M1d array correction

H

Helge A. Bentsen

Senior
Jan 19, 2011
1,016
63
48
44
Oslo, Norway, Norway
drbentsen.no
When using a Gallileo 616 with M1Ds, what is the preferred way of using array correction on a 9 box hang?

1. Running all 9 linked and enter 9 in the M series array correction?
2 Subdividing in long/mid/short-throw and using 3 different array corrections?
3. Rolling your own array correction based on the actual measured frequency response of the array in the room?
4. "It depends"?

I'm leaning towards option 2 with some system EQ based on measurements and taste, but I would like some input on this from experienced users :)

Note that this is for a touring application.
 
Re: M1d array correction

I don't know anything about Galileo or how it manages arrays, but I do know that in the very least you want to keep all LF subsections of the array gained the same. If the gain is changed to any LF section, array coupling will be compromised and all sorts of lobing can occur. Of course, there may be situations where you may want to direct bass to a specific location, but I doubt that can be done without a computer simulation.

No, I am not an experienced user of large arrays, so take my advice with a grain (or two) of salt. :lol:
 
Re: M1d array correction

Helge Dr. Bentsen said:
When using a Gallileo 616 with M1Ds, what is the preferred way of using array correction on a 9 box hang?

1. Running all 9 linked and enter 9 in the M series array correction?
2 Subdividing in long/mid/short-throw and using 3 different array corrections?
3. Rolling your own array correction based on the actual measured frequency response of the array in the room?
4. "It depends"?

I'm leaning towards option 2 with some system EQ based on measurements and taste, but I would like some input on this from experienced users :)

Note that this is for a touring application.
Click to expand...

Helge,

I'd always recommend a minimum of three zones.
 
Re: M1d array correction

The standard approach is to have several zones on the array, and assign all of those outputs to a single array correction. You don't need to link the outputs, the array correction assignment is separate from the link group assignment.
 
Re: M1d array correction

Helge,

Nine box M1D array is, in my opinion, still rather small sound system, dimension (length) wise.
Array correction, if you meant low frequency buildup, is probably best handled for the whole array. This WILL provide the best result as well as remove a temptation to fiddle in this particular band pass.
On the other hand, you may wish to apply some sort of HF shelf or other sort of HF boost EQ for units intended to cover further most seats.
Those will probably be the top ones.
Array correction on G616 will address the whole hang. You may chose a different box 'count' when deciding about the setting. The decision will be partially influenced by type of act, hence available LF headroom as well as personal taste.

Apart from respecting MS, I'm not affiliated with them in any other particular way, so please take my thoughts as a general type of 'advice'.

With respect
 
Re: M1d array correction

Phil Graham said:
There's little need to worry about gaining the various array section the same, especially at low frequencies.
Click to expand...

Hey Phil,

Could you expand on that?

As far as I knew, the LF in an array isn't very directional, so turning down boxes, say, at the bottom of the array, are certainly not going to reduce the LF heard in the close seats. Instead, it's just going to totally screw up the array's overall performance.

If I wanted to reduce LF in close seats, I'd want the array to be as long as possible (read: no particular boxes turned down making the array effectively shorter) to keep the highest vertical directivity possible, thereby allowing the large forward-facing lobe to 'miss' the seats below?

Again, I've never really done this in real life, I'm just thinking aloud based on what I've read.

Edit: Phil, I think you're referring to something Galileo-specific, while I'm probably using the wrong terminology to refer to general array behavior?
 
Last edited:
Re: M1d array correction

Silas Pradetto said:
Hey Phil,

Could you expand on that?
Click to expand...

Not today, sorry. Eventually, probably.

As far as I knew, the LF in an array isn't very directional, so turning down boxes, say, at the bottom of the array, are certainly not going to reduce the LF heard in the close seats. Instead, it's just going to totally screw up the array's overall performance.
Click to expand...

I would never suggest that turning down the nearest boxes is going to reduce the local LF level, nor would I suggest that is a viable means for managing the local LF levels. The wavelengths are very long, so the turning down of the bottom array boxes in the name of even HF balance is going to have minimal effect on the overall array performance and may actually result in smoother side lobe behavior.

