Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Robert Lunceford]

A single cabinet designed for use in a line array can have FIR filtering applied in order to improve the performance and  to reduce unwanted characteristics inherent to that speaker. What happens when we then combine multiple speakers to form a line array and these speakers then acoustically interfere with each other?
Can FIR filtering be used to reduce the unwanted effects inherent to a line array. If so, would you need a different set of filters for a 6 box hang vs a 12 box hang?

In this video, there is a comparison of finger lobing between a Meyer line array and a Danley Jericho synergy horn. The video demonstrates that the lobing is much more prominent in the line array vs the single point source. Would it be possible using FIR filters to reduce the finger lobes in the line array so that it performs equal to the point source cabinet?
https://www.youtube.com/watch?v=cPHpusei0jQ

Does Meyer use FIR in their Mina/Galileo?

[David Sturzenbecher]

A single cabinet designed for use in a line array can have FIR filtering applied in order to improve the performance and  to reduce unwanted characteristics inherent to that speaker. What happens when we then combine multiple speakers to form a line array and these speakers then acoustically interfere with each other?
Can FIR filtering be used to reduce the unwanted effects inherent to a line array. If so, would you need a different set of filters for a 6 box hang vs a 12 box hang?

In this video, there is a comparison of finger lobing between a Meyer line array and a Danley Jericho synergy horn. The video demonstrates that the lobing is much more prominent in the line array vs the single point source. Would it be possible using FIR filters to reduce the finger lobes in the line array so that it performs equal to the point source cabinet?
https://www.youtube.com/watch?v=cPHpusei0jQ

Does Meyer use FIR in their Mina/Galileo?

You are venturing into the realm of Martin's MLA and EAW's Anya at this point, no? In which case different FIR filters are used on each driver in each box.  While I don't have any direct experience with either outside of casual demos, their sales pitch would lead you to believe these unwanted effects are greatly reduced.

[Luke Geis]

The short answer is no.

The problem is the physical location in relation to the listener. So while you could use time and phase to make it all work for one location, as soon as you move, you are now out of the position in which all that correction was optimized for.

The often expected function of a LA system is that all the speakers act as one unit. The issue is that they do not. The reason an LA works at all is because of the physicality and point of incidence. In the near field, the majority of the boxes are high above and blowing over the listener. The listener only truly hears the lower 1/3 of the speakers and the resultant effect is that in the near field the LA almost acts like a point source system with a rather crappy impulse response and the phase issues, but it works. In the far field, the listener is far enough away that they can hear all the boxes at once, and the difference in path lengths between each box is close enough to not be of huge detriment to the sound quality. If you are 150' away from a 16 box hang, the path lengths may only be off by a few feet from top to bottom depending on venue layout. Again the issues with phase still exist, but it can be lived with. It is not ideal, but sound quality can be made to be good enough at near and far distances and have less change in SPL over that distance while doing it.

Some systems have moved to using phase as the actual tool to improve the performance of said speaker. Martin and EAW are the biggest players in that concept right now. They use a large number of drivers in each speaker with processing that controls each " cell " in order to steer the sound in a particular way. This can help keep energy off of walls, or direct it in a certain way to improve the listeners experience. FIR or not, this can already be done.

FIR should not be confused with time and phase alignment of speakers as a whole. It is a function mostly used for crossover points of drivers within a single speaker, not speaker to speaker so to speak. The delay / time alignment of two speakers can already be controlled at the DSP level without any FIR algorithms.

FIR is not a magic bullet. It comes at the cost of latency. As of right now, the lowest practical crossover point that fir can be utilized and be within reasonable latency is around 300hz with about 10ms of latency!!!!!! Keep in mind that FIR filters have to stack, so if you have two FIR filters in one 3 way speakers crossover network, the delay of one must stack up with the other. This may mean going from an overall latency of a couple of ms. to 10ms or more depending on crossover points and filter shelf order!!! 10ms is roughly equivalent to 10' in distance!

So using FIR filters in an LA sytem won't greatly improve upon its function. The individual cells will be improved, which means that as whole, the system will improve too. Beyond that, it is just DSP to control what goes where. FIR will likely not be a major factor in improving a LA performance beyond what they already do.

I am in the camp that a point source is the ideal approach to solving audio issues. The challenge is trying to make a point source box work much like a LA system does. Even coverage from front to back and side to side in both the frequncy and SPL domain. At the moment a line array system, despite its operation practices, is still not beating physics, it is simply bending the rules a little.

[Ivan Beaver]

If you are 150' away from a 16 box hang, the path lengths may only be off by a few feet from top to bottom depending on venue layout.

