Forum Home » Sound Reinforcement » LAB: The Classic Live Audio Board » Nexo GeoS 12 and sub
|
| Re: Nexo GeoS 12 and sub [message #320100 is a reply to message #319968 ] |
Tue, 06 May 2008 14:04   |
HarryBrillJr.
Messages: 592
Registered: April 2007 |
Has No Life |
|
|
| Rasmus Rosenberg wrote on Tue, 06 May 2008 07:12 |
Hey Harry,
What's program is this? Also can you make a model of 6 high in one stack compared to two stacks of 3 high. Im testing it out this weekend.
RAsmus
|
I can show only 2 dimension. I can however model them both ways. DO you want the 2 stacks close together or separated by some distance? Is the distance center to center or side wall to side wall?
|
|
| |
| Re: Nexo GeoS 12 and sub [message #320295 is a reply to message #320248 ] |
Tue, 06 May 2008 22:26 |
HarryBrillJr.
Messages: 592
Registered: April 2007 |
Has No Life |
|
|
|
You should measure it. With a spectrograph you can even make polars.
|
|
| |
| Re: Nexo GeoS 12 and sub [message #321524 is a reply to message #320881 ] |
Sat, 10 May 2008 18:01 |
HarryBrillJr.
Messages: 592
Registered: April 2007 |
Has No Life |
|
|
| Lou Castellucci wrote on Thu, 08 May 2008 12:27 |
So, for just a little bit more info, what speakers did you use in MAP to configure the CD18s? Did you use HP700s? or a real CD18? If you did use HP700s what did you do model the CD18?
|
CD18s are not available in MAPP unfortunately. Meyer measures the loudspeakers and uses that data. Meyer does not have CD18s to measure, and would probably not bother if they did.
A CD18 is basically 2 bandpass subs back to back. At 63Hz, a sub is basically a point source, and is nearly omni. ANY 2 subs with reasonably close dimensions to a single CD18 would work to model the CD18 for the sake of demonstrating array performance. A single "modeled CD18" with a cardioid pattern should suffice as an appropriate model in our array model. I did not model the SUPER cardioid model, but that doesn't change the relationship between 2, 3, 4 subs spread vs tight packed, cardioid or not. If you spread them out, it narrows the pattern, because it's creating a longer line array, albeit horizontal. The ground plane effectively doubles the length in the vertical. If you stacked them high and tight pack them you will begin to lose the affect of the rear drivers on the front and vice versa. I'm not convinced you will notice in the front since the large frontal area will serve to narrow the pattern anyway. However in the rear you would start to notice some increased gain. I cannot predict this since MAPP is 2D, not 3D.
Is it possible the sense of wider pattern is due to the left and right arrays interacting less, and therefore canceling less off center, due to the tighter individual patterns? So each array when spread gives a narrower pattern, avoiding interference with it's neighbor on the other side. See pictures at the bottom. Notice the BLUE steaks.
IMO the best way to array them is stacked high in a vertical line. If you are in an arena or other space with flat floors and balconies, then you might want to consider flying the subs behind the main array. This will probably give better imaging at the expense of the coupling you get on the floor.
Regardless of how you choose to array your CD18s, you want to avoid putting the back of the cabinet near a boundary, and side walls will need special attention too. They are honestly very unforgiving in some situations.
[Updated on: Sat, 10 May 2008 18:02]
|
|
| | |
| Re: Nexo GeoS 12 and sub [message #321916 is a reply to message #321883 ] |
Mon, 12 May 2008 03:42 |
HarryBrillJr.
Messages: 592
Registered: April 2007 |
Has No Life |
|
|
The only concern we have is that the MODEL has the SAME PATTERN.
The levels in the processor may not be matched, but other than that it's just delay and polarity as far as I can tell. I'll measure the processor next time I get a chance and post the results. Any other processing would be common to front and rear.
