Rational Acoustics



Kip Conner
December 13th, 2008, 01:35 PM
Here's a complete new one for me.

I've been working with a new 3 way box and sub that I have tons of shows with over the last few years. The 3 way box was designed with LR 24 slopes in all pass bands according to the manufacturer. There was no other para-metric information available to me other than the XO points.

The mixer wanted this system as flat as possible because he's a taper. He records all of his mixes and honestly, he spends more time with ear buds in and put more effort into show documentation than for those in the room. But that's just an opinion.

To make the system flat from 20 to 200 I had to make an interesting choice in XO points.

All slopes are L-R 24:
20->60 / 80->250 / 250->1.25 / 1.25->Out

I would expect these type of results if there was poor roll off on each side of the XO. For instance, if there was hump down out of view of the window and you didn't bother to scroll down the magnitude screen. But the roll offs looked good. Bands 1&2 are comprised of 2 18's and 2 15's drivers so there's no shortage of low frequency reproduction.

PaulTucci
December 16th, 2008, 05:53 PM
Here's a complete new one for me.

I've been working with a new 3 way box and sub that I have tons of shows with over the last few years. The 3 way box was designed with LR 24 slopes in all pass bands according to the manufacturer. There was no other para-metric information available to me other than the XO points.

The mixer wanted this system as flat as possible because he's a taper. He records all of his mixes and honestly, he spends more time with ear buds in and put more effort into show documentation than for those in the room. But that's just an opinion.

To make the system flat from 20 to 200 I had to make an interesting choice in XO points.

All slopes are L-R 24:
20->60 / 80->250 / 250->1.25 / 1.25->Out

I would expect these type of results if there was poor roll off on each side of the XO. For instance, if there was hump down out of view of the window and you didn't bother to scroll down the magnitude screen. But the roll offs looked good. Bands 1&2 are comprised of 2 18's and 2 15's drivers so there's no shortage of low frequency reproduction.


Kip,
You couldn't flatten the system to his liking with mere EQ?
Are you saying that opening up the crossover between the subs and lows yielded better results than the overlapping xover in that frequency range of 60-80 Hz? Perhaps there is destructive interference causing cancelation.

Reflection-free measurements on your part can indicate if it's a design issue or an environmental one.

PT

Kip Conner
December 18th, 2008, 08:52 AM
These were ground plane measurements done outside (about 10 ft. from source) to try to minimize de-constructive interference. I saved various ref files with different scenarios as we ran through the various options for XO points and orders. The only way I would know how to describe it would be that harmonics filled in the gap. Everything about it screams measurement error or possible processor error.

Admittedly I think that he me go way too far. It falls into the "just because you can, doesn't mean you should".

PaulTucci
December 19th, 2008, 05:38 PM
In the original measurement with an 80 Hz crossover point a dip at xover existed.
Did you try polarity flipping the sub band?
I would try that one first because polarity reversal of the low driver would change the low to mid acoustic crossover area.

In your many other shows did you sense an issue or not?

PT

Kip Conner
December 22nd, 2008, 06:16 AM
The phase alignment was perfect. I'm tempted to go back and go to an 8th order and see what happens then.

I'm wondering is there's a way to post some screen shots of the various XO settings that I used. I saved all of the ref files.

PaulTucci
December 22nd, 2008, 11:56 AM
In the advanced reply you will find an insert image tool. Save your screen shot, have it hosted and you can refer us to it as below.
In these threads pictures are invaluable.

PT

Harry Brill Jr.
December 22nd, 2008, 08:43 PM
You can even upload an image. Just re-size it first please. There is no reason to have it more than 100k or so. What I do is click Print Screen on the keyboard then I open a new bitmap in Paint then Ctrl V to paste the image. Then I re-size it and save it as a .jpg.

Kip Conner
December 24th, 2008, 09:34 AM
Here is an image of the transfer functions. I did four scenarios:

White Trace- what the client decided he wanted and what he kept. for this situation there's a void between 60 and 80 Hz

Green Trace- the XO is set for 60 Hz LR24 for both sub and low mid driver

Purple- the XO is set for 80 Hz for both sub and low mid driver

Blue- the XO is set for 90 Hz for both sub and low mid driver

PaulTucci
December 24th, 2008, 05:08 PM
Kip,

I do not see a void or dip at 60-80 Hz. What I do see is a sub / low system that (if only the crossover frequency is being changed and not the levels) combines better, more constructively around the crossover frequency the higher the crossover frequency is. Obviouslythe sub is providing more energy because as you drop the crossover frequency, the area above the xover level is dropping.
I'm more confused than before abou this "void" you mention. Somebody help!

PT

Kip Conner
December 24th, 2008, 10:31 PM
The void is in the electrical domain, within the XO. The subs top end stop at 60Hz and the Low Mid do not begin until 80Hz. The two drivers cross at 70Hz but at a very low amplitude point compared to a -6db point in the LR24 filter set.

The choice to dump all of those frequencies and create a void gave me the white trace- which was the flattest response.

PaulTucci
December 25th, 2008, 12:45 PM
Kip,

Alright, I now understand the "void." I was thinking it lived in the acoustic domain.
Would it not be easier to just turn down the sub level and leave the mfr's crossover point intact? By the looks of your screen shots, I'd say the sub is louder. By lowering the crossover point you are merely getting rid of some energy and lowering the acoustic crossover point. This may not be the correct thing to do for your low drivers. An important point to consider here is even if the phase traces of the sub and low drivers matched, that condition is overridden if the amplitudes of the two don't match. Equal amplitude and in phase for constructive addition is a two parameter condition.

To achieve the engineer's desired response curve, I think you just EQed with the crossover. That is different than matching the phase/amplitude trace at the sub crossover and tailoring the sub's frequency response to his liking. That last part is an EQ issue.
PT

Harry Brill Jr.
December 26th, 2008, 10:50 AM
Perhaps the sub is more sensitive in the upper part of it's range. A post crossover EQ is one possible solution to flatten it out. Then you could align the crossover.
ANother possibility, which we cannot tell without seeing the individual traces, is that the combination of the low and sub gives increased output through a bit of overlap. This could be addressed with an EQ BEFORE the crossover. This has the added benefit of giving you additional headroom and directivity (if you need it).