[BMWturbo]

Not really sure where to put this one. I've been contemplating this for a little while now, and after some success with a new location I'll probably look at making another set of pods over the next few months, when some other projects are complete.

I know the best way to 'test' something is to 'try' it, but before I spend time building some pods I wanted to get some thoughts on whether it will be worthwhile or a waste of time.

What I'm looking to do is make some dash pods, on-axis to listener, that will sit tightly up near the A-pillar, so essentially an A-pillar location, but mounted from the stock location in the dash.

On a previous Sail panel midrange install I noticed that the reflections from the side glass wreaked havoc on the stage definition/focus and I couldn't get around this, so coupled with a loss of depth, these were removed.

I'm thinking that with these new pods, I'd build an absorbant 'wall' or shroud on the outside of each midrange to try and absorb the off-axis sound rather then let it travel and reflect off the side windows. I'm hoping this might have a positive effect on the stage focus, as I know the depth will be okay in the proposed location.

Has anyone had eny experience with this before? I can't seem to find much reference to trying to absorb off-axis waves by narrowing the drivers dispersion.

Thanks in advance.

 

[thatvan]

Man I was thinking the same thing only yesterday!

 

[br85]

If they're hugging the a-pillar that much, try glassing a waveguide out to the front of the pillar (where it meets the door, and steer it away from the window a bit. Hard to describe, but basically you should use as many reflections to your advantage as you can, and then minimize the others after the fact. In that location you're making something of a waveguide anyway, just with a poor throat, which can wreak havoc on the response. If you end up with air gaps in the throat, shove some NHMC (plasticine) in the gaps to seal it.

Still running the 3" tangbands with no tweets Ben?

 

[t3sn4f2]

If I'm not mistaken, there is no practical way to absorb midrange the way you want to, high probably, but midrange would take something very significant.

How about aiming more between the seats making your off axis your on axis (after some EQ) and letting the more pronounced on axis get lost down the center between the 2 seats. That would effectively help take away the reflections from the window thanks to the drop in output at the extreme off axis then.

 

[BMWturbo]

BR 85 :- Still using the little TB's without tweeters. They don't sound too bad on-axis and have reasonable extension, this is the reason I want to try and run these MR's only, and not have the need for a tweeter.

I'm almost certain the reason my sail pods didn't work was the reflections, but I couldn't form any sort of guide with these as they'd impact on the side mirror vision, so I was stuck.

I'm thinking it might be worthwhile to limit as much reflection as possible and see how that sounds.

t3sn4f2 :- My MR's are playing from 315hz 24/dB up , with some form of absorbant section, I'll have to research to see what would be the lower limit of absorbrtion of different materials. I was thinknig some foam, or possibly a dynxorb type material.

I didn't want to have to compensate for off-axis response by using EQ and I think that if I was to try and aim the Drivers side MR at the centre of the car I'd be introducing more reflections from the instrument cluster binacle.

 

[br85]

Haha, I forgot about that, yeah mate you need to consider on-axis to be an advantage right up until the point where the reflections are hitting your ears at -6db or so (you'll have to make up your own mind about that). If you've still got your test equipment, try and measure your response at different axis' (try 30,20,15,10,0 degrees for example) and see which yields the flattest/smoothest response. Can you do csd plots at all to measure "time smear"?

 

[t3sn4f2]

BR 85 :- Still using the little TB's without tweeters. They don't sound too bad on-axis and have reasonable extension, this is the reason I want to try and run these MR's only, and not have the need for a tweeter.

I'm almost certain the reason my sail pods didn't work was the reflections, but I couldn't form any sort of guide with these as they'd impact on the side mirror vision, so I was stuck.

I'm thinking it might be worthwhile to limit as much reflection as possible and see how that sounds.

t3sn4f2 :- My MR's are playing from 315hz 24/dB up , with some form of absorbant section, I'll have to research to see what would be the lower limit of absorbrtion of different materials. I was thinknig some foam, or possibly a dynxorb type material.

I didn't want to have to compensate for off-axis response by using EQ and I think that if I was to try and aim the Drivers side MR at the centre of the car I'd be introducing more reflections from the instrument cluster binacle.[/QUOTE]

True.

 

[BMWturbo]

Haha, I forgot about that, yeah mate you need to consider on-axis to be an advantage right up until the point where the reflections are hitting your ears at -6db or so (you'll have to make up your own mind about that). If you've still got your test equipment, try and measure your response at different axis' (try 30,20,15,10,0 degrees for example) and see which yields the flattest/smoothest response. Can you do csd plots at all to measure "time smear"?

Arguably a 3-6dB roll-off from 2-3k up won't be too much of concern to me, I really should drop a spare set of drivers in some pods and do some off-axis response testing. I'm not certain how Andy's 'Target Curve' comes in to play when you're reducing the reflectins though.

I've had a play with CSD plots a while back, but am not really versed with the gating etc and seemed to get some funny results.

 

[ErinH]

waveguide?

 

[Fast1one]

waveguide?

Thats what I was thinking...

 

[thehatedguy]

You know how big a waveguide would have to be to provide pattern control down to 300 hertz? BIG, real BIG...atleast in free space.

 

[ErinH]

Nope. No idea. I've never even looked into waveguides. That's why I had the "?" after it. I leave the waveguide stuff for people who know about it.

 

[thehatedguy]

Considering Dr. Geddes's Summa speakers use a 15" round waveguide to provide pattern control down to 900 hertz, you can imagine how large you must get to control down to 300 hertz.

But that might not be that important since you would probably be most concerned with controlling your 1-4k the most.

 

[squeak9798]

You know how big a waveguide would have to be to provide pattern control down to 300 hertz? BIG, real BIG...atleast in free space.

In an automotive environment.....how big? Seriously.

 

[thehatedguy]

Too big for a car at those frequencies.

 

[BMWturbo]

So is there another possibility to be able to create a controlled dispersion pattern?

I'm already using a cone driver, so no gains to be had there.

 

[npdang]

Have you considered an electronic correction solution, such as an impulse response convolver?

 

[BMWturbo]

Does that need a flux capacitor also?

*Edit* I did a quick search on this and from the little information I could find, it seems to be a 'hall' type DSP algorithm? I'm unsure how that could in effect modify the dispersion of a conical driver?

 

[t3sn4f2]

Does that need a flux capacitor also?

*Edit* I did a quick search on this and from the little information I could find, it seems to be a 'hall' type DSP algorithm? I'm unsure how that could in effect modify the dispersion of a conical driver?

Apline Imprint - Auddessy Multi EQ XT

 

[SSSnake]

I'm with Jason on the size of a 300Hz waveguide BUT why try down the 300Hz. If you are running a low freq driver (midbass) on the front stage that can reach up to the 1khz area cleanly then dispresion control (above 1K) becomes more feasible (especially if you use the car as part of the waveguide). My goal is to get as much dispersion control as I can and then match the radiation pattern where I switch over to the low freq driver. Even these goals are pretty tough. THEN once you get as much physical correction as possible the convolver becomes a very viable option.