2.1

7- M. M. Louden: 1971: 4th best ratio - 1:1.5:2.2

8- M. M. Louden: 1971: 5th best ratio - 1:1.2:1.5

9- M. M. Louden: 1971: 6th best ratio - 1:1.4:2.1

10- C. P. Boner: 1942 - 1: 2^(1/3):4^(1/3) - This equals Id. O - 1:1.26:1.587

Note: this is a mathematical very interesting ratio to study. This

ratio is exclusively calculated on the axial modes guaranteeing the

mathematical maximum ratio between x, y and z axis (22.62%).

This however causes some less desirable coincidences with 3 Tangential an one Oblique mode.

11- J. E. Volkmann: 1942 (later discussed by H. Bolt) - 2:3:5

12- Rounded C. P. Boner ratio - 1:1.26:1.59

13- Derived from C. P. Boner (stylized ratio) - 1:1.25:1.6

14- Golden rule ratio: 1968 - 1.236:2:3.236

15- IEC 268-13: Recommendation for listening room: 1987 - 2.8:4.2:6.7

16- IEC 60268-13: Recommendation for listening room:1998 - 2.7:5.3:7

17- Dolby's optimum ratios for Film & Music Room - 0.67:1:1.55

18- Origin unknown: equals ± 1:1.6:2.56 - 0.625:1:1.6

19- Origin unknown: Resembles Id. 18 - 1:1.618:2.588

20- Origin unknown: meant for small room - 1:1.5:1.6

21- Origin unknown: normal - often used ratio - 1:1.6:2.5

22- Origin unknown: meant for long rooms. - 1:1.25:3.2

Obviously the steeper the ratio... the easier it is to integrate to a

standard ceiling... for example #16 ratio of 2.7:5.3:7 could give a 10'

x 19'8" x 26' room.

ROb

PS: for the ratios that contain no 1's... you could convert them by

deviding all the numbers by the smallest digit... EG: 2.7:5.3:7 could

be devided by 2.7 which gives 1:1.963:2.593

[artto]

beer budget basically said it right. However, "technically speaking", it really doesn't matter which dimension is which, the combination of dimensions will all affect the room modes the same way. What differs, is what dimensions are available for speaker placement (and listening locations), the longest one being preferable for the widest "stereo" listening coverage area. With Khorns I prefer the increased breath, depth & imaging qualities that wide-stage stereo provides. However, if your're room is not deep enough, it may be preferable to use the shorter wall so you're not backed up against the back wall where a lot of excessive bass can build up. I generally don't use the "half-room principal" on all three dimensions, just the ceiling height, simply because for the most part, most of us have to deal with an 8' ceiling. Using 16' for the ceiling height in room mode calculations can often give you a slightly different perspective on how the room will perform, albeit, not the same thing as the real dimension, but somewhere in between. In some situations this can also be shown to work against you instead of in your favor. There are also other methods: http://www.acoustics.salford.ac.uk/acoustics_world/room_sizer/room_sizing2.htm

Here is the one I use for use in an Excel spreadsheet, look for modes1.xls http://www.linkwitzlab.com/rooms.htm

[JBryan]

Hey Coytee - me again. I promise I'll stay 'On Topic' with this thread.

I tried several times to calculate the room nodes using the Golden Mean but my room is not very accommodating so it proved to be futile. The room is 'L'-shaped measuring 30' and 25' from the apex and the ends of the 'L' are 15' and 18' respectively. To complicate the calculation, the ceiling is not the same height - instead it has several planes that measure between 7'3" to just under 8' as the acoustic tiles cover the various-sized HVAC venting. There's also a stairway that occupies a portion of the room. Trying to figure in all the odd measurements became tedious and I finally abandoned the the attempt out of frustration.

Would it be easier to use a CD or such that produces a sweep of the frequencies (20Hz-20kHz or so). If it did it slow enough, one could simply note the resonant nodes using a SPL meter and address them accordingly. I imagine that your wouldn't even have to calibrate the CD as the resonances would be reproduced whether you have the actual frequency or not.

3 things - first, is this a viable approach for those of us who have to work with the rooms we have? If so, are these sweep CDs available and where would I find one? Finally, once the nodes have been identified, what can be done? Is it just a matter of damping the reflective points and corners? Where do I find simple instructions dealing with room resonances and treatments?

Thanks, Bryan

[DrWho]

Coytee, if you're going to be building a room from scratch then I highly highly recommend that you build a room with non-parallel surfaces. That means splayed side walls, a gradually rising cieling and a slanted rear wall. Not only does it drastically improve the acoustics, but it also adds to the whole theatre feel as well. The basic rule of thumb is one inch for every foot - making a room that is 20 feet long, 40 inches wider in the rear than the front (20 inches for each wall). Or an 8 foot cieling is 8 inches further back at the top.

