SPL Max vs. SPL sensitivity

[Samwise]

Hello All,

 

I am looking at two subwoofers:

 

One brand (d&b audiotechnik) gives its SPL as a max figure - 139 dB @ 1 Meter (Source - Pink noise with crest factor 4) - Power handling 600W RMS 2x18" drivers 37 - 90 Hz

 

The other (Logic Systems) gives its SPL as a sensitivity figure of 102 dB @ 2.83V - 1M (Source - Band limited pink noise) - Power handling 1000W RMS 2x15" drivers 40 - 150 Hz

 

Is there a way to convert one of these SPL figures into the same type or this a silly question and I'm missing something? Which is the most efficient? Sorry too if this is old ground but I had a look on the forum search and couldn't spot anything.

 

Regards, Sam.

[nb705]

Hello All,

 

I am looking at two subwoofers:

 

One brand (d&b audiotechnik) gives its SPL as a max figure - 139 dB @ 1 Meter (Source - Pink noise with crest factor 4) - Power handling 600W RMS 2x18" drivers 37 - 90 Hz

 

The other (Logic Systems) gives its SPL as a sensitivity figure of 102 dB @ 2.83V - 1M (Source - Band limited pink noise) - Power handling 1000W RMS 2x15" drivers 40 - 150 Hz

 

Is there a way to convert one of these SPL figures into the same type or this a silly question and I'm missing something? Which is the most efficient? Sorry too if this is old ground but I had a look on the forum search and couldn't spot anything.

 

Regards, Sam.

 

I would say it's not an entirely Apples to Apples comparison. But FYI this is how you calculate max RMS power handling from sensitivity.

 

P = V^2/R (assuming the speaker is an 8 ohm nominal load for this) shows that the speaker has a sensitivity of 102dB 1w 1m

 

1000w is 30dB more power (1E3) than 1w

 

Simple addition gives us a max of 132dB

 

What I've just done is extraordinarily inaccurate and vague. But it gives a basic grounding in the kind of maths you should use to avoid marketing pitfalls.

[Simon Lewis]

To calculate maximum sound pressure level from the sensitivity figure and power handling figure, use:

 

SPL = ref sens + 10log x power handling

 

For the logic loudspeaker:

102 + 10log 1000 = 132dB

 

for the d&b loudspeaker, we are told its max SPL. If we work backwards and say that 10log x power handling (10log600) = 27.78 and subtract from 139 we arrive at 111.22 dB @ 1W @1m. This appears to be rather high (it's the sort of figure seen with compression drivers or in multiple coupled loudspeakers). However, the SPL figure is described as "max" and the test signal is pink noise with a crest factor (or peak to mean ratio) of 4. If we take 10log (peak/mean) we see that a crest factor of 4 equates to 6dB. If we now subtract 6dB from our calculated sensitivity (so we are using a 'mean' power rating), we arrive at a figure of 105dB. This seem slightly high, but not impossible.

 

The next problem is that the test bandwidth for the two units is not the same. Some standards call for filtered pink noise between 125Hz and 8kHz, others for filtered pink noise over a decade in the device's passband. The narrower the passband, usually the more efficient it can be, so this may explain something of the extra performance the d&b appears to have.

 

Another issue is that the thermal management of the two units may be different, and the effect of heat on the voice coil is to increase its temperature - and therefore its resistivity - thus causing less power to be "drawn" from the amplifier). Overall, the harder the speaker is pushed, the less output it delivers! This is termed power compression.

 

On paper then, the d&b seems to give very slightly more output for less input power. However, this may not be the whole story - we do not have any further information about power compression, comparative power bandwidth, distortion products, loudspeaker loading, impedance curves etc., etc.

 

As usual, the best advice (still bearing in mind what the paper based comparison suggests) is to listen to both - preferably side by side!

 

Simon

[Simon Lewis]

Just for accuracy, I should have typed SPL = ref sensitivity + 10log power handling. Fingers and brain need to co-operate better!

[Mark Payne]

Just for accuracy, I should have typed SPL = ref sensitivity + 10log power handling. Fingers and brain need to co-operate better!

 

Brilliant stuff from Simon as always.

 

I have been doing a lot of measurement of subs and LF contour as we work with L-Acoustics on the KARA and SB18 pilot programme.

 

I am more than ever convinced that SPL max and Sensitivity (worked forwards or backwards) tells you not a lot! What is harder to quantify (and more important) is actual transient response and the behaviour of the amp/sub system when it goes to limit.

 

We tend to drive our sub systems to the edge so what happens at the edge in terms of cone excursion and thermal protection limits within DSP/Amp is IMHO more important that the Sensitivity or Peak SPL on paper. Way more important.

 

You are going to have to listen to them as you rag the system! Its the only way!

 

M

[Samwise]

Thanks guys,

 

After a quick math revision that all makes sense. I have of course heard the d&b subs (B2's) in action in action and been moved by them. I'm looking at trading in some gear for a pair of the Logic Systems subs. I'm sure I'll hear some in action at one of the festivals this summer!

 

Sam.

[Dave SA]

We've got some D&B B2's at the unit Sam. When you come up to grab the Logic bins you can have a play with them if you like!