If I wanted to reduce LF in close seats, I'd want the array to be as long as possible (read: no particular boxes turned down making the array effectively shorter) to keep the highest vertical directivity possible, thereby allowing the large forward-facing lobe to 'miss' the seats below?
Click to expand...

A long array is only (about) half the battle. The real issue is the low frequency azimuth to the array vs. the HF azimuth. A longer array doesn't fix this issue. Array length only really determines how effective other means of addressing the difference in azimuth will be.

Edit: Phil, I think you're referring to something Galileo-specific, while I'm probably using the wrong terminology to refer to general array behavior?
Click to expand...

No, this is in no way Galileo-specific.
 
Re: M1d array correction

Helge Dr. Bentsen said:
When using a Gallileo 616 with M1Ds, what is the preferred way of using array correction on a 9 box hang?

1. Running all 9 linked and enter 9 in the M series array correction?
2 Subdividing in long/mid/short-throw and using 3 different array corrections?
3. Rolling your own array correction based on the actual measured frequency response of the array in the room?
4. "It depends"?

I'm leaning towards option 2 with some system EQ based on measurements and taste, but I would like some input on this from experienced users :)

Note that this is for a touring application.
Click to expand...

Sorry you're stuck with the M1Ds. Not my favorite speaker. You can use as many outputs from the Galileo as you want, but apply the LF compensation to them as a group. Start with the number of boxes you have, but you may find you want slightly less compensation, if so, just reduce the number of boxes in the compensation. The factory recommendation is pretty good. This lets you break up the array into zones if you want to do some HF compensation in the near and far field, but eqs the LF as a whole.

Unlike many line array elements, the Meyer speakers are pretty much full range, and do not count on the array to provide LF build up, so the LF roll off provided by the Galileo across the array is critical to them sounding right.

Mac
 
Re: M1d array correction

Thanks for your answers, as always a lot of usefull reading :)

I came across this article while googling the M1Ds:

http://bobmccarthy.wordpress.com/2010/06/29/sim3-optimization-design-seminar-at-uc-irvine/

I'm thinking that if you want to use 606s ABC method you can't have a 9 box array correction on all boxes since that would make your individual zones sound "wrong" when you eq/level them independently since the correction curve is based around the total summation of 9 boxes. So if I use a 9 box array correction and subdivide it into three zones, I can do some slight level tampering and HF compensation in zones, but I have to treat the LF level on a global scale. If I want to use the full ABC method I'll have to treat each zone independently first, and then combine and EQ/level them into a complete system.

Or am I missing something important here? Maybe a third way?
 
Re: M1d array correction

The Meyer array compensation only deal with low frequencies - say 500 Hz and below. I would leave all the boxes at the same gain setting and part of the same array correction, and use wide filters to make changes in the high frequency shading in your different zones. Say if you wanted the bottom boxes to not tear faces off quite so much, you'd use a fairly gentle shelving high filter starting quite low (but not into the realm of the array correction).

There's a little setting which usually takes me a while to find, but it lets you see the effects of the array correction (and also atmospheric compensation) in the output EQ window.

EDIT: Now that I'm thinking about it, I wouldn't do a whole lot of EQ the way I suggested - phase between zones can get screwed up a bit when using EQ that way.
 
Last edited:
Re: M1d array correction

Helge Dr. Bentsen said:
Thanks for your answers, as always a lot of usefull reading :)

I came across this article while googling the M1Ds:

http://bobmccarthy.wordpress.com/2010/06/29/sim3-optimization-design-seminar-at-uc-irvine/

I'm thinking that if you want to use 606s ABC method you can't have a 9 box array correction on all boxes since that would make your individual zones sound "wrong" when you eq/level them independently since the correction curve is based around the total summation of 9 boxes. So if I use a 9 box array correction and subdivide it into three zones, I can do some slight level tampering and HF compensation in zones, but I have to treat the LF level on a global scale. If I want to use the full ABC method I'll have to treat each zone independently first, and then combine and EQ/level them into a complete system.