Such a configuration also provides room for an inwall component rack as well as plenty of volume for IB subwoofers too [] Or you can simply take advantage of the extra sound transmission reduction.

To get an idea of the room modes for such a room, you would just use the average width and heights of the room.

Btw, the goal is to maximize the modal density of the room. With the splayed wall example, you can model all the various dimensions of the room to get an idea of how the density is increased (and the overall magnitude of each mode is reduced too, but that won't show up in the modelling programs).

[Coytee]

*, if you're going to be building a room from scratch then I highly highly recommend that you build a room with non-parallel surfaces*

You saw where I was going.

Wife happens to like higher ceilings (current house has 9') so it wouldn't be a stretch to use 9' as a starting point and if other factors allow ($$) investigate 10 or perhaps 12 max. Once that hurdle is hammered out in principle, then the splaying of the walls & ceiling would be my next questions.

I might put it this way (if Artoo is still reading and as I understand, is an architect?), if you were going to design a room from scratch for a 7.1 Khorn (or mix of heritage) system, using 9' or up to 12' max ceilings, how would you design YOUR dream room from scratch? I could easily see this room being connected to the house by a hallway and otherwise being disconnected on 4 sides. Since the wife & I are at discussion stage only, it's easy to ask for the moon so that when I have to 'settle' on what's acceptable to HER, I might still be within specs. [6]

[artto]

My "dream room" is to be constructed as part of an underground home built into the slope of hill. The room's ceiling slope would generally conform to the slope of the ground surface above, with the main speakers at the end of the room with the lower part of the ceiling, so the room expands, or opens up the further back you are. As Mike suggested, it's a good idea to have non-parallel surfaces. It doesn't really take too much too have a substantial impact on the diffusion of the sound. Does this negate or eliminate the necessity for additional diffusive surfaces? Absolutely not! (just look at any fine auditorium) The room would also be segregated from the main part of the house as Coytee mentioned. Using non-parallel wall surfaces would require (IMO) very large corners for the Khorns to still fuction "optimally". I'm still considering using what amounts to very large free standing false corners, much larger than what you would typically find in most any "rectangular" domestic room. Ceiling height would probably vary from at least 10' to a maximum of 16-18'. I figure the Khorns will probably be spaced about 30' - 35' apart with the "real" room walls even wider. I've played around with some 5, 6 or 7 sided trapezoidal shapes but the corner geometry always seems to be a problem so I still might end up with some kind "room-in-a-room" configuration.

[DrWho]

Or you can just use jubilees all around and not have to worry about room corners at all []

[WMcD]

If I may offer a comment based on my unsophisticated undertanding. My thought is that it makes more sense when some basic issues are discussed, in an orderly way.

1) There are going to be resonances or standing wave patterns based on the distance from one wall to the other opposite wall. We;ve got three pair to work with, considering the ceiling and floor as the third pair.

2) These resonances occur where the walls are spaced one-half or 0.5 wavelength. Lets just consider a pair of walls as a tube with two sealled ends. This just so we have a single effect to work with.

3) The pressure peaks occur at the walls. The sound wave is normally made up of particle velocity peaks alternating with pressure peaks. Each drives the other. The energy is alternating between those two means of storing energy in free space, but at a solid wall, particle velocity is zero (it hits the wall) and pressure is high.

4) Of course the wall now becomes a big speaker. There is a pressure peak there and launches particles with velocity. This is, essentially why echos occur. the exchange of energy is altered because of the barrier.

5) There is a similar condition for at a frequency which is 1.0 wavelengths. There is a pressure peak in the middle of the room but also at the walls . . . again.

6) We can continue with that progression, again just on a wall pair, every half wavelength up, at 1.5, 2.0, 2.5 wavelengths.

7) From the above, you can imagine we've got a tube with resonances up the spectrum with successive peaks. A comb filter. (It is called this because tthe peaks look like teeth on a comb.) Question, what to do to fill in the peaks?

8) If we had only one other pair of walls, we'd try to make them resonate at a frequency someplace halfway between the original peaks. I'd have to look at the math but this might be a wall spacing which is square root of 2 or 1.412 of the firsts pair.

9) Of course we really have two more pair. So we want the added peaks to be 1/3 and 2/3 up from the original peaks. The walls must be 1 times cube root of 2 and 2 times cube root of 2 multiples to get the effect

10) The effect of 9) is that we can "start" with wall spacing being the ratios of cube root of 2.

11) I'll belabor the above. We've got three wall pairs. The room is like a 1/3rd octave equalizer with every third slider ganged together. But we can adjust the center frequency of the ganged sets. Or maybe a pipe organ with just three notes with lots of overtones (harmonics).

