Just wondering how you determine if you have peaks and nulls and how to deal with them

One of the reasons I bought the Legatias is to see how a different set of drivers fares in my vehicle. The result is they have the same nulls as the Focals so that tells me its the vehicle (Yes folks I spent $1500 just to experiment with accoustics)

My graphs start with a peak at 25Hz, a null at 80Hz, peak at 160Hz and another null at 315 - 400Hz. Then rise to 800Hz then its quite flat

So anyway I've tried the drivers in different positions and angles but it does almost nothing. When I took readings with the opposite door open, one peak dropped and another increased. It leaves me wondering if they are peaks and nulls or just peaks or just nulls (Btw The graphs may not be accurate but its all I've got to go by and besides I can hear it as well)

The difference from peaks to nulls is about ten decibels so I don't want to use EQ to flatten it out untill I get the difference to be a lot less

Anyone got suggetions on how to tell whats what, and what to do about it?

thanks


Doug

how do i work out whether i have peaks or nulls? with a calculator and pen and paper.

what do i do about them? well changing angles is only going to change the drivers response. therefore, if u have a peak/null with the driver response, changing the angle will get rid of them, but it will also change how the driver sounds. i tend to measure the response of my own drivers before trying to address the room issues.

then with room related issues, it depends on the frequency. if its bass/midbass, they are a bit difficult to deal with acoustically as we dont really have any room to fiddle.

null at 80Hz isn't the driver rolling off is it?

160Hz is probably door related and the higher freq's could be related to say a centre console issue.

start with pen and paper, work out which node ur sitting in, and start fiddling

This possibly one of the better questions I have acknowledged on this forum for a long while.

What head deck/ processor/ drivers are going to give you anomalies like this? (10 dB difference)

This why using an RTA for a flat(?) response doesn't work.
(lets be honest here most people use an "A" or "C" weighted response which is NOT flat)

Why doesn't a flat response produce the ideal sound?
It does, but using an equaliser to cure it, produces one of the poorest listening reproductions available.

Why can't you EQ one speaker to sound like another of a different brand?

Point
The solution is the problem. The answer is not the use of an EQ.

As you possibly realise, the response from your drivers hardly resembles the measure response at your listening position.
Therefore it is a reasonable conclusion that your space is the the primary actor for distortion.

What to do?
Sound waves propagate in spheres, so unless you have a baffle step response, angling your drivers will achieve little relief (Off axis response aside)
Use a wave guide to reduce baffle step response, increase nearfield listening and GUIDE the wave away from time medium reflections. (0.1-0.15 Milli-seconds)
For late reflections, (0.2-0.3 Milli-seconds) my normal recommendation would be arrays which can control dispersion into these time zones, however this is something you would of allowed for in the original installation.
http://www.mobileelectronics.com.au/forums...showtopic=87723

You are now limited to absorption and diffusion, both of which at these frequencies require considerable width to be effective. In most cases this will be unsuitable for in car use.

The only plausible resort left is helm-holtz resonators, which can be useful for problem frequencies.

unless of course he owns a DEQX that is.

interesting that u suggest the use of an array. i just cant see the effectiveness of an array within the cabin sapce IF he were to be using stereo. it would appear to be to well confined and would add a degree of complexity to the issue which may or may not be practical.

i think he has to start from the beginning tho and define which particular node he is sitting in. 25hz could be box related, so that may be easy to deal with. 80Hz could be as simple as moving the sub around, or moving his drivers seat further forward/back. the 160hz anomaly is going to have the most severe impact on bass and is going to be the most difficult to deal with. could it be infact derived from a poorly constructed baffle/box?

Arrays though, decrease directivity in one direction and increase directivity in the other. So a peak or a null, caused by an environmental factor, can possibly be eliminated by reducing directivity and preventing the acoustic wave from entering that 'space', or increasing the directivity which will allow you to change firing angle. This in turn may eliminate the problem.

Both the 25Hz and 80Hz points wont be fixed by this though, as the space is being treated as a pressure zone, and not an environment affected by acoustic propagation (evidenced by the cabin gain). Would possibly help the higher frequencies though.

