Can't speak for everyone's experiences, but here in my house zero feedback goes 'thru' a stone when honing hand-held. I do actually do this quite often, usually when I am in a rush and the counters around the sink are cluttered. Or when the stone's form-factor fits the task nicely.
Long time ago, I put a mic below a Coti and one above. Did the same with a vibrometer. What I found was that the applicable feedback, tactile, audible and visual, occurs above the stone's working surface. My honing hand on the blade's tang is the source of steel/stone feedback. While I like honing hand held, I find working on the bench to be more accurate.
Maybe someone using thinner stones, or with a heavier hand might gather different data. Hard to see how but who knows what is possible sometimes?
My left (stone holding) hand reacts to the bevel's angle and adjusts position accordingly. It's more than one extra force vector actually, because there is a flexible wrist involved. My right wrist guides the bevel on it's own though. so it's not something I need to have happen. At the end of the day though, no one needs to hone one way or another, so long as they're happy with the results and process. All roads lead to Rome, sorta.
The point I was trying to make initially though, wasn't about feedback or ultimate sharpness, or honing preferences so much. It's about isolating the issues that may be playing into OP's uneven bevel. When learning, or troubleshooting, I have always found it best to 'zero' the equation and work up slowly without adding unknowns and variables into the mix haphazardly.
In a situation where the users force/pressure might be a player in the storyline, it stands to reason that adding an unknown variable to the scenario while trying to solve the riddle is counter-productive. It's like making too many changes all at once and having success, you never know which change, or what combination of changes corrected the problem. And then it's hard to be repeatable. It's also possible that the problem wasn't actually solved or corrected; sometimes one action might negate another and zero it out. Even though that might yield the expected result, the actual problem was never identified and corrected - and that's all the difference in the world sometimes.
Long time ago, I put a mic below a Coti and one above. Did the same with a vibrometer. What I found was that the applicable feedback, tactile, audible and visual, occurs above the stone's working surface. My honing hand on the blade's tang is the source of steel/stone feedback. While I like honing hand held, I find working on the bench to be more accurate.
Maybe someone using thinner stones, or with a heavier hand might gather different data. Hard to see how but who knows what is possible sometimes?
My left (stone holding) hand reacts to the bevel's angle and adjusts position accordingly. It's more than one extra force vector actually, because there is a flexible wrist involved. My right wrist guides the bevel on it's own though. so it's not something I need to have happen. At the end of the day though, no one needs to hone one way or another, so long as they're happy with the results and process. All roads lead to Rome, sorta.
The point I was trying to make initially though, wasn't about feedback or ultimate sharpness, or honing preferences so much. It's about isolating the issues that may be playing into OP's uneven bevel. When learning, or troubleshooting, I have always found it best to 'zero' the equation and work up slowly without adding unknowns and variables into the mix haphazardly.
In a situation where the users force/pressure might be a player in the storyline, it stands to reason that adding an unknown variable to the scenario while trying to solve the riddle is counter-productive. It's like making too many changes all at once and having success, you never know which change, or what combination of changes corrected the problem. And then it's hard to be repeatable. It's also possible that the problem wasn't actually solved or corrected; sometimes one action might negate another and zero it out. Even though that might yield the expected result, the actual problem was never identified and corrected - and that's all the difference in the world sometimes.
