[WIP] World of Aduhr

[Charerg]

Right, it's been a good long while since my latest update, so I guess it's time to post another one. Although this time around I don't have a new area to post per se. I've been working very sporadically on the height map for Central Eocidar, but it's only maybe 20% done.

All in all, I was getting a bit fed up with the rather painstaking method of doing it all by hand and decided that a change of tactics was in order. I've dabbled a bit with sculpting programs (Autodesk Mudbox, more specifically) in the last year or so. With those, the workflow is usually first starting off with a piece (say, a head) at low subdivision (low resolution) and basically first sculpting the shapes and adding the major details in that resolution. Then you move to a higher subdivision, adding more and more detail as you go, all the way to the pores on the skin and what have you.

With that in mind, I was inspired to attempt a new technique. I realised that mapping a piece of 10 000x5000 map is completely unnecessary for creating the general details of the topography (where are mountains, highlands, basins). In fact, since you're tempted to work while zoomed in a lot, I think it's more of a hindrance than a benefit. So, I've decided to adopt a more "subdivision type approach", where I start creating the topography at a very low resolution and then scale upwards, adding details. Also, I will be creating a greyscale height map, since that works a lot better with this sort of process because blurring isn't a big issue with greyscale.

In order to test the method, I decided to start off by re-working Akanrias a bit. There were a few things I was a bit unhappy with regarding the subcontinent's shape:

Akanrias Original.PNG

So, I fixed those up and turned my previous elevation map into greyscale. I use a RGB-scale btw, with 25 metres per point to convert this into metres. The values 0-5 are below sealevel (-150 to 0 m), while the values 6-255 represent the height range (0 to 6250 m).

Akanrias New.PNG

I also made a nice gradient for this, so I can straight up convert this into a layered elevation map like I used previously. To actually work on the piece, I considered if I should use a different projection. However, GProjector only accepts Cylindrical Equal-area as input out of the equal-area projections, and while the area is displayed correctly, the shapes are distorted a lot. So, all in all, I decided for a compromise: I'm going to use Equirectangular Oblique, with the map centered on the piece I'm working on. That should get rid of the worst projection distortion. And the resolution limits of GProjector are less of an issue when working in greyscale.

After creating the initial topography and running through Wilbur once, this is what the low-res map I'm working on looks like. Btw, this is clipped from an Equirectangular Oblique projection centered on 30 W 20 S. Comparing this to the 3D view I think this works pretty well (not too much distortion).

With the layered gradient:
Akanrias Layered.png

Gradual transition gradient:
Akanrias Gradual.png

I'd include the elevation key, but with this resolution it would cover half the map (or be totally unreadable if I downscaled it). So here it is separately, same as used previously:
EleKey.png

I also made a few changes to the elevations from the previous version. I decided to remove one of the LIPs on the southern coast. That is because I'll probably update the tectonics at some point, and I'm planning to change the position of Nomune somewhat (so that LIP doesn't make sense anymore). I also added a large lake and changed the drainage pattern as an experiment. Right now, both of the intra-cratonic basins of Akanrias drain into the "south sea". However, I'll probably change the drainage of the western basin back into its old course (drain into the southwestern sea).

So, that's what I have for this update. So far, I'm pretty satisfied with this new method. I created the above maps in a few days, and it's a refreshing change to work on a greyscale height map. Also, this method allows me to use Wilbur to full effect, "Wilburifying" the map each time I increase the resolution. This is still very much a WIP of course, but I thought I'd post this since there are quite a few world-mappers likewise working on elevations right now, and someone might find this useful.

All in all, I have high hopes that this method offers a good compromise of achieving good production speed without sacrificing quality in the process. But we'll see how the final results turn out. As usual, feedback and comments are welcome!


Edit:
Oh, btw I also incorporated Naima's feedback a bit and changed the costlines so they're not so "over-fractalized" in this version (although that's not really apparent in the low-res map).

[Charerg]

I'm trying to keep up a bit more frequent stream of updates, so here's something of a mini-update.

I developed my 500px elevation map to a point where I think I've exhausted the possibilities of the low-res map. So, I've upscaled the 500px map up to 1000px. The final will be about 1550 px in width (I had to drop the resolution of the world map to 5800x2900 due to GProjector's resolution restrictions). Here's what the 1000 px map looks like (this is just the 500px upscaled, haven't worked on this yet other than establishing the land mask).

Akanrias 1k Gradual.png

Here's the same map split to elevation zones as in Azelor's climate guide (although the lowest level is 0-50 m here):

Akanrias 1k AzelLayered.png

I also extracted myself a reference map of Earth using the same gradient (25 m per RGB value). So, in this case I capped the elevations at -150 minimum and 6250 maximum. Here's what the reference map looks like (had to downscale it somewhat to fit into the forum's file size restrictions, still pretty huge though):

 

Earth Gradual DScaled.png

So far, I'm pretty happy with how the map is turning out.

[Charerg]

To continue my attempt at semi-regular updates, I finished the 1000 px map. I ended up doing some pretty big changes in some regions, so it was a bit more time consuming than anticipated. The final results look good though.