Having seen Doug's van though, I have a feeling the 160Hz point is a quarter wave null due to speaker placement and firing angle.

your going to have to fill in the gaps for me on this one

how is the firing angle related other than driver response in this particular case?

Abmolech thanks for reminding me of that thread that Proclass put up (I've got some reading to do) and thanks to all for your words. I still have a lot of experimenting to do and a lot of what has been said goes straight over my head but I think I'll get there. I can't seem to upload pics any more so I can't show you the graphs, would make things a bit easier

To help answer a few questions, the drivers I have are Infinite Baffle
8" subs 4" mids and 1" tweets

Crossovers are
Subs TH - 160
Mids 160 - 6.3k
Tweet 6.3k - TH

(btw the crossover points are what Scott Buwalda recommended and he made them so I take his word for it)

So the peak at 160 is a crossover point, and it is the one that increases with the opposite door open. The peak at 25Hz is the one that decreases with the door open and as mentioned I also believe it is intallation - related. I may need to vent the doors more (but I don't want to drill holes until I'm sure about it)
The big bad null at 400Hz is the real killer I reckon because 440 = A and a lot of music is played in that key. To my ears there are holes in the sound in particular tracks

What do you do with a pen and paper? calculate wave lengths?
I've done a bit of that but I don't know what to do with the information. I worked out the 400Hz wavelength is about the distance from door to door and another is the distance from driver to corner of door and roof.

As mentioned, sound propogates in a sphere so I can't see how changing the angle of drivers would help much (and my recent efforts prove this) So I think the baffle idea must be the way to go but how do you do it? Are there tutorials on the topic that you could point me towards? Or if you disagree with the baffle idea what are the other methods mentioned, they've gone over my head

One thing I have tried is to cover the cargo barrier with egg-crate accoustic foam. All it did is reduce 800Hz - up by about three decibels

Well I'll see a few people tomorrow and maybe put some ideas together. Might get the G5 out of retirement and join in who knows? But one thing I got to do is

DON'T FORGET TO VOTE!!!

cheers


Doug

lol, i've forgotten about that several times.

thats exactly what u do with pen and paper. work out the wavelength. this is useful because it will then tell you what is causing the null. remember, strictly speaking, only waves of the same wavelength can cause destructive interference. from the information u have obtaned, u can then work out what could possibly be causing the second wavelength and then treat it. the treating is the hard part, including the 440hz anaomaly due to the energy and wavelength involved.

i haven't read the article below

http://www.fesb.hr/~mateljan/arta/download...user-manual.pdf

but it could help in ur paritcular instance. its a manual for the ARTA, but it goes into alot of detail. good background info that may help u draw some conclusions from the results u have obtained.

It is very refreshing to acknowledge peoples maturity in thinking about problems such as this.

The 80 Hz will be from an acoustic impedance change.
Air has less mass than your driver cone (well doh), and the greater the mass difference the more of an an impedance mismatch, which results an efficiency drop off. The impedance mismatch causes an efficiency of less than 5%, IE 95 % is turned to heat. Any change in mass of either the cone or the AIR will therefore produce a significant change.

In your vehicle the air is "loaded" (wave guide) because the wave is constrained to expand in less than a sphere. For every wave length 6 dB is lost in acoustic energy into heat, IF is expands in a sphere. IE For every diameter factor increase, the volume of a sphere increases by 4. If we were to produce a flat baffle (simplest wave guide) large enough so the wave is constrained to form only half a sphere, it would only use 3 dB for every expansion factor. (2 times the volume)

In your vehicle you gain 12 dB over an ordinary room. This is because your vehicle is a wave guide, in this case it is constrained to expand in under 4 times.
<a href="http://www.trueaudio.com/st_spcs1.htm" target="_blank">http://www.trueaudio.com/st_spcs1.htm</a>
Often this can cause a "sudden" unloading (same as the resonant frequency of vented enclosure) if the steradians rating drops.
If there is simply a null at this frequency, and 100 Hz etc are fine, we may have a standing wave from the rear doors to the front or your mid bass are causing cancellation with the sub. Try the sub woofer on its own and note the RTA. If we still have a definite notch, it is a standing wave.

