I can’t remember when this started happening, as it has been many months that this has impacted me. But I recall distinctly that when adding imagery, it would be aligned with the default axis perfectly. Now when I add a location, the imagery imported is not aligned with the axis. This is very annoying and creates some problems when changing the image texture, which I need to do now that the new imagery source is awful.
I can’t recall if this problem started occurring with the new imagery source or if the problem predated this change. I could have started when I upgraded to SU Pro 2017.
You can see this clearly in the attached image where the locked component is aligned with the axis, but the underlying image is skewed clockwise by ~0.4 degrees. I’m using a default template and I’ve tried other templates as well just to be sure it wasn’t a changed axis.
This isn’t anything new. It has always been like this. You might find some locations on earth that have the image boundary according to the (locked) group boundary but they are rare. Also (if you are able to, with SU pro) if you swich on solar north (used for shadows) you’ll see that it most likely is slightly off of the the green axis (system north).
The group boundary is according to the systems axes, aligned to green and red. The imported image and terrain are not.
It looks like the imagery is imported aligned to solar north, which is good. I guess I just need to update the axis in my template to align with solar north to make this annoying problem go away, but it will still be skewed using the standard top view. I wonder why this issue exists. Maybe I’m missing something with respect to different parts of the world.
This did happen in some areas when Google let us use their Maps, too. It’s due to the differences in the various North alignments. Unless all your sites are in the same location, I wouldn’t change the model axes. You’ll forever be chasing the correct alignment.
This has always been the case with Terrains. However I still haven’t got an answer as of why. Does anyone know if green axis in the model or the edge of the terrain (Y axis in the terrain imagery) is the actual geographic north?
which states:
Aligning the drawing axes with the cardinal directions
Each axis has a solid line on one side of the origin and a dotted line on the other side of the origin. The solid blue line leads up from the origin and the dotted blue line leads down. The remaining lines correspond to one of the cardinal directions (north, south, east, west). The following table outlines how each line corresponds to a cardinal direction, leading from the axes origin point.
Line Direction
Solid green line North
Dotted green line South
Solid red line East
Dotted red line West
Yet SketchUp sets another solar north when the model is Geo Located. The solar north used for shadows matches the Y axis in the imagery. Which north is the geographic north? Green axis or the imagery/solar north?
the geographical north (north pole on the globe, which I assume is what SketchUp’s green is, a fixed location)
magnetic north, which direction doesn’t match at all with the north pole. It varies (wanders) in time
solar north to which the imported location is referring. It does not coincide with either of the above.
Solar north is a very complex direction (the axis of which is describing an 8 shape in space over the year). I have no clue how far SketchUp is taking this into account over the year, if at all.
I guess that due to the fact that both axes (geographical north-south v.s. solar north) do not coincide we see these slightly rotate imported imagery with “Add Location”.
The side of the imported image group (hidden edges) has the direction of solar north. The bounding box has the direction of the system’s axes red and green.
Just my impression.
I think magnetic north can be ruled out in a world where no longer magnetic compasses are how you tell north.
Astronomic north also seems a bit irrelevant. Maybe it is used to calculate shadows but that could simply be done internally with no need to expose it as a separate direction to the user. The “solar north” you get when geo-locating a model seems to be defined by what Google Earth/Google Maps considers north in its 2D imagery. I can’t see however how this could be anything other than geographical north. Maybe “green north” is after all the magnetic north and not the true north?
Found the answer in another thread. Apparently the north from terrain is true north (otherwise the imagery used on Google Maps and Google Earth line up properly). The green axis is however not true north! It represents the UTM grid north which is the vector that would be true north in the middle of each cell. The closer to the grid lines you get the more it deviates from true north.
