Why does OLC display wrong altitudes?

On Tuesday, February 12, 2019 at 1:38:28 PM UTC-5, wrote:
On Tuesday, February 12, 2019 at 1:09:57 PM UTC-5, Tango Eight wrote:
On Tuesday, February 12, 2019 at 12:52:05 PM UTC-5, Ramy wrote:
I’ve seen this occasionally in my flights if the takeoff altitude was not determined correctly, or if I forgot to start the logger before the tow started. I dont recall if it is OLC or the logger which adjusting your takeoff altitude to airport altitude, but if there is mismatch there, the rest of the altitude trace will be wrong.
The solution is to edit the flight in OLC and manually adjust the start to your release time, which seem to fix this issue.

Ramy

Ah. On the flight log Moshe linked, the "flight" starts in the air. That must be it. I had some sort of pre start screw up that caused me to have to restart the flight recorder and (unusually) it did not do its normal thing and give me a take off to landing log with a short gap.

T8

A couple of other example such logs I've looked at also are missing the takeoff. So that may be part of the problem. But missing the takeoff is no excuse to mess up the altitudes in the fixes that do exist. Here's another example, and this one is from a logger (Dell Streak running XCsoar, AFAIK) that does not have a pressure sensor, both altitude fields are identical and contain the GPS altitude (which should never be "corrected"?):
https://www.onlinecontest.org/olc-3....l?dsId=4917557
- the altitudes in the first hour are displayed way too low in the OLC web page graphics.

And here's another example (one of my own flights this time), this log also is missing the takeoff, and in this case the whole (fairly short) flight displays with wrong altitudes. Near the end of the flight the OLC web page graph shows negative altitudes, landing at -561m - while the lowest altitude in the actual log (as downloaded from OLC) is +165m. It's all GPS altitudes, no pressure sensor in my Nook.
https://www.onlinecontest.org/olc-3....l?dsId=6600502