400 (440) Hz
Let us ASSUME the problem is a door to door standing wave (likely). This will could be caused by the glass on either door.
Wind one the window down, if it is still present, it is the drivers themselves.

Why can array's and/or wave guides help?
Most of us are aware of off axis response, and the main causation is the driver diameter relationship to the frequency. (example an 8" cone starts to beam at 1725 Hz)

What if we were to use two 8" drivers on the same baffle in close proximity?
We would still have off axis response narrowing at 1725 Hz, however in ONE direction. Across the TWO drivers we would have a narrowing of the off axis response Much earlier, VIZ 1725/2= 860 Hz. If we had a standing wave at 900 Hz (floor to roof), we could use this off axis response, to narrow the dispersion away from the roof/floor to a much later time arrival.
You may now grasp the predilection for array's extending from the floor to the roof in home audio. IE the floor to roof standing waves are a non issue until much later.
We could have even more "fun" by using a wave guide to narrow this further, or using time delay on ONE driver to change the onset of cancellation.

http://www.sengpielaudio.com/calculator-roommodes.htm

deadening, absorbtion surfaces and such, and careful listening.

if there really is a null at 400Hz or so, i would look at things like console, or width of the vehicle. (mine was 600Hz, re-aligning the woofer angles fixed it)

lower than that, you're getting to front/rear standing waves and panel absorbtion. sound deadening, time alignment and path convolution can help here.

EQ is really a last resort, and should be used for more than +/- 3dB adjustments.

I thought this would be good place to explain absorbers, reflectors and the combination of the two, "diffusers".

If we look of the coefficient of absorption of a material we make get a factor from 0 to 1.
What does this tell us?
1 is = completely absorbed and 0 is completely reflected (blocker). This is likened to an open window having a factor of 1 (fully absorbed) and a lead wall fully reflected.(0)
If we use rock-wool at 100 mm thick it would suggest a rating of 1 at 100 Hz, versus rock-wool of 25 mm thick giving us a factor of 0.5.
If we therefore use 100 mm thick rock-wool we will therefore fully absorb a 100 Hz wave, and none will reflect.
If we use this in a vehicle doors will we get standing wave at 100 Hz across our vehicle?
Quite probably
The coefficient of absorption only tells us if it will reflect or absorb, what it does not tell us how much acoustic energy is converted to heat (attenuation)
Example if we had 100 thick rock-wool with a speaker on the other side with 100 Hz sine wave how loud would it be?
If it was 100 dB, we would hear none of it? what if it was 200 dB?
Clearly we need to know the coefficient of attenuation.

So why do you always talk about how those diffuser pads are ineffectual for lower range frequencies, and what is this "magical" 1/4 wave point?
Velocity
(Direction of travel in a certain time)
Imagine a wall, if we throw a tennis ball against it, what is the velocity when it is deforming at the wall?
Zero
It has no direction, IE at that instant it is stationary.

When a sound wave hits a wall, the velocity is zero (same as radiating driver cone exciting an air molecule) at the quarter wave the velocity is the highest, (in fact every 1/4 there after)
For an absorber to turn acoustic energy into heat, it is required to induce the wave to diffract (bend around corners). You can try this yourself, by bending a wire in the same place several times and observing how hot it at the bend. If the diffraction is too sharp (too tight a corner) it will reflect, and therefore this determines the maximum density the substance can have before it shifts from an absorber, (least dense) to a diffuser (middle dense) to a reflector (most dense). This is your coefficient of absorption.

Therefore the most efficient place to turn acoustic energy into heat is on the quarter wave, it is also the point at which a wave of the same frequency will sum or cancel, because the velocity (interchange) is at its peak.
To get a measurable change we need to be therefore a minimum of 1/10 of the wave length away from the point of reflection, to get an audible change, consider 1/4 wave length mandatory.

Sorry but diffusers need to be 1/4 deep to have any real use.

So we are suggesting an absorber 215 mm thick for a 400 Hz wave?

absorbtion also need not be done perpendicular to the direction from the source to the listener.

by having absorbent materials along the surface of the 'wave guide' (any interior surface of the car) then some attenuation will be realised, and more of the direct signal will be heard, and less reflected sound.