S15: Econowave DSP - a Constant Directivity vs Dipole study

[dantheman]

I'd bet most people listen to music on their iPod or car stereo the most and that is their reference then they go to the concert. I have a hard time going to live concerts of the music I like b/c the sound is often bad--same goes for the performance. Ouch. Live music has other appeals for me however.

Dan

[dantheman]

Originally Posted by john k...

No offense taken, the Note is what it is and I tried to document it as thoroughly as possible so that thoose who build it will not suffer any surprises. I'd be happy to provide the in room response at different positions but unfortunately I don't have the Notes set up at the moment. Perhaps the next time I have them set up I will make the measurements and add them to my site.

I'm glad you didn't take offense. I'd say your documentation may be the best around--at least it's in the same league as the best around and the end user really shouldn't have surprises. The in your room response doesn't mean a whole lot to me. Believe you me, I'm not saying your speakers measure poorly or anything, the measurements just suggest how it should best be used. More vertical polar data would be more useful as far as I'm concerned. The "in room" above the modal region has less usefulness than the anechoic stuff. Below the modal region I'll take in my room. I actually think that possibly extend that up to 500Hz or so.

Thanks,

Dan

[dewardh]

Originally Posted by dantheman

I'd bet most people listen to music on their iPod or car stereo the most and that is their reference then they go to the concert.

I'm sure it's true that more people listen to more music on the ubiqutous earbuds than at $100 a ticket "concerts" . . . but I'm not sure that means that the earbud sound then becomes their "reference" for what the music is supposed to sound like. If it is there is no "standard" by which loudspeakers or any other part of the reproduction chain can be judged, and "turn up the treble, turn up the bass, love them MP3s" is the new "high fidelity", and it's horns all the way, 'cause dude, they're *loud* . . .

[c2cthomas]

Originally Posted by dewardh

I'm sure it's true that more people listen to more music on the ubiqutous earbuds than at $100 a ticket "concerts" . . . but I'm not sure that means that the earbud sound then becomes their "reference" for what the music is supposed to sound like. If it is there is no "standard" by which loudspeakers or any other part of the reproduction chain can be judged, and "turn up the treble, turn up the bass, love them MP3s" is the new "high fidelity", and it's horns all the way, 'cause dude, they're *loud* . . .

Well - have a GOOD set of headphones can serve as a good yardstick by which to compare speakers to - and some earbud phones are highly regarded - YMMV.

The "'cause dude, they're loud" thing has always been around - and I don't think that segment of listeners will ever go away. Some do mature and refine their taste - hey we all start somewhere and it usually isn't at the top. It's our mission in life to overcome the hype that most consumers have been brainwashed with and introduce 'em to the "real world" of "real music". The sad truth is - most of them don't want to bother with it.

[gainphile]

Except for spatial distortion, headphones are more accurate method to listen. Even cheap ones. Remember there is no room or reflection to worry about. No horizontal/vertical response, no crossovers. Sometimes I check my speakers by comparing with headphones. For subjective tonal balance at least.

[dewardh]

Originally Posted by gainphile

Sometimes I check my speakers by comparing with headphones. For subjective tonal balance at least.

Oh yes . . . and the greatest virtue of headphones is their portability, since our auditory memory is, typically, terrible. They are not without their own problems, though, including the same suseptibility to the upstream chain as any other transducer. But they also have to be "callibrated" to . . . something. I do, occasionally, get to get to flip the phones off my ears and compare the sound in the hall (one place in the hall, anyway) with what's coming down the wire from the pair of NT1000 (which, of course, are capturing the sound in a different place from where I'm listening). Orchestral balances will be different, the direct/reverberant ratio will be (often very) different, but the "overal "tonal balance" tends to be preserved. I can then take the recording home and do the inverse . . . flip the phones on and off while comparing to the "in-room" sound of my loudspeakers, or any other speakers for that matter. And ultimately any other source, since having convinced myself that the headphones "get it right" in comparison to the original I can say that any speaker that doesn't match the headphone's "tonality" are getting it wrong, at least compared to the recording. It's better than doing it by "memory", anyway.

That doesn't work so well, though, for "studio" recording, or amplified electronic performances, because "what's on the wire" is *not* the same as "what's in the room". And even for "live acoustic" recordings the headphones don't transport much of anything about "image" or "spaciousness" or a whole raft of other semi-intangibles . . . that's all still left to some vague subjective sense of "what sounds right" . . .

[ScottG]

Originally Posted by bwaslo

I can't agree with that at all. I've been toeing in speakers for years and can say that without CD it works only a little. With cone and dome or wide panel speakers, one good seat might be widened to two almost-as-good seats, but after that the character of the two speakers change. Moving closer to one speaker off-axis doesn't resemble the sound of moving further from the other on-axis and any image falls apart. I spent my time endlessly fooling with the balance knob. Not so with CD waveguides.

