RFZ Calculation Assistance?

[tubeman]

Hi everyone, I was hoping to find some advice on how to properly do some calculations in measuring and creating a Reflection Free Zone.

I don't know how to take the "average absorption coefficients" into account when doing the calculations. Using the Everest and Howard/Angus books for some formulas, I've managed to figure out that the earliest reflection (which are from the sidewalls) will be 2.615 meters, coming out to 7.60196 ms of delay. The delay of the direct sound to my monitoring position (1.0668 M) is 3.101 ms.

From my monitoring position, the difference between the direct sound and the reflected sound is 7.788 dB with a 4.50079 ms delay difference. But the dB measurement doesn't take into account any absorption coefficients. I have FuzzMeasure to help...

[Ethan Winer]

You might be over-thinking this. What you want is highly absorbent material that's thick enough to get down to at least the low mids. Then put that at the reflection points on the side walls and ceiling. By "highly absorbent" I mean 1.0 at mid and high frequencies. And by "reflection points" I mean an area a least a few feet square.

--Ethan

[avare]

Ethan's advice is sound (pun intended). If you want to do the engineering/calculation route, which I love, The loss from the absorbent material is:

L=10log(1-abs)

where
L= dB loss
abs=absorption coefficient

If you take a moment to think about the equation, it is taking the dB value of the reflected sound.

If you are doing calculations, do not use the average absorption coefficient. Look at the absorption at lower frequencies. The ITU spec for control rooms specifies reflections within 15 ms be at least 10 dB down for frequencies above 500 Hz. Personally I use 20 ms.

Good luck!

Andre

[tubeman]

thanks for the tips - i realize that i'm being awfully cerebral about all this, but i have to provide hard numbers for the project. i see that 2" 703 is significantly more absorbent at the low mids, so looks like i'll pick some of that stuff up! just to clarify, that equation calculates dB loss at the absorption point?

i also found that to incorporate the intensity with respect the monitoring position, that would mean also subtracting 20 log (Path Length) from the peak intensity.

also, if 1 is the highest coefficient, that means that 10 dB is the most that someone can decrease in terms of reflection absorption at that point. I imagine that, considering that at least 15 dB is recommended for an RFZ, some additional calculating needs to be done with all the early reflection points to get a total dB reduction number?

and Ethan, i saw some diagrams on your site that test the frequency response of a couch, using an omni about 4" away. what type signal do you recommend sending out of the monitoring system run those sort of tests? A fast-ish sinesweep at 100 dB or something like that?

[gullfo]

4. FORMING THE ACOUSTIC EQUATIONS

interesting information... :-)

depending on the mounting the absorption for a given panel will change, as well as the boundary wall reflection/absorption effects.... plus incidence angle and other factors so you would want to measure to be sure as to what the values really are that you need to attenuate.

or just cheat like i do

Porous Absorber Calculator V1.52

[Ethan Winer]

Originally Posted by tubeman

and Ethan, i saw some diagrams on your site that test the frequency response of a couch, using an omni about 4" away. what type signal do you recommend sending out of the monitoring system run those sort of tests? A fast-ish sinesweep at 100 dB or something like that?

You mean near the bottom of this page, yes?

RealTraps - Creating a Reflection-Free Zone

I did those tests with the ETF software and its swept sine wave.

More here:

RealTraps - Optimizing Acoustic Treatment using ETF

--Ethan

[tubeman]

great, all this advice and directions is really helpful! btw, how long should the sine sweep be?

[Ethan Winer]

Originally Posted by tubeman

how long should the sine sweep be?

Whatever the software outputs will work. ETF offers fast and slow sweeps, with the slow sweep having more resolution. Since the difference is only 1 second versus 4 seconds I always use the slow sweep.

--Ethan

[tubeman]

hey guys, i'm working towards analyzing my measurements and wanted to see how acousticians read their results.

the program i'm using runs only mono tests. if i window a particular reflection and see that a ceiling absorber reduces a reflection at 1 Khz by 9 db in the left channel, and the same in the right, should i take away from these measurements that i'm actually getting

1 - a 12 dB reduction in sound power/intensity in that reflection (9dB + 9dB = 12 dB)
2 - a 15 dB reduction in sound pressure (9 + 9 = 15)
3 - stick with the 9 dB mono result?
4 - all of the above?

[Ethan Winer]

I don't think like that - I usually measure both speakers playing the same mono test signal. Measuring both in mono is especially important for low frequencies, though it's a good idea to also test each speaker separately in case something is seriously out of whack on one side only.

--Ethan

[avare]

Originally Posted by tubeman

hey guys, i'm working towards analyzing my measurements and wanted to see how acousticians read their results.

the program i'm using runs only mono tests. if i window a particular reflection and see that a ceiling absorber reduces a reflection at 1 Khz by 9 db in the left channel, and the same in the right, should i take away from these measurements that i'm actually getting

1 - a 12 dB reduction in sound power/intensity in that reflection (9dB + 9dB = 12 dB)
2 - a 15 dB reduction in sound pressure (9 + 9 = 15)
3 - stick with the 9 dB mono result?
4 - all of the above?

None of the above. The value you see is how much the reduction is from the absorbent material. Is that 9 dB compared to no absorption or 9 dB below the initial impulse level? If it is the former, then you have met the spec, because there is also a loss from the longer path.

