Home High Efficiency Speaker Asylum

Need speakers that can rock with just one watt? You found da place.

Re: Nice ending indeed

Hi Flyingsod,

Okay, on the radiation pattern thing:

Assuming you're in a non-anechoic room and sitting more than about 5 feet away from the speakers, most of the sound that reaches your ears is reflected rather than direct sound. The farther back you are, the more the reverberant sound dominates. We get directional cues primarily from the direct sound, but the reverberant sound contributes to loudness, timbre, and a sense of spaciousness.

So if you accept that the reverberant field matters (and not everyone accepts this), it would be desirable to "get it right". In my opinion "getting it right" has two parts.

First, we want to delay as much as possible the arrival of the first reflections, as early-arriving reflections are more likely to be detrimental to imaging or timbre or both. Typical first reflection zones are floor and ceiling bounce, and a bounce off of each side wall. We'd like a path length difference of at least 10 feet between the direct and reflected sound for those first reflections, and we'd like to have a time delay of 10 milliseconds or more before the onset of reflections. I like to use diffusion at the early reflection zones when it's practical.

Second, when the reverberant sound does arrive, we want it to be spectrally correct (not have a significantly different tonal balance than the direct sound) and be highly diffuse (which prevents image skewing and promotes spaciousness). We'd also like for the reverberant field to decay slowly (which means we don't want the room to be overdamped).

Now some of this is obviously room-related, and some is loudspeaker-related. A speaker with a well-controlled radiation pattern can be "aimed" to minimize early sidewall reflections. I like to criss-cross the radiation patterns slightly in front of the listening position, as this not only minimizes that first sidewall reflection but also gives a wider sweet spot. Let me digress here for a moment.

The ear localizes sound by two mechanisms: Arrival time and intensity. If the speakers are pointed straight ahead or only toed in a little, the image is pulled to one side pretty severely for an off-centerline listener because the same speaker wins both arrival time and intensity - the latter because he's moving more on-axis of that nearer speaker. Now if we criss-cross the speakers' axes in front of the listening position, an off-centerline listener still gets pretty good soundstaging because one speaker wins arrival time but the other speaker wins intensity, since he's now more on-axis of that farther speaker. With a well-controlled radiation pattern, this extreme toe-in also virtually eliminates that early same-sidewall reflection.

Okay the next thing a fairly uniform radiation pattern does is maintain good spectral balance in the reverberant field. Most speakers have poor spectral balance in the reverberant field, so the perceived tonal balance is skewed accordingly. Below is a link to SoundStage measurements of a 6.5" two-way. Its response is amazingly flat on-axis, but off-axis we see a dip where the woofer starts beaming and then a broad peak at the lower end of the tweeter's range because the tweeter has such a wide radiation pattern. The problem area is that lower treble region (3-5 kHz) where the tweeter is putting out a lot of extra energy into the reverberant field. This is right where the ear is most sensitive, so there's a pretty good chance this speaker would sound a bit bright and forward despite its excellent on-axis performance.

The ear is constantantly analyzing incoming sounds to see if they are reflections of a recent signal or brand-new signals. If it's a reflection, then the directional cues are ignored - this is how we tell the direction of a sound source in a reverberant environment. The ear makes this comparison based on spectral content. So if the spectral content of a reflection is way off, the brain has to work a bit harder to correctly classify it as a reflection. I believe that a long-term result of this "increased CPU usage" can be listening fatigue.

So in my opinion radiation pattern matters for natural timbre, sweet spot width, and reducing listening fatigue.

Duke


This post is made possible by the generous support of people like you and our sponsors:
  Kimber Kable  


Follow Ups Full Thread
Follow Ups


You can not post to an archived thread.