[fight4yu]

Hi. I am using REW v5.0, UCA202, and regular RS SPL digital meter. i am using the hometheatershack correction files.

Attached are the graph for 3 settings I use:
First setting is with 75dB calibrated, using level -10dB/FS
Second setting is with 85dB calibrated, using level -10dB/FS
Third setting is with 85dB calibrated, using level -3dB/FS

What I cannot understand is why my lower end (10-20Hz) looks like this.. I think it started to roll off pretty quickly at 30Hz or so, and it should KEEP decreasing. Why is the "pick-up"?
It seems that REW always start at what I calibrated at...? In setting1, it starts at 75dB, and in setting2, it starts at 85dB..
I am puzzled by this behavior. Anyone can help me understand, and if I did anything wrong?
On the higher end, there is no issue... It keep dropping to 30dB after the xover.. .So, the meter can definitely detect less than 75dB...

 

[Wayne A. Pflughaupt]

What I cannot understand is why my lower end (10-20Hz) looks like this.. I think it started to roll off pretty quickly at 30Hz or so, and it should KEEP decreasing. Why is the "pick-up"?

It’s nothing, really. Seriously, it’s nothing. It’s generated by the calibration file when the meter doesn’t “see” an signal that low. So, you can ignore it.

Regards,
Wayne

 

[fight4yu]

Thanks Wayne.
However, even if I "clear" the calibration file, the basic shape of the graph still looks like above. I can only make it "looks normal" by unchecking C weighted SPL... I vaguely understand the weighting is to do with compensation.. but I don't seem to understand well what it is compensated for and why... especially if there is nothing, why compensate?

And at what point can I determine that it does not see any signal? Is that the point when I see my graph make an infection point?

Thanks Wayne!

 

[laser188139]

... And at what point can I determine that it does not see any signal? Is that the point when I see my graph make an infection point? ...

As you go to lower frequencies, you start to see more manufacturing variation between microphones. They don't all have exactly the same response that low. To the extent your particular microphone differs from the one used to build the generic RS calibration file, or the built-in C-weighting when you cleared the calibration, you are going to see odd results at the low end. If you look at the generic RS calibration file, you will see it is adding a large value at the low end, 37dB in mine, so a very weak measurement can still appear as a high value.

Wayne did not precisely mean no signal at all, just no signal above what it is being measured by the meter as background noise at that frequency. So where you see the curve turning upward, you are seeing the calibration file/C-weighting compensation being added to a probably flat background signal. In your case, I would ignore everything below the inflection point at 20Hz. If you want accurate measurements down to 10Hz, you need to buy a calibrated microphone, i.e., a microphone with its own individual calibration file. As your measurement has already fallen off quite a lot by the time you get to 20Hz, I don't expect there is anything below there of interest to you right now.

Bill