Rational Acoustics

February 8th, 2011, 05:30 PM
Lately I’ve run into a lot of split mains systems in various Dutch theaters. A high mounted speaker for the balcony and a low mounted or stacked speaker for the floor seats. When trying to achieve equal level amongst others, the latter is mostly turned down in level due to the decreased distance it has to cross or spill from the upper speaker and delayed accordingly. In turn it’s reasonable to assume that the phase and level alignment point will likely end up somewhere below the imaginary center line between the speakers. In this example mic position one. After alignment both speakers have figuratively become one, especially because they share recourses at lower frequencies due to their increasingly omnidirectional behavior. Now it’s time to align the combined mains or uncoupled line array to the subs, which are 9 out of 10 times stacked on stage. With a single point source system, I like to apply Bob McCarthy’s approach and align the mains to the subs half way in the audience. Dividing the “error” or lobing equally across the audience. However as you can see in my 100 Hz example at mic position 2 this could mean that you end up in a null or ‘power-valley”. Assuming that you’re trying to cross the system at 100 Hz. For evident reasons this phenomena becomes worse as you lower in frequency. From a timing point of view the only logical point seems to be to go back to phase aligned mic position one. In most cases however, this position tends to be relatively far to the back of the audience. Resulting in even more bass at the front rows. I’ve used G.P.A. 2.48 to create these predictions because it allowed me to quickly illustrate the problem which doesn’t require a lot of accuracy. It’s the physics that are the cause of the problem. When the displacement between the main speakers increases more than halve a wavelength at the crossover frequency, lobing becomes an issue. These lobes then get steered because of the delay offset. In turn it becomes harder to find a valid and democratic point to align the mains to the subs. Speaker repositioning would solve a lot of problems but isn’t an option in practice due to limited mounting/rigging hardware and points. But more so, a lack of time and being house P.A.’s also often not allowed or appreciated. Bringing our own P.A. is neither an option. Yet this specific split mains scenario is something I face very often. These predictions are of course very clean and free of reflections. Being indoors camouflages these effects to some extent, but still. Probably TANSTAAFL. I’m curious to your experiences.



Rasmus Rosenberg
February 9th, 2011, 03:39 AM
Hey Merlijn,
Nice post. Design to align!

February 9th, 2011, 08:46 PM

First, in your second picture of GPA it seems that the sub is OFF.
Have you tried to put it on, just to see ?

Like you said, in this kind oh theatres, " In turn it becomes harder to find a valid and democratic point to align the mains to the subs".

Exactly ! Even in a one level audience setting delay is a compromise just because
the delay between mains & subs is changing depending of your location.
So just imagine with 2-levels ......
But there is always a specific setting of the delays who gives you the best alignement FOR A MAXIMUM OF SEATS.

In my modest opinion, i will do something like that:

-First adjust the lowers speakers with the subs (with the phase trace or other method) for the maximum of floor seats (i think it's easy and just need a little delay since speaker & subs seems to be physically aligned on your picture)

-Second, place mic at balcony seats. Record IR with ALL the system ON.
Now, note the 3 timing positions of T0=upper speaker, T1=lower speaker, T2=subs.
I will now adjust 'by ear' the delay of the upper speakers with a value comprising between T1 and T2.
Or maybe if you have really good IR of the subs at this location, try to adjust delay with the phase trace.

Hope that helps....

February 15th, 2011, 06:34 AM
Hi Patrice,

Thank you for your answer. I've left the sub turned off deliberately to illustrate the lobing of the spilt mains system. As you can see picking a spot for aligning the subs isnít as easy compared to a single point source system. Last week I worked at a theatre with 2 balconies and a 3-speaker spilt mains system again with a single sub on stage. Iíve included a picture. Even more interaction at the crossover frequency. Iím still hoping on some more feedback and experiences from the visitors of this forum.