My suggestion creates less, virtually no, interaction between the narrow front radiator and the side firing unit. Tweeter in WG there on a separate board allows control of the incident angle at the listening position as well and does not "destroy" the main speaker.
I have no narrow dispersion speaker on hand otherwise I would have tried it already.
Drivers in a single box will always create some level of interaction. Furthermore I highly doubt that the side driver should deliver only content >1kHz. In my experience it should be full range.
Similar but please not so similar in price
I've been thinking about a DIY version for a while, but it's beyond my skills. One could also imagine such a 3 speaker setup with the sides toed out, as you mention. Best on all fronts?
For what ? For "envelopement" ? That still could potentially come sufficiently from the main spaker, depending on it's actual dispersion width right below 1K.
You mean a matrix decoded version? I suspect side sources would sound a lot more spacious than mid sources. Matrix decoded reproduction really doesn't add any realism, it "just" adds increased stability to localization.
By the way, there will be a software matrix decoder available soon. Please see:
AcourateConvolver - AudioVero
This is what it will look like:
Attachments
Last edited:
Power response will be off with a side firing mid/high. There's also spatial information <1kHz.
It is more off from constant without it I'd say
What is the optimal power response?
What is the optimal magnitude response of the first spaciousness effective reflections?
What is the optimal D/R ratio?
This determines the appropriate solution.
Sure the pattern in the figure is a sketch. The actual number of lobes depends heavily on the frequency. Onmi with one lobe at low freqs, and unnumerous amount of lobes at the treble.
The frequency dependent lobing is the key to achieve room reflections with low correlation.
So your conclusion is that room reflections should be spectrally as different as possible?
That is not the question here. You were suggesting a model with a high D/R for clarity and added reflections for spaciousness.
"The D/R ratio should be rather high."
"Reflections for added spaciousness and realism should come from only a few distinct locations in order to maintain clarity."
I guess the suggested model does that and naturally the power response of the combo will increase above the suggested cross-over and the directivity index will go down for the same matter.
Depends finally on the room, the setup and the desired D/R, which you did not specify. But is adjustable in the model. One can find it per remote control.
Per my understanding (from Klippel) D/R "is preferred" between critical distance and roughly twice the critical distance of the used speaker in a given room (-5dB). But I guess that is to much reflected sound for your taste and goal. So use the remote control.
You can't discusss the topic by omitting power response. As a measure it's in the same category as reverberation time. It does not tell you how the speaker-room interface has to look like but it tells you if you're in the ballpark or not.
I am not omitting it. But it is a result of the D/R YOU want to achieve. Higher D/R to begin with means higher DI means lower sound power.
I am happy with -5dB D/R.
Not necessarily should they be spectrally as different as possible, but they should be decorrelated in time and phase.
In technical measures there will be spectral comb filtering, but in perception it does not matter.
Power response is pretty useless parameter, as it is a feature of a speaker and not related to room interaction. For example you can get ideally constant power response only by adjusting the freq response to provide constant power output ! It does not indicate in which direction the energy goes etc.
Are we talking about the same thing? I was talking about overall power response.
But Elias, isn't that what exactly a dipole achieve, IF the rear wave is diffracted well? Let's imagine a LEDE room, but the other way around ,with the diffusers behind the speakers.
I believe decorrelation has to follow strict rules in order to create a realistic sound event because our hearing is based on pattern recogition.
Who said power response is the only criteria? Nevertheless power response is important. It definitely is room dependent.