This is a great video on the technical side of map-making, but I've been wondering if you're planning on discussing the actual science of it more. Things like: - Where to put wetlands, fjords, lagoons, deltas etc. - How the age of a mountain range affects its topography - The effect of post-glacial rebound on geography - Where to put endorheic basins
deltas go at the end of a river, near its outlet. this is mainly caused by something that appears at the end of a rivers path or something that was stopping the river from going all the way out to sea. after this, water quickly occupies whatever it can.
RE: computer-generated coastlines using noise and thresholding. This can be manipulated somewhat by putting a couple gradients down as a mask for the noise layer, so that the noise layer will become transparent where coastlines are supposed to be smoother, and more opaque where coasts need to be more rough. Of course, there are still other issues (generated noise won't create features like deltas, lagoons and bays and such) but I think there's certainly room for computer-generated content, provided you put in the work to make the things it comes up with realistic. And that can be done, even though it's easier not to. I think both hand-drawn and computer-generated approaches are valid. And just like you need to get good at hand-drawing a map, you need to get good at telling the computer how to generate maps. It's just a different skill to the same end.
Yeah. If you're going the lazy route, you'll get bad results no matter which approach you take. Computer generated can be just as valid as handdrawn, but of course just slapping some noisefunctions together is not going to cut it. The actual point is that programming an algorithm to spit out realistic maps is only worth it if you plan on making hundreds of maps. If you only need one, handdrawn will yield satisfactory results faster.
@@TheAgamemnon911 Well, if you're into worldbuilding, it might be worthwhile. I know I've been working on a procedural plate tectonics simulator... Though I guess that's as much about programming practice as it is about worldbuilding for me.
@@TheAgamemnon911 I think there may be some confusion regarding what is meant by "computer-generated" in this sense. Artifexian (and by proxy, Gamesaucer) isn't talking about writing a program in a coding language that generates landforms. He's mostly talking about things like adding random noise to an image, and then messing with thresholds to turn that noise into a fractal coastline. And when it comes to time investment, I find that even when you're being really picky and finicky with computer-generation, as Gamesaucer is talking about, it's still going to be less time consuming and require less precise manual dexterity than hand drawing a map. The barrier to entry for computer-generation is mostly just understanding the tools at your disposal and how you can apply them to achieve the same end result as hand drawing.
@@Rotaretilbo Expanding on dexterity: I want to mention a lot of people do have dexterity issues. I swear half of the people I’ve met in my life talk about how they can’t draw bc of their hands being shaky, nerve damage, etc. I’m personally disappointed by the idea of computer generated being somehow lesser but I think that’s because I’m an artist first, not a scientist, and I accept a lot of “well that probably wouldn’t exist” happening. It’s EXTREMELY hard to get into world building as a community if you’re someone who’s more sensitive to things like criticism due to the weirdly clique-like nature of groups within it. Every subgroup seems to have a blacklist of creators that are seen as bad information sources (which are always big names) and you get scolded for daring to watch the content. Otherwise you’re treated as if you’re already supposed to know all the same information. This is just one example of how hard it is to get into the culture, let alone the practice (the focus on worldbuilding for Tolkien-esque settings also cuts out a large amount of people too). These people are the most informed about how to do this building, and usually won’t ‘look' at you strangely for wanting to know how to do something well. Generally at least there is a mutual admiration for the love and work that goes into this, but that doesn’t balance it out entirely.
When I first saw part 1 of this go up, I was definitely intrigued. I love learning new tricks for getting various "looks" in map designs. Initially, I was a little bummed that now that we're actually *doing the thing* in episode 2 and laying out the topography that it's effectively a straight-up "hand drawn by educated eye" approach. I feel like the style amounts to "get good, son." That's not a criticism, though. To steal your words in a different part of the video, I think that's a feature, not a bug. Your comment about something worth doing being worth taking the time to do properly struck a chord and is something I think I kinda needed to hear. I love your channel and love your presentation style! Keep rocking!
Sidenote: Your expression during "EDGAR OUT" makes my day every time. I don't know what it is, but something about it is just the perfect ending to your educational style.
One (or several) criticisms I have with the map you are building here: Your coastlines are all equally jagged with lots of bays and little islands. There is nothing wrong with having straight sections every now and then, especially where the currents would deposit sediment. And on the other hand, your mountain range is not jagged enough - meaning that I miss the signs of erosion in the form of long and branched out valleys that rivers typically create. Also mountain ranges along faultlines would typically have multiple rows of peaks running in parallel.
Maybe he will add erosion due to wind, rain and rivers in another video. Notice that he added lakes, but no rivers to feed them (so, I think, those lakes are depresions filled with rainwater, and they would not exist in the rainshadow given by the mountain, but I think he skipped that detail for the sake of a short video).
@@evecol4525 Oh, in that case: Look at the Andes, especially in Bolivia, but any currently geologically active mountain ranges will work: Himalayas, Alps, California...
I know this series is about illustrator, but with what you said about drawing by hand; I draw all my maps in MSPaint and try to get something i like the look of, and its so rewarding to see the worlds i created, and trying to make it feel lived in is so rewarding knowing that i used a mouse in a crappy microsoft app to draw these land forms
I just randomly started this series because I found it interesting and oddly satisfying, but I just wanna say your little speech about getting better with practice and "you 100% will get there" was really nice to hear. :) I've been having a lot of writers block so thanks for the motivation!
I went back and looked at your original video on tectonic plates and it's really cool to see how much better this is. The original video was good, but this one is amazing, and that's reassuring because it means that we can eventually produce some really awesome stuff too.
Wow, wow, wow. This is both facinating and overwhelming at the same. Ill have to watch this when i can fully give my attention. Making a planets worth of height maps has become an incredibly larger task than my already present understanding of geography, geology, ecology, and climate will allow. 😖😄🙃
I think the best way to go about it is by using tools like World Machine which can simulate realistic terrain including things like erosion, currents, beaches and more
I mainly use G Plates to visualise prehistoric Plate movements in my world. I love to form a world from some starting landmasses colliding, new land forming and interacting with the other ones. I feel like it makes the world more vivid when i have a clue about how tectonics in my world work rn. Also having mountain ranges at places elsewhere then the center or coast of a continent is quite nice.
A trick I have learned when trying to make coast lines is to take paper get part of it wet and you should see rough edge let it dry and you can trace that that boom coast line.
This video was/is a HUGE help to clear up where my next diirection should be. Ive already made my general drafts on paper, next is building the heightmap itself on the computer. So glad to see that I was on the right path regarding realizing i will have to hand create elevation by elevation.
this video in essence: "draw some circles. Now draw the rest of the owl". All of these crevices, and peninsulas and islands and stuff, how can you come up with them on a spot? I know what you are gonna say, study a lot of maps, but you are supposed to get something from these maps, right? Some principles. I wouldn't even guess that ocean currents can affect the global shapes of continents, even if I would study maps to the end of times, for example.
He has a video on ocean currents i think, so if you're unsure how they effect landmass, check that out. Alternatively, look at cool coastlines that you want to replicate and research why they're like that, it'll help inform where the right place to put them on your map would be. It basically comes down to research i guess. Looking things up so that knowledge can inform how you go about making the shapes.
Mostly looking at atlas maps, fantasy maps (depending on your needs), how currents and tectonics affect the coastlines and height shapes and get a feel for it. Having made several maps by hand; the crevices, plateaus, peninsulas and island shapes are something you mostly come up on the spot, either because you feel like it would look better, or because you want to make a purpose out of it (world-building reasons). However with prior knowledge of maps, you can make them look realistic because what you will come up with are the random patterns you have been seeing on other maps. "It just works" My workflow, like Artifexian has always been to make rough shapes, according to my needs, then procedurally add more details to the shapes. You will never get a scientific, down-to-the-maths method of making continents due to so many factors being at play, it's just more complicated to do it with maths rather than randomness and patterns, so it's sticking to the feel and how good/realistic it looks coupled with map knowledge. I'd say having a general grasp of how it works is what makes or break your map, not because it isn't realistic, but rather because it looks like so. If anything regarding map learning, look at how random most of the shapes are, you don't need to be intensively scientific about it, like I said before getting a grasp of some concepts is enough. After that, it's mostly a matter of practise. That said though, be iterative! A map is not set in stone you can always improve it later on if you learn a couple things along the way. To me it has always been a back and forth process between looking at maps and making some.
there are some other videos where he explains this kind of things. In this one, the intention is to show hoe to draw the maps using the softwares he's showing us. If you want the answers for the questions you have, I suggest you to look for the other artifexian's videos.
