Room dimensions

[William Lemmerhirt]

Ok. So per @TheWarrior requesting room dimensions to calculate room modes and then what to do with them, here are mine. The proportions aren’t exact but the measurements are. Hope the picture is clear enough to see my scribbles. I can color it or highlight it if necessary. Also, here are some pics of the spaces to help fill in the gaps that I can’t provide in the sketch. They are of my house in different states of living and remodeling etc, so some aren’t flattering. Kids... lol
Also, the PC’s are in the front corners and the 3rd is in the left rear by the fireplace. So keep that in mind. Mains are 14-1/2 from LP, front subs are 16’ and the rear sub is 9’ away. I tried posting my mdat so anyone can look at any measurements I’ve made. I’ve swept with each sub individually and together, and w/mains and Direct. File is too big, trying to figure it out. Mdat is easier than uploading all the dumb screen caps. Ok, 1,2,3, go!
Need anything else? Just ask.

 

[yepimonfire]

It looks like the right side of the living room opens to the kitchen? That can totally nullify any cookie cutter calculations.

Also might want to spread the fronts out more. Should be 22-30 degrees. Hard to tell without a measurement but it looks like you’re what, 12’ from the speakers? Should be 10’-12’ apart if that’s possible.

Sent from my iPhone using Tapatalk

 

[William Lemmerhirt]

Yep, the LR does open to the kitchen foyer a hallway and two other rooms. The pic with the mains you’re referring to is old, and just to give a feel for size and airspace. The mains are actually centered at 10’ currently. 14-1/2’ from LP.

 

[TheWarrior]

Length:
22'= 51 hz
23'= 49 hz
32'= 35 hz

Width:
29.5'= 38 hz
42.5'= 26 hz

Height
9'= 125 hz
11.5'= 98 hz

Two things first: I love your room and the stealth guard kitteh!

Mitchibo PM'd me a subs only response graph - my thoughts:

First: the pics and the map don't line up in regards to your sub locations.

Discrepancy centered @ 35 hz = two modes stacking on each other. The dip at 32 hz is the length mode that your rear subs (are there two in the rear?) should be able to handle. The dip at 37 hz may reduce after first filter is applied.

Steady rise from 37 hz to 46 hz= mode between your seated front and rear boundaries. Need to reduce amplitude of this mode, need to know where your subs are located in the measurement you sent.

The graph peaks @ 46 hz and then rolls off from there.

Please provide a 15-400 hz 1/24 octave no smoothing graph WITH PHASE at all 3 seats of the couch, with the subs and speakers together. Specify LFE crossover, please. And please specify where your subs are, and what they are, is there 3 or 4?

 

[agarwalro]

Cute cat!

 

[William Lemmerhirt]

Ok. I’ll need a little time. Thanks for the nice words. That’s my killer cat Sully. Biggest baby in the world. Perhaps I wasn’t clear enough. The map locations are correct for subs with the pictures to fill in the blanks and show ceiling slope, and sunken LR etc. So, two subs in front corners, one sub in rear left corner. Distances in first post. I’ll dig my gear out shortly.

 

[William Lemmerhirt]

Cute cat!

Thanks. He’s a “rag doll”. Turns 4 very soon. Great puppy dog lol.

 

[William Lemmerhirt]

Length:
22'= 51 hz
23'= 49 hz
32'= 35 hz

Width:
29.5'= 38 hz
42.5'= 26 hz

Height
9'= 125 hz
11.5'= 98 hz

Two things first: I love your room and the stealth guard kitteh!

Mitchibo PM'd me a subs only response graph - my thoughts:

First: the pics and the map don't line up in regards to your sub locations.

Discrepancy centered @ 35 hz = two modes stacking on each other. The dip at 32 hz is the length mode that your rear subs (are there two in the rear?) should be able to handle. The dip at 37 hz may reduce after first filter is applied.

Steady rise from 37 hz to 46 hz= mode between your seated front and rear boundaries. Need to reduce amplitude of this mode, need to know where your subs are located in the measurement you sent.

The graph peaks @ 46 hz and then rolls off from there.

Please provide a 15-400 hz 1/24 octave no smoothing graph WITH PHASE at all 3 seats of the couch, with the subs and speakers together. Specify LFE crossover, please. And please specify where your subs are, and what they are, is there 3 or 4?