And while "a few feet" different may not appear to be much, it is a HUGE difference in terms of combfiltering.

Even inches make a difference.

Try this little test.

Take a small SINGLE DRIVER (preferred) cabinet (like a 3" or 4" full range), put them on top of each other and apply pink noise to them.

Stand a little ways back.  Have a friend slide the top one back a couple of inches.

If you can't hear the difference, I suggest lighting as a career choice

[Lee Buckalew]

A single cabinet designed for use in a line array can have FIR filtering applied in order to improve the performance and  to reduce unwanted characteristics inherent to that speaker. What happens when we then combine multiple speakers to form a line array and these speakers then acoustically interfere with each other?
Can FIR filtering be used to reduce the unwanted effects inherent to a line array. If so, would you need a different set of filters for a 6 box hang vs a 12 box hang?

In this video, there is a comparison of finger lobing between a Meyer line array and a Danley Jericho synergy horn. The video demonstrates that the lobing is much more prominent in the line array vs the single point source. Would it be possible using FIR filters to reduce the finger lobes in the line array so that it performs equal to the point source cabinet?
https://www.youtube.com/watch?v=cPHpusei0jQ

Does Meyer use FIR in their Mina/Galileo?

I cannot speak to Anya as I have not used or measured it enough.  As far as MLA is concerned the answer is yes but not quite in the way that you are asking. 
The finger lobes are a result of comb-filtering due to vector differences between and among multiple drivers and each listening position. 
The comb-filtering is not affected merely by good filter design, FIR or other, but also relies on proper driver spacing and carefully controlled interactions between interacting drivers.  Proper acoustical design is also critical.

In MLA there are very steep filter slopes utilized to control the range of frequencies that each driver type produces so that not only can each cell be controlled but each passband does not interact with any adjacent passband.  By stopping interaction with adjacent passbands much of the comb-filtering that line arrays produce is eliminated.  Through proper acoustic design of the driver cells/cabinets/array positioning many of the comb-filtering interactions that many line arrays create is eliminated.  You will still have interactions from cabinet to cabinet as the elements are curved.

At that point the remaining issues are proper coverage which, in MLA, is created by proper intercabinet angles of the array elements.  After this step filters are applied to each cell to create proper phase and frequency coherent interactions in the plotted listening plane.  At this stage the position of each cell within the array has been accommodated and corrected for in terms of phase interaction. 
I have not experienced a properly deployed MLA configuration that produced audible combing effects with pink noise either in the vertical or the horizontal coverage either on axis or off axis.  I say properly deployed for two reasons; first, it must be hung at the height and intercabinet angles that you used in the calculations, second, it must have enough cabinets to be effective at what you are asking the system to accomplish.

I would also point out that not all comb-filtering is observable by walking through coverage.  Multiple drivers combining in a single manifold (EV MT Systems anyone) or single drivers feeding multiple horn exits (many line array HF designs) often produce significant comb-filtering due to the multiple path lengths involved on a frequency by frequency basis.  What you hear as a result is the combination of all elements producing both constructive and destructive interactions but you do not hear the swishing, jet engine at takeoff sound that many people assume accompanies all comb-filtering.

Lee

[Nathan Riddle]

Try this little test.

Take a small SINGLE DRIVER (preferred) cabinet (like a 3" or 4" full range), put them on top of each other and apply pink noise to them.

Stand a little ways back.  Have a friend slide the top one back a couple of inches.

If you can't hear the difference, I suggest lighting as a career choice

You mean this video

The same sound from multiple sources is an unnatural event
https://www.youtube.com/watch?v=VHjdh-Vka-g

[Tim McCulloch]

You mean this video

The same sound from multiple sources is an unnatural event
https://www.youtube.com/watch?v=VHjdh-Vka-g

Politicians, used car salesmen and booking agents all talk out of both sides of their mouths and we don't think it unnatural at all...

[Peter Morris]

A single cabinet designed for use in a line array can have FIR filtering applied in order to improve the performance and  to reduce unwanted characteristics inherent to that speaker. What happens when we then combine multiple speakers to form a line array and these speakers then acoustically interfere with each other?
Can FIR filtering be used to reduce the unwanted effects inherent to a line array. If so, would you need a different set of filters for a 6 box hang vs a 12 box hang?

In this video, there is a comparison of finger lobing between a Meyer line array and a Danley Jericho synergy horn. The video demonstrates that the lobing is much more prominent in the line array vs the single point source. Would it be possible using FIR filters to reduce the finger lobes in the line array so that it performs equal to the point source cabinet?
https://www.youtube.com/watch?v=cPHpusei0jQ

Does Meyer use FIR in their Mina/Galileo?