13:35 Re: hand-drawn vs. computer-generated topography, I think a reasonable combination of the two methods is probably the best way to get realistic shapes, landforms, etc. In my own conworld mapping efforts, I've used a global plate tectonics sketch to get the general placement of landmasses, then used something like the "clouds filter + threshold" technique you referenced to get semi-detailed coastlines that I can then further refine manually to suit local climates and ocean currents. Similarly, when designing topography, I'll start from a hand-drawn rough draft of where I want the major elevation levels to be, export it as a series of masks, and use them in something like Wilbur to get good-looking erosional features, river networks, etc. P.S. 15:53 I actually pronounce QGIS the same way you do, so no complaints from me. :D
It’s really not necessary. You can do fine with a mouse! Deff would not recommend a laptop trackpad though. If u do have a laptop u can just buy a cheap USB mouse to use with it, which is a lot more affordable than buying a drawing tablet, although u can get those for pretty decent prices nowadays too. But I totally understand that it’s not possible for everyone, or myself even, currently.
12:47 There is a snapping function. When you're in edit vertex mode, under the list of the coordinates of your points is an option called "snap nearby vertices." Sadly it only works with already extant vertexes, so you have to create your shapes first and then snap them to each other.
in gplates if you use the polyline feature (L) instead of the polygon feature it will not force a close loop, meaning you only have to connect 2 points to a plate, then when you create the feature the 2 plates will share the boundary.
Just a suggestion, you might want to highlight the differences between the technique you're using to draw your map and how it differs from Nate's as well as the pros and cons of both methods. Where Nate uses photoshop and masking in layers of elevations using the lasso selection tool and the fill command using raster graphics (unless he has changed his workflow recently?), you're using vector and creating a bunch of shapes for each elevation before merging them and arranging them. There are so many ways to do it, and each person may prefer a different method, or one might work better for their work flow than the other.
Hi Conrad :) He does touch upon the differences in this video - I think the main difference is that the vector-based programs like inkscape or illustrator are less of a strain on your computer due to them working with equations rather than pixels - I use photoshop, and take it from me - large raster maps with multiple layers will cause your machine to lag like hell. Though I use mask you can easily just use the pen tool, filling in the selections with the desired colour, eliminating the mask - its just a matter of preference.
@@nathanMangion_Elyden I really like the idea of using the mask... but also the idea of using vector for maps. I guess it's one or the other though haha
First of all, you are a really great resource and inspiration, thank you for that. Over the years I always worked at my world bit by bit, but never was quite content and reworked it a lot. That process led to trying out various methods and styles, and was all about learning and learning. I did a rough world map with plates, currents and weather zones etc., but figured in the end, that it would be just too much, because my story is based on just one sub continent. But it helped nonetheless. I have a rough idea of the surroundings and influences on my continent. Right now I redrawing and reshaping the map as lineart inspired by old maps and the maps of middle earth. There will be places, that will be seem wrong, but old maps of our world looked far worse, so I don't worry to much. And I may do another map with this method you showed, because it looks quite neat and more accurate. But brawing by hand gives me time to reflect the history of some landscapes, how they formed and what took place there. So even it is time consuming, it helps me a lot to figure out, in what a world my story will be situated. For research I love to go to bookshops and just let me surprise what I'll find there. So I got a huge book of the history of catography and also an atlas of imaginary places - and a lot more of various sort. Wall of text, check.
So something that comes to mind for realistic, if not improbable, world building is how many a person think having 4 and 6 limbed vertebrates would be unrealistic due to most if not all vertebrates coming from a common ancestor. But if one where willing to go into their worlds history and depart drastically from how earth evolved, you could have early vertebrates from 2 different groups be the primary colonizers of 2 different continents. Evolve separately for millions of years and inevitably the continents will come together and separate many times redistributing organisms. Making for an interesting biosphere. Of course this applies to more than just animals and limb count. Plants and fungi could also be quite different. It would be like having two planets in one.
I would suggest looking up the program Wilbur, as it doesn't seem like your method will be using it. It can be used to model rainfall and erosion. I have used it to place rivers and it seemed fairly realistic, though it struggled with lakes.
6:46 I did that with my world Etsaiñ. I visualized it in an Equirectangular and didn’t use gplates except for plate details. I just made things wider near the poles
It’s funny. I love this channel because you like to have realism integrated into your builds. I love science so I know the realism, and I’ve wanted to make a fantasy world for a long time. I have even conducted research on unknown topics to get a more realistic result. I then integrate everything in a way where I start from creation of a world, then jump to different maps to show the progression of time. I try to get the most realistic ways to make a map.
Great video with a sexy sexy map. What are your thoughts mythological world building? It's a totally different style than what you seem to find enjoyable, but like Middle Earth doesn't take into account any plate tectonics or ocean currents- the entire thing is created by gods screwing with each other. I hate the "A wizard did it!" thing, but it seems like a narrative approach to land formation totally contrasts doing it scientifically.
Oh My God! Just discovered this, it is amazing. I now feel inspired to make on myself. Coming from a D&D background this would be really helpful but also a cool way to wow my friends. Wish me luck!
It's amazing how worldbuilding can suck you in. I was just looking for a way to add some realistic-ish climates to a fantasy world I'm developing for writing, and now I'm seriously considering making an atlas map even though I'll never once need it to tell the story I wanna tell, and no reader will ever give a damn. But this. This is too cool to pass.
I would absolutely recommend using Blender to draw on a 3D sphere if you don't already know how to use Blender. It is never a bad idea to learn how to use a new program.
I'm trying to see what Earth would be like if the geographical poles moved to different locations. In my case one is parked over eastern Brazil and the other on the opposite end in SE Asia(two strong magnetic anomalies), creating an equatorial and subtropical ocean system that would go all the way around(eg use the trade winds to go completely around the world). I have the 2nd largest tropical rainforest on what is in our world the Canadian Prairies, Northern US plains, and the foothills of the Rockies. The largest being a strip down what is now Africa tilted 90 degrees on the map. since the sea(and also the land to less extent) is pulled out more at the equator, the sea levels are around 500m higher. This forms a large Shallow sea. The "Saskatchewan sea" extends from tropical islands(formally high arctic mountain ranges) down to the coast that runs through what used to be Alberta, Saskatchewan, Dakota, Wisconsin etc. Sea level drops closer to the poles has cut off the Gulf of Mexico turning it into a highly saline inland sea and desert that sits below sea level. Cities that "survived" whatever cataclysm caused the changes(whether by time or not) and the oldest cities in the world are ones that were well established before, during, and after the transition, or left behind valuable ores to mine. Likely what was formally Alberta would be rich in resources would be one of many, the shallow Saskatchewan Sea bringing a bounty of fish. However the tropical climate would limit the kids of crops grown, however the abundance of metals, oil, fish, and fruits can be traded for wheat, corn and other colder climate crops grown further south. The new nature of the world affords easy sailing anywhere without the need for massive canal systems, global trade is considerably more robust due to this. There is less extremes due the tropical and subtropical waters being uninterrupted and both poles sit on continental plates. Australia would have transitioned from the arid desert it is today, to a more mid latitude breadbasket, the outback is green and full of farming.
I love the way you pronunced my last name hahaha. Also, you conviced me to buy a drawing tablet. I've always wanted one but given that I can't draw for the life of me, I've never seen the point of it. Now I do. Also Also: there is a way to "snap" points in GPlates. With the "move vertex" (V) tool activated, there is a box below the list of points that make up the shape that reads "snap nearby vertices" together with a threshold in degrees and a restriction by plate ID.
