[skrivis]

The original question, was "What speaker Types image most realistically".

While many "do the deed" in different ways, my suggestion was that by attempting to reproduce the "recorded ambient detail" and limiting the detracticting sonic "overlay" of the listening room, you would acheive a greater degree of realism.
...

Which kind of eliminates dipoles, bipoles, omnidirectional designs...

Panel speakers are bad because you get time smear from the large sound source. I feel the same is true of line source speakers (ribbons and such).

I was fortunate enough to live for a number of years in a rural area, where I was able to take my stereo outdoors during the summer. It was very interesting to hear different speaker systems without any room effects at all. Perhaps the worst of the bunch was a pair of Magnepans.

[John Casler]

Which kind of eliminates dipoles, bipoles, omnidirectional designs...

Panel speakers are bad because you get time smear from the large sound source. I feel the same is true of line source speakers (ribbons and such).

I was fortunate enough to live for a number of years in a rural area, where I was able to take my stereo outdoors during the summer. It was very interesting to hear different speaker systems without any room effects at all. Perhaps the worst of the bunch was a pair of Magnepans.

Yes it does, when talking about the mid and upper frequencies, which is where the detail I'm talking about is more recoverable.

Funny you mention taking your system outside, since I was looking at my smaller secondary system only hours ago and thinking how I could get it out on my patio for a good listen.

I have done this many times, and the lack of "room boundaries" and the subsequent interaction being "gone" is very nice.

[Redbone]

What's weird to me is that the RM40s are recommended to be crossed over in front of a person. One would think that the ideal situation would be to beam the right speaker into a person's right ear and the left speaker in a person's left ear. If the waves cross in front of the listener, I would think that this would be bad, but my own tests suggest this is good and not bad. I'm not sure why.

Very good question, John, care to take a crack at it ?

[John Casler]

What's weird to me is that the RM40s are recommended to be crossed over in front of a person. One would think that the ideal situation would be to beam the right speaker into a person's right ear and the left speaker in a person's left ear. If the waves cross in front of the listener, I would think that this would be bad, but my own tests suggest this is good and not bad. I'm not sure why.

Very good question, John, care to take a crack at it ?

Yeah, throw that at me

Well the VMPS ribbon "can" be converged on axis to the ear with great results, "BUT" most people will find the "balance" slightly off.  It may be "too" precise and edgy for those who still have sensitive hearing.

But there are many conditions that prompt Big B, to suggest the frontal convergence.

1) It pulls the ribbon "slightly off axis" allowing it to soften slightly and blend better with the woofers.

2) It provides a slightly larger sweet seat for those who "do" like to listen two abreast.

3) It moves the "outward firing" dispersion off to a more favorable angular incidence to the side walls for less room interaction.

I use a nice little trick to get the speakers aligned with "laser" precision without the expense of a laser.  I attach little hand mirrors to the top half of the neopanel, and sit in a chair 1-2 feet in front of my listening spot.  I thne move the speakers until I can see my right ear in the right mirror and left ear in the left speaker.  Bada Bing, precision alignment for the cost of two 39 cent mirrors and some masking tape.

I might mention that "on axis" might be a little edgy for "extreme nearfeild" but can be used effectively as you move more into the farfield.

[csero]

Actually I don't think that headphones would offer the idea, since they don't present sound to the "pinna" in a way that will allow it to provide the directional and localizational information to form a soundstage and imaging.

...

That's why they were using those "heads" to do binaural recording. The shape of the head and it's "ears" mimic the human head, therefore providing the cues we need.

Two huge problems here:

The pinna, and the related directional filtering is highly individual. That's why generic dummy hear binaural recording is fail to externalize properly for lot of listeners.

The other is the stage is moving with the head. Small head movements are very important in localization, and the lack of the effect makes it unrealistic.

You are right, that there should be only one set of ITD, ILD and pinna related filtering in the whole recording/reproduction chain, but the later better be yours. That's why I use transaural reproduction for the front stage.

[skrivis]

Two huge problems here:

The pinna, and the related directional filtering is highly individual. That's why generic dummy hear binaural recording is fail to externalize properly for lot of listeners.

The other is the stage is moving with the head. Small head movements are very important in localization, and the lack of the effect makes it unrealistic.

You are right, that there should be only one set of ITD, ILD and pinna related filtering in the whole recording/reproduction chain, but the later better be yours. That's why I use transaural reproduction for the front stage.

Very interesting. So this includes things like Hafler's difference hookup for loudspeakers, Carver's Sonic Holography, and the MARS system.

I experimented with some of these a number of years ago. I came up with one method myself. I simply cut the profile of my head from a sheet of foam rubber, and wore the sheet on my head while listening. It did a reasonable job of eliminating higher frequency crosstalk, but was not very comfortable.

I also tried bisecting my listening room with a divider of foam rubber, and that worked well, but was also not a convenient solution.

I concluded that the brute force approach wasn't practical.

[skrivis]

yours. That's why I use transaural reproduction for the front stage.

A further question... how do you use transaural reproduction for the front stage? Is this a commercial product you use?

I noticed that WinDVD employs technology from SRS Labs. I'm going to have to experiment with this to see what it does.

[csero]

Those older units were seriously fawled. Now with DSPs it is quite possible to have good result. Correct commercial implementations are rare. What you find in HT receivers ( NIRO, VMAx, Dolby) mostly add an inverse hrtf processing on top of the crosstalk cancellation to position virtual speakers (very bad).

For pure crosstalk cancellation I know about 2-3  VST plugns on PC, the bruteFIR on linux, the AD eval board with the Filmaker shareware, the Yamaha Yss901 chip or the DSP from the Creative Playworks PS2000 in HW.