View Single Post
Old 04-22-2012, 10:08 AM   #37
planetnine
Human being with feelings
 
planetnine's Avatar
 
Join Date: Oct 2007
Location: Lincoln, UK
Posts: 7,924
Default

-Use REW to store named responses of your room, and then as you move/change things around you can compare the graphs directly.

-Read the setup instructions on REW and use your SPL meter to measure at 85dB.



Remember that the nodes (cancellation point) of any room dimension's f1 axial mode is in the middle of that dimension (antinodes or max SPL are at the dimension boundary);

f1




the nodes (cancellation points) of any room dimension's f2 axial mode is at a quarter of the dimension in from each boundary (antinodes or max SPL are at the dimension boundary and at the dimension's centre);

f2




the nodes (cancellation point) of any room dimension's f3 axial mode is in the middle of that dimension and 1/6th of the dimension in from each boundary (antinodes or max SPL are at the dimension boundary and at 1/3 (2/6) of the dimension in from each boundary).

f3





the nodes (cancellation point) of any room dimension's f4 axial mode is at 1/8th and 3/8ths of the dimension in from each boundary (antinodes or max SPL are at the dimension boundary, in the middle of that dimension and at 1/4 (2/8) of the dimension in from each boundary).

f4





I've linked to this from the German Sengpiel Audio website -a very good resource for calcs and diagrams for audio physics.



These Axial Mode nodes and antinodes are actually planes, parallel to the dimension boundaries. Think of these when moving your head around your room at the problem frequencies. You should get REW to produce a sinewave and drag the frequency cursor to the peaks and the troughs of your response curves and then listen to see if the dropout (node) or hotspot (antinode) is in a plane (sheet area) -it will be parallel to the walls it is bouncing between and will need broadband trapping on those surfaces. This is a glimpse into the fun techniques you will use to get rid of the extremes of those graphs. If you get within 12dB you will have a good room to mix in; within 6dB it will be a very good room to mix in


Once you've got your head around axial modes, try tangential and oblique...



With tangiential, the sound waves use two of the three pairs of boundary surfaces in a room to reflect from. This can be in a quadralateral pattern as shown in the diagram (equal mode numbers for each) or can actually be a first mode in one dimension and a second or third mode in another. nodes and antinodes tend to be at boundary intersections and in lines at the intersection of planes parallel to the dimension boundaries being used by the standing wave..

Oblique modes are where all three pairs of boundary surfaces are used in the standing wave. The sound path is a three-dimensional twisted figure of eight around the room for equal modes across dimensions, or even more manic 3D lissajous figures for unequal mode numbers across dimensions. The good news is that only the very lowest mode numbers and simplest paths are of actual significance in real rooms, or when they tend to combine with other modes reinforcing certain frequency ranges. Nodes and antinodes are in corners and at points formed by constructive and destructive interference created by the diferent modes on all six boundaries.

The three pairs of numbers in mode calculators are for the modes in the three dimensions: 1 0 0, 0 1 0 and 0 0 1 are the first axial modes in each dimension (x, y and z for length, width and height). 2 0 0 would mean the second axial mode for the length, 0 3 0 would be the third axial mode for width. 1 1 0 would be a tangential mode across the walls (length and width) involving first mode or f1 frequencies for the tangential wave path. 1 2 0 would be first mode of length and second mode of width (a sound path going twice across the width and once along the length) the sound path length is a function of the squares of the multiples of each dimension (based on multiple being the mode number).
Still following?

Oblique modes use all three dimensions and so use all three numbers of the matrix. 1 1 1 would be an oblique path using each surface once, and its sound path length is a function of the squares of all three room dimensions. Use a dimension more than once and the mode number increases to that multiple and the factors go up in calculating the path length. You need computerised 3D plotting applications to show the nodes and antinodes for these (or lots of pretty paints and some hallucinogens).

Most good mode calculators use all combinations of all three dimension mode numbers up to about 10, calculate the modes, order them by frequency and disregard any with frequencies over about three or four hundred Hertz. Other useful information is mode stregth and a graphical display of mode clustering, where modes gang up on you at close frequencies. See Ethan Winer's DOS MODECALC.EXE for an example, which uses different colours for axial, tangential and oblique modes. There is also something similar on Bob Goulds online calculator.


Oher useful information is a trace called a Bonello Distribution graph. This shows how modal density in your room is distributed across the concerning bottom 300Hz frequency range. It is often mapped against an ideal distribution curve and the idea is to get your line below the other, an even distribution of modes with no bumps which would mean clumps of modes within a frequency band giving modal combinations that might be hard to treat for.

Bonello Distribution



This leads us on to ideal ratios of dimensions for a rectangular room. Certain ratios of length, width and height dimensions give desirable spacings of standing wave modes. Your better mode calaculator will calculate the ratios of your space and point out how non-ideal your room dimension ratios are

Remember through all this that these calculations are for an ideal reflective room of uniform construction with no doors and windows, and the introduction of these and such frivolities as carpets, furniture, desks and audio equipment will mask and smear the real values, hiding the modes and their effects. Add in some broadband traps and other acoustic treatment and it can be hard to compare the theoretical and the empirical.

If you ever get to build a studio from scratch, make sure there are no parallel walls and the roof is at an angle!!


I am having some fun with you here, but I hope you do actually follow some of this, it will open your eyes (ears?) and make some of it mean something while you are plodding about with your traps and measuring gear. I hope it contributes and gets you great results in the end



>
__________________
Nathan, Lincoln, UK. | Item Marker Tool. (happily retired) | Source Time Position Tool. | CD Track Marker Tool. | Timer Recording Tool. | dB marks on MCP faders FR.

Last edited by planetnine; 04-22-2012 at 10:15 AM.
planetnine is offline   Reply With Quote