Wave cancellation!

[austi15]

I had some sibilance after I moved my speakers. Everything else was better (soundstage particularly, I moved the speakers farther apart), but the annoying peak around 9-10khz was driving me crazy on vocals (particularly vocals on trance songs, which often emphasize really high frequencies).

My speakers are stacked B&Ws, with the upper speakers bi-amped (HF driven with 2216, LF a Yamaha A-700). No adjustment of the treble could make it right. Wiring the speakers out of phase (upper speaker's HF or lower speakers overall) messed up the treble even more. I found that the wavelength of about 9-10khz was between 1.35" and about 1.5"

Pushed the upper speakers back 1.4 inches after a bit of experimenting and the sibilance was gone.

I think it was sort of targeting that range of frequencies, but it may have simply avoided other acoustic phenomena caused by the tweeters 3 inches apart and lined up exactly. Any other ideas on why this worked?

 

[toxcrusadr]

How did you know it was 9-10 KHz? Were you feeding it from a sig gen to find out?

But to answer your question, stacking spesakers is what causes these kinds of problems, or so I've read here at AK. By moving one of them back, you corrected for it, at least enough to make it tolerable. It happens precisely because the wavelengths are comparable to the difference in distance between the two tweeters and your ear, which causes comb filtering as the two wavefronts cancel. Search AK for 'comb filtering' and you'll find some interesting stuff. In particular there was a detailed test done by Zilch using stacked Advents, I believe.

 

[toxcrusadr]

Ah, here we are.

http://www.jblpro.com/pub/technote/fat_wht.pdf

See Page 3 for a nice graphical illustration of comb filtering.

 

[austi15]

How did you know it was 9-10 KHz? Were you feeding it from a sig gen to find out?

But to answer your question, stacking spesakers is what causes these kinds of problems, or so I've read here at AK. By moving one of them back, you corrected for it, at least enough to make it tolerable. It happens precisely because the wavelengths are comparable to the difference in distance between the two tweeters and your ear, which causes comb filtering as the two wavefronts cancel. Search AK for 'comb filtering' and you'll find some interesting stuff. In particular there was a detailed test done by Zilch using stacked Advents, I believe.

It wasn't the most accurate method (lol), but I fed it through my ADC Model Twenty eq, boosting and dropping different frequencies until I had roughly identified that the obnoxious sound was coming from that range. Using a wavelength calculator I found the wavelength and set it back the proper amount.

Ah, here we are.

http://www.jblpro.com/pub/technote/fat_wht.pdf

See Page 3 for a nice graphical illustration of comb filtering.

Thanks! So the reinforcement occurring at that range was the issue I was hearing. Could I use this to my advantage? The article outlines different ways of avoiding problems with it and helping response.

Initially this was a broad question, but now that it is focused solely on speakers perhaps it would be better to have this discussion in the speaker forum. I'm not sure.