Final map pre-Wilbur:
Akanrias 1k Gradual II.png

After Wilbur:
HWbur2 Gradual.png

And the final map split into layers by elevation:

 

Layered as per the old elevation system I used (see previous posts for elevation key):
HWbur2 Layered.png

Layered as per the "Azelor system" (see Azelor's tutorial for the key):
HWbur2 Azel.png

So far I'm really happy how the map is turning out. It's inevitably going to look a bit more "messy" than the previous maps where each elevation layer was sort of hand-tailored instead of being generated from a greyscale height map, but I feel this method results in a much more organic-looking map. So, now a short break and it's on to creating the final map in full res (1576x710 in this case)!

[Harrg]

Very nice! How you split your map into layers by elevation? You use "magic stick" in Photoshop(or alternative tool in Gimp )? Because some time I very want correct my height map after Wilbur in different layers for each scale.

[zhar2]

Very nice! How you split your map into layers by elevation? You use "magic stick" in Photoshop(or alternative tool in Gimp )? Because some time I very want correct my height map after Wilbur in different layers for each scale.

In gimp you can use "posterize"

[Harrg]

Thank you zhar2) I forgot about this.

[Charerg]

Very nice! How you split your map into layers by elevation? You use "magic stick" in Photoshop(or alternative tool in Gimp )? Because some time I very want correct my height map after Wilbur in different layers for each scale.

I have basically divided up the RGB scale to 25 metre intervals. So basically 256 possible values, that represent an elevation range from -150 metres to 6250 metres. It goes as follows:

0 RGB: -150 m
1 RGB: -125 m
.
.
7 RGB: 0 m
.
.
255 RGB: 6250 m

Then I've created a few gradient maps which divide the RGB scale into appropriate elevation layers. So, it's essentially just a Gimp Gradient Map. I'm not sure what the equivalent is in PhotoShop, though.

It's worth noting that when you export a height map from Wilbur, it automatically scales it to RGB values based on the Min. and Max. Elevations you've defined. In my map, the highest elevation is about 2750 metres, so I have to sort of rescale the exported height map back into the usual range using the "Levels" in Gimp.

Sorry that I can't provide more detailed advice, but unfortunately I don't have any experience using PhotoShop (as I use Gimp exclusively).

[Harrg]

Thank you Charerg. I present an alternative it in photoshop)) Ph. and wilbur use similar tools.

[Pixie]

The general direction of this project is.... THE RIGHT DIRECTION.

Take some rep! Well deserved.

[Charerg]

Right, so the final high-res map is finished (if you consider 1576x710 high anyway...). I made quite a few revisions to the interior, so it took me a bit longer than anticipated. And of course you could keep fiddling with the details ad infinitum, but in an attempt to rein in the perfectionism a bit, I've decided to call this version finished.

So, here's the map with Azelor's elevation key:
AzelHMap.png

And with the "Char elevation key" seen previously (this time there's enough resolution to even include the key with the map):
HMap Layered.png

The above with gradual transitions:

 

HMap Gradual.png

Note that the lower elevation has been capped at -150 metres min. and 6250 m max. here, so the lower (and uppermost) layers of my elevation key are unused in this scheme.

Here's how the greyscale height map looks on the globe:
AkanNova Orthogonal.PNG

As I mentioned, I made a few updates to the geology of the subcontinent's interior as well. I'll add that to this post when I have the time.


Edit: the Akanrian Geology

About that geology update, my previous geologic update had the ages of the older fold belts as follows:

~1100 Mya: Central Akanrian Fold Belts
~850 Mya: East Akanrian (aka Tamanaro) Fold Belts

I expanded this somewhat by splitting the West and East Akanrian crations into two component blocks. The W. Akanrian Craton's southern portion is now a separate block (called the Uiruno Terranes) that was accredited to the craton perhaps ca. 1500 Mya. So, there's the remnant of an ancient fold belt there. The E. Akanrian Craton has been split into two cratons: the southern one is called the Ingilo Craton. The two pieces collided ca. 1300 Mya, forming the Ingilo Fold Belts.

In addition, there is an ancient failed rift responsible for Akanosuwa, the large interior lake. The Akanosuwa Rift formed at an age close to the Tamanaro Fold Belts, so perhaps 900 Mya. All these ages are extremenly rough of course, as I haven't modelled the tectonics beyound 200 Mya (which was covered in the tectonics update about a year ago). As a real world inspiration I had the North American Midcontinent Rift System (about 1100 Mya), responsible for creating the basin occupied by Lake Superior.

So, here's the updated geologic stuff with a nice relief map from Wilbur (I omitted most of the younger formations from this as those were covered in the previous geology update):

GeoMap.png

Oh, and I should mention that I changed the former East Akanrian Shear Zone into a more complex fault complex. Previously I had it as something similar to the Alpine Fault. I realised that this kind of primarily transpressional transform faulting doesn't make any sense in these circumstances. So now the faulted area has pull-apart basins in areas where a "releasing bend" causes extensional stress (transtension) and mountains where a "restraining bend" causes compressional stress (transpression). Ofc this transform fault became inactive about 150 Mya, so I'm a bit on the fence whether I should've eroded the region into little lower elevations. But like I said, I don't want to get too stuck with the details, so I decided to leave it as it is.