Or am I missing something important here? Maybe a third way?
Click to expand...

I think the 2 methods are not mutually exclusive. Listening will tell you right away if there is a problem, but doing the kind of subtle (1 or 2dB) level shading he talks about in that article, and some high frequency eq to compensate for air loss, or near field brightness, should not be effected by the array compensation. The Galileo certainly is designed to let you do both. You can play around with it in MAPP Online since it has a virtual Galileo.

Mac
 
Re: M1d array correction

Scott Helmke said:
The Meyer array compensation only deal with low frequencies - say 500 Hz and below. I would leave all the boxes at the same gain setting and part of the same array correction, and use wide filters to make changes in the high frequency shading in your different zones. Say if you wanted the bottom boxes to not tear faces off quite so much, you'd use a fairly gentle shelving high filter starting quite low (but not into the realm of the array correction).

There's a little setting which usually takes me a while to find, but it lets you see the effects of the array correction (and also atmospheric compensation) in the output EQ window.

EDIT: Now that I'm thinking about it, I wouldn't do a whole lot of EQ the way I suggested - phase between zones can get screwed up a bit when using EQ that way.
Click to expand...

Just FYI - the array correction filters can start significantly higher than 500Hz. Look at the filters for Milo at 8 cabinets, and M1D's for 6 which starts at 8K.

Regardless, the approach I always take (be it wrong or right) depends on what my program material is going to be. For the live television events that I do and corporate events I do I typically will leave the array correction set to the same number of cabinets in the array. The end result is a very linear PA which for that stuff is good. Realistically, very few people are there to hear the canned music, or the bump-ins, they are there for the talent or the CEO to be heard clearly.

For the music stuff I do I will start with the array correction set to the number of cabinets in each zone of the array and adjust accordingly based on SMAART (or SIM), and my ears. Sometimes I leave it there, sometimes I adjust it one way or the other, but it all depends on what i hear.

I hope that helps.
 
Re: M1d array correction

Scott Helmke said:
The Meyer array compensation only deal with low frequencies - say 500 Hz and below. I would leave all the boxes at the same gain setting and part of the same array correction, and use wide filters to make changes in the high frequency shading in your different zones. Say if you wanted the bottom boxes to not tear faces off quite so much, you'd use a fairly gentle shelving high filter starting quite low (but not into the realm of the array correction).There's a little setting which usually takes me a while to find, but it lets you see the effects of the array correction (and also atmospheric compensation) in the output EQ window.EDIT: Now that I'm thinking about it, I wouldn't do a whole lot of EQ the way I suggested - phase between zones can get screwed up a bit when using EQ that way.
Click to expand...

Hi Scott

Here is an acoustic measurement of a line array element showing the effects of a 9 dB shelving filter at about 4K5.

- The purple is the original "flat" measurement.

- The blue is the 9dB shelf – note there is about 45 degrees of phase shift at 4K.

- The green is a modified preset to correct for the phase difference so that it is within 10 degrees or less so everything will still sum nicely

Peter
 

Attachments

  • high shelf plot 1.jpg
    high shelf plot 1.jpg
    252.6 KB · Views: 0
Last edited:
Re: M1d array correction

Brad Galvin said:
Just FYI - the array correction filters can start significantly higher than 500Hz. Look at the filters for Milo at 8 cabinets, and M1D's for 6 which starts at 8K.
Click to expand...

OK, I finally got around to checking - in the Compass software, output processing window (output EQ display), there are three little tabs over on the upper right of the screen. Click on the "Response" tab and check the "Array Correction" box. Then you can see what the array correction is doing.

Anyway, it's more than I thought - I think I last seriously looked at it using SMAART on a LD-3 processor. I'm not sure I'd agree with "up to 8k", since usually you see EQ frequency defined at the -3dB point or similar. But higher than I'd said - full effect of the filter at 500Hz and below, maybe. Very gentle slopes, -3dB down point for eight M1D looked around 1.5KHz.
 
Top Bottom