Thanks. I should have narrowed your options to consider appreciably by saying something like “holding all elements constant such as the recording, source media (vinyl, digital, PCM, DSD, etc., resolution) room, listener, listening position and volume (essentially holding all parameters constant except for the electronics and transducers employed), then wouldn’t it follow that any audible distortion imparted by the electronics or transducers used to reproduce the recording could affect how we experience or perceive soundstage?
That more narrow question helps, although I’d also have to assume we’re going on sound alone (no other sensory cues or cognitive biases) but then, we’ve still got some “it depends” answers in there. For example, it depends on the exact nature of the audible distortion and also of the electronics. In the case of transducers, then yes, sure, they certainly can and commonly do make a difference to soundstage but that’s largely to do with their interaction with room acoustics (their dispersion field for example) with speakers or interaction with individual users’ heads (the “fit”) in the case of HPs and particularly IEMs. When it comes to other electronics, DACs or amps for example, there shouldn’t be any audible distortion so the question of some audible effect on soundstage is moot. In those very rare, very low fidelity DACs and amps, some tube ones for example, then the answer is “possibly”, you could have a spray of harmonic distortion or IMD that might impact soundstage in some cases.
I don't remember group delay being mentioned in books on sound localization. …
Now, as with anything that's been discussed so far, if it becomes audibly different, it is always possible for our interpretation of the sound to be affected by it. But is it more meaningful as a variable for soundstage than the color of the pillows on the couch, or the size of the room I'm in while listening? I'm not sure.
I seem to remember something about it but not more than a mention and that was many years ago, most likely Olive or Toole. I did some experiments myself but that was nearly 30 years ago and concluded that it was an issue I could effectively ignore in my work except in very specific and very rare circumstances. Namely an isolated impulse type sound (quite tightly gripped Claves for example), very closely mic’ed or in near anechoic conditions and when applying steep EQs/filters for freq dependent processing, a set of circumstances I don’t recall ever actually encountering in practice, despite working with percussion extensively. I don’t remember now what it was I actually heard when testing but vaguely recall it sounded like a “loss of focus” (when the mono sound was centrally panned). That could be described as a “soundstage effect”, although there’s a good chance I’m misremembering what I heard (confusing it with something else) and I can’t rule out some other acoustic cause any way. “
I’m not sure” is a good answer IMHO, although the lack of evidence to support the proposition, despite a reasonable amount of study over the decades, indicates that “probably not” would be more correct.
Not related to group delay but nevertheless an interesting effect is “stereo shuffling”, which is a stereophonic (soundstage) effect controlled by relative level and/or EQ, but it’s only relevant to M-S processing. It is possible to widen the stereo image illusion using “shuffling”, beyond the width of physical speakers.
G