I love the idea of you commenting on the previous video's comments - thanks for addressing mine btw. 8:55 all those 'little' islands between those two continents, is that really how they can grow? I don't really see anything like it on earth. Maybe something like how Iberia presses up to the Atlas area or Indonesia or, maybe the best example is Nunavut in Canada. But they all seem quite different. What is your idea here? 17:13 haha got surprised by that little lake there? xD
Hey Artifexian, an amazing video, thank you. Re: Photoshop - I am not entirely sure what is the advantage of doing this in vector to be completely honest, apart from the ability to easily toy with the scale of the landforms, which is something you don't really want to do on a map anyway, as it would distort the complexity of the for example coastline. Yes, you can zoom in here, but I guess it kinda depends on what the map is supposed to be for. If all you want is just a map everyone will view at a certain, fixed scale, without interactively zooming in or out, I am not sure vectors have any big advantage. For the purpose of worldbuilding itself, sure. Anyway, you can for example make a brush in Photoshop which will make semi-random, jagged coastlines on its own, and it will be far more simple than trying to make all the shapes happen just with a manual pencil, which is in my mind a prohibitively tedious way.
Hello. This video has been such a great help, I don't know where I'd be if I hadn't seen it. I wish I had the app you were using (I'm using autodesk). Your a continuous source of help and inspiration. Thanks a lot, Artifexian!
Couldn't agree more with the criticism of the fractal land generation, one thing those tutorials seem to do well though is recommend the program Wilbur for simulating erosion which seems to do a really excellent job.
@9:20, my first map was in paint, so I didnt have any layers, or any projection tools. I was finally able to project it onto a globe using Google Earth Pro, and its a bit wonky but its still fun.
One detail I noticed: the strokes of your various landmasses and plates (actually, all the strokes you imported from the snapshot), were scaled up from 1 pt to 12.8 pt, giving you space to add detail inside those lines. If you scale the snapshot without also scaling the strokes, then the strokes mantain a value of 1 pt, and that space disapears.
On the software and such, using a program like Maya where you can get it to update textures to PSD files (no clue if blender has that? Probably does?) could be a cool way to train yourself to draw polar distortions (As in, draw what you think would work, see how it maps, analyse, change, update the 3D program and repeat until it looks cool. Would probably also help for knowing how to draw in coastlines in the polar regions while already accounting for distortions)
PS has some 3D functionality now, you might be able to do it there. Though switching back from 3D to 2D via undos is a pain. There's probably a better way in there that I don't know about yet.
How do you recommend dealing with the problem where parts of your maps are at different fidelity than other parts? The "resolution" in your mountains seems less than that of your coasts.
Try keeping the zoom steady. Don't create part of your map with 300% and another with 150%. The bigger the zoom, the more detail you tend to add. Also, don't use too big of a zoom, otherwise you lose the greater scope of your map.
Always drawing the same zoom level is a good tip. As for my mountains in particular, they aren't done yet. Their resolution will be matched to that of the coastlines.
When you mentioned blender i realized that you're actually drawing a type of elevation map, which can be used in building a 3D version of your map. Not only would you have a story for your world but you could also walk and run around in this world too.
One thing you might want to look into for computer-generated geography is a program called Songs of the Eons ( look on reddit.com/r/SongsOfTheEons ). It's primarily a large-scale RTS sort of video game, but the interesting thing is that the devs have been building their own world-generating system from scratch, taking into account stuff like plate tectonics, ocean currents, even wind erosion and other things that would be too tedious to do manually. I haven't tried myself (it's still technically in development and unfinished) but I think you could probably control certain factors to guide the program toward the sort of world you're looking for. Once it generates the globe for you, you could use it to make any number of maps. I would highly recommend that anyone interested should give it a glance.
There was a video or something a while back where someone made a 3D "ball" out of hexagonal shapes. So if someone is creative they can make a globe map in this way.
There's a number of ways you can project a sphere onto a grid of hexagons of triangles - look up 'geodesic sphere' for some explanations and imagery, but in essence you start with a regular polyhedron - an icosahedron is a popular choice. Then there is a method for subdividing it until you get a resolution that satisfies you, and then it will unfold easily along the lines of those triangles, and you can use that to independently render continents with minimal distortion.
for other people who don't have illustrator but have photoshop: the pen tool in photoshop is the equivalent of the pencil in illustrator, you can either do it by point or freeform too
The physical globe method is what Chris Wayan of World Dream Bank / Planetocopia have done for some of his planets. But he also often does preliminary drawings on paper.
i definitely did my last map by generating clouds in photoshop and adjusting the threshold. It didn't even occur to me that I was ignoring things like the effects of ocean currents. Lucky for me, I've become completely disenchanted by my map and I'm eager to start another... and then I peeked over here, and spotted some mapping videos. B) I felt like a hack at just the right time.
I think the question with using gplates to just draw on a sphere comes up because some versions of illustrator already have that and for the ones that don't you can download free scripts and plug-ins to do it. If you're using inkscape, you'll have to use something like gplates, but if you're using Illustrator version 10 or newer, you have the option to do that in-app.
If you a re looking for realistic you should draw the highest peaks then create river systems to guide you, this will also place your major lakes and determine coast lines where river deltas will create many islands. Also temperate climates will tend to less jagged train as erosion is temperate - not harsh like deserts or tropics
There is in fact a snapping function in GPlates, although it's a bit finicky and i don't know how exactly you access it. But there defo is one, i used it alot to sort out my plate tectonics.
Regarding Strider‘s comment about not drawing the topology by hand: I agree that using fractal generators, perlin noise or similar isn‘t a good solution. What would be useful (and I think that is what they were getting at) is a tool that actually simulates plate tectonics and erosion in a realistic, physics-based way. That probably still wouldn‘t be super accurate but a good starting point.
Well, if you do a Q&A for the next one: any advice for making maps based on existing topography? I'm working on a project that takes place in the area of my home state several thousand years in the future, and I was wondering what you'd recommend for creating accurate "future" maps for that area.
Looking at maps of earth we can see plenty of mountain ranges and other geological features created by plate tectonics, which are based on historic plate activity from plates which no longer relate to the current arrangement shaping earth (e.g the Urals in Russia or the Great Dividing Range in Australia). Do you have any advice for placing topography based on no longer present plates. Do you have a notional idea of craton/basin likely placements in your head when setting out topography?
i showed off the map i drew and someone linked the 1st vid of this tutorial to me, a wasnt plan on usign this guide cuz im just gonna make it simple as a guide for my story but i got intrigue, and wow!! i wanna try to do this but due to my laptop being a potato, i cant use Adobe photoshop/illustrations at all! tho im gonna take notes on how to so when i got a better laptop, imma do this
You are obviously drawing like the continents are flat. It is there any noticeable inaccuracies that anyone should take notes? I mean, take the final map and load it in G-plates and I am sure that some coast lines or rivers will look weird, specially close to the poles. I love your channel and your work!
how i managed to avoid the whole "get a tablet" problem: my world is impossibly small (r = 0.22R) and so all those details are bigger and therefore take less time to draw in with a mouse.
I might be the closest thing to a GPlates expert (with regards to world building) there is at this point. I discovered GPlates a long time ago, and have since started a very large project involving it in order to make a realistic world. Here are a few of the things I've learned and the tips I can give: - To make things align, there is a way to sort-of align things. When you click "Create Feature", you can assign a conjugate ID and enable "Create Conjugate Feature", which will make two features, one with one ID, and another with the other (e.g., you make a line with ID 1 and conjugate ID 2, GPlates will spit out 2 identical lines, one with ID 1, the other with ID 2). This works best with polylines. - You can save a whole lot into one feature collection. In my project, I have a file for continental plates and a file for oceanic plates. Continental has over 100 individual IDs (I wanted to use microplates), and Oceanic has around 20. IDs are used primarily for rotation in the .rot file, features with the same ID will use the same rotation scheme. I noticed in the original GPlates tutorial, he saved the different continents into different files-- I don't think I can stress enough how useless that is to do. Put them all into the same collection, but ID them differently. - The .rot file does not care what feature collection something is in-- anything with an ID that has a rotation in there will move according to the file. - Feature collections are largely for coloring things differently and hiding layers, I think of them more as layers than files. - The best way to align things as far as I am aware is to zoom all the way into a point and draw the shapes directly on the points. If you're *really* OCD, you can "Edit Feature Properties" and edit the coordinates themselves (it's under "gpml:unclassifiedGeometry"), but honestly, that's way too overkill IMO. - Oo, shortcut keys: General tools: F = select tool, D = drag tool (hold shift to rotate, hold control when using other tools), S = measurement tool (you can change the radius of the planet on the side) Geometry tools: G = polygon tool, L = line tool, M = point tool (when drawing geometry you can switch between the three), Editing tools: I = add point (works with any geometry, existing or being drawn), V = move point, X = delete point, T = scissors tool (splits polylines into two segments), Rotation tool: P = edit rotation (changes euler coordinates in .rot file when applied) Shortcuts that aren't tools: Ctrl+D = assign anchored ID (holds an ID in place and shows everything rotating relative to it), Ctrl+E = edit feature properties (really useful for changing how long a feature lasts, when it starts and ends), Ctrl+R = query feature properties (I use this to see its euler coordinates, which are shown relative to the anchored ID) Feel free to ask any questions if you care, I am way too familiar with GPlates by now.
