AVS Forum banner

Status
Not open for further replies.
1 - 1 of 1 Posts

·
Registered
Joined
·
864 Posts
Discussion Starter #1
This is the October 19th - 2001 version of the ReadMe file for the Mediaroom spreadsheet.


The Mediaroom spreadsheet (MRSS from now on) was made for a group called Mediaroom I created at yahoogroups.com in April 2000. This group died of inaction so after learning about the several WWW sites that discuss home theater design, I started to regularly post messages offering this tool to anyone that wants it.


These notes and the spreadsheet itself are work in progress. I hope they are useful and that someone (perhaps me!) builds in the future a mediarooom based on this and the work of other enthusiasts that may participate in the development of the design.


If you are new to acoustics, I suggest you read some books and papers on the subject to gain the necessary background information. The Everest book (F. Alton Everest - Master Handbook of Acoustics) and the SGHT issue indicated below are good starting points. Explaining all the concepts behind MRSS is beyond the scope of these brief notes. Explaining Microsoft Excel is also way beyond the scope of these notes.


I believe you need at least Excel 97 for Windows, as it was initially developed using it. I'm currently using Excel 2000. It also works on Excel 2001 for Macintosh (with VB macros enabled).


MRSS is work in progress and has nothing to do with my professional activities and official fields of expertise so it is the chance that it can be improved or even corrected. The concepts and criteria applied could also be non-ideal for some cases. Consider it pure speculation and use it at your own risk, though I would build a mediaroom for me based on this work RIGHT NOW! if I had the opportunity.


The energy behind this work comes from a love for music and HiFi equipment, and also sheer escapism, as it relaxes me and provides opportunities for fun research, and also for meeting and getting into interesting discussions with smart and knowledgeable people.


Suggestions, improvements, and corrections to MRSS and these notes are welcome. I urge you to join the HT/music forums on the internet, and to use these groups to post comments so they have a chance of being discussed by more people.


As MRSS is based on the public work of others, little here is of my own creation. Only the Excel template design and lots of trials to reach a dimension set is my work. My gratitude to the writers and researchers whose work is used as the basis of the Mediaroom spreadsheet.


Precaution: Keep a virgin copy of MRSS somewhere and change the content only of the cells in grey unless you really know what you are doing.


Please let me know if you find this work useful, specially if you make practical use of it. Also let me know if these notes could be improved.



Sheet 1 - "Dimensions"


The core concept of the Mediaroom spreadsheet comes from one of Bonello's criteria, which states that we want to use room dimensions that would create main resonant nodes that are 5% apart or more (See reference 5 below). The discovery of such dimensions using MRSS is a trial and error process and I found several sets that are adequate. The one I like the most is entered as the default dimension set in MRSS (6.26 x 5.67 x 2.98 meters). There could be more good dimension sets and it would be great if others do their own trials, discover and publish additional ones.


What you do is introduce the candidate room dimensions (in meters, sorry) in the grey cells (G6, G7 and G8) in Sheet 1 and press Control-F or click the button labeled "Calc". MRSS will calculate the room nodes and show in red the node frequencies that are closer than 5% to the following one. The node frequencies are color coded and shown in the first grid, so it is easy to determine which two dimensions can be modified to widen the gap between two nodes.


Also, when room dimensions meet criteria established in a recent paper by Walker (See reference 6 below), cell B28 will indicate "TRUE". That paper/article also provides the formula for the ideal RT60 for the room which is shown in cell B30.


According to experts (See reference 1 below), above a certain frequency (room dependent) the nodes become numerous and their effect on the "room sound" become less pronounced. This is the frequency indicated as F2 and it is not a hard limit. According to reference 1, all the range between F2 and F3 is a mixed range where nodes are less important but still relevant. Also, from around 200Hz upwards, you can use conventional diffusors to break the resonant nodes so they again become less important. For that reason, it seems to be desirable that the upper limit of the critical range where resonances are adequately apart reaches upwards of 200Hz.


The critical range is F1 to F2 which in the default case is 26 to 204Hz (this last number is factored by two, as explained below). F0 is the lowest frequency that fits in the corner-to-corner length of the room. F1 is the first resonance across the length of the mediaroom.


As it proved achievable for the default dimension set, I used F2 x 2 as the design F2 in order to go as high as possible keeping the nodes 5% or more apart. In the default dimension set, I am able to reach 204 Hz complying with the 5% criterion and 233.7 Hz having a single frequency-pair gap fall to 3.5%.


To make it easier for people testing dimension ratios published by writers in the field, it is possible to have MRSS calculate the length and width of the room using only the room's height and the ratios. For that, check the option labeled "Use ratios?". MRSS will ignore the length and width entered in the grey cells and will use instead the ones derived from the height. Length/height ratio and width/height ratio must be entered in cells F6 and F7 respectively.



Sheet 2 - "Diffusors"


While diffusing low-frequency resonances is a valuable improvement to the Mediaroom design, you may simply use conventional absorbtion and transmission to tame them.


You may want to break the resonant nodes at low frequencies (F1 to F2), where conventional diffusors might not be useful at all. The way to do this is to use low-frequency diffusors (LFD) that, because of their size, have to be built as the room walls (See references 2 and 4 below). Resonant nodes create frequency response peaks and also RT60 peaks. Diffusing the nodes allows for improvement in both aspects.


In order to create a squarish free-space area, the side LFDs must be less deep that the back one. As depth and length are related in a LFD, this allows for having two shallow LFDs in each side wall. An squarish room is not a goal by itself but for the default dimension set shown, the alternative is to make a narrow, long room which seems undesirable and might not meet Walker's criteria.


In theory you just need one LFD for any pair of parallel surfaces but the side walls have to be symmetrical so three is the right number.


