What are the benefits of rear-firing tweeters?

[Bon]

I found the thread
http://www.diyaudio.com/forums/multi-way/6299-rear-firing-tweeter-why.html
which was a while ago. The few responses were skeptical to say the least but it was a decade old. My question was prompted by a visit to the recent Sydney Audio Show, where rear-firing tweeters were very common with the high-end loudspeakers (e.g. Hansen, Opera, Sonus Faber). Is it just fashion or is there some real physical basis for rear-firing tweeters (and mid-range)? If the on-axis in-room frequency response is relatively flat, smooth, well-damped, where are there benefits in adding reflected, non-coherent HF to the on axis? Why is adding extraneous delayed and phase shifted HF a good thing? So many questions. I am excluding OB speakers from my comments since I have no experience of them.

 

[a_tewinkel]

You seem to be suggesting that reflected sound is bad. It is not always positive, but a complete lack of reflected sound is probably the worst of all situations. We hear not only the direct sound, but reflected sound also contributes to timbre. Rear tweeters are attempts to improve the timbre of the reflected sound. In some cases perhaps, the aim might (also) be to attempts to change the spatial aspects of the sound. If a rear tweeter is desirable on a standard box type speakers is questionable however. I would say it's not.

 

[powerflux]

Messing up correct phase & stereo-placement and therefore introducing an extra "spatial" sound.

in my opinion.

 

[Bon]

You seem to be suggesting that reflected sound is bad. It is not always positive, but a complete lack of reflected sound is probably the worst of all situations. We hear not only the direct sound, but reflected sound also contributes to timbre. Rear tweeters are attempts to improve the timbre of the reflected sound. In some cases perhaps, the aim might (also) be to attempts to change the spatial aspects of the sound. If a rear tweeter is desirable on a standard box type speakers is questionable however. I would say it's not.

It is my understanding that a good recording contains encoded information about the acoustics of the recording venue . This is what needs to be preserved in the replay. The listening room acoustics are supposed to interfere as little as possible with the recorded acoustic. Surely the goal is accurate reproduction of what is captured on the recording? Rear-firing tweeters appear to add room reflected versions which have random phase and delay distortion of the original room acoustic and music.

 

[Melo theory]

Adding a rear firing tweeter is done to promote even-power response in a room.....which is very, very important.
But IMO, even-power response is better represented with an omni or an OB.

 

[a_tewinkel]

It is my understanding that a good recording contains encoded information about the acoustics of the recording venue . This is what needs to be preserved in the replay. The listening room acoustics are supposed to interfere as little as possible with the recorded acoustic. Surely the goal is accurate reproduction of what is captured on the recording? Rear-firing tweeters appear to add room reflected versions which have random phase and delay distortion of the original room acoustic and music.

Keep in mind that stereo can never accurately render the acoustics of the recording venue. It is inherently a compromise. No room reflections at all will approach headphone listening, may lack any spaciousness due to high IACC (depending partly on the recording) and will make the cross talk dip very audible. Some amount of reflections is good, but a rear tweeter will probably still leave a hole in the power response, it may distort imaging and cause colouration. Have you read http://www.diyaudio.com/forums/multi-way/195124-what-ideal-directivity-pattern-stereo-speakers.html? It is a very long, but also a very good debate on the directivity requirements for stereo.

 

[dewardh]

It is my understanding that a good recording contains encoded information about the acoustics of the recording venue . This is what needs to be preserved in the replay. The listening room acoustics are supposed to interfere as little as possible with the recorded acoustic.

"Supposed to" being the operant words in this regard. There is going to be reflected sound in any tolerable listening room . . . it is important to our perception that the direction of the early reflections be reasonably consistent with the "image" being presented. Otherwise the mind struggles between hearing what is (two speakers in a room) and what we're trying to convey (the acoustics of the recording venue, assuming it's not a studio).

But simply adding rear tweeters is not really the answer . . . omni or dipole loudspeakers is . . .

 

[dewardh]

But IMO, even-power response is better represented with an omni or an OB.

Make that "omni or dipole" . . . a lot of "OB" speakers don't have either the power response or the controlled directivity that one might (should) want.

 

[Telstar]

Very few dipole tweeters exist, so adding a rear-firing tw is a good solution (i'm talking of a dipole speaker already).

 

[Bon]

Keep in mind that stereo can never accurately render the acoustics of the recording venue. It is inherently a compromise. No room reflections at all will approach headphone listening, may lack any spaciousness due to high IACC (depending partly on the recording) and will make the cross talk dip very audible. Some amount of reflections is good, but a rear tweeter will probably still leave a hole in the power response, it may distort imaging and cause colouration. Have you read http://www.diyaudio.com/forums/multi-way/195124-what-ideal-directivity-pattern-stereo-speakers.html? It is a very long, but also a very good debate on the directivity requirements for stereo.

http://www.diyaudio.com/forums/multi-way/195124-what-ideal-directivity-pattern-stereo-speakers.html is a very interesting read but goes beyond my question which is about adding deliberate room dependent reverberant energy to apparently normal forward firing loudspeakers.

For domestic listening to reproduced sound it is usual to assume that the signal source will provide the appropriate level of reverberant sound, and therefore the listening room should be fairly ‘dead’, with adequate sound absorption provided by carpets, curtains and upholstered furniture.

 

[davidsrsb]

I have noticed many speaker tests recently are off axis and still ruler flat, with a note that on axis the treble is several dB lifted. The trend these days seems to be for very strong treble (I'm looking at you Focal) to compensate for us all getting old. Rear facing tweeters is just another way to further increase treble levels, while sharing the power dissipation across two drivers.