Elevation Gain Issues

I believe that you are talking about preparing the course in a computer.

In your computer the elevation is calculated by using DEM (digital elevation model). There are a lot of errors since: 1 – The DEM data is not always correct; 2 – The DEM data as a granularity bigger than you watch.

In your watch the elevation is calculated by using a barometer. The main source of error is the barometer pressure drift in time and places.

The elevation gain in your watch is calculated with the difference between two elevations between each second (??). Don’t know how it is in your program.

The differences are perfectly normal and none are completely right but I trust the watch and not the programs.

Agree with previous but 1000 feet is a lot. Can you set your course as public and give us a link to the course and a link to an activity where you ran the course?

For me I never get the same elevation if I compare the course and the activity but it's never a large difference. Sometimes more, sometimes less.

One example:
Course: https://connect.garmin.com/modern/course/30822146
Activities: https://connect.garmin.com/modern/activity/4821829461 https://connect.garmin.com/modern/activity/4795939736 https://connect.garmin.com/modern/activity/4654265079

Thanks for your help. So trust the watch more than the course then? Here is my latest example... I ran the course 99% the same as the planned course I made on the computer.

https://connect.garmin.com/modern/course/32356734

connect.garmin.com/.../4842512624

Thanks for your help,

Basil

Thanks for your help. So trust the watch more than the course then? Here is my latest example... I ran the course 99% the same as the planned course I made on the computer. I have altimeter and barometer both set to continuously update.

https://connect.garmin.com/modern/course/32356734

connect.garmin.com/.../4842512624

Thanks for your help,

Basil

So trust the watch more than the course then?

Yes, I would. There is one setting you can check though. In the watch you can configure auto calibration of the altitude, options are off, at start and continuous. I have it configured to calibrate at start of an activity but I don't trust the GPS position, and DEM info enough to use continuous calibration. A few meters off in a steep canyon will calibrate wrong.

Here is my latest example...

That is a tough example and I think it is hard to judge any of them. It is a canyon and if the courses passes a elevation contour it will add up and result in a lot in the end. 
I've never used the course builder for trails when I have to draw free hand.

What is your personal feeling. Was it close to 3200ft or was 1650ft more realistic?

For about four weeks I have similar problems with my Fenix 6 Sapphire (latest firmware version 8.10; no beta; no additional watchfaces and apps):

The watch shows far too many vertical meters in total (compared to Fenix 5s, Etrex 30 or Komoot - up to about 40 to 60 percent deviation !!!). For example on one of my last bike rides, the Fenix 6 sapphire shows a positive difference of 394 m, although in reality it should be around 250 m. This is beyond good and bad - and meanwhile it pi...s me off! Calibrating does not help. 

As you can see in the screenshot attached, the vertical meters in the graph are "unsteady", i.e. they go up and down nervously. The absolute values (at the start, in breaks and at the finish) are largely correct. But the real profile should be smooth.

My guess is that potentially after the last firmware update, something goes wrong with the accumulation of the altitude/elevation. Does anyone know what could be the cause?

So trust the watch more than the course then? Here is my latest example... I ran the course 99% the same as the planned course I made on the computer

Oone specific  thing I can see is the bridge at maybe 5.75 miles. The DEM line does not know the bridge is there and follows the elevation down to water level and back up, which is about 50m of the difference. 

It could be something blocking the barometer hole or a problem in the temperature sensor. Sometimes the hole could have soap from cleaning the watch. I had some barometer problem in the fenix 5 and they were cleared with a warm water bath. 

How is the temperature graph? 

Thank you for your answer, Jose. I do not think there is a problem with the barometer (hardware), because the current altitude at the start, in breaks and at the finish is measured by the watch (more or less) perfectly. The striking deviations occur with the cumulative altitude difference / elevation over the whole activity.
Though, you find the temperature graph attached (I wore the watch on my wrist, and the track was partly sunny and partly shady).

For about four weeks I have similar problems with my Fenix 6 Sapphire (latest firmware version 8.10; no beta; no additional watchfaces and apps):

This thread is about the elevation calculated by Garmin Connect when creating a course vs the elevation reported by the watch when running the course and most of us agree that in that case the watch is probably the most reliable.

The problem you have is something else and has been discussed in other threads (if I remember correctly you posted in one of them as well). Please keep thread discussions to the topic or question asked by the author.