Wait a second. I just reread this. Is that whole second half for mitchibo? The peaks and such you’re referring to seem to line up with his graph more than mine. Just checking.

 

[ATLAudio]

Please provide a 15-400 hz 1/24 octave no smoothing graph WITH PHASE at all 3 seats of the couch, with the subs and speakers together. Specify LFE crossover, please. And please specify where your subs are, and what they are, is there 3 or 4?

Why should you want to review the phase? How do you apply this measurement?

 

[TheWarrior]

Wait a second. I just reread this. Is that whole second half for mitchibo? The peaks and such you’re referring to seem to line up with his graph more than mine. Just checking.

Lol! You're totally right, I was waiting for Mitchibo to respond and then I saw this thread and my brain just overlapped...

So both Cylinder subs are front and rear, left side? With a sealed sub front right?

Why should you want to review the phase? How do you apply this measurement?

Thanks for asking!

The audibility of phase is apparent below the transition frequency of the room due to the wavelengths involved. You either will, or will not hear a specific frequency. As our hearing is not linear, like an omnimic, a frequency response graph alone is not entirely representative of what we're hearing.

Ported subs have the advantage over sealed, using this metric, in that you are able to clearly see the in-phase sound wave followed by the out of phase port response. When time domain issues are present, the phase response will show it, usually as an interruption or flat spot in the parabolic swings of the phase response. Sealed subs are a lot more discrete, but the same inferential knowledge can be applied.

This is why I criticize waterfalls as those issues can vanish, or be exaggerated, depending on the resolution chosen by the user. You have to choose resolution in either time or frequency, you can't have both, so it becomes easy to think there is a real issue, when the actual issue has been masked by the compromise in resolution.

 

[William Lemmerhirt]

Nope. 2 cylinders up front. I ported sub left rear.
So I didn’t have time to dig out my stuff but now at lunch I’m trying to find my screen captures that have been placed who f’n knows where. Damn REW.... making new ones I guess.

 

[ATLAudio]

The audibility of phase is apparent below the transition frequency of the room due to the wavelengths involved. You either will, or will not hear a specific frequency. As our hearing is not linear, like an omnimic, a frequency response graph alone is not entirely representative of what we're hearing.

It's my understanding that humans do not respond to phase shift (without amplitude shift, but that's a given) - at most hearing a small "difference" in anechoic tests using real and contrived signals, but not able to express a preference. In normally reflective rooms there is no response.

 

[William Lemmerhirt]

ok couldn't dig out stuff but here's some more graphs for now. I'll take a new set when I get a chance...

this is each cylinder. audyssey, and minidsp off.

pc's only, and the green is all 3 subs+mains after audyssey and minidsp

stereo and direct

and final, 3 subs only

probably useless, but I after f'n around saving them, i'm posting anyway lol!!!
edit: XO at 80hz

 

[TheWarrior]

@William Lemmerhirt Thanks for these, but yeah, they don't tell much.

For when you have time, so you don't need scroll back up: "Please provide a 15-400 hz 1/24 octave no smoothing graph WITH PHASE at all 3 seats of the couch, with the subs and speakers together. "

Measuring each sub individually is helpful for level matching, but as we're talking about them as pressure sources, they have to be all run together so we can see how they are driving the modes/standwaves. I am assuming they are all level matched and delays have been set accordingly. If not, do that first.

Getting the whole couch covered makes sure we're getting the same sound to everyone.

Recording higher frequencies ensures any changes or DSP inputs are truly effective by the reduction or elimination of higher order modes - as the room dimensions influence the sound waves contacting them, the sound waves do not reflect once and dissipate.

Because there are so many large dimensions, the lower frequencies involved will show their higher order reflections with in a 400 hz window. Example:
Length:
22'= 51 hz - 2nd order 102 hz - 3rd order 153 hz - 4th order 204 hz
23'= 49 hz - 2nd order 98 hz (which is also a height mode, with luck, they'll cancel each other out)
3rd order 147 hz - 4th order 196 hz
32'= 35 hz - 2nd order 70 hz - 3rd order 105 hz - 4th order 140 hz

It's my understanding that humans do not respond to phase shift (without amplitude shift, but that's a given) - at most hearing a small "difference" in anechoic tests using real and contrived signals, but not able to express a preference. In normally reflective rooms there is no response.