The fourth (front) wall is not a LFD as it will have the door(s), A/C-heating vents and other elements such as a RPTV or a front-projection screen. It might have also sound treatment on it or close to it, including perhaps conventional diffusors, adequate to the particular main speakers used. In-wall/soffit mounting of the main speakers is a design I'm particularly interested in.


At this point, a LFD for height-related resonances is not part of the Mediaroom design. Building LFDs in the ceiling or floor is likely to be very expensive. It is likely too that the LFDs in the walls would diffuse and absorb some of the energy of the floor-ceiling resonances.


The complex-numbers math to do the full design of non-integer-based LFDs is beyond me and I'm looking for someone to help with it. But the main features of the diffusors can be established without knowing their full design (See references 2 and 4 below). For the default dimension set and the initial LFD design, the combined LFDs can diffuse frequencies from about 74Hz up to about 324Hz. As this range overlaps most of the main resonant nodes (range F1-F2) for the default case, most of the main nodes would be diffused and the hot spots in the room would be reduced. Lovely, isn't it?


Some nodes below the LFD-diffused range and also the first height-related node can be tamed using free-standing absorbers placed in strategic spots in the mediaroom. They can double as tables or similar. It seems possible to reach a solution in which the only untreated nodes would be the lowest ones, where there is little content (26.5 and 29,2 Hz in the default case). Bass sounds of the lowest frequencies and also some above them might partially escape the room through the walls, floor, and specially the ceiling, diminishing their effect on the room's sound.


The average-depth line of the LFDs would be located at the nominal dimensions of the room so, although the wells would be visually perceived mostly as open space, the truly free or truly open dimensions of the room will be smaller that the design dimensions. Additionally, the floor plan would be larger too, a result of the depth of the diffusors. With the preliminary LFD dimensions shown in Sheet 2, the truly open space is about 5 x 4.5 meters though again, it would be perceived as larger.



Sheet 3 - "Reverb"


This was an afterthought (prompted by a query in AVSForum, and debugged with help from AVSForum members) as my initial concept was to establish the room's RT60 doing direct measurements and adjusting it by trial and error, starting with the low frequencies. But it convenient to have an idea of what is required to reach the ideal RT60s and that's the purpose of Sheet 3.


It uses a second spreadsheet with lots of data for acoustic surfaces. I don't remember the WWW site where I downloaded it from but would love to credit it here and in the MRSS itself if someone provides the information. The identification number for the wall coverings --as established in the second spreadsheet-- is entered in the appropiate cells in the MRSS ("Reverb" tab, column D, grey cells).


In the wells of the LFDs and in other places in the room, there would be sound absorbing devices in order to bring RT60 for low frequencies within the desired range. I believe diffusing would not be enough to tame low-frequency energy unless the transmission characteristics of the walls and ceiling are carefully chosen and implemented. That may not be achievable in practice so absorbers are probably required. These LF absorbers would affect higher frequencies too, bringing RT60s for these frequencies to very low numbers. RT60s would need to be raised by using conventional diffusors in front of the LFDs.


Sheet 1 shows the ideal RT60 for the room (per Walker's paper) and Sheet 3 additionally shows the target number (midpoint) and the permisible limits for standard frequencies, also per Walker. Additional criteria related to the maximum RT60 differences between neighboring frequency ranges are indicated in the paper. Ideally, a balance could be reached where RT60s for all frequencies are within the allowable limits. Given the ranges established by Walker, this goal seems achievable.


At this point I like a diffusing solution for the ceiling where it is covered by patches of absorbing material placed according to a calculated distribution. Covered area is one-fourth of the total ceiling area. This comes from a picture in a paper (See reference 3 below). The area shown in the paper is square and that's another reason for a squarish room.


For those wanting a mostly dead room --as they will be recreating the ambiance electronically-- it is likely that once the bass absorbers are in place inside the LFDs and absorbers are placed so all first reflections are killed, the room would be dead enough for that purpose. It may perhaps require inexpensive --but perhaps a lot of-- additional absorbing material in front of the LFDs and specially in the ceiling. The most exciting approach I've seen for this is Ambiophonics ( http://www.ambiophonics.org ).


The floor would be carpeted. Anything fancier, such as a LFD built in the floor, seems to be unnecessary for RT60 purposes and/or likely to be too expensive to build.



Main sources of information used for developing MRSS are:


1) F. Alton Everest - Master handbook of acoustics, Third edition. Fourth edition was recently released.


2) Non integer based diffusers. James A. S. Angus. AES Preprint 5064


3) Optimized planar and curved diffsorbors. Cox and D'Antonio. This comes from the RPG WWW site, http://www.rpginc.com/news/library.htm. Look for the link "Optimized Diffsorbers"


4) Control room design incorporating RFZ, LFD and RPG diffusors. Peter D'Antonio. Db magazine.September-October 1986


5) Stereophile guide to home theater Vol. 5 No. 3. Article by Russ Herschelmann.


6) "Listening conditions for the assessment of sound programme material",Walker. EBU Tech 3276-E available from the EBU WWW site at http://www.ebu.ch/tech_32xx.html. A supplement indicating the requirements for multichannel systems is available from the same URL.


Note: MRSS is work in progress. I might be adding additional information and corrections to this ReadMe file in the future. The latest MRSS package is at:

http://www.aretecnica.com/fam/oc.htm


If this site cannot be reached, you can request a copy from me at the e-mail address below. Please let me know if you find this work useful, specially if you made practical use of it.


Pablo Roufogalis L.
[email protected]
http://www.aretecnica.com/fampg.htm
 
1 - 1 of 1 Posts
Status
Not open for further replies.
Top