What does the -/+ 3 for speakers measure?

[DubPlate]

I though I understood what -/+3 meant, but while reading a review for a sub (SVS SB 2000), the commentator stated something to the effect that: the output below 30hz drops off by about 10db. However the sub is rated at -3 at 19 hz.

How can that be? If the output drops off at 30hz by 10db, shouldnt 30hz or higher comprise the -3 spec?

Also at what volume are the speakers typically measured? Reference or max?

Thanks in advance.

 

[William Lemmerhirt]

Basically that measures amplitude across a speakers passband. The less variation, the better it sounds. Theoretically...
The sub in question, I believe is spec’d at -3db “typical in room response”. The actual measurements were probably done anechoically or close mic’d etc. Without seeing the article I’m only guessing here. The sb2k(which I have experience with) is a small sealed sub and like most sealed subs roll off higher and usually take into account some room gain. That is sometimes overstated as room gain is room size/dim dependent and takes a lot of power and excursion to be beneficial.
If there’s a graph, it will likely have the dB level of the sweep.

 

[shadyJ]

SVS specs their subs from groundplane measurements. In a +/- 3 dB window, the '-3 dB' point can not be lower than 6 dB from the highest peak in the response of that frequency band window.

However, different measurement techniques can measure different responses. Close-mic measuring will give you a different response than groundplane quasi-anechoic. There is also free-air quasi-anechoic which can also produce a different response. None of these are incorrect responses, they are all just products of their individual measuring technique. None of those are less valid than the others. The commentator could simply have measured a different response than what SVS measured due to a different measurement technique.

Most subwoofer testers aim for a 90 dB base line response. Josh Ricci and myself use 90 dB at 50 Hz as a reference point. Maximum output will usually give you a substantially different response than nominal output levels. It is only a good metric if you want to see how much headroom the sub can give you.

 

[DubPlate]

SVS specs their subs from groundplane measurements. In a +/- 3 dB window, the '-3 dB' point can not be lower than 6 dB from the highest peak in the response of that frequency band window.

However, different measurement techniques can measure different responses. Close-mic measuring will give you a different response than groundplane quasi-anechoic. There is also free-air quasi-anechoic which can also produce a different response. None of these are incorrect responses, they are all just products of their individual measuring technique. None of those are less valid than the others. The commentator could simply have measured a different response than what SVS measured due to a different measurement technique.

Most subwoofer testers aim for a 90 dB base line response. Josh Ricci and myself use 90 dB at 50 Hz as a reference point. Maximum output will usually give you a substantially different response than nominal output levels. It is only a good metric if you want to see how much headroom the sub can give you.

Got it