Columns Retired Columns & Blogs |
Loudspeakers Amplification | Digital Sources Analog Sources Featured | Accessories Music |
Columns Retired Columns & Blogs |
Loudspeakers Amplification Digital Sources | Analog Sources Accessories Featured | Music Columns Retired Columns | Show Reports | Features Latest News Community | Resources Subscriptions |
You can't calculate RT60 with any accuracy in the presence of resonant modes. Your measurements are only accurate from the lower midrange upward.
John Atkinson
Editor, Stereophile
> Equalization can correct room mode peaks and nulls but what about the resonance times? Are the measurements accurate? Is this a real problem? <
First, EQ cannot really correct for peaks because the peak levels change all over the room. Over a span of only four inches the response can change significantly, even at very low frequencies.
EQ cannot practically correct for nulls either, both because of the response changes over small distances as noted above, and also because raising a null requires huge amounts of amplifier power. I measure nulls 30 dB deep or even more all the time. To raise a 30 dB null to be only 20 dB deep requires a huge amount of amplification. That strains the power amps (if they can handle it at all), and also increases loudspeaker distortion significantly. And you still have a 20 dB null.
As John explained, RT60 is mostly irrelevant in small rooms such as found in most homes. The decay time below around 300 Hz is dominated entirely by individual room resonances. Even at higher frequencies RT60 is mostly irrelevant. Yes, there are individual reflections whose decay times can be measured. But this is not true reverb that first swells and then decays.
All of that said the RT60 display in ETF is not accurate anyway. I've been working on this with ETF author Doug Plumb for a while, and Doug tells me he has no plans to fix RT60 in ETF. Now that he has the newer R+D program, he's offering that for free to all ETF owners in an attempt to move them over to this newer and better program. RT60 is still a little "touchy" in R+D, but it does mostly work now.
--Ethan
> what about the resonance times? <
(I failed to address that fully above.)
This is another failing of trying to use an EQ to correct room acoustic problems. Your observation is correct - room mode decay times are at least as big a problem as the peaks and nulls themselves, and these resonances are the main contributor to the effect known as "one note bass."
--Ethan
Thank you both for your replies. They were very helpful.
Ethan, even though you acknowleged the effect of low frequency reverberation times caused by room modes any ideas about fixes? And if ETF doesn't do a good job of measuring any suggestions on how to check? If the RT60 plot doesn't do a good job, what about looking at the 3-D plots to identify long reverberation times or do these calculations have the same inaccuracies?
Sorry for the delay. I usually look in every day, but I've been swamped lately.
> if ETF doesn't do a good job of measuring any suggestions on how to check? <
The important metric here is modal ringing, and ETF does that very well. Like RT60, this is also a way to assess decay times, but it displays the decay times at the specific frequencies that actually matter.
--Ethan
I really appriciate the help, but as a newby I need some hand holding. Modal Ringing? Don't see that listed as a plot in ETF or don't recall reading that in the Master Handbook of Acoustics. Besides those two sources and CARA that is the limit of my knowledge on the subject of acoustics. I would assume that it would refer to a extended decay time that repeats itself at multiple frequencies based on room modes.
So anyway, what is it and what do I do with it?
> Modal Ringing? <
That's what you see in ETF's low frequency waterfall plot, as shown below. Each "mountain" along the rear wall is a peak in the response, and the decay times (the mountains come forward over time) show that the peaks are due to room modes. This is an important distinction because not all peaks are caused by room modes. Some are the result of comb filtering, which happens any time sound reflects off a boundary back into the oncoming wave. If the reflected wave is out of phase with the source you get a null, and when they're in phase you get a peak.
BTW, don't be shy about asking for hand-holding. That's what I'm here for.
--Ethan
So, both room modes and comb filtering will display as a periodic event, but the room modes will have a long decay and the comb filtering will not? Am I even close or am I just throwing a bunch of terms out there that make no sense?
Close. Both room modes and comb filtering will yield peaks and nulls in the response. But only those peaks that correspond to room modes will sustain and artificially extend the decay time.
This is a major cause of the effect known as "one note bass," where all bass notes sound more or less the same. The extended decay also makes the peaks sound louder than a raw response reading would indicate. This is because the sound sustains for longer and thus has more total energy at those frequencies.
--Ethan
One note bass, no highs no lows, must be BLOSE