You're coastlines are sexy. I like. Also, hadn't thought about using a vector prog vs a raster, but its definitely true I've had issues with pixel scale using clipart studio to draw my maps. I'll have to look into that.
Not sure if you have or will adress the following.. when you start drawing landmasses closer to the upper parts of the 2:1 rectangle, how do you account for the distortion. Example: the polar landmass, how in the world can you effectively or accurately draw landmasses there? Thanks, LOVE these videos.
About lining up tectonic plates in GPlates. I haven't had a chance to mess with it yet, but from watching you, is it possible to copy existing plate-lines, then copying and pasting them in place and modifying the newly pasted plate-line as a new object? Sort of like copying-pasting things in place in Illustrator.
It would be nice to know how to build climate for a greenhouse earth kind of state You explicitly ommited it at the hot and cold earth video, but the earth has spent the majority of its existence in a greenhouse state, so knowing how to build climate for it might be useful.
Don't make lakes an opacity mask! Make them a separate layer. It will help you later on when you start adding effects to the different topographies. Plus, Artifexian does this anyways in a later step, so go ahead and save yourself the trouble.
Between you, Nathan, and @Tom NixNachname, my brain is kinda spinning on the possibilities of vector-based fantasy cartography (that isn't Campaign Cartographer...). My goal is to have a nice big vector file for my planetary "master map," and then zoom in for continental / regional maps... but it's looking like I can't get the scale consistency one could achieve in QGIS by using Illustrator / Inkscape. Is that correct..?
@@Samuel_J1 i'm guessing it's like 4-5 hours in time lapse. it's probably all done in one day. so it's up to 10 hours. probably spent most of the time editing the video.
Interesting video as always, especially the logic behind vector-based drawing programs as opposed to pixel-based such as Photoshop (especially the one that came with MY computer which has one do so, SO many loops to simulate mask layers since it doesn't come with the ability to do mask layers....) Speaking of which, I'm starting to get the feeling that my initial topography inquiry is less "big picture" planning and more like "let's just draw something that looks interesting and fill in the blanks later" approach. Granted, it does make for a bit more creative freedom, but I can't help but fear that later on in the process when I do something unique and interesting and a subsequent video gives a similar example of how implausible, if not impossible, that particular elevation and coastline design to such a degree that, well, many an hour of work is lost in said process before the correction. As for said workflow, I assume that there are hotkeys on that program that allows one to shift the vector layers between one another without having to manually move them in the layers menu? I don't recall you doing the latter in the speed draw portion. As always, thanks for the upload, Edgar.
Quick question based on your comment on computer generated coastlines: How do you feel about programs designed specifically for terrain creation (the one I'm most familiar with being World Machine)? Specifically when they have realistic erosion filters? I'm in the middle of digitizing a map I've hand drawn using World Machine, and once I got the mountains where I had them on the hand drawn map, the rivers, lakes, and valleys that World Machine put in based on the landscape as it was matched up almost exactly with the features I'd hand drawn. It can also be a great tool for making maps like this, as you can export (raster) color maps that literally just have the topography lines drawn in against a grey background, which is super useful if, like me, you don't quite have a handle on how to figure out slopes.
My hunch for rifts, ravines, and valleys is twofold: first if you're working at whatever elevation, say 1000 m, then maybe those depressions are at 500m. Secondly, I'd look to see how much below sea level depressions on land can get, which, with a quick google search, is a bit complicated. The shore of the Dead Sea is about 400 m below sea level and the deepest part of the floor of Lake Baikal is about 1200 m below sea level. So it seems that it averages to about 700 m below sea level. That could be "realistic". However, if you wanted to add a bit of the fantastic to it (a nice fantastic depth to go with some fantastically higher mountains, say), then maybe your depressions can be as deep as 2000 meters below sea level? These are my guestimates. Cheers!
How did you do the mountains top down? you drew the highest elevation colour and then the second highest, but somehow ended up with the white colour on top? what buttons were pressed there
Interesting I like how it is developing in particular how you only partially followed your boundary guidelines both extending outside your boundaries and into your boundaries as you accounted for coastal features. The one detail I was a little confused on is why did you cut out your lakes at sea level? After all IRL lakes tend to be at far higher elevations than seas. I'm guessing the areas along the convergent boundary where you lack land are going to be underwater volcanoes? As for the view point on randomized methods while it certainly would be problematic if applied haphazard what about processes which naturally tend to be fairly "haphazard" geologically speaking? By that I mainly mean impact based events or hot spot driven magmatic provinces Also are you going to include any sunken continents such as Zealandia or to a lesser extent a microcontinent such as the Kerguelen Plateau? That might be a way to break up those very fast equatorial currents
why did you put the currents below the template layer, and adjust the opacity to be able to see it. why not just put it above it and toggle it visible when u want to see them?
Regarding the elevation colour-scheme: did you know that the earth's elevation distribution is bimodal? MinuteEarth explains: th-cam.com/video/KOv3FGVmRcA/w-d-xo.html Also, you had a question about computer generated landscapes, and your answer assumed a simple noise function. I'm not sure what the state of the art is 10 years later, but RealmOfTheMadGod had a really neat system that shows off a computer's ability for carefully designed random map generation: www-cs-students.stanford.edu/~amitp/game-programming/polygon-map-generation/ The Stanford link is a surprisingly good read, and doesn't overburden the technicalities. MinuteEarth I hope speaks for itself.
One program that does make more 'realistic' maps is World Machine, there is a free version on their website, I actually found it as a tool for making Minecraft maps of all things. Using a flow chart style system that I really like, it does far more than random noise, it can model various terrain types, and different natural processes like erosion and sediment deposition. With manual shapes, you can define the broad strokes of the land you are making, and let it simulate the details of it being a real place. I think it is a very happy "best of both worlds" when it comes to using computer generated topography. Only problem is it is windows only :P I found a basic intro to it, I think it'd be really useful for mountain making and such! th-cam.com/video/tTY_vj4UIEo/w-d-xo.html EDIT: just saw the more recent video, where you mentioned it already, oops!
Great video series :D Why dont you draw in rivers from mountains first? that way you can form valleys with the elevation levels. Or is this something you do automatically?
Thanks for the video! I may be a bit behind, but I was trying to follow your plate tectonics video from a while back and I can't really wrap my head around how to create a map in Gplates with more than two plates? Do they somehow connect at points or are they drawn as completely different shapes? did I miss something in the video?
Line up continents / snap Could you work on the prime meridian side of the map import it back into g plates and move the prime meridian 180* and export it again to work on the other side of the map?
How does he get the level of detail with his coastlines? Whenever I try the same technique on inkscape, all it comes out with is massive curves, despite removing the smoothing
This is a great video on the technical side of map-making, but I've been wondering if you're planning on discussing the actual science of it more. Things like:
- Where to put wetlands, fjords, lagoons, deltas etc.
- How the age of a mountain range affects its topography
- The effect of post-glacial rebound on geography
- Where to put endorheic basins
I'm sure that's in his other videos, he has a lot.
deltas go at the end of a river, near its outlet. this is mainly caused by something that appears at the end of a rivers path or something that was stopping the river from going all the way out to sea. after this, water quickly occupies whatever it can.
He literally has multiple playlists dealing with exactly that, have fun!
@@spcxplrr ....they didn't say all rivers?
RE: computer-generated coastlines using noise and thresholding. This can be manipulated somewhat by putting a couple gradients down as a mask for the noise layer, so that the noise layer will become transparent where coastlines are supposed to be smoother, and more opaque where coasts need to be more rough.
Of course, there are still other issues (generated noise won't create features like deltas, lagoons and bays and such) but I think there's certainly room for computer-generated content, provided you put in the work to make the things it comes up with realistic. And that can be done, even though it's easier not to. I think both hand-drawn and computer-generated approaches are valid. And just like you need to get good at hand-drawing a map, you need to get good at telling the computer how to generate maps. It's just a different skill to the same end.