As for the Meyer subs, Mapp is 2D, if you stand the subs up so you are looking DOWN on them, you are only going to see the response of a single driver. If you rotate them 90 degrees you will see the response of 2 drivers. Still the drivers are mostly omni at 60 hz, the wavelengths are long, and they couple quite well so in reality it wouldn't make much difference for this discussion.
The 2 deep end fire is really a vastly superior way to achieve a cardioid pattern, and 4 deep is even better, especially if you aren't flying them.
|
|
|
| Re: Nexo GeoS 12 and sub [message #322693 is a reply to message #321916 ] |
Wed, 14 May 2008 10:01 |
 Bob gardam
Messages: 42
Registered: August 2005
Location: Tasmania, Australia |
Should Get Out More |
|
|
| HarryBrillJr. wrote on Mon, 12 May 2008 18:42 |
The only concern we have is that the MODEL has the SAME PATTERN.
The levels in the processor may not be matched, but other than that it's just delay and polarity as far as I can tell. I'll measure the processor next time I get a chance and post the results. Any other processing would be common to front and rear.
As for the Meyer subs, Mapp is 2D, if you stand the subs up so you are looking DOWN on them, you are only going to see the response of a single driver. If you rotate them 90 degrees you will see the response of 2 drivers. Still the drivers are mostly omni at 60 hz, the wavelengths are long, and they couple quite well so in reality it wouldn't make much difference for this discussion.
The 2 deep end fire is really a vastly superior way to achieve a cardioid pattern, and 4 deep is even better, especially if you aren't flying them.
|
Strangely enough I actualy spent some time looking at a 241 with smaart today, being curious about what went on.. there is different processing one the front and rear drivers apart from polarity and delay, theres also a lot more going on in there, some of which I don't think can be measured with smaart
when I measured it I noticed the delay setting alway seems to be wrong for some frequencies but right for others, it almost seems like they have different delays for different frequencies, though I can't think of a way to achieve that without the use of fir filters or multiple crossovers.
So I'm not entirely sure what is going on in the processors, but what I can tell you is that it's more than just delay and polarity!
|
|
| | |
| Re: Nexo GeoS 12 and sub [message #322974 is a reply to message #322693 ] |
Wed, 14 May 2008 22:30 |
|
| Bob gardam wrote on Wed, 14 May 2008 11:01 |
| HarryBrillJr. wrote on Mon, 12 May 2008 18:42 |
The only concern we have is that the MODEL has the SAME PATTERN.
The levels in the processor may not be matched, but other than that it's just delay and polarity as far as I can tell. I'll measure the processor next time I get a chance and post the results. Any other processing would be common to front and rear.
As for the Meyer subs, Mapp is 2D, if you stand the subs up so you are looking DOWN on them, you are only going to see the response of a single driver. If you rotate them 90 degrees you will see the response of 2 drivers. Still the drivers are mostly omni at 60 hz, the wavelengths are long, and they couple quite well so in reality it wouldn't make much difference for this discussion.
The 2 deep end fire is really a vastly superior way to achieve a cardioid pattern, and 4 deep is even better, especially if you aren't flying them.
|
Strangely enough I actualy spent some time looking at a 241 with smaart today, being curious about what went on.. there is different processing one the front and rear drivers apart from polarity and delay, theres also a lot more going on in there, some of which I don't think can be measured with smaart
when I measured it I noticed the delay setting alway seems to be wrong for some frequencies but right for others, it almost seems like they have different delays for different frequencies, though I can't think of a way to achieve that without the use of fir filters or multiple crossovers.
So I'm not entirely sure what is going on in the processors, but what I can tell you is that it's more than just delay and polarity!
|
Different delays for different frequencies is trivial, all IIR filters behave in this manner. You would not be using FIR filters for LF bandpass, due to the excessive latency penalty!
Phill Graham
Doctoral Candidate
Georgia Tech
|
|
|
| Re: Nexo GeoS 12 and sub [message #322993 is a reply to message #322974 ] |
Wed, 14 May 2008 23:30 |
Bob gardam
Messages: 42
Registered: August 2005
Location: Tasmania, Australia |
Should Get Out More |
|
|
Okay, I'm heading a bit out of my depth here I suspect.