Yes you can but its complicated ...

https://www.youtube.com/watch?v=ljhmCsdtTSM

[Mark Wilkinson]

The short answer is no.

The problem is the physical location in relation to the listener. So while you could use time and phase to make it all work for one location, as soon as you move, you are now out of the position in which all that correction was optimized for.

The often expected function of a LA system is that all the speakers act as one unit. The issue is that they do not. The reason an LA works at all is because of the physicality and point of incidence. In the near field, the majority of the boxes are high above and blowing over the listener. The listener only truly hears the lower 1/3 of the speakers and the resultant effect is that in the near field the LA almost acts like a point source system with a rather crappy impulse response and the phase issues, but it works. In the far field, the listener is far enough away that they can hear all the boxes at once, and the difference in path lengths between each box is close enough to not be of huge detriment to the sound quality. If you are 150' away from a 16 box hang, the path lengths may only be off by a few feet from top to bottom depending on venue layout. Again the issues with phase still exist, but it can be lived with. It is not ideal, but sound quality can be made to be good enough at near and far distances and have less change in SPL over that distance while doing it.

Some systems have moved to using phase as the actual tool to improve the performance of said speaker. Martin and EAW are the biggest players in that concept right now. They use a large number of drivers in each speaker with processing that controls each " cell " in order to steer the sound in a particular way. This can help keep energy off of walls, or direct it in a certain way to improve the listeners experience. FIR or not, this can already be done.

FIR should not be confused with time and phase alignment of speakers as a whole. It is a function mostly used for crossover points of drivers within a single speaker, not speaker to speaker so to speak. The delay / time alignment of two speakers can already be controlled at the DSP level without any FIR algorithms.

FIR is not a magic bullet. It comes at the cost of latency. As of right now, the lowest practical crossover point that fir can be utilized and be within reasonable latency is around 300hz with about 10ms of latency!!!!!! Keep in mind that FIR filters have to stack, so if you have two FIR filters in one 3 way speakers crossover network, the delay of one must stack up with the other. This may mean going from an overall latency of a couple of ms. to 10ms or more depending on crossover points and filter shelf order!!! 10ms is roughly equivalent to 10' in distance!

So using FIR filters in an LA sytem won't greatly improve upon its function. The individual cells will be improved, which means that as whole, the system will improve too. Beyond that, it is just DSP to control what goes where. FIR will likely not be a major factor in improving a LA performance beyond what they already do.

I am in the camp that a point source is the ideal approach to solving audio issues. The challenge is trying to make a point source box work much like a LA system does. Even coverage from front to back and side to side in both the frequncy and SPL domain. At the moment a line array system, despite its operation practices, is still not beating physics, it is simply bending the rules a little.

Hi Luke,  I have different experiences with FIR than you describe.

You make distinction between system time and phase alignment, and FIR use for driver crossover points.
They are the exact same task IMO.  When we align subs to mains, we are really aligning subs to the lowest frequency driver in the mains...which makes it driver-to-driver.
It just so happens the rest of the main's driver-to-driver is supposedly already done for us.
You can just as easily use FIR to provide the time and phase alignment between subs and mains as you can with traditional IIR DSP,
AND with FIR you can make the phase alignment's lay on top of each other through a much wider critical summation region, fairly easily.

FIR filters do not need to stack.
Drivers are driven by channels operating in parallel, not series (stacked).
Delay induced by FIR is that of the longest filter used, .....which simply translates to how low in freq, how much correction, are you trying to work in. 
So if it's 4ms to 300Hz on a 3-way, it's 4ms total.

I also have to say, I think line arrays are growing up, addressing the problems of using multiple sources.
Because I think as much attention is being given to the acoustical design like Lee mentions, as is to steering the output with DSP.
The acoustic design work is helping fulfill classic line theory requirements with regard to driver-to driver spacing requirements,
and better vertical directivity control is occurring....even without DSP.

I certainly do not know what kind of DSP Martin and Anya are using, but I'd almost bet my ass it has to include FIR at the HF/VHF end.
Looking forward to learning !

[Ivan Beaver]

You mean this video

The same sound from multiple sources is an unnatural event
https://www.youtube.com/watch?v=VHjdh-Vka-g

The first time I saw/heard that demonstration was at a SynAudCon class almost 2 decades ago.

When combined with some speech tracks and different short (a few ms) delays, I got real interested in the detrimental effects of different sources arriving at the same location at different times (combfiltering), and how to combat it.

It was one of those "pivotal moments" in my life.

[ProSoundWeb Community]