Sat front row at the symphony...

Each concert hall has their own sonic signature as I am certain you must know. 
 

That said, as a general rule, a few rows back will give the most balanced sonic experience. 

The idea that “the conductor’s perspective” (location) is best is flawed. Good orchestral composers and orchestrators are keenly aware of and exploit the fact that instrumental sounds blend to create specific desired colors and textures when heard from a distance. Just one example, a unison musical line played by a clarinet and a flute will sound very different and with a unique tonal color when heard from a distance than it will heard up close. The “detail” and separation heard up close may not be at all what the composer intends for the composition.

@mikeydee Glad you enjoyed it. A great Orchestra and venue. A  close colleague  of  mine plays with the PSO.

Also, their association with Reference Recording gives us wonderful things to listen to at home.

frogman,,

You're right that instrumental sounds have different colors, textures, tonal balance at close vs further distance.  But look into the mind of the composer.  He/she thinks of a melody in the mind, tries it on the piano, then writes it down.  Often the first complete work is solo piano, or piano 4 hands, or 2 or 3 pianos.  Later the piano work is arranged for orchestra.  A good example is Mussorgsky's original solo piano version of Pictures at an Exhibition.  Later, Ravel orchestrated it, which is the most popular version heard, but there were other composers who had different orchestrations.  

On youtube, there are recordings with simultaneous views of the complete score (written music).  As a musician, I want to know what the score contains.  Even a solo piano score contains details that I didn't realize were there even after thinking I knew the work well.  As for more complex orchestra works with many different instrumental groups, the score shows that a typical audience listener is missing the majority of what the composer had in mind.  An example is the prelude to Act 1 of Wagner's Lohengrin.  The score shows EIGHT violin parts.  I had thought there were only 2-4 after 50 years of thinking that I knew the piece well.  Only a 1st row listener (better yet, a stage performer) can hear much more of the details in the score.  From the 1st row at a performance, I could appreciate much more of the truth, which is in the score.  From greater distances, it is hopeless if you want to hear all the detail.  But the stage is best.  Mercury Living Presence recordings have the main mikes 10 feet over the conductor's head, to capture the most detail, with a good amount of space.  Most other commercial recordings are far inferior.

fleschler,

The most accurate and natural speaker is plasma. Totally massless, small driver the size of a tweeter, very efficient. Unfortunately, they are dangerous for ozone and other noxious gases, fire hazard from the burning flame from high voltages. Nelson Pass was hospitalized for an asthma attack after using a plasma speaker.

The next best transducer is the electrostatic principle. The lowest moving mass, total control from the membrane/stator sandwich. But all commercial stat speakers have severe flaws. To make up for the inefficiency and need to be near full range, large panels are needed. Even STRAIGHT large panels deliver smearing, due to the different distances to the listener ears from thousands of locations on the large panel. In this regard, the WORST speaker I ever heard was the Dayton Wright XG8 (10?) I heard in 1980. It was a 4 foot square panel.

The next meaningful experience was with Art Dudley when he worked for Edison Price. I heard the small Stax F81 and F83 speakers there. The F83 was a double stacked F81. I loved the midrange/HF purity of the F81 which was less than 3 feet tall. I had hoped the larger F83 would overcome the severely low 73 dB efficiency of the F81. It did, but unfortunately the 6 foot height caused severe rolloff of HF compared to the F81. I later figured out that the larger panel area, the more multipath time smearing occurs--worse time alignment. The best stat for tonal purity remains the original Quad 57 whose tweeter panel is very skinny and only about 30" tall. Later Quads are veiled by comparison, utilizing the flawed concept of time delay and much larger panels.

Putting all this together, I have a concept for the best possible stat speaker. If I were a famous audio designer, I could charge $ 1 million for this concept. But there is no market for accurate stat speakers in an a-phile market that cares more for boom boom loud dynamics and deep bass. So I reveal it here, in the hopes that some manufacturer who cares more for sonic accuracy and purity takes notice. Here goes--a large enough panel handles a wide freq range with reasonable SPL capability. But the panel is curved concave to the listener instead of convex like ML, Soundlab. The panel is a slice of a sphere whose radius of curvature is the listener distance. Say the distance is 8 feet. The slice might be 1 foot wide and 4 feet tall. The circumference of an 8 foot radius sphere is 8 x 2 pi = 26 feet. So this is like the curved edge of a 55 degree pie slice vertically, and a nearly straight 1 foot horizontally. The only listener requirement is to sit at the exact focal center of this spherical slice. That way, the direct radiation path which has the most HF extension reaches the listener from all parts of the panel with perfect time alignment. There is still a flaw from off axis parts of the panel reaching the listener with different freq balances, similar to a cardioid mike with rolled off HF off axis. This design is still better for accuracy than any large stat ever made. The smaller, the better, as in the Stat F81 and original Quad, if your music requirements are up to 80 dB. For me, I don’t want to hear junk from today’s speakers designed for loud SPL’s. I’m not impressed by 100 dB of junk when those speakers are badly veiled at 20-60 dB.

Another big problem with large panels is the bloated image, totally unnatural. A singer delivers sound from a mouth about 1-2 inches in diameter. A trumpet is like a 1 inch diameter tube whose horn flares to only a few inches, etc. So a wide range dynamic tweeter that goes down to 1 kHz can do a reasonable job for accuracy and proper image size. Dynamic tweeters are reasonably low mass and much more accurate than dynamic midrange and LF drivers. But all current panel speakers deliver bloated images. In my focused design, the image would be more true to life. The only instruments that are properly reproduced by current panel designs are large ones like pianos and pipe organs.

My Audiostatic 240 from 1980 is 2 straight panels mounted on a dummy support, so you can angle the 2 panels any way you want. The best results are from concave angling, with both panels beaming to each ear. I got the most bass, HF and SPL with this arrangement. But for best focus and purity in midrange/HF with admitted sacrifice of bass, I only use 1 panel which is 5" wide x 48" tall. Beamed right to my ear, it is the closest to my concept of a better design. On audiostatic.com, the MDi is shown for 3000 euros, although there is no opportunity to hear it before you buy, unless you travel to his suburb of Amsterdam, Netherlands.  The designer, Ben Peters is old, so I don’t know the delivery details. The panel is 11" x 44" so it looks like a smart design with the least compromises.