Would you think that cancellation would be the culprit even if I measure well at the listening position? Would it then be a matter of specific frequencies being lost when being reproduced by both channels? Something not revealed when measuring with pink noise?

TIA