while a speaker can score high spl at a single frequency it's meaningless....without bandwidth and tolerance there's no way to categorize it as listenable or even usable without response correction....
unless your simply interested in a contest about sensitivity.
unless your simply interested in a contest about sensitivity.
Last edited:
To continue with mbrenna's and turk182's comments, SPL at 1 meter sensitivity needs to be specified as either @ a given voltage, or as @ 1Watt.
Both are common forms of specifying sensitivity.
Voltages used for the @1W form, are 2V for 4 ohms nominal impedance; 2.83V for 8 ohms, and 4 V for 16 ohms.
The @ 1W form can be directly converted to efficiency. https://sengpielaudio.com/calculator-efficiency.htm
The @ a given voltage form cannot, unless nominal impedance is also provided.
Since nominal impedance is often a bit hokey, often coming from an eyeball assessment of the impedance curve,
simply using 2.83V for all speakers, as well as showing the full impedance curve is best practice, imo.
Ok, I put up that info (which no doubt many of you know quite well)...so I can ask...
How are you getting your sensitivity spec?
If store bought, from a manufactures spec sheet? Their single number spec?, or assessing the response curve (if provided) on your own?
Or your measurements?
If DIY, going off driver specs? Or measured?
If measured, at a set frequency? Or an average though the entire frequency range?
The entire frequency range is easy enough for a passive speaker, but how are folks measuring sensitivity for active muti-ways?
Each muti-way section by itself is straightforward to measure, but how do you aggregate the sections to determine an overall sensitivity?
Both are common forms of specifying sensitivity.
Voltages used for the @1W form, are 2V for 4 ohms nominal impedance; 2.83V for 8 ohms, and 4 V for 16 ohms.
The @ 1W form can be directly converted to efficiency. https://sengpielaudio.com/calculator-efficiency.htm
The @ a given voltage form cannot, unless nominal impedance is also provided.
Since nominal impedance is often a bit hokey, often coming from an eyeball assessment of the impedance curve,
simply using 2.83V for all speakers, as well as showing the full impedance curve is best practice, imo.
Ok, I put up that info (which no doubt many of you know quite well)...so I can ask...
How are you getting your sensitivity spec?
If store bought, from a manufactures spec sheet? Their single number spec?, or assessing the response curve (if provided) on your own?
Or your measurements?
If DIY, going off driver specs? Or measured?
If measured, at a set frequency? Or an average though the entire frequency range?
The entire frequency range is easy enough for a passive speaker, but how are folks measuring sensitivity for active muti-ways?
Each muti-way section by itself is straightforward to measure, but how do you aggregate the sections to determine an overall sensitivity?
A good example is the Visaton B200, rated at 95 dB — the driver has a significantly rising response — they chose to pick the sensitivity number higher up in the curve. In reak=lity 89-90 dB is more accurate. The FR issue (and the HF laser-like dispersion) can be largely fixed with phase plugs.
dave
dave
the best way I know to most accurately measure sensitivity, is with all crossovers, EQ's in place, use pink noise as the stimulus and measure its average voltage over a time period, while simultaneously averaging SPL produced (Z flat). Then do a little math to bring it to @2.83V, or 1W, your choice.
The technique can measure a true average impedance (as opposed to conventional nominal values) , by averaging current over the same period.
A Umik-1 and trueRMS meter will work fine, provided the meter has averaging capability. REW's meter averages via Leq-Z.
I use the technique all the time. On my current 5-ways, 2.83V sensitivity per section is: sub 100dB, low 100, mid 103, HF 104, and VHF 102dB.
The first three sections are all 4 ohm, so their 1W spec decreases -3dB. And the HF and VHF are 16 ohm compression driver, so theirs goes +3dB.
Which of course means for level frequency response, voltages are nowhere near equal....
Which also means how do you average what's going on? I really don't have a firm technique yet, for determining overall sensitivity of multi-way actives.
Maybe it's simply an irrelevant concept for an overall spec....
84dB here. On a related note, max SPL is not sensitivity driven but purely swept volume. Sensitivity just drives how much power is required to achieve max volume.
Not that I use the figure for anything, really, other than having a ballpark idea of where I am compared to other builds, but the way I "decide" the sensitivity of the speakers I make is to average between the two frequencies where the FR is mostly flat, i.e. somewhere above the start of the bass roll-off and somewhere below any significant top octave ripples and/or roll-off. Edit: and I always go with X dB @1 m / 2.83 V.
Last edited:
It depends, I’ve got lots of speakers. I’d guess 85 on the low end and 97 for the high end. Just ballpark non engineering wags. I listen to my surround system mostly and that‘s around 91, except the sub which is at 88.