Is it a right way to find room modes with REW?

[FargateOne]

Hi,
At the end of the chap 8 of Dr Floyd E Toole "Sound reproduction" 3rd ed., fig 8.26 shows room modes identified in the FR graph.

I know there is mode calculators and i know Rew has one. But in case of irregular room dimensions or, walls with door or corridor or different length like in my room, those help more or less. And I would like to really learn how to find them on a FR graph.
Or I would like to be sure that it is impossible with a mic and REW.

I have tried the same method to find the modes resonnances in my room.
For each dimension, I put the mic on the wall at the center of it and took measurements one foot away at a time.

Dimensions are 471cm length, 350cm width, 230 cm heigth; the front wall is the long one. There is a corridor at the left side of the front wall and a door at the right side of the back wall.

I attache the mdat file. Maybe with screen captures someone could help me to identy the modes? Or is my method not good for it?

For example, in height dimension (230cm) I started with the mic up to the ceiling at middle of it and took one measure each 30cm (1 foot) down. The graph shows what seems to me a mode at 38,5 Hz which is 890cm wavelenght half of it is 445cm. The heigth is 230cm so it can't be a 0,0,1 axial mode, what mode is it?


Attached Thumbnails
[IMG alt="Click image for larger version
Name: Width modes.jpg Views: 32 Size: 412.1 KB ID: 2659690"]https://www.avsforum.com/forum/attachment.php?attachmentid=2659690&stc=1&thumb=1&d=1577470421[/IMG] [IMG alt="Click image for larger version
Name: Height modes.jpg Views: 19 Size: 379.0 KB ID: 2659692"]https://www.avsforum.com/forum/attachment.php?attachmentid=2659692&stc=1&thumb=1&d=1577470421[/IMG] [IMG alt="Click image for larger version
Name: Lenght modes.jpg Views: 18 Size: 457.7 KB ID: 2659694"]https://www.avsforum.com/forum/attachment.php?attachmentid=2659694&stc=1&thumb=1&d=1577470421[/IMG]

 

[Marc Lombardi]

It's interesting what you're trying to do but I'm not sure it will get you what you want. I believe the intention is to see how the room modes affect the sound at the listening position. So if you just make a measurement of each speaker and left/right combined at the listening position you will see the combined effect of all modes. See the attached image ....

This is combined left/right in my room with no treatment. You can see the combined modal resonances at 40, 70, 85, 200 where there are large peaks and nulls.

 

[DanDan]

Place your source speaker(s) on the floor in corner(s). This will fire up all modes. Place your mic in an opposing corner. The resulting Waterfall graph should contain all the modes. Signal Generator, Sine, tick frequency follows cursor. Slide to cursor over a mode of interest. While it is playing walk the room noting the hots spots and nulls. Mark the floor with masking tape or make a mode map. This will prove extremely useful warning you of no go zones in terms of speaker and listening positions.

 

[FargateOne]

It's interesting what you're trying to do but I'm not sure it will get you what you want. I believe the intention is to see how the room modes affect the sound at the listening position. So if you just make a measurement of each speaker and left/right combined at the listening position you will see the combined effect of all modes. See the attached image ....

This is combined left/right in my room with no treatment. You can see the combined modal resonances at 40, 70, 85, 200 where there are large peaks and nulls.

Hi,
so I did (see mdat) . I wanted to see if I could find the 3 axials modes of first order (1,0,0 ; 0,1,0; 0,0,1 if I am correct) and doing this being sur to EQ the most importants peaks. But I can't do the math. For instance I see 3 peaks 43Hz, 72, 177 half wavelength of which give approx 4m, 2,35m and 1 m. Those are not related to the room dimensions.(467cm L x 350cm H x 230cm W) . It is not a matter of national security though!

What bothers me the most is the peak at 1090Hz (31cm wavelength) that seems to come from nowhere to me!

Place your source speaker(s) on the floor in corner(s). This will fire up all modes. Place your mic in an opposing corner. The resulting Waterfall graph should contain all the modes. Signal Generator, Sine, tick frequency follows cursor. Slide to cursor over a mode of interest. While it is playing walk the room noting the hots spots and nulls. Mark the floor with masking tape or make a mode map. This will prove extremely useful warning you of no go zones in terms of speaker and listening positions.

The only way I could do that it is to move the sub ( a down firing one which is good for the exercice). Afraid that I have only only one corner available ( a piece of furniture at one corner, a door at an other one, a corridor at an other one, ouch) and to put the mic at an opposite corner will be more difficult. I will report later.

 

[DanDan]

Well, LF is not too specific about location. Near a corner.... Also remember all floor ceiling junctions are corners. Similarly just do your best to use double or triple boundary locations to stimulate and receive all modes. This aspect of the exercise is just to be complete, to catch all, including odd ones. But the kicker is the exact sine wave. You could just slowly sweep a sine and stop when you hear a major resonance. With speakers where they are. As you move up in frequency the pattern of peaks and nulls and the number of modes increases exponentially I guess, so up at 1K....? Sheesh.... different techniques...... Look for a corresponding reflection in the ETC. If you suspect a particular source, say the ceiling, move the mic towards it. See if the reflection moves in time. This would confirm it. But as I say exponential complexity so a deceptively crude test might work better. i.e. Block using a MiniTrap or rigid batt or even something more solid, a sheet of ply.
Perhaps worth keeping in mind that Room Modes may not be the biggest player at the table by a long shot. Particularly in complex shaped room.

RealTraps - Room Modes

You might want to try sweeping from 10Hz or lower, but that extreme LF may be just noise or another room even. The 40 and 60 don't look very threatening. Really, as can happen, the short term peaks and devastating nulls would appear to be the bigger culprits here.