Above the transition frequency of the room, you are exactly right! But you have to respect the size of the sound waves being propagated at bass frequencies. Any sound wave has positive and negative phase, right? Well if a 20 foot diameter wave is bouncing off one wall, there is a high probability that not everyone on the same couch will hear that frequency due to one or more seats being out of phase. (hence why multi subs help smooth out the response, seat to seat)

Sometimes 'out of phase' conveniently measures as a null in the frequency response. But sometimes the issue is small and that couple of milliseconds that are being induced by the room means your ears will miss that sound. So you need to be able to compare the frequency response to the acoustic phase response to be certain.

 

[ATLAudio]

@William Lemmerhirt

Above the transition frequency of the room, you are exactly right! But you have to respect the size of the sound waves being propagated at bass frequencies. Any sound wave has positive and negative phase, right? Well if a 20 foot diameter wave is bouncing off one wall, there is a high probability that not everyone on the same couch will hear that frequency due to one or more seats being out of phase. (hence why multi subs help smooth out the response, seat to seat)

Sometimes 'out of phase' conveniently measures as a null in the frequency response. But sometimes the issue is small and that couple of milliseconds that are being induced by the room means your ears will miss that sound. So you need to be able to compare the frequency response to the acoustic phase response to be certain.

If you're saying that phase matters with an associated change in amplitude, I agree, and long wavelengths in small rooms will do that. However, this correlates more or less with your reservation with waterfalls; that FR can show you most if not all you'll need. That said, I can also see how phase and waterfalls (when properly viewed), can have reliable secondary value, to "check your work" so to speak.

 

[yepimonfire]

Length:
22'= 51 hz
23'= 49 hz
32'= 35 hz

Width:
29.5'= 38 hz
42.5'= 26 hz

Height
9'= 125 hz
11.5'= 98 hz

Two things first: I love your room and the stealth guard kitteh!

Mitchibo PM'd me a subs only response graph - my thoughts:

First: the pics and the map don't line up in regards to your sub locations.

Discrepancy centered @ 35 hz = two modes stacking on each other. The dip at 32 hz is the length mode that your rear subs (are there two in the rear?) should be able to handle. The dip at 37 hz may reduce after first filter is applied.

Steady rise from 37 hz to 46 hz= mode between your seated front and rear boundaries. Need to reduce amplitude of this mode, need to know where your subs are located in the measurement you sent.

The graph peaks @ 46 hz and then rolls off from there.

Please provide a 15-400 hz 1/24 octave no smoothing graph WITH PHASE at all 3 seats of the couch, with the subs and speakers together. Specify LFE crossover, please. And please specify where your subs are, and what they are, is there 3 or 4?

I’m not sure where you are getting your numbers, almost every method used to calculate room modes gives numbers about half of yours, perhaps missing something.

Generally, the fundamental room modes occur at frequencies that are double the room dimensions in length, so for example, my 20x12 room should have modes at 28hz and ~50hz, measurements show this to be the case. Placing my sub in a corner gives 6dB bumps at 28hz and 47hz. Taking measurements at the sofa along the side wall shows an8 dB bump at 47hz as well. My 12’x11’ room has a 12dB bump 50hz.

Both REWs room sim, and RealTraps calculator give modes much lower than the ones you quoted. I get ~17hz and 25hz for 32’ and 23’.



Sent from my iPhone using Tapatalk

 

[TheWarrior]

I’m not sure where you are getting your numbers, almost every method used to calculate room modes gives numbers about half of yours, perhaps missing something.

Generally, the fundamental room modes occur at frequencies that are double the room dimensions in length, so for example, my 20x12 room should have modes at 28hz and ~50hz, measurements show this to be the case. Placing my sub in a corner gives 6dB bumps at 28hz and 47hz. Taking measurements at the sofa along the side wall shows an8 dB bump at 47hz as well. My 12’x11’ room has a 12dB bump 50hz.

Both REWs room sim, and RealTraps calculator give modes much lower than the ones you quoted. I get ~17hz and 25hz for 32’ and 23’.



Sent from my iPhone using Tapatalk

Because I've tried to simplify this by only discussing the first part of the equation, which gets you a half wavelength, which is still useful information because of the phase issues of bass frequencies. Otherwise, you are correct, the full calculation is to multiply the dimension by 2, and then divide in to speed of sound.

Where I am wrong is in calculating higher order modes with that half wavelength.