Also can't agree with all of your list of weaknesses of CD/toe-in. No image outside the speakers? Maybe I haven't heard the speakers you have, but the CDs currently in use here do better on that than anything else I've used. And image size? In most live venues, as said, there really is NO image other than from the people sitting around you and room ambience. With close seating with quartets and small groups there is of course an image, but in recordings of those, the image size from the CDs seems pretty close to my ears. Not perfect, but compared to stats or minimonitors? No contest, except maybe from the line source planars (needing a bigger room). All this is subjective, of course, so no one much is likely to agree...

"Beaming" is relative. It really depends on the loudspeaker's dispersion pattern. As it happens, "full-range" drivers tend to have a horizontal polar distribution that often works well. Add-in your typical dome tweeter and it's usually no longer beaming, OR is only beaming in certain freq.s (usually higher than 5 kHz and perhaps at lower freq.s due to poor integration with the midrange's off-axis response). No equalization filter will improve this situation.


How far does an image extend beyond the speakers with the CD design you are used to? How far on other designs. Where do you sit relative to the loudspeakers (each design)? How close are each to walls in your listening room?

Generally the smaller the speaker (baffle included), the closer you should be to the loudspeaker. Move farther away and everything "shrinks" up. Move closer and you experience an "expansion". In this respect there are several advantages to a small source loudspeaker (..with a low diffraction baffle). Moving loudspeakers closer to you allows for:

1. increased image size due to an apparent intensity vs. sd.
2. increased sound stage due to a decrease in cross-correlation (particularly as a component of "head-shading").
3. a potentially increased sound stage and "external imaging" due to higher higher spl's off-axis (left of left speaker right of right). (again, a lower intensity for cross-correlation, though irrespective of "head-shading").
4. an effective increase in "head-room" (particularly for "point source" loudspeakers).
5. potentially a significant improvement in the reduction of apparent wall reflections (..assuming that when you move the loudspeakers closer to you they are moved further from room walls).
etc..

(..there are of course several *disadvantages* as well.)


BTW, if you have a loudspeaker similar to the Behringer (that Dan has), you can try all this out for yourself on a "like-kind" basis. In other words the Behringer is at least moderately small in size, yet still has a very uniform polar pattern that is very close to being "CD" (..in the range most people associate with a "CD" design).

[dantheman]

Deward, I'd actually say that it(the iPod, car and even their home stereo/boombox) is the standard that the concert sound is judged--and it typically fails by a mile. You can assume these are not classical music listeners--I'm talking about music that often regretfully makes money. What it's supposed to sound like is not a consideration. What it does sound like is. Typically the concert is not about the sound for most people. No one would go if it was. It is about the music, the band, the show and the atmosphere of the show. Lady Gaga has to be seen to be enjoyed. Ha ha Try not to quote me out of context if you reply.

Dan

[Elias]

Hello,

Thanks to Gainphile for the impulse data. Normalised wavelet responses this time For better comparison!

Here was wavelet plots for the dipole: S16 - Constant Directivity Dipoles

Again dipole responses normalised for left and right speakers:







Now for the CD speakers, left and right!






I still think the top end above 2-3kHz room reflections are lacking. Comparing to the polar plots (see below quote for CD speaker and above link for dipole), polar plot does not predict lacking signal! Room must eat it! Due to normalisation of the plots frequency response effects do not play a role here.


Let's see room response for a 150ms period, for dipole and for CD:






Top end reflections rappidly disappear.

Secondly we see dipole works best in the lower midrange compared to box speaker, can see faster decay below about 500Hz.


For 300ms window, dipole and CD:






Again frequency dependent decay of the room reflections.

Seems to be no matter how constant is the directivity the room will dominate the time-frequency pattern !

Instead of constant directivity, maybe the speaker directivity should be invert of room response decay? That turns out widening dispersion with frequency. Not many does that however! Obviously it's not easy, it goes all against the convenient physics of radiation principle. Anyhow the idea is interesting, don't you think so?


edit: And all of these are measured at the listening position, which was about 2.8 m according to Gainphile.

- Elias

Originally Posted by gainphile

I have since incorporated the waveguides to my dipoles with somewhat success.

...

I pulled out my fingers and built these.

[mige0]

Originally Posted by Elias

Normalised wavelet responses this time For better comparison!

Muuuch better


Agree with your conclusions - despite the tweeter range, room seem to dominate.

The horn of the CD seems to be better than the WG of the dipole (faster decay) - possibly less refections from the front wall is mixing in here too (good to see in the short period plots).

Michael