Are you using 2" 703? If you are and want to lower the levels even more try doubling up the thickness of the 703.

You may be wondering how important is 1 dB. If I was building the space for myself, I would not worry about it and be happy. If I was building the space for someone else, or for my commercial studio, I would worry about it. I would not want to say "almost meets international specifications."

Andre

[tubeman]

haha, yea, "almost meets international specifications" isn't the way for a job!

Basically, I compared the untreated measurement to the measurement with the 703 on the ceiling. I windowed a reflection that I saw was being attenuated by the 703 in both measurements, and used the "difference sum" option. The graph showed a 9 dB difference at that reflection. hopefully i did that correctly.

all measurements were taken at the monitoring position with nothing changed, so i think the answer is compared to no absorption. so if shortest distance from the monitor to the ceiling and ceiling to my ear is 3.2 meters, then travel time means about loss of 20 log 3.2 = 10 dB already and 19 dB total...i think...

[tubeman]

argh, i didn't think about running the mono signal through both monitors simultaneously. i just did one at a time.

one thing i've noticed...for me, the earliest reflections are coming from the desk, keyboard, etc. all within 3 or 4 ms. i dropped 4 inches of 703 on top of the desk, and the reduction was enormous.

makes me think about all the studios out there that have nearfields on top of a console, especially the one on the cover of Everest's How to Build a Small Budget Studio book.

[avare]

Congratulations! You have met international standards!

You mentioned board reflections. That is one of the reasons why a PROPER mastering is good.

Andre

[Ethan Winer]

Originally Posted by tubeman

argh, i didn't think about running the mono signal through both monitors simultaneously.

Most music, especially pop music, has bass instruments panned to the center. So testing both speakers in mono simulates how bass will actually be reproduced in the room.

--Ethan

[tubeman]

fuzzmeasure has a similar graph (log-squared) to ETF's energy time curve that everst talks about. This is the impulse response of my room before any treatment - does it look like the first wall reflection, at 41.5ms (the direct is at 35ms) is already about 20 dB below the direct sound's amplitude?

[Ethan Winer]

Originally Posted by tubeman

does it look like the first wall reflection, at 41.5ms (the direct is at 35ms) is already about 20 dB below the direct sound's amplitude?

Yes, but I just emailed Chris Liscio (FM author) to comment too since it's his program and he's the expert.

--Ethan

[liscio]

Originally Posted by tubeman

fuzzmeasure has a similar graph (log-squared) to ETF's energy time curve that everst talks about. This is the impulse response of my room before any treatment - does it look like the first wall reflection, at 41.5ms (the direct is at 35ms) is already about 20 dB below the direct sound's amplitude?

To comment on the Log Squared Impulse - I'm pretty sure it's the same as the ETF function you describe.

I also have an Energy Decay Curve (a.k.a. Schroeder Plot) display option, which would have been confused with an Energy Time Curve if I listed both in the list of graphs. I could have gone with Schroeder and EDC as graph titles, but I wanted to take the route of 'proper names' rather than 'common names' for those.

Schroeder called his plot the Energy Decay Curve in his paper on calculating Reverberation Time, so I ran with that. Can't tell you where I first saw 'Log Squared Impulse Response' off the top of my head, though - you'll have to take my word for it. ;)

Hope this helps!

[tubeman]

cool - when i calculated the reduction of dB from path length and drywall abs, the difference b/t the calculation and the FM reading is almost 13 dB.

when i compare the effect of each absorber's IR relative to the initial IR (with unchanged measurement settings), the dB reduction at the reflection points are much closer to predicted.

i took 4 measurements at each stage using a 4 second sweep topping out at 100 on the level meter, with 1.5s and 3s beginning and end silence. the direct peak and noise floor difference is something like 100 dB so i think that's enough signal to noise, unless something is wrong in the signal path.

[liscio]

It looks like you do have decent SNR in that measurement. Perhaps we're wrong about when the first reflection really is. Maybe the little peaks at 35.5ms or 36.1ms are the actual first reflection.

What do your calculations tell you about when the first reflection should arrive from the treated surface?

I only chose the one with the most distinct peak, but that could easily be caused by a different reflection point.

[tubeman]

hey chris,

well, the distance from monitor to mic/me is 45" (3.75 feet, 1.143 meters)
the rough shortest path length from monitor to desk to mic/me is 55" (4.583 feet, 1.397 meters)

i figured that, using the 1128.6 ft/s number

3.3227 ms is the delay from monitor to me
3.913 ms is the delay from monitor to desk to me
.5903 ms is the difference in delay

using the 20 log (path length) formula for intensity loss:

1.16 dB loss from monitor to me
2.903 dB loss from from monitor to desk to me

so i think there should be about a 1.7 dB reduction in the Log Squared graph, but it looks more like -12 dB or so if using the 35.5 ms peak. i don't think the desk's absorption coeffcient can produce a -11 dB reduction unless it's average coefficient is .75 or so.

[avare]

At what angle is the sound impinging on the desk? How wide is the desk?

Andre

[tubeman]

hey, i'd say maybe 105 degrees, but that's an estimate. the monitors and my ears are both 48" off the ground and the desk surface is 30"

the desk is 31" wide and 24" deep.