Under an assumption you are reading this on a phone, take your phone and rotate it counter-clockwise 45 degrees and then look again at that rendering: is the light now the top-left face or the bottom right face? The issue for me is not that the light is really telling me the orientation, but the slope... and so I feel like the light is coming from the bottom right here (preferentially: I can force myself to consider either).
In the original orientation, if I convince myself the light is at the bottom of the cube the surface feels really awkward as it is receding away from me top to bottom, not due to light but simply geometry, which is not common: things that are stacked that way tend to collapse into things that are stacked the other way (and so hills are also like that).
I would like to have seen this explored from other directions as well.
It's clear that given a choice between top and bottom, light is expected to come from the top.
However, in real life, the sun crosses the sky from eastish to westish. Could both accuracies be better satisfied by placing the light source to one side?
Would it be less wrong and less confusing to just put South at the top, and make sure there's a prominent compass rose?
For prominent shapes like your country or state, or a map you might hang on the wall, we're very used to seeing them with North at the top. But when using a map for navigation (either a paper one or an electronic device) it's quite common to have other orientations.
Convention and predictability are strong reasons for maps to be north-oriented. A different orientation increases the cognitive load of using the map, which goes against its usual design goals.
The exception are small hyper-local maps where orienteering by the use of compass, far away points of interest, and the sun are not part of the map's purpose, and reorienting is helpful or aesthetically pleasing because of the shape or layout of the area shown; think shopping malls, amusement parks, or the immediate environs of a point of interest.
I don't know if it's just "small hyper-local" maps. The standard map of Manhattan, for example, is normally oriented with "uptown" at the top which is about 40°, closer to NE than N. Mainly because it fits on rectangular displays better that way.
I understand your points about convention, of course, but the trade-off against convention (or in favour of establishing a new local convention) is perhaps made more often than you think.
Does same applies for maps of New York City in general? That is those showing larger area than just Manhattan?
Manhattan is good example of hyper-local map. Even if area is relatively big. As the city planning leads to such format. That is mostly straight roads going in two directions, where cardinal direction isn't so important for navigation.
> think shopping malls, amusement parks, or the immediate environs of a point of interest.
Fire escape maps in buildings.
I find this annoying that most are north-up or some other 'typical' convention, where IMHO you want them to be so that things are oriented relative to how the viewer is reading them.
This must vary a lot worldwide. I've always seen them with the wall the map is placed on representing 'up', which makes sense when you are looking at in a hallway, because then 'left' on the map is also left in actuality (and 'right' obviously right).
Humorously, the term "orientation" comes from when Christian Europeans traditionally had Jerusalem - in the east ("orient") - at the top of the map. It wasn't until around the time of Mercator that they started putting North at the top.
But sunlight comes from the north...here in the southern hemisphere. It would be pretty confusing to have two conventions for the different hemispheres, and would make it weird in whole-world views. I never rotate maps when I'm navigating either; it seems like that would make it harder to keep track of which way things are.
As long as you take care to keep rotating the map while you move, the end result is that you can local directions like left and right directly when reading the map.
Why are there symbols that are far bigger than what they represent? Why are there weird lines all over the hills? Why are there giant letters spelling out place names?
The goal of a map is to convey information, like writing. It's supposed to give you a realistic idea of what to expect but it is not a substitute for reality. I've never thought of the light on a shaded relief map being "the sun". Why would a map need to tell me which way the sun is shining? That's not its purpose.
Nice website! Do you know if there is a place where I could easily calculate how many hours of sunshine do I get at particular location during the year assuming that I have lower x degrees of my horizon are occluded (i.e. I live next to a taller house)?
This got me confused as well. At noon it’s almost always to the south but at sunrise and sunset it will be to the north during the summer. Play with shademap for a bit and you’ll see this effect.
Edit: Think of the sun as following a V shape over the course of a summer day. It rises in the NE moves to S and sets in the NW.
I’m at about 50N and the Sun definitely rises and sets North of the east-west grid roads here in the summer. Unless it’s some crazy optical illusion, but I don’t think so.
Edit: playing with the mbostock Solar Path visualization linked by a sibling comment confirms that it starts rising and setting north between March 21 and Sept 21.
Snow lasts longer on the north face, making mountains look much lighter on the northern sides. The eye can be trained to accurately reconstruct terrain regardless of the lighting direction. Shading from the north is more natural, especially when shadows are not cast.
26 comments
[ 3.2 ms ] story [ 66.7 ms ] threadThis is a classic case of mistaking the map for the territory.
In the original orientation, if I convince myself the light is at the bottom of the cube the surface feels really awkward as it is receding away from me top to bottom, not due to light but simply geometry, which is not common: things that are stacked that way tend to collapse into things that are stacked the other way (and so hills are also like that).
It's clear that given a choice between top and bottom, light is expected to come from the top.
However, in real life, the sun crosses the sky from eastish to westish. Could both accuracies be better satisfied by placing the light source to one side?
For prominent shapes like your country or state, or a map you might hang on the wall, we're very used to seeing them with North at the top. But when using a map for navigation (either a paper one or an electronic device) it's quite common to have other orientations.
What it is meant to look like is a small 3D model of the landscape, not the planet itself.
The exception are small hyper-local maps where orienteering by the use of compass, far away points of interest, and the sun are not part of the map's purpose, and reorienting is helpful or aesthetically pleasing because of the shape or layout of the area shown; think shopping malls, amusement parks, or the immediate environs of a point of interest.
I understand your points about convention, of course, but the trade-off against convention (or in favour of establishing a new local convention) is perhaps made more often than you think.
Manhattan is good example of hyper-local map. Even if area is relatively big. As the city planning leads to such format. That is mostly straight roads going in two directions, where cardinal direction isn't so important for navigation.
Sometimes, yes. The most iconic example is probably the MTA subway map which is visible in almost all subway cars.
Fire escape maps in buildings.
I find this annoying that most are north-up or some other 'typical' convention, where IMHO you want them to be so that things are oriented relative to how the viewer is reading them.
https://www.bbc.com/future/article/20160614-maps-have-north-...
The goal of a map is to convey information, like writing. It's supposed to give you a realistic idea of what to expect but it is not a substitute for reality. I've never thought of the light on a shaded relief map being "the sun". Why would a map need to tell me which way the sun is shining? That's not its purpose.
Disclaimer: I made shademap
Edit: Think of the sun as following a V shape over the course of a summer day. It rises in the NE moves to S and sets in the NW.
Edit: playing with the mbostock Solar Path visualization linked by a sibling comment confirms that it starts rising and setting north between March 21 and Sept 21.