Yeah. If you're going the lazy route, you'll get bad results no matter which approach you take. Computer generated can be just as valid as handdrawn, but of course just slapping some noisefunctions together is not going to cut it. The actual point is that programming an algorithm to spit out realistic maps is only worth it if you plan on making hundreds of maps. If you only need one, handdrawn will yield satisfactory results faster.
@@TheAgamemnon911 Well, if you're into worldbuilding, it might be worthwhile. I know I've been working on a procedural plate tectonics simulator... Though I guess that's as much about programming practice as it is about worldbuilding for me.
@@Gamesaucer Sure, if you want to do it for the programming practise, that's perfectly fine, too.
@@TheAgamemnon911 I think there may be some confusion regarding what is meant by "computer-generated" in this sense. Artifexian (and by proxy, Gamesaucer) isn't talking about writing a program in a coding language that generates landforms. He's mostly talking about things like adding random noise to an image, and then messing with thresholds to turn that noise into a fractal coastline. And when it comes to time investment, I find that even when you're being really picky and finicky with computer-generation, as Gamesaucer is talking about, it's still going to be less time consuming and require less precise manual dexterity than hand drawing a map. The barrier to entry for computer-generation is mostly just understanding the tools at your disposal and how you can apply them to achieve the same end result as hand drawing.
@@Rotaretilbo Expanding on dexterity: I want to mention a lot of people do have dexterity issues. I swear half of the people I’ve met in my life talk about how they can’t draw bc of their hands being shaky, nerve damage, etc. I’m personally disappointed by the idea of computer generated being somehow lesser but I think that’s because I’m an artist first, not a scientist, and I accept a lot of “well that probably wouldn’t exist” happening.
It’s EXTREMELY hard to get into world building as a community if you’re someone who’s more sensitive to things like criticism due to the weirdly clique-like nature of groups within it. Every subgroup seems to have a blacklist of creators that are seen as bad information sources (which are always big names) and you get scolded for daring to watch the content. Otherwise you’re treated as if you’re already supposed to know all the same information. This is just one example of how hard it is to get into the culture, let alone the practice (the focus on worldbuilding for Tolkien-esque settings also cuts out a large amount of people too). These people are the most informed about how to do this building, and usually won’t ‘look' at you strangely for wanting to know how to do something well. Generally at least there is a mutual admiration for the love and work that goes into this, but that doesn’t balance it out entirely.
When I first saw part 1 of this go up, I was definitely intrigued. I love learning new tricks for getting various "looks" in map designs.
Initially, I was a little bummed that now that we're actually *doing the thing* in episode 2 and laying out the topography that it's effectively a straight-up "hand drawn by educated eye" approach. I feel like the style amounts to "get good, son."
That's not a criticism, though. To steal your words in a different part of the video, I think that's a feature, not a bug. Your comment about something worth doing being worth taking the time to do properly struck a chord and is something I think I kinda needed to hear.
I love your channel and love your presentation style! Keep rocking!
Sidenote: Your expression during "EDGAR OUT" makes my day every time. I don't know what it is, but something about it is just the perfect ending to your educational style.
"it's not a bug, it's a feature" - me in a nutshell
One (or several) criticisms I have with the map you are building here: Your coastlines are all equally jagged with lots of bays and little islands. There is nothing wrong with having straight sections every now and then, especially where the currents would deposit sediment. And on the other hand, your mountain range is not jagged enough - meaning that I miss the signs of erosion in the form of long and branched out valleys that rivers typically create. Also mountain ranges along faultlines would typically have multiple rows of peaks running in parallel.
Maybe he will add erosion due to wind, rain and rivers in another video. Notice that he added lakes, but no rivers to feed them (so, I think, those lakes are depresions filled with rainwater, and they would not exist in the rainshadow given by the mountain, but I think he skipped that detail for the sake of a short video).
@@evecol4525 You mean the coasts? Best example of sedimentary depositing by currents: Baltic sea: Coasts of Latvia, Lithuania, Poland.
@@evecol4525 Oh, in that case: Look at the Andes, especially in Bolivia, but any currently geologically active mountain ranges will work: Himalayas, Alps, California...
you inspired me to create my world
I know this series is about illustrator, but with what you said about drawing by hand; I draw all my maps in MSPaint and try to get something i like the look of, and its so rewarding to see the worlds i created, and trying to make it feel lived in is so rewarding knowing that i used a mouse in a crappy microsoft app to draw these land forms
The map looks so realistic
I just randomly started this series because I found it interesting and oddly satisfying, but I just wanna say your little speech about getting better with practice and "you 100% will get there" was really nice to hear. :) I've been having a lot of writers block so thanks for the motivation!
The map looks really good!
I went back and looked at your original video on tectonic plates and it's really cool to see how much better this is. The original video was good, but this one is amazing, and that's reassuring because it means that we can eventually produce some really awesome stuff too.
Wow, wow, wow. This is both facinating and overwhelming at the same. Ill have to watch this when i can fully give my attention. Making a planets worth of height maps has become an incredibly larger task than my already present understanding of geography, geology, ecology, and climate will allow. 😖😄🙃
I think the best way to go about it is by using tools like World Machine which can simulate realistic terrain including things like erosion, currents, beaches and more
I mainly use G Plates to visualise prehistoric Plate movements in my world. I love to form a world from some starting landmasses colliding, new land forming and interacting with the other ones.
I feel like it makes the world more vivid when i have a clue about how tectonics in my world work rn. Also having mountain ranges at places elsewhere then the center or coast of a continent is quite nice.
A trick I have learned when trying to make coast lines is to take paper get part of it wet and you should see rough edge let it dry and you can trace that that boom coast line.
I've heard of people using the rust of metals, but this is new
This video was/is a HUGE help to clear up where my next diirection should be. Ive already made my general drafts on paper, next is building the heightmap itself on the computer. So glad to see that I was on the right path regarding realizing i will have to hand create elevation by elevation.
this video in essence:
"draw some circles. Now draw the rest of the owl".
All of these crevices, and peninsulas and islands and stuff, how can you come up with them on a spot?
I know what you are gonna say, study a lot of maps, but you are supposed to get something from these maps, right? Some principles. I wouldn't even guess that ocean currents can affect the global shapes of continents, even if I would study maps to the end of times, for example.
as he said, look at real maps and get a feel for it, doesn't have to be realistic, just look realistic.
He has a video on ocean currents i think, so if you're unsure how they effect landmass, check that out. Alternatively, look at cool coastlines that you want to replicate and research why they're like that, it'll help inform where the right place to put them on your map would be. It basically comes down to research i guess. Looking things up so that knowledge can inform how you go about making the shapes.
What I’ve taken from looking at these is be precise, not accurate. Meaning be specific regardless of if it’s realistic
Mostly looking at atlas maps, fantasy maps (depending on your needs), how currents and tectonics affect the coastlines and height shapes and get a feel for it. Having made several maps by hand; the crevices, plateaus, peninsulas and island shapes are something you mostly come up on the spot, either because you feel like it would look better, or because you want to make a purpose out of it (world-building reasons). However with prior knowledge of maps, you can make them look realistic because what you will come up with are the random patterns you have been seeing on other maps. "It just works"
My workflow, like Artifexian has always been to make rough shapes, according to my needs, then procedurally add more details to the shapes. You will never get a scientific, down-to-the-maths method of making continents due to so many factors being at play, it's just more complicated to do it with maths rather than randomness and patterns, so it's sticking to the feel and how good/realistic it looks coupled with map knowledge. I'd say having a general grasp of how it works is what makes or break your map, not because it isn't realistic, but rather because it looks like so.
If anything regarding map learning, look at how random most of the shapes are, you don't need to be intensively scientific about it, like I said before getting a grasp of some concepts is enough. After that, it's mostly a matter of practise.
That said though, be iterative! A map is not set in stone you can always improve it later on if you learn a couple things along the way. To me it has always been a back and forth process between looking at maps and making some.
there are some other videos where he explains this kind of things. In this one, the intention is to show hoe to draw the maps using the softwares he's showing us. If you want the answers for the questions you have, I suggest you to look for the other artifexian's videos.
