wattage

I agree with Onhwy61's comments. I'll add the following:

First, the 86 and 98 db numbers you cited are presumably at 1 meter. Neglecting room effects, the SPL of dynamic (box-type) speakers will fall off at about 6 db per doubling of distance, meaning a reduction of perhaps 10 db or more at typical listening distances. (The rate of falloff will be significantly less than that for planar speakers, though). Add 3 db back, however, to take into account that two speakers are being driven.

Second, keep in mind that speaker sensitivities are often specified on the basis of an input of 2.83 volts, rather than 1 watt. For an 8 ohm speaker impedance, that makes no difference, since 2.83 volts into 8 ohms corresponds to 1 watt. But 2.83 volts into 4 ohms corresponds to 2 watts, so subtract 3 db to get efficiency on a per watt basis if the speaker has a nominal impedance of 4 ohms and is spec'd on the basis of 2.83 volts.

Third, as can be seen from measurements provided in conjunction with reviews in Stereophile and elsewhere, it is not uncommon for speaker sensitivity and efficiency specs to be optimistic by 2 or 3 db or more.

Fourth, and perhaps most importantly, while the doubled power capability you referred to will provide for transient peaks on many recordings having narrow dynamic range, keep in mind that SOME recordings will have enormously greater differences in volume between the peaks of brief transients and their average volume. One common example being well recorded minimally compressed classical symphonic music. I have more than a few such recordings in my collection which when listened to at average levels of about 75 db at my listening position reach brief peaks of 100 to 105 db at my listening position. A 30 db difference (105-75) between the volume of brief peaks and average volume means that 1,000 times as much power is required for those brief peaks as for that average level.

Finally, see my post here for a description of the detailed calculations that are involved.

Regards,
-- Al

P.S: I hope you won't mind if I point out that it's "strictly," not "strickly." A common mistake.

1 watt at 86 db,2 watts for 89db,4 watts for 92, 8 watts for 95, 16 for 98db, 32 watts for 101, 64 watts for 104 and so on. So yes,40 is usually sufficient but keep in mind that thx is 110db and ultra thx is 112db.

Extra wattage is always a good thing with transient peaks (symbol crashed or bass drums).
Finally, you should have a "perceived" doubling of volume for every 10db increase.

All I can say is 'quality over quantity'. In my experience most audiophiles buy a lot more power than what is really needed. Spend the money on what makes your system sound 'good' as opposed to what makes it sound 'loud'.

Almarg, would you please elaborate on why SPL does not decrease at 6 db per doubling of distance on planar speakers? I thought that figure was a constant for propagation in Earth's atmosphere rather than being related to the type of device producing the sound?

77jovian, the main reason that SPL decreases as distance increases, in a reasonable home listening environment at least, is not related to atmospheric attenuation. It is the result of the sonic energy "spreading out" over a progressively larger cross-sectional area as distance increases, resulting in a given cross-sectional area receiving a progressively smaller fraction of the total energy.

In contrast to a typical box speaker, though, where a given frequency is likely to be emitted by one or two or three relatively small drivers, a large panel speaker will tend to "beam" most frequencies, resulting in a lesser degree of "spreading" as distance increases. Or putting it another way, as the listening distance increases the listener will tend to be "on axis" with a greater amount of the panel's area. And also more equidistant to the various sections of the panel's area, resulting in less difference in the arrival times of the sound emitted by the various sections, resulting in more coherent summing of the different arrivals.

Regards,
-- Al