rdfb wrote:I consider manually winding the altimeter, without a specific setting, to have too much of a risk of getting it wrong. I'll only do it on the ground, setting it to zero or a known aerodrome elevation as appropriate. In the air, I won't ever adjust the altimeter except to a known setting.
... which could have been misheard, or misremembered, or miswritten, or misread. I'm not sure winding off a height that you've planned is any less reliable.
Ways in which winding an altimeter could go wrong:
- You wind it the wrong way by accident.
- You wind it twice because you were already on the QFE (eg. you did some circuits before departing so you were on the QFE and forgot to set it to a QNH).
- You forgot to ensure you were flying straight and level before you started to wind it.
- You weren't trimmed well so you didn't hold straight and level while you wound it.
- You mishear, misremember, miswrite or misread the aerodrome elevation.
- You get your arithmetic wrong (eg. altitude 2300 feet minus aerodrome elevation of 145 feet = ???)
- You misread the scale (eg. move it 800 feet instead of 80 feet).
You might think that the above won't happen to you. But mistakes usually happen when there's something else not going to plan, and you follow habit even though it's no longer appropriate because of what has changed (yet due to high workload, you are unable to identify that you need to adjust your habit, or what you need to do differently this time).
Not that setting QFE is so important when flying VFR, because as others have said, I don't look at it when below circuit height except for flaps/turn on climbout. But if I do an instrument approach I'll do it on the QNH for this reason, and have my DA/MDA written down in front of me as well as set on the altimeter bug. So I prefer not to be in the habit of winding the altimeter.
The QNH could indeed be misheard, or misremembered, or misread, but that's only one thing. Your method to get the QFE will add to any QNH mis-setting error, since you'll likely have received QNH updates enroute. Also, if you stick to the QNH, you'll generally only hear one or two hPa difference every time; if you hear a big change you can be in the habit of confirming it. That provides some protection against gross errors.