I've been revisiting some recordings produced by John Culshaw, and once again I'm reminded that I love, absolutely love, "stereo staging" in recordings of theatre pieces (musicals, plays, operas).
When stereo came in, there was a lot of question about whether voices needed to move across the stereo arc -- did listeners need to hear a voice moving when a character moves, or would that just be distracting? Culshaw was one producer who was certain that, as he wrote, "if the baritone needs to flee from the tenor's wrath, he must obviously move away from him." So he and his other producers prepared for each opera recording by writing out all the movements that the characters would make, and then actually rehearsing the singers in where they were supposed to move, even putting numbers on the floor so they'd know where to stand at a given moment. (The singers had to move while singing, because the early stereo setups meant that you couldn't achieve a convincing movement by just panning the voice; it had to be the real sound of moving from one microphone to the next.) His recording of Georg Solti conducting Wagner's Rheingold, the first huge hit in stereo opera -- it stunned everyone by breaking onto the Billboard pop charts -- was full of this stuff.
As Culshaw explained in his liner notes, he not only staged operas from left to right but from front to back, so the voices would move backwards to indicate that they were at the back of the stage. In the first scene of Rheingold, when Alberich the dwarf is chasing the Rhine Maidens around the river, the voices move while the chase goes on, and when one of the maidens swims away to a higher point, she moves to the extreme right side and also moves back a bit (so her voice will be in a different acoustic for her next line). With the recording of Britten's Peter Grimes, even though Culshaw didn't have time to produce it himself, he and Benjamin Britten prepared for the recording by plotting out every movement and perspective: where the door would be in the pub scene, where Grimes' window would be. You can actually hear Peter Pears' footsteps as he walks from left to right, which is certainly something you don't usually hear in a studio recording.
Anyway, while some companies took up this idea for stereo demonstration purposes, not all of them really embraced it, and a few just flat-out rejected it. (When Herbert Von Karajan went to Deutsche Grammophon after recording operas with Culshaw for several years, he tried to talk his German producers into using Culshaw's techniques, but they talked him out of it, preferring to have the voices remain static.) And as time went on and stereo became commonplace rather than a gimmick, most opera recordings no longer had a lot of movement; they'd use such effects occasionally for entrances and exits, but they would not have a singer's voice actually move across the stage to correspond with some movement that would occur in a stage production. Culshaw's successors at Decca kept the technique going for a little while (the first recording of Britten's Death in Venice has the voices moving around more or less as they did in the first stage production), but mostly gave it up after a while. You'll almost never hear an opera recording with a true stereo staging plan.
I understand why this died out, because it's gimmicky and sometimes counter-productive. (Culshaw would sometimes have a whole aria sung in one speaker because the character was supposed to be sitting or standing at the left side of the stage. But as Culshaw's assistant and later senior Decca opera producer Christopher Raeburn pointed out, when someone sings an aria, you should have them dead center; it's just not effective to have the voice at one extreme or the other.) And there are a lot of late '50s recordings of operas, musicals and spoken-word performances that use stereo not just as a gimmick but a really pointless one: they'll have voices bouncing back and forth between right and left for no clear reason, just to show that they can do it. That's irritating and gets in the way of appreciating the performance.
But the basic principle behind the "stereo staging" approach -- that if you're recording a theatrical work, there should be some sense of where people are and why -- is one that I find makes for more exciting listening. Not only because you can follow the singers' movements but because it creates more variety not to have every singer in one position and one acoustic for an entire scene. If a movement is important to the scene, then it makes sense to put it in the stereo spread, the way producers put in important sound effects. (In The Marriage of Figaro, for example, every producer of an audio recording knows he's supposed to put in the sound of Susanna slapping Figaro. Yet they'll leave Susanna in one place for the whole scene, even though she's supposed to come in from offstage, approach Figaro, and then approach other characters.) It's not a big deal, obviously, but it does enhance the experience if done right; it helps make an audio opera recording into a "theatre of the mind" experience.
I think the last recording to really use this technique elaborately, and certainly one of the best, was the 1976 recording of Porgy and Bess by the Houston Grand Opera company (the best of the three recordings of Gershwin's complete score). The producer, Thomas Z. Shepard, who was RCA's head of classical and Broadway recordings, not only filled the recording with sound effects, but incorporated all kinds of audio movements to stand in for movements he'd seen on stage. It's one of the things that makes that recording feel much more theatrical and alive than other recordings of this piece. (Shepard also sometimes used this in his Broadway recordings -- in Sondheim's Pacific Overtures, you can hear the two characters in "Poems" walk from speaker to speaker as they walk across the stage.)