Dip in FR for Centre Speakers

[Hifisound]

For some of the centres a distinct dip is seen at ~900-1000Hz

KEF Q900 Speaker System HT Labs Measures | Sound & Vision
Polk LSiM707 Surround Speaker System HT Labs Measures | Sound & Vision
Revel Ultima2 Salon2 Speaker System HT Labs Measures | Sound & Vision
KEF R900 Home Theater Speaker System HT Labs Measures | Sound & Vision (slight one here)
Emotiva Airmotiv 5CH Speaker System Review Test Bench | Sound & Vision
https://www.soundandvision.com/content/wharfedale-jade-7-speaker-system-ht-labs-measures

Now it may not be always there , eg .
https://www.soundandvision.com/content/psb-imagine-x-speaker-system-review-test-bench

And this has more in addition too
https://www.soundandvision.com/content/psb-imagine-t3-speaker-system-review-test-bench

But in the cases above where its present, the fronts/L&R are amazingly flat
So find it hard to believe that its bad design/mistake
Is it intentional when designing a good horizontal centre channel ? given the way it placed ? Or some measurement issue when measuring a centre ?

An opposite example is that of KEF X300a
http://www.connect.de/testbericht/kef-x300a-im-test-1506924.html
Now whats that rise at ~900-1000Hz ? Esp when this is the active version of Q100 ?

 

[eriksquires]

I think the issue is most likely in the baffle step.

I took a look at my data for the LM-1 and LM-1C (center) and the woofers seem to suffer a lot more interference around that area (600-2kHz) in the center channel vs. the L/R versions, despite using the same model drivers.

Since a center channel is almost always used in HT, it's most likely the room correction software will take care of it pretty seamlessly.

It may be worthwhile to use a baffle step diffraction calculator such as The Edge to see what's going on.

Best,


E

 

[Hifisound]

I think the issue is most likely in the baffle step.

I took a look at my data for the LM-1 and LM-1C (center) and the woofers seem to suffer a lot more interference around that area (600-2kHz) in the center channel vs. the L/R versions, despite using the same model drivers.

Since a center channel is almost always used in HT, it's most likely the room correction software will take care of it pretty seamlessly.

It may be worthwhile to use a baffle step diffraction calculator such as The Edge to see what's going on.

Best,


E

Thanks, will dig more into this

Though interesting thing is, the one example I gave of centre not having an issue is a 2-way ! which known to be bad design (lobing) as compared to other 3-ways which have the dip

 

[eriksquires]

The LM-1C is a two way also, and while it did not have the same amount of problems your "bad" examples showed, they were definitely there.

You may want to go looking for D'Appolito 2-way designs and see if you discover the same patterns there.

Best,

E

 

[eriksquires]

Another possible effect here is the crossover points. A 3-way implies a midrange with a lower cutoff around 200-500 Hz and upper around 3-5kHz.

The LM-1C has a crossover point of around 2 kHz instead.

All of this would be interesting to measure and simulate.

It very well could be that the midrange position and size, combined with the baffle, is causing an unusual amount of interaction at 1 kHz.

Best,

E