Dipole front wall reflection EQ

Not everybody can afford placing dipole speakers far enough from the front wall. Mine need to stay 60-70 cm close. An speaker with free-field linear frequency response will now show an obvious, rather broad and clearly audible peak at the wavelength that is 4 times the distance to wall - in my case, right in the midbass at ~140Hz . I am sure I'm not the only one having this issue :)

Options:
1. absorb the rear wave - not very practical or aesthetic
2. toe-in the dipole: it helps a bit, but not quite that much, because even at 45 degrees there is still significant output at the back due to the small distance to the wall
3. EQ the "hump": that's what I ended up with and so far did not see any issue.

Do you see any issues with option 3 ? Does anyone use it as well ?

I am actually thinking if in such cases it would make more sense to target a flat response of the SPEAKERS+WALL ensemble rather than speakers alone. One could use room EQ as in some newer receivers (Audissey EQ)
 
Do you see any issues with option 3 ? Does anyone use it as well ?
I am actually thinking if in such cases it would make more sense to target a flat response of the SPEAKERS+WALL ensemble rather than speakers alone.
At 140 Hz and your small wall distance, you can't hear the speaker alone. It will always be the combination of speaker and early reflection. So it is perfectly ok to EQ for a flat response at the listening position.
I hope the speaker distance to the side wall is not in the 50-80 cm region. :rolleyes:
While the midbass peak can be equalized, you are stuck with the early (4-5 ms) reflection throughout the complete frequency range. This will result in imaging problems in the 0,5-5 kHz region. What are you intending to do about that?

Rudolf
 
4. Convert your dipole to a wide-baffled closed box? I'm a fan of solving problems acoustically. ;)

That's not exactly a solution, rather a replacement of the problem and the introduction of others ;)

At 140 Hz and your small wall distance, you can't hear the speaker alone. It will always be the combination of speaker and early reflection. So it is perfectly ok to EQ for a flat response at the listening position.
I thought so too. Thanks !

I hope the speaker distance to the side wall is not in the 50-80 cm region. :rolleyes:
Yes, there's significantly more space in lateral, some 1.5 m (they're on the long side of the room)

While the midbass peak can be equalized, you are stuck with the early (4-5 ms) reflection throughout the complete frequency range. This will result in imaging problems in the 0,5-5 kHz region. What are you intending to do about that?

Now that's an interesting topic.
As you say and according to the theory (Toole's book) there should be definitely some imaging shifts due to the front wall reflections. However, I subjectively find the imaging of the speakers very good to my taste:
- the center image is VERY well defined. This actually make sense, because the front wall reflections should cause an image shift TOWARDS the center.
- the sound stage is broad, stable and well defined in the whole range between the 2 speakers. This could be related to the particular polar response of my mid-high panels, which are a 2D array of 3.3" fullranges crossed over to a Neo 3 PDR. This makes them rather directional even above the dipole peak (>~500 Hz).
- the only impairment I could observe so far was the "depth" dimension: if I want depth, I need to roll-off everything > 5Hz at the back side. Which I do, being helped by the nature of the small fullrange drivers which roll off quickly on the back.

So I guess that the imaging could be "wrong" in terms of absolute source positions compared to, say, the same material on 2 normal speakers in anechoic conditions. But OTOH, imaging is something rather subjective and the absolute positions of sound sources are not really that important.
 
+1. Last week end I had a real epiphany. I pushed the dipoles as far out in the room I could, 2.5m from the front wall and 1m away from the side walls ( I managed to keep 3m separation). Well, blown away is certainly an understatement :). The rear wave was somewhat absorbed in the highs, but too much to my taste, it lacked "air" and "sparkle" on cymbals for ex (I know, how original is that!).
I would just not put a dipole in a small room, a small omni, nearfield, sounds way better to my ears.
 
Last edited:
However, I subjectively find the imaging of the speakers very good to my taste:
If, and how early reflections interfere with imaging will highly depend on the local situation - listening geometry, front wall condition etc. What I experience (without proper attenuation) is this:
front wall refl.gif
Phantom sources near left and right from the reflection point tend to be drawn to that point. The amount of draw can change with frequency.
Best strategy would be to keep everything far from the walls, as lolo suggests. Does he live in one of those beautiful spacious Loire chateaux? ;)

Rudolf
 
Well, your picture suggests two strategies:

- put speakers as far to the walls as possible (the obvious choice if you have Lolo's castle :D)

- second best: put speakers as _near_ to the wall as you can so you can still call it a dipole: in this case, the reflection point is very close to the speaker itself and the shift will be reduced. EQ might be needed again and there's some combing unless care is taken to absorb part of the back wave.
 
I think people tend to forget that monopoles radiate as much to the rear as dipoles up past the mid-bass. If you are getting reflections at 140hz with a dipole, you also will from a monopole.

That's true, but instead of a peak, which is quite disturbing, with monopoles you'll get a dip, which is likely less audible and will make your speakers sound as if they had more deep bass.

