Been getting this consistently for about eight months now with my Panasonic RR-DR60 and a custom noise floor setup I built in the spare room. It's not just the odd syllable you can stretch into something - these are full, contextual responses.
Last Tuesday I asked ". How many people are in this room with you?". And got back a clear ". Four, maybe five". With what sounds like genuine uncertainty in the tone. That hedging detail is what gets me. A random audio artefact doesn't express doubt.
The pattern I've noticed is that direct questions get better responses than open-ended ones. Something about specificity seems to sharpen whatever's coming through. Ask ". Give me a sign". And you get noise. Ask ". What year did you die?". And suddenly the session gets interesting.
I've cross-referenced several sessions with the Zoom H5 as a secondary recorder, running simultaneously in a different corner of the room. Same responses captured on both units. That rules out the recorder itself generating the artefact, which was always my sceptic brain's first objection.
What I can't explain is the emotional colouring sometimes present - the ". Four, maybe five". Response genuinely sounded tired, almost resigned.
Curious whether anyone else has found that question specificity matters, or whether you're getting direct answers through other methods - spirit box, Estes, whatever your setup is. Also whether anyone's tried triangulating with multiple recorders simultaneously, because I think that's an underused verification method in this community.
