SPL Calibration value - 75dB (too low?) for RT60 meas. in a classroom

[sergioTOOL]

Hi

I still haven´t used the REW, but planning on using it fairly soon.
The help files point out that a SPL calibration of 75dB must be made at the MIC position using white noise. Once the calibration has been made, can the level of the sweep signal be increased?

I´m going to measure RT60 (ISO 3382) and freq response in a classroom of about 9mx9mx4m. For this I´ll need a fairly big loudspeaker to reproduce a loud sweep signal to cover the size of the room, and 75dB at the receiver point is certainly not loud enough for RT60.

Could you help me out on this one? Would you recommend I use REW´s sweep signal method to measure RT60 in this room? Can I adjust the levels, whilst making sure no clipping takes place. Should I repeat the sweep signal several times per measurement to probably obtain better results?

Many Thanks for your help.

SERGIO

[brucek]

You're a bit off the mark in some of your assumptions.

When REW takes a measurement, it results in an impulse response. All the graphs and plots are derived from the single impulse response (including RT60).

RT60 is not really meaningful below ~200Hz (use waterfall and decay for that band), and should be only considered in a large room (that you obviously have).

You should place your microphone in an area, or multiple areas where you want to take the RT60 reading. You will do a normal response measurement at 75dB at that listening position. You will then apply the filtering and decay variables on the impulse page and calculate the RT60. Then you can examine the results on the RT60 tab.

You really need to play with REW in situation a bit to get a feel for how it works. You'll drive yourself crazy with questions until you get your hands wet .......

And yeah, you can change any level you wish in REW.

brucek

[sergioTOOL]

Thanks for the quick reply!!!
Will begin with some trial measurements tonight, hopefully I´ll get the feeling for the program.
One other thing about the levels which I didn´t understand. You wrote that when carrying out the measurement, levels at the mic had to be 75dB. However at then end of your reply you state that they can be changed (in my case increased to cover the room´s large volume).

thanks again

sergio

[brucek]

You wrote that when carrying out the measurement, levels at the mic had to be 75dB. However at then end of your reply you state that they can be changed (in my case increased to cover the room´s large volume).

I simply meant that the levels are all adjustable to get a proper 75dBSPL at the listening/mic position.

There's no advantage to measuring higher or lower than this.

brucek

[sergioTOOL]

Sorry to be so stuborn about the 75dB SPL at the mic position... but isn´t it too low to measure RT60. Energy will not built up sufficiently in the room with that sort of level (doing an analogy with impulse or noise measurements).
Im not familiar with sweep signals that´s why im a bit wary about the level here.
What if I measure say 95dB at the mic position and then set the calibrationb value to those 95dB. Will I get different readings? (i.e. different RT60s).

thanks

[JohnM]

You shouldn't see any difference in the RT60 values unless you drive some part of the system into heavy distortion. Remember that the RT60 calculation is based on at most 30dB of decay of the impulse response, unless your ambient noise level is high you should get good results at 75dB. Sweep measurements are not like using a pistol or cannon and capturing the impulse directly, the effective energy fed into the room is the accumulation of the 75dB level throughout the duration of the sweep so very good signal-to-noise ratios can be achieved at modest listening levels even with fairly short sweeps, for better signal-to-noise increase the sweep length rather than the level and/or use the multiple sweep averaging feature.