How about a setup like this: L-wings from speakers to the wall on the inner side ? This would make sure no reflection occurs on the inner part of the front wall, at the expense of an asymmetric dipole response.

Anyone tried this ?
 

Attachments

  • Lwings.png
    Lwings.png
    3.6 KB · Views: 201
Nah, I don't have a castle dear Lords, sorry to dissapoint! :cool:

the room is 7.3x5.2m, but the open kitchen kicks in a 4.5m, otherwise I would have tried to push them even further!

Sure, omnis will give more reflexions, but you can listen to them really close, like 1m away or even closer, and then manage a larger ITD gap, that's the key!

I recently built an el-cheapo Demokrit speaker, they image better than the dipoles with much, much less room treatment. Sure, the SPL is not the same..nor do they reach deap bass but hey, they are dead fun to listen too! Very close only though..
 
The rear wave was somewhat absorbed in the highs, but too much to my taste, it lacked "air" and "sparkle" on cymbals for ex (I know, how original is that!).
I would just not put a dipole in a small room, a small omni, nearfield, sounds way better to my ears.

Sure, omnis will give more reflexions, but you can listen to them really close, like 1m away or even closer, and then manage a larger ITD gap, that's the key!

I recently built an el-cheapo Demokrit speaker, they image better than the dipoles with much, much less room treatment. Sure, the SPL is not the same..nor do they reach deap bass but hey, they are dead fun to listen too! Very close only though..
Love to have that confirmed by independent ears ! :)
So get a sub and the tweeters and the sparkle and air is even more shiny but more natural at the same time compared to the reflector.
 
So, german philosophy = pleasure anticipation, french philosophy = pleasure delaying ? :D
It all lies in the point of view :D

More on topic: do you absorb/diffuse the front wall reflections ?
My frontwall is reflective and the middle part is diffusive because there is a folding door with a zig zag pattern. With the dipoles it sounds a bit better when it is closed. With the omnis there is no preference.
But I have the space to setup the dipoles up ~1.5m from the front wall. And for me and my room this seems to be a good distance. I can cater for 1m from the side walls as well.
 
Well.. German debt is 80% of GP and going down, France is 90%, and going up.. I know where the Philosophers are, I can tell!

There is one thing this audio business taught me, is patience. I am improving every day! :D

On topic: I absorb the rear wall as much as I can, 20 BBC modules and I have 12 extra "uber" ugly (but effective!) absorbing pads I can place where I want, +10 light 1" pyramidal pads for flutter echo. Nothing is fixed yet so I am playing every time with them, a bit more here, a bit less there.. tweaking still.
This far, the best results I got was with a slight absorption of front wall in the corners only, light pads on the sides to avoid flutter echo, and rear wall as dead as possible. I am still to get some properly measured Basotec to try Rudolf's "trick", and build some diffusers.
I am also quite close to think that a rather dead room as a whole, but with a fairly reflective front wall is the way to go. Reverb compensation for music recorded in real spaces would be given from rear surround side speakers and matrix decoding when needed. That way you could get both of best worlds for any genre. Not a conclusion yet, just a thought..need to gather more knowledge!
 

sraudio

Member
2012-11-26 3:14 pm
That's true, but instead of a peak, which is quite disturbing, with monopoles you'll get a dip, which is likely less audible and will make your speakers sound as if they had more deep bass.

How about a setup like this: L-wings from speakers to the wall on the inner side ? This would make sure no reflection occurs on the inner part of the front wall, at the expense of an asymmetric dipole response.

Anyone tried this ?

You misunderstand something.Not peak, not dip.There are many peaks and dips. This depends on distance from the wall and frequency. One difference is that monopole backwave is 180 degrees different from dipole. So at same distances the pattern will be different because of that. Also keep in mind that with dipoles you get less reflections in room at low frequencies, only at higher frequencies you get more reflections than monopole.
 
You misunderstand something.Not peak, not dip.There are many peaks and dips. This depends on distance from the wall and frequency. One difference is that monopole backwave is 180 degrees different from dipole. So at same distances the pattern will be different because of that.


I don't think I misunderstood anything and I am aware that there is a complex interference pattern and also room modes, but we're talking about a primary strong reflection off the wall at a short distance, which makes it very strong.

Precisely because the monopole backwave is in-phase (while the dipole is 180 out of phase), you have:
- for dipole: you get a first and strongest peak at the 4 x distance_to_wall wavelength (the back wave travels 1/2 wavelength to the wall and back and adds in phase to the front wave)
- for monopole, you get a dip at 4 x distance_to_wall wavelength (the "back" wave travels 1/2 wavelength to the wall and back and adds out of phase to the front wave)

Sure, this is all superimposed to the other peaks and dips of the complex interference pattern plus room modes, but with a strong reflection those first peaks/dips are very dominant.

Also keep in mind that with dipoles you get less reflections in room at low frequencies, only at higher frequencies you get more reflections than monopole.

I don't see why you get less reflection at low frequencies. You probably mean you excite less room modes, which is a different story.