12) The progressive spacing of the wall pairs at cube root of 2 ensure that each ganged set of peaks do set at 1/3 octave points. If, IF,the resonances are fairly broad, we have uniform gain from resonances up the spectrum with no holes. I think that this is why a classic way of designing a rectangular room is with the cube root of 2 progression of wall spacing.

12.5) My math may be wrong, but the basic concept is there, and important.

13) It is a good starting point. But there are other effects. The reflections can occur between the corner pairs we draw on the floor plan. Call one the NE, SW corner pair. The other the NW and SE pair. Pressure antinode, (high pressure) occur because air pressure is trapped in the tall wedge, running from floor to ceiling. So we have two pair of those in the rectangular room.

14) If we turn the room on the side, conceptually, and draw a floor plan, there are another set of wedges. And we can turn it tipped forward, there are yet two more..

15) In 12) we just looked at flat walls as pressure sources. In 13 and 14 we looked at wedge shaped boundries. The last, discussed below, are potentially more effective. They are the trihedral corners.

16) The trihedral corners are the intersection of two walls and a floor, or walls with ceilings. They are starting to work to trap pressure, like an old time hearing aid, and also radiate it, like the extension of a K-Horn. Even worse, or better depending upon your point of view, they are looking at a similar structure across the room.

17) The stuff in 16) describes that we have eight (8) trihedral corners in any room. Each are functioning as hearing aids when the sound goes in, and loudspeaker horns when it comes out! To some extent this is also going on in the wedge shaped "floor plan" boundries.

17.5) Note that puting a KHron in a corner or a tube trap in a corner are effort to exploit the effects. The pressure from the KHorn, or any speaker has a load to work against against trapped pressure. The tube trap has pressure to absorb so it does not have to be as big..

18) We have not discussed real rooms, or the exact spacing of the wedges or trihedral corners. One thing is that the wedge effect and trihedral corners also fill in the gaps of our 1/3 octave room, perhaps in unpredictable ways. In a real room there is a door or window which remove some of the components.

19) There is no real reason to wring hands over not having the perfect wall spacing because the it is difficult to predict the actual effect of the many compexities of the wedges and corners. This is why rooms which depart from classic measurements can still be good. Irregularities are good as compared to regularities. You might blunder into an issue, none the less.

20) So what about polycylinders? They reflect at least higher frequency sound in a broad pattern. Is that good or bad? Could they not do more harm than good? My thought is that the regularity of rectangular rooms reflections are much more likely to cause problems than randomness.

Naturally, I invite comment.

Best,

Gil

3) The .

[Champagne taste beer budget]

WOWSA!!

I may not know the complexity of all the variables to know if your take

on this is correct or not, but you do lay down a dang good explination

of why it should be. Everything made sense, based on a few assumptions.

I'll assume those assumptions are correct, but look forward to the

other replies sure to come.

Well done, Gil.

[WMcD]

Thank you for the kind comments. Much of my approach is because my high school physics teacher gave a very good explanation of wave mechanics. He was one of those guys who change your appreciation of the world.

I also generally dislike science and math teachers. The only good ones are geniues who use their intellect to make complicated things, simple. Most fall short and make simple things complicated.

But going back to the original subject. I do doubt that a room having the magic proportions will be good, or one which does not is bad. The problem is that our rooms have acoustic features of carpet, furnature, doors, and windows.

The ratios work on the assumption that walls are perfect reflectors. But even that leads to bad side effects. Sometimes.

An intersting experiment is to stand in a room and clap you hands, sharply, once, and listen.. In some you will hear a slap echo. There is a "boyoyoining" sound like a sound effect from a sci fi movie. Try it. It is a no cost experiment.

I'm a bit mystified at the inconsistencey of the effect. One space at the office has a lot of acoustic objects (stacks of boxes, cubicals, carpet, and suspended ceiling). It does have the echo.

OTOH, I tried this in a diniing room a bit larger with tile floor, plaster otherwise, some wood furnature, a smaller rug, big windows on either end. No echo effect.

In each case there are large expanses of walls.

Of course, you're experiences may vary.

Gil

[TommyC]

Note to self: Go back and read Gil's post when it is Not 1:15 am!!

I recently had a chance to visit a purpose built listening room that was pretty much exactly as Dr. Who described above. The only difference was that the back wall wasn't a straight wall, it was 3 sided, the two ends angled away from the front wall to the door in the middle, which was parallel to the front wall. the Ceiling got higher from front to back. There was also minimal treatment on the walls and ceiling. The effect was breathtaking! I got the opportunity to take my Heresies over and try them in this room and they sounded amazing! I will probably never be able to afford to build something like that, but if I ever can, that is exactly how I will do it!