Rational Acoustics

October 26th, 2016, 02:51 PM
Relatively new to Smaart and was wondering if anyone could give me some advice about the average trace shown below.

I took 6 Snapshots at various locations around 1 half of the room. The System is 8 DVDOSC Stacked at 100hz Preset (HI)on top of SB28.

My question is if the coherence looks generally ok, why the big dip between 2-4k?

And then what measure would one take to rectify the problem.

Many thanks in advance :)


October 26th, 2016, 11:26 PM
Hi, I'm not all that experienced with measuring larger sound systems in rooms and buildings, as I personally measure in the automotive interior.

That said, the principles are of corse the same, and I am very experienced with dual FFT including Smaart v7&8. I use it almost daily.

So that's my credentials. Take what I say with a grain of salt.

The first question I would ask is, what exactly is the dB per line/division of your graph set to? This can be found in transfer function options.

The second question, what smoothing level is being applied?

The third question, is there some known response of this particular loudspeaker system that leads you to believe that the dip you see is even a "problem" per se ?

Maybe this is normal behavior, ie, to be expected? This certainly appears relatively well behaved to me, HOWEVER, without seeing phase, Live IR, and the rest of the Smaart screen to be able to see a lot of the relevent settings, making any kind of determination based on magnitude alone is near impossible.

Also, understand that Smaart's TF averaging (spatial) is a dB average, that can either be coherence weighted, or not.

I've personally not seen instances where dB vs. power averaging results in grossly different magnitude responses, but the. again, like I said, I'm not experienced with measuring large speakers systems in rooms/buildings.

I guess the last thing I would recommend is to look at the individual traces that make up the average, and see which ones, if not all of them, have this trend (the dip) in common. Then, by knowing which individual measurements have this characteristic, and where these individual traces were taken (in relation to the coverage pattern of the loudspeaker and position within the room) you can better determine why the trend is there in your average.

The school of though that many guys have, or so I've read, when talking about spatial averaging is that one needs to know what trends the individual measurement positions have in common, and what are position dependent anomolies. Now, spatial averaging by its nature is SUPPOSED TO be a way of kind of "automatically" reducing the position dependent anomolies. However, depending ON the position of each contributing measurement, one may or may not manifest this "idealized" scenario, ya know?

I don't know if any of that babble was helpful, so to sum up, be sure to show the entire Smaart screen when asking questions about measurements :)