I was discussing linear amplifiers being linear on another board and one of the participants mentioned that amps sound different at different power levels. He wasn't talking about the linearity of the amp, but the response of the human ear. He then posted the following link;

Equal loudness curves:
http://hyperphysics.phy-astr.gsu.edu/hbase/sound/eqloud.html

I think that we all realized that the ears response changes as a function of frequency, but, I did not realize that it also changed as a function of sound intensity. Look at the curves at both the top and bottom of the audio spectrum.

I suspect that these are generalized curves and everyones may be a little bit different. Possibly major differences occur if there is some hearing impairment.

Wow, something else to think about when discussing audio equipment.

Link: http://hyperphysics.phy-astr.gsu.edu/hbase/sound/eqloud.html

So what does all of this mean? That a speaker will sound different driven at 80 dB vs 100dB? That the frequency curve will actually look different?

I guess this is even more ammunition to support dialing in your system and critically listening at reference levels - whatever that may be.

I don't think it needs discussing relative to audio equipment, except for the fact the the ear's 'response' becomes more linear as the 'input' gets louder. The rest of the natural and unnatural world does not tailor its sound to the response curve of our ears, so tweaking an amp or speakers to do so would sound unnatural to us.

Well, folks, what struck me when I reviewed those charts is what about those A/B speaker tests? You know, the ones where you have a simple switch and first listen to one speaker and then the other speaker. If there is a significant difference in the speaker efficiency and it is not compensated, then there may be a difference induced by the ear.

Same goes for auditioning other audio gear. If the sound levels are not adjusted, what happens to the results?

In any case, what we have is another complex factor complicating an already complex subject.

That's why it's important to level match the speakers to a reference.

Yes, I think that you are correct. However, I do not recall that the tests that I remember did that referencing. Guess that I will have to re-read the articles.

Bernard, a basic requirement for a blind listening test of electronic equipment is that the levels be matched to within 0.1dB. The reason for this is that in some tests using test tones an overall difference of as little as 0.2dB has on occasion been detectable, not as a loudness difference as such, but possibly perceived as more clarity, better imaging, etc. This also explains some differences reported in casual non-blind listening where the listening levels weren't carefully matched.

John: I'm surprised that as little as 0.1 dB would make a difference. The shape of those curves in the charts, referenced above, did not suggest to me that it was that critical.

I am now beginning to wonder how many of these listening tests are actually calibrated that well? That is something for me to look for next time I read a comparison article.

Thanks.

It even lends more credence to the rumor that inexpensive receivers have a non-linear vol control giving lots more volume in the early stages of travel.

So that when the "Great Unwashed" while standing there twiddeling controls at the local B&M can get a good impression of the very "Powerfull" amp that gets so loud so fast and outdoes all the others!

Must really, really be loud at max!!

Hey,....a good marketing gimmic!!

LOL!!! I have not heard of that before! But, if actually true, I would not be terribly surprised.