2.2 System Question

[Chris A]

I pulled it all out and started from scratch tonight with REW sweeps. I found no matter what I did with the center, it had a lot of dips and peaks. The 7ii mains have a very good freq response from the get go, so much of my up and down FR after Audyssey runs is probably due to the center.

 

Most of what Audyssey does in terms of time alignment and EQing is mostly below 200 Hz (IIRC).  Above that point, very small differences in path lengths between loudspeakers to your listening position will show up as comb filtering when they are all playing the same (mono) signal during a single sweep.  This is okay, since human listening system is generally very tolerant of comb filtering effects.  Additionally, just the differences in the first reflection arrival times will create large differences in the REW sweeps using more than one loudspeaker during the sweep.  For instance, a graphic that I found:

 

 

All of this is typical for combined measurements.  When you add the center channel to a REW sweep, you will get more comb filtering, but you also will get better stability of the perceived stereo image with the center of the image locked near the center loudspeaker, and the center channel will fill in the 1.8 kHz hole that you will find with stereo-only sound images.

 

If you did REW sweeps for each loudspeaker from your listening position (or from a listening position next to your prime listening position), then displayed the impulse responses from each individual loudspeaker sweep within REW, you might see something like this (given that Audyssey hasn't been used to match delays).  Even small differences in time delays, as little as 1/10th the wavelength at the frequency of interest, can be detected as a "worsening" of the sound field image playing back with the center:

 

 

I've also found that any differences in FR between each front loudspeaker and the center will lead to a "timbre mismatch", and this can be exaggerated by reflections in the near fields of each loudspeaker (different delayed reflections generally within 3 feet [1 metre]) of the centerline of the speakers laterally or vertically.  So the recommendation to place absorbent material around sound-reflective objects on the front or side walls next to your fronts or centers will also improve the center channel (3.2) listenability.  All this assumes, however, that Audyssey has been used to set channel delays carefully for the listening position from which you will judge the 3.2 array performance.

 

Generally speaking, getting the center channel set up to match the fronts, especially for "synthesized center channel from stereo" performance, is actually more involved than most people want to acknowledge.  This means that getting surround sound arrays (e.g., 5.1, 7.1, etc.) working properly in-room is much more difficult than stereo arrays on average, and why I believe so many people prefer stereo (even with its 1.8 kHz notch due to interaural hearing time delay) to 5.1 arrays: multichannel arrays must be set up well to compete well with even casually set up stereo arrays.

 

Chris

Edited June 16, 2015 by Chris A

[gadgtfreek]

Chris, I think I know the answer, but is aiming speakers based on FR tested a better move, or actually aiming them considering the MLP?

 

What I saw yesterday is some changed to the center made more sense on the graph, but did not make sense from a standard aiming of the mains and center.

 

I almost think you properly aim things (aim center up or down at ears and mains toed in gradually) and "it is what it is".

[Chris A]

I'm guessing that "aiming" in this context is where to position the centerline of the midrange/HF horn relative to the listening position. 

 

Well designed midrange/HF horns will have a fairly wide polar coverage so that you shouldn't have to sit in a particular spot in order to be "on axis", i.e., +/- 15 degrees should have about the same polar coverage in both axes (V, H).  If the midrange/HF horn in your center loudspeaker is the same geometry (same width and height) as your front loudspeaker midrange horns, then I'd recommend trying to achieve about the same off-axis angle to the listener on the center vs. fronts, to within about 15 degrees.  They really don't have to aim directly at your listening position at zero degrees, but I'd be a little careful about aiming the center at the ceiling.

 

The dips in frequency response due to air gaps around the center (under, over and sides) is something that you have some control over even after  adjusting the aiming angle at your listening position--just fill in the holes, or by moving the entire center speaker toward the front wall to eliminate gaps behind the back of the speaker.

 

You can also eliminate nearfield reflections around the midrange horn mouth within about 3 feet by laying down some absorbing material, if possible, like a thick fuzzy tablecloth or small mat on flat horizontal surfaces, or pieces of an acoustic absorption square cut to cover any prominent vertical surfaces, such as an equipment rack, etc. This will coalesce the center midrange imaging and flatten its FR at your listening position.

 

Chris

[gadgtfreek]

Thanks. I usually toe the mains in towards a spot 2 feet or so behind the MLP, so it is not too aggressive IMO. I will aim the center upward properly at the MLP and then try moving in and out and see what happens.

 

Since it sits on top of a wide tv stand (much wider and deeper than it is) do you think covering the stand top is a worthy experiement?

[Scrappydue]

do you think covering the stand top is a worthy experiement?

cover with what?

 

wonder if lifting it up a tad would change the peaks and nulls? perhaps a very low center stand?

 

something similar to this size?

 

http://www.hometheaterplus.com/l-ml2cb-1887.html