13:35 Re: hand-drawn vs. computer-generated topography, I think a reasonable combination of the two methods is probably the best way to get realistic shapes, landforms, etc. In my own conworld mapping efforts, I've used a global plate tectonics sketch to get the general placement of landmasses, then used something like the "clouds filter + threshold" technique you referenced to get semi-detailed coastlines that I can then further refine manually to suit local climates and ocean currents. Similarly, when designing topography, I'll start from a hand-drawn rough draft of where I want the major elevation levels to be, export it as a series of masks, and use them in something like Wilbur to get good-looking erosional features, river networks, etc.
P.S. 15:53 I actually pronounce QGIS the same way you do, so no complaints from me. :D
Your map is simply georgeous, I would not have your patience
When he said get a drawing tablet I was sad 😔
It’s really not necessary. You can do fine with a mouse! Deff would not recommend a laptop trackpad though. If u do have a laptop u can just buy a cheap USB mouse to use with it, which is a lot more affordable than buying a drawing tablet, although u can get those for pretty decent prices nowadays too. But I totally understand that it’s not possible for everyone, or myself even, currently.
this is old but for people in the future, "one by waycom" is like 35 pounds for the small version, and it's more than enough for drawing maps
12:47 There is a snapping function. When you're in edit vertex mode, under the list of the coordinates of your points is an option called "snap nearby vertices." Sadly it only works with already extant vertexes, so you have to create your shapes first and then snap them to each other.
I know it’s from 4 years ago but thank you! Im gonna try this out
omg it works!! thank you so much it saves me so much time!!
@@flet7355 glad I could help
in gplates if you use the polyline feature (L) instead of the polygon feature it will not force a close loop, meaning you only have to connect 2 points to a plate, then when you create the feature the 2 plates will share the boundary.
Just a suggestion, you might want to highlight the differences between the technique you're using to draw your map and how it differs from Nate's as well as the pros and cons of both methods. Where Nate uses photoshop and masking in layers of elevations using the lasso selection tool and the fill command using raster graphics (unless he has changed his workflow recently?), you're using vector and creating a bunch of shapes for each elevation before merging them and arranging them. There are so many ways to do it, and each person may prefer a different method, or one might work better for their work flow than the other.
Hi Conrad :) He does touch upon the differences in this video - I think the main difference is that the vector-based programs like inkscape or illustrator are less of a strain on your computer due to them working with equations rather than pixels - I use photoshop, and take it from me - large raster maps with multiple layers will cause your machine to lag like hell.
Though I use mask you can easily just use the pen tool, filling in the selections with the desired colour, eliminating the mask - its just a matter of preference.
@@nathanMangion_Elyden I really like the idea of using the mask... but also the idea of using vector for maps. I guess it's one or the other though haha
First of all, you are a really great resource and inspiration, thank you for that.
Over the years I always worked at my world bit by bit, but never was quite content and reworked it a lot. That process led to trying out various methods and styles, and was all about learning and learning. I did a rough world map with plates, currents and weather zones etc., but figured in the end, that it would be just too much, because my story is based on just one sub continent. But it helped nonetheless. I have a rough idea of the surroundings and influences on my continent.
Right now I redrawing and reshaping the map as lineart inspired by old maps and the maps of middle earth. There will be places, that will be seem wrong, but old maps of our world looked far worse, so I don't worry to much. And I may do another map with this method you showed, because it looks quite neat and more accurate. But brawing by hand gives me time to reflect the history of some landscapes, how they formed and what took place there. So even it is time consuming, it helps me a lot to figure out, in what a world my story will be situated.
For research I love to go to bookshops and just let me surprise what I'll find there. So I got a huge book of the history of catography and also an atlas of imaginary places - and a lot more of various sort.
Wall of text, check.
So something that comes to mind for realistic, if not improbable, world building is how many a person think having 4 and 6 limbed vertebrates would be unrealistic due to most if not all vertebrates coming from a common ancestor. But if one where willing to go into their worlds history and depart drastically from how earth evolved, you could have early vertebrates from 2 different groups be the primary colonizers of 2 different continents. Evolve separately for millions of years and inevitably the continents will come together and separate many times redistributing organisms. Making for an interesting biosphere. Of course this applies to more than just animals and limb count. Plants and fungi could also be quite different. It would be like having two planets in one.
You have to make sure that those continents will never get closer
I would suggest looking up the program Wilbur, as it doesn't seem like your method will be using it. It can be used to model rainfall and erosion. I have used it to place rivers and it seemed fairly realistic, though it struggled with lakes.
6:46 I did that with my world Etsaiñ. I visualized it in an Equirectangular and didn’t use gplates except for plate details. I just made things wider near the poles
It’s funny. I love this channel because you like to have realism integrated into your builds. I love science so I know the realism, and I’ve wanted to make a fantasy world for a long time. I have even conducted research on unknown topics to get a more realistic result. I then integrate everything in a way where I start from creation of a world, then jump to different maps to show the progression of time. I try to get the most realistic ways to make a map.
Great video with a sexy sexy map. What are your thoughts mythological world building? It's a totally different style than what you seem to find enjoyable, but like Middle Earth doesn't take into account any plate tectonics or ocean currents- the entire thing is created by gods screwing with each other. I hate the "A wizard did it!" thing, but it seems like a narrative approach to land formation totally contrasts doing it scientifically.
Oh My God! Just discovered this, it is amazing. I now feel inspired to make on myself. Coming from a D&D background this would be really helpful but also a cool way to wow my friends. Wish me luck!
It's amazing how worldbuilding can suck you in. I was just looking for a way to add some realistic-ish climates to a fantasy world I'm developing for writing, and now I'm seriously considering making an atlas map even though I'll never once need it to tell the story I wanna tell, and no reader will ever give a damn. But this. This is too cool to pass.
I would absolutely recommend using Blender to draw on a 3D sphere if you don't already know how to use Blender.
It is never a bad idea to learn how to use a new program.
I'm trying to see what Earth would be like if the geographical poles moved to different locations. In my case one is parked over eastern Brazil and the other on the opposite end in SE Asia(two strong magnetic anomalies), creating an equatorial and subtropical ocean system that would go all the way around(eg use the trade winds to go completely around the world). I have the 2nd largest tropical rainforest on what is in our world the Canadian Prairies, Northern US plains, and the foothills of the Rockies. The largest being a strip down what is now Africa tilted 90 degrees on the map. since the sea(and also the land to less extent) is pulled out more at the equator, the sea levels are around 500m higher. This forms a large Shallow sea. The "Saskatchewan sea" extends from tropical islands(formally high arctic mountain ranges) down to the coast that runs through what used to be Alberta, Saskatchewan, Dakota, Wisconsin etc. Sea level drops closer to the poles has cut off the Gulf of Mexico turning it into a highly saline inland sea and desert that sits below sea level. Cities that "survived" whatever cataclysm caused the changes(whether by time or not) and the oldest cities in the world are ones that were well established before, during, and after the transition, or left behind valuable ores to mine. Likely what was formally Alberta would be rich in resources would be one of many, the shallow Saskatchewan Sea bringing a bounty of fish. However the tropical climate would limit the kids of crops grown, however the abundance of metals, oil, fish, and fruits can be traded for wheat, corn and other colder climate crops grown further south. The new nature of the world affords easy sailing anywhere without the need for massive canal systems, global trade is considerably more robust due to this. There is less extremes due the tropical and subtropical waters being uninterrupted and both poles sit on continental plates. Australia would have transitioned from the arid desert it is today, to a more mid latitude breadbasket, the outback is green and full of farming.
I love the way you pronunced my last name hahaha.
Also, you conviced me to buy a drawing tablet. I've always wanted one but given that I can't draw for the life of me, I've never seen the point of it. Now I do.
Also Also: there is a way to "snap" points in GPlates. With the "move vertex" (V) tool activated, there is a box below the list of points that make up the shape that reads "snap nearby vertices" together with a threshold in degrees and a restriction by plate ID.
"I did a bit of work off camera to save time"
We all know you used creative mode to get those diamonds, man
Your channel is so cool! I don't you to be such an inspiration! Please keep doing what you do!! 😁
I love the idea of you commenting on the previous video's comments - thanks for addressing mine btw.
8:55 all those 'little' islands between those two continents, is that really how they can grow? I don't really see anything like it on earth. Maybe something like how Iberia presses up to the Atlas area or Indonesia or, maybe the best example is Nunavut in Canada. But they all seem quite different. What is your idea here?
