Brandon, you used the phrase "before distortion" as the standard, but as you know, speakers exhibit substantial distortion at a point where the electronic components still have distortion levels far below audibility. For example, note that the M80 test at 95dB by the NRC showed that at 200 and 2000HZ distortion was beginning to appear, being about 40dB(about 1%)down. There don't appear to be any published tests of distortion at still higher levels or any definitive reports of the levels at which it becomes clearly audible. Certainly it would be higher than 95dB and much higher levels are possible before the speaker would suffer permanent damage. It can be noted that, driven by 1,000 watts, which has been exceeded in testing, the M80s would have an output of about 120dB, apparently without that permanent damage occurring.

Another point raised by your question is that hopefully those 100dB plus levels you mentioned were experienced only briefly, since generally it's cautioned that levels over about 85dB on a long-term basis(e.g. even when using power lawn mowers)make hearing protection advisable to prevent permanent damage.

Yeah the 105db levels I didnt listen to for long. Its really not all that comfortable I probably do listen to music louder than alot of folks do. At least those of us not in the hobby. Personally, the only system I can listen to comfortably at higher volumes is my own, Any car stereo, or boombox or some movie theaters are completely unbearable. I will get a headache and just rather have pure silence. And its not the db levels I think its just the harshness and really unbalanced nature of those kind of systems. And for the record I dont find my 80's "bright" at all.

Im thinking before distortion like Grossly distorted where you can easily tell the speakers are no longer within their comfort zone.

Yes, but how "gross" it would have to be before it was actually detectible when listening to music is still a question. For example, Axiom itself ran a distortion study a couple years ago which indicated distortion varying from about 2% at 10KHz to 100% at 20Hz before it became audible. The follow-up study mentioned there apparently hasn't been done yet.

I will attempt to condense this one down to the Coles notes (Cliff notes to some I believe) for the purpose of this post. But we should think about doing a mini series in our newsletters about it some time. There are many different types of distortion and they are all audible at different levels down from the fundamental depending on frequency and type of distortion. For instance harmonically related distortion is very hard to detect if it is the second harmonic but gets progressively easier as you head towards the fifth harmonic at which time it is as audible as straight noise. Low frequency noise distortion is very hard to hear where as high frequency is easy. Getting to the original question here; speakers, if properly designed and this is a big if out there, can have very low audible distortion at high power. They will exhibit compression which is comparable to just running out of volume or perhaps soft clipping. Inevitably this compression does not happen equally at all frequencies so you have a level below which there is no compression and then a level at which the compression would start to become worth noting which should be considered the max level the speaker can play to. A single M80 in an anechoic chamber starts to exhibit compression at about 118 dB. You can keep pumping more power in but you won’t be getting all frequencies back out in a linear fashion. The other very important fact to keep in mind is that even though 118 dB sounds well above any sort sane listening level; music and movies are dynamic. So if you want really clean reproduction at moderately high listening levels, or normal levels in a large room, you have to consider those peaks. This where lots of power and lots of speaker output without compression becomes heavenly.

JohnK,

I don't understand how to read that curve. For example, at 60Hz I see a peak of 69dB. What does this mean?

Mo, it's how far below the top curve representing the speaker output that the bottom curve representing measured distortion is. At the 60Hz frequency you mention the difference is roughly 21-22dB. Distortion down that amount is about 8-9%.

Ok, I get it. Thanks.

Wow! That means at the levels I listen to music (about 85 to 90dB), I experience about 1.5% THD max (room acoustics excluded)...and that's over a very narrow range of frequencies. Over the majority of the range, THD is about 0.5%. That's incredibly impressive actually. It would be even more impressive if it's all second harmonics .

I'd be curious to know how THD varies at lower SPLs.

Well, as to distortion at lower levels, the usual 90dB NRC graph is also shown and the distortion is low enough that at even lower dB levels it can be assumed that it wouldn't be audibly significant. Of course, note in the NRC measurements description that they point out that "90dB measured anechoically is very loud and considered far beyond normal listening levels..."

Incidentally, the formula for calculating the dB down distortion level is 20log(distortion percentage/100).