Audiophools...

[Kingston]

I really doubt that one could figure out anything going on around 55Hz from a recording made with a microphone in a room and analyzed with a FFT analyizer but what do I know.

Besides, if anything, lowpassing (as you yourself said) caused by the ball thingies would have affected phase relations/group delays around the fs and not in what definetly is the stopband for them (as such small objects would have nothing to do with frequencies of wavelength several times their size).

I'd also like to thank the OP of this thread for participating in my first KvR signature.

I'm sure you're new to measuring studio acoustics. There are microphones that are near perfectly flat from at least 15hz to 20khz from (at least) dpa or behringer. I used this one: http://www.behringer.com/ECM8000/index.cfm?lang=ENG

They are actually designed for just these applications in acoustic design. FFT spectrum analysers are the main tool in measuring frequency response in these environments. Impulse response tools are used for time domain measurements. I suggest you look at http://www.winmls.com/ if this is too baffling to you.

I used SPAN as it's more than appropriate for quick measurements in frequency domain. As you can see in the graphs there are several things going on in there that are less than ideal.

The bigger bass dips are caused by the 'wall boundary effect': http://www.genelec.com/support/faq/faq23.php#formula

When you know that my monitors were were about 40cm away from the wall when measuring these graphs, you can prove the bass dips mathematically as well.

The rest of the bigger bass and low mid mess is due to room modes. There are several free programs for calculating those as well and they correspond to the FFT graphs EXACTLY.


Also, let me point this out once more as it seems to fall on deaf ears and blindness: THE RUBBER DE-COUPLING SQUASH BALLS HAVE NOTHING TO DO WITH THE PROBLEMS IN THOSE GRAPHS. THERE ARE BEFORE AND AFTER GRAPHS AND AS YOU SEE, THINGS STAY NEARLY THE SAME IN BOTH. And in fact they cause "evening out of the response", which can be heard. I think I'll trust genelec on this (as they sell these things as well).

Excuse the caps, but there has been extreme stupidity and inability to read present in this thread several times already.

And sorry about the rant, I'll stop now and move all my acoustics related discussion to proper forums. I really don't want to take this any further with people who know f**k all about acoustics and base their knowledge on guesswork.


(and nuffink, thanks again for that helmholz info link, it was enlightening)

[peejunk]

FFT analysis has a precision that is fixed in hertz. It means that there is a step at which FFT modes "resonate" with spectral content in the incoming signal (and yes, you can f**king look at fourier transform that way). This step will depend on the size of the window. So if a frequency (say 55Hz) falls between two modes of FFT we will witness something known as spectral dispersion. A sine at an "inbetween" frequency wouldn't cause a straight line but a peak-like function affecting a lot of modes surrounding itself.

Windowing functions are used, among other things, to somewhat remedy this. The keyword being somewhat. Be it anyway, broadband FFT analysis is not a proper tool for analyzing very low frequency content (as this effect is rather irelevant in high frequencies where a range of 10 or 20Hz doesen't even equal to a pixel in a logarythmic view of spectre).

Using a Behringer anything for precise task as acoustic meausrement is slightly pushing it. I'm not saying that they suck or won't do or anything, just that Behringer aren't known for being the high-end, ultraprecise, being-the-shit manufacturer.

As for the room acoustic bits, you oobviously misread my post since I actually did agree with you (and thusly, disagreed with nuffink).