However, It's my understanding that an IIR filter won't cause any *delays* at different frequencies, but it will cause phase shift at different frequencies (they can cause a change in the phase of the signal without the signal being changed from a time perspective, which get quite complicated when you look at it from a waveform level, but the way I'd qualify that is that while the peaks and valleys of the wave form would appear at slightly different times due to the phase shift, the start of the sound and end of the sound would occur at the same points)
You're probably right about not using FIR filters on subs due to latency, however I don't know of any other way to explain what I measured (which definitely doesn't mean that another way to explain it doesn't exist) Perhaps the are only using Fir filters on the from driver (which would need significant delay on it anyway to aligned with the back driver and this would possibly lower the impact of the fir filters latency) I don't remember if I noticed the strange delays on both drivers or only one.
Once again, to anyone reading, I'm out of my depth here, so I'm mainly writing this so someone smarter than me can either confirm my thoughts or tell me how wrong I am in order to learn, so don't trust anything I've just said!
|
|
|
| Re: Nexo GeoS 12 and sub [message #323004 is a reply to message #322993 ] |
Thu, 15 May 2008 03:21 |
HarryBrillJr.
Messages: 592
Registered: April 2007 |
Has No Life |
|
|
|
Like I said, the next time I get a chance, I'll measure it and post the results here. And it still doesn't make a bit of difference for the model. Cardioid is cardioid is cardioid. Why are we even having this discussion? I never debated the processing in the NX241.
|
|
|
| Re: Nexo GeoS 12 and sub [message #323082 is a reply to message #322993 ] |
Thu, 15 May 2008 11:42 |
|
| Bob gardam wrote on Thu, 15 May 2008 00:30 |
Okay, I'm heading a bit out of my depth here I suspect.
However, It's my understanding that an IIR filter won't cause any *delays* at different frequencies, but it will cause phase shift at different frequencies
|
An IIR filter most definitely causes a change in delay, and that does depend on frequency, and is different at each frequency. That change in delay is due to phase behavior of the filter, which is minimum phase in the case of an IIR.
Phase are sort of two sides of the same coin. A shift and in phase causes a delay, and pure delay can be modeled as a linear phase shift with frequency.
| Quote: |
(they can cause a change in the phase of the signal without the signal being changed from a time perspective, which get quite complicated when you look at it from a waveform level, but the way I'd qualify that is that while the peaks and valleys of the wave form would appear at slightly different times due to the phase shift, the start of the sound and end of the sound would occur at the same points)
|
This is a bad description of a real effect. There is local frequency delay from the phase shift behavior of a say, a parametric filter, but the group delay of a broad range of frequencies would not be shifted. That does NOT change the fact that the waveform envelope of this range of frequencies is indeed affected in time by the phase of the IIR filter.
You are confusing the point though. An FIR takes a fixed delay penalty to allow you to manipulate the phase at any point independent of the amplitude of the filter. To be able to say ANYTHING about the phase of a given frequency, you need at least one period of that frequency to occur. Thus the delays inherent in FIR filters get very long at low frequencies.
An IIR works just like an analog filter. It has the same amplitude and phase behavior as the textbook transfer function.
Cardiod isn't magic. Physical spacing, plus delay+polarity swap. The delay can be done digitally, or it can be done in the analog domain by stringing together allpass filters.
Your "Crazy phase" is more than likely a function of the box's acoustic phase, couple with the high and lowpass filters' phase response.
Phill Graham
Doctoral Candidate
Georgia Tech
|
|
|
| Pages (3): [ 2 ] |
  |
Goto Forum:
Current Time: Sat May 17 02:26:59 CDT 2008
Total time taken to generate the page: 0.29171 seconds |
Forum Home
Login
Forum Search
Register
FAQ