17:13 haha got surprised by that little lake there? xD
Hey Artifexian, an amazing video, thank you. Re: Photoshop - I am not entirely sure what is the advantage of doing this in vector to be completely honest, apart from the ability to easily toy with the scale of the landforms, which is something you don't really want to do on a map anyway, as it would distort the complexity of the for example coastline. Yes, you can zoom in here, but I guess it kinda depends on what the map is supposed to be for. If all you want is just a map everyone will view at a certain, fixed scale, without interactively zooming in or out, I am not sure vectors have any big advantage. For the purpose of worldbuilding itself, sure. Anyway, you can for example make a brush in Photoshop which will make semi-random, jagged coastlines on its own, and it will be far more simple than trying to make all the shapes happen just with a manual pencil, which is in my mind a prohibitively tedious way.
Hello. This video has been such a great help, I don't know where I'd be if I hadn't seen it. I wish I had the app you were using (I'm using autodesk). Your a continuous source of help and inspiration. Thanks a lot, Artifexian!
Couldn't agree more with the criticism of the fractal land generation, one thing those tutorials seem to do well though is recommend the program Wilbur for simulating erosion which seems to do a really excellent job.
Heat death of the universe joke. I love it.
Good morning!
Good afternoon!
Good night!
@9:20, my first map was in paint, so I didnt have any layers, or any projection tools. I was finally able to project it onto a globe using Google Earth Pro, and its a bit wonky but its still fun.
One detail I noticed: the strokes of your various landmasses and plates (actually, all the strokes you imported from the snapshot), were scaled up from 1 pt to 12.8 pt, giving you space to add detail inside those lines. If you scale the snapshot without also scaling the strokes, then the strokes mantain a value of 1 pt, and that space disapears.
On the software and such, using a program like Maya where you can get it to update textures to PSD files (no clue if blender has that? Probably does?) could be a cool way to train yourself to draw polar distortions
(As in, draw what you think would work, see how it maps, analyse, change, update the 3D program and repeat until it looks cool. Would probably also help for knowing how to draw in coastlines in the polar regions while already accounting for distortions)
PS has some 3D functionality now, you might be able to do it there. Though switching back from 3D to 2D via undos is a pain. There's probably a better way in there that I don't know about yet.
It would've been really cool to see this process done with your map from the tectonic and climate tutorials, see the whole thing from start to end.
How do you recommend dealing with the problem where parts of your maps are at different fidelity than other parts?
The "resolution" in your mountains seems less than that of your coasts.
Try keeping the zoom steady. Don't create part of your map with 300% and another with 150%. The bigger the zoom, the more detail you tend to add.
Also, don't use too big of a zoom, otherwise you lose the greater scope of your map.
Always drawing the same zoom level is a good tip. As for my mountains in particular, they aren't done yet. Their resolution will be matched to that of the coastlines.
When you mentioned blender i realized that you're actually drawing a type of elevation map, which can be used in building a 3D version of your map. Not only would you have a story for your world but you could also walk and run around in this world too.
One thing you might want to look into for computer-generated geography is a program called Songs of the Eons ( look on reddit.com/r/SongsOfTheEons ). It's primarily a large-scale RTS sort of video game, but the interesting thing is that the devs have been building their own world-generating system from scratch, taking into account stuff like plate tectonics, ocean currents, even wind erosion and other things that would be too tedious to do manually. I haven't tried myself (it's still technically in development and unfinished) but I think you could probably control certain factors to guide the program toward the sort of world you're looking for. Once it generates the globe for you, you could use it to make any number of maps. I would highly recommend that anyone interested should give it a glance.
There was a video or something a while back where someone made a 3D "ball" out of hexagonal shapes. So if someone is creative they can make a globe map in this way.
There's a number of ways you can project a sphere onto a grid of hexagons of triangles - look up 'geodesic sphere' for some explanations and imagery, but in essence you start with a regular polyhedron - an icosahedron is a popular choice. Then there is a method for subdividing it until you get a resolution that satisfies you, and then it will unfold easily along the lines of those triangles, and you can use that to independently render continents with minimal distortion.
for other people who don't have illustrator but have photoshop: the pen tool in photoshop is the equivalent of the pencil in illustrator, you can either do it by point or freeform too
The physical globe method is what Chris Wayan of World Dream Bank / Planetocopia have done for some of his planets. But he also often does preliminary drawings on paper.
i definitely did my last map by generating clouds in photoshop and adjusting the threshold. It didn't even occur to me that I was ignoring things like the effects of ocean currents. Lucky for me, I've become completely disenchanted by my map and I'm eager to start another... and then I peeked over here, and spotted some mapping videos. B) I felt like a hack at just the right time.
I think the question with using gplates to just draw on a sphere comes up because some versions of illustrator already have that and for the ones that don't you can download free scripts and plug-ins to do it.
If you're using inkscape, you'll have to use something like gplates, but if you're using Illustrator version 10 or newer, you have the option to do that in-app.
what is the option to draw on a sphere in illustrator?
If you a re looking for realistic you should draw the highest peaks then create river systems to guide you, this will also place your major lakes and determine coast lines where river deltas will create many islands.
Also temperate climates will tend to less jagged train as erosion is temperate - not harsh like deserts or tropics
There is in fact a snapping function in GPlates, although it's a bit finicky and i don't know how exactly you access it. But there defo is one, i used it alot to sort out my plate tectonics.
Regarding Strider‘s comment about not drawing the topology by hand: I agree that using fractal generators, perlin noise or similar isn‘t a good solution. What would be useful (and I think that is what they were getting at) is a tool that actually simulates plate tectonics and erosion in a realistic, physics-based way. That probably still wouldn‘t be super accurate but a good starting point.
Well, if you do a Q&A for the next one: any advice for making maps based on existing topography? I'm working on a project that takes place in the area of my home state several thousand years in the future, and I was wondering what you'd recommend for creating accurate "future" maps for that area.
2:59 can definitely be used in more than one context
I've used gplates to create topography in the past by creating new features, filling the polygons, and making them different colours.
Looking at maps of earth we can see plenty of mountain ranges and other geological features created by plate tectonics, which are based on historic plate activity from plates which no longer relate to the current arrangement shaping earth (e.g the Urals in Russia or the Great Dividing Range in Australia).
Do you have any advice for placing topography based on no longer present plates.
Do you have a notional idea of craton/basin likely placements in your head when setting out topography?
I'm loving this series. Thank you so much, Edgar.
i showed off the map i drew and someone linked the 1st vid of this tutorial to me, a wasnt plan on usign this guide cuz im just gonna make it simple as a guide for my story but i got intrigue, and wow!!
i wanna try to do this but due to my laptop being a potato, i cant use Adobe photoshop/illustrations at all!
tho im gonna take notes on how to so when i got a better laptop, imma do this
You are obviously drawing like the continents are flat. It is there any noticeable inaccuracies that anyone should take notes? I mean, take the final map and load it in G-plates and I am sure that some coast lines or rivers will look weird, specially close to the poles. I love your channel and your work!
This cannot come out fast enough.
how i managed to avoid the whole "get a tablet" problem: my world is impossibly small (r = 0.22R) and so all those details are bigger and therefore take less time to draw in with a mouse.
"I think you'll find it a hugely rewarding process"
My thought process through watching this: Man this looks so fun. Where's my tablet?
I might be the closest thing to a GPlates expert (with regards to world building) there is at this point. I discovered GPlates a long time ago, and have since started a very large project involving it in order to make a realistic world. Here are a few of the things I've learned and the tips I can give:
- To make things align, there is a way to sort-of align things. When you click "Create Feature", you can assign a conjugate ID and enable "Create Conjugate Feature", which will make two features, one with one ID, and another with the other (e.g., you make a line with ID 1 and conjugate ID 2, GPlates will spit out 2 identical lines, one with ID 1, the other with ID 2). This works best with polylines.
- You can save a whole lot into one feature collection. In my project, I have a file for continental plates and a file for oceanic plates. Continental has over 100 individual IDs (I wanted to use microplates), and Oceanic has around 20. IDs are used primarily for rotation in the .rot file, features with the same ID will use the same rotation scheme. I noticed in the original GPlates tutorial, he saved the different continents into different files-- I don't think I can stress enough how useless that is to do. Put them all into the same collection, but ID them differently.
- The .rot file does not care what feature collection something is in-- anything with an ID that has a rotation in there will move according to the file.
- Feature collections are largely for coloring things differently and hiding layers, I think of them more as layers than files.
- The best way to align things as far as I am aware is to zoom all the way into a point and draw the shapes directly on the points. If you're *really* OCD, you can "Edit Feature Properties" and edit the coordinates themselves (it's under "gpml:unclassifiedGeometry"), but honestly, that's way too overkill IMO.
- Oo, shortcut keys:
General tools: F = select tool, D = drag tool (hold shift to rotate, hold control when using other tools), S = measurement tool (you can change the radius of the planet on the side)
Geometry tools: G = polygon tool, L = line tool, M = point tool (when drawing geometry you can switch between the three),
Editing tools: I = add point (works with any geometry, existing or being drawn), V = move point, X = delete point, T = scissors tool (splits polylines into two segments),
Rotation tool: P = edit rotation (changes euler coordinates in .rot file when applied)
Shortcuts that aren't tools: Ctrl+D = assign anchored ID (holds an ID in place and shows everything rotating relative to it), Ctrl+E = edit feature properties (really useful for changing how long a feature lasts, when it starts and ends), Ctrl+R = query feature properties (I use this to see its euler coordinates, which are shown relative to the anchored ID)
Feel free to ask any questions if you care, I am way too familiar with GPlates by now.
You're coastlines are sexy. I like.
Also, hadn't thought about using a vector prog vs a raster, but its definitely true I've had issues with pixel scale using clipart studio to draw my maps. I'll have to look into that.
How do you do such detailed and accurate coasts and topographies? Mine are either overly smooth or ridiculously jagged.
7:30 thats a really interesting feature, not hear for how to make an atlas just studying more realistic land masses😊
Not sure if you have or will adress the following.. when you start drawing landmasses closer to the upper parts of the 2:1 rectangle, how do you account for the distortion. Example: the polar landmass, how in the world can you effectively or accurately draw landmasses there? Thanks, LOVE these videos.
About lining up tectonic plates in GPlates. I haven't had a chance to mess with it yet, but from watching you, is it possible to copy existing plate-lines, then copying and pasting them in place and modifying the newly pasted plate-line as a new object? Sort of like copying-pasting things in place in Illustrator.
It would be nice to know how to build climate for a greenhouse earth kind of state
You explicitly ommited it at the hot and cold earth video, but the earth has spent the majority of its existence in a greenhouse state, so knowing how to build climate for it might be useful.
Don't make lakes an opacity mask! Make them a separate layer. It will help you later on when you start adding effects to the different topographies. Plus, Artifexian does this anyways in a later step, so go ahead and save yourself the trouble.
Between you, Nathan, and @Tom NixNachname, my brain is kinda spinning on the possibilities of vector-based fantasy cartography (that isn't Campaign Cartographer...). My goal is to have a nice big vector file for my planetary "master map," and then zoom in for continental / regional maps... but it's looking like I can't get the scale consistency one could achieve in QGIS by using Illustrator / Inkscape. Is that correct..?
Artifexian please show us how to make Hybrid type star systems with multiple planets having multiple moons in Universe Sandbox
Also I'm pretty sure it's not impossible to write a map generator that takes into account ocean currents and plate tectonics.
Out of curiosity: How long did it take you in real-time to do all the drawing that we see happening in this video?
I also want to know the answer to this. Looks like it took hours just to do that one island :/
@@Samuel_J1 to make something yourself that's everything you want, it will take time. Art requires time investment to produce something
@@AimlessSavant oh, I know. I've drawn my own map in Photoshop and it's taken hours and hours and its still not as detailed as this.
@@Samuel_J1 i'm guessing it's like 4-5 hours in time lapse. it's probably all done in one day. so it's up to 10 hours. probably spent most of the time editing the video.
Interesting video as always, especially the logic behind vector-based drawing programs as opposed to pixel-based such as Photoshop (especially the one that came with MY computer which has one do so, SO many loops to simulate mask layers since it doesn't come with the ability to do mask layers....)
Speaking of which, I'm starting to get the feeling that my initial topography inquiry is less "big picture" planning and more like "let's just draw something that looks interesting and fill in the blanks later" approach. Granted, it does make for a bit more creative freedom, but I can't help but fear that later on in the process when I do something unique and interesting and a subsequent video gives a similar example of how implausible, if not impossible, that particular elevation and coastline design to such a degree that, well, many an hour of work is lost in said process before the correction.
As for said workflow, I assume that there are hotkeys on that program that allows one to shift the vector layers between one another without having to manually move them in the layers menu? I don't recall you doing the latter in the speed draw portion.
As always, thanks for the upload, Edgar.
*time to pirate illustrator to make hot maps*
2:17 Actually, I'm perfectly fine with using my computer pad, that's what I've even been using for years now.
this is awesome
Quick question based on your comment on computer generated coastlines: How do you feel about programs designed specifically for terrain creation (the one I'm most familiar with being World Machine)? Specifically when they have realistic erosion filters? I'm in the middle of digitizing a map I've hand drawn using World Machine, and once I got the mountains where I had them on the hand drawn map, the rivers, lakes, and valleys that World Machine put in based on the landscape as it was matched up almost exactly with the features I'd hand drawn. It can also be a great tool for making maps like this, as you can export (raster) color maps that literally just have the topography lines drawn in against a grey background, which is super useful if, like me, you don't quite have a handle on how to figure out slopes.
thank you so much
How would you represent land lower than sea level like a rift, ravine, and possibly a valley
My hunch for rifts, ravines, and valleys is twofold: first if you're working at whatever elevation, say 1000 m, then maybe those depressions are at 500m. Secondly, I'd look to see how much below sea level depressions on land can get, which, with a quick google search, is a bit complicated. The shore of the Dead Sea is about 400 m below sea level and the deepest part of the floor of Lake Baikal is about 1200 m below sea level. So it seems that it averages to about 700 m below sea level. That could be "realistic". However, if you wanted to add a bit of the fantastic to it (a nice fantastic depth to go with some fantastically higher mountains, say), then maybe your depressions can be as deep as 2000 meters below sea level? These are my guestimates. Cheers!
Very cool! Do you draw with the mouse?
I already have an illustrated map but I need to generate topography onto an existing bitmap image.... how would I go about doing that?
How did you do the mountains top down? you drew the highest elevation colour and then the second highest, but somehow ended up with the white colour on top? what buttons were pressed there
17:15 the new Blender 2.8 is relatively simple to use.
"You could go out and buy a cheap ball..."
Or you could use the one already in your basement.
Interesting I like how it is developing in particular how you only partially followed your boundary guidelines both extending outside your boundaries and into your boundaries as you accounted for coastal features. The one detail I was a little confused on is why did you cut out your lakes at sea level? After all IRL lakes tend to be at far higher elevations than seas. I'm guessing the areas along the convergent boundary where you lack land are going to be underwater volcanoes?
As for the view point on randomized methods while it certainly would be problematic if applied haphazard what about processes which naturally tend to be fairly "haphazard" geologically speaking? By that I mainly mean impact based events or hot spot driven magmatic provinces
Also are you going to include any sunken continents such as Zealandia or to a lesser extent a microcontinent such as the Kerguelen Plateau? That might be a way to break up those very fast equatorial currents
why did you put the currents below the template layer, and adjust the opacity to be able to see it. why not just put it above it and toggle it visible when u want to see them?
Cool Video.
Regarding the elevation colour-scheme: did you know that the earth's elevation distribution is bimodal? MinuteEarth explains: th-cam.com/video/KOv3FGVmRcA/w-d-xo.html
Also, you had a question about computer generated landscapes, and your answer assumed a simple noise function. I'm not sure what the state of the art is 10 years later, but RealmOfTheMadGod had a really neat system that shows off a computer's ability for carefully designed random map generation: www-cs-students.stanford.edu/~amitp/game-programming/polygon-map-generation/
The Stanford link is a surprisingly good read, and doesn't overburden the technicalities. MinuteEarth I hope speaks for itself.
Thanks for that link, that was interesting!
(I mean the Minute Earth video)
One program that does make more 'realistic' maps is World Machine, there is a free version on their website, I actually found it as a tool for making Minecraft maps of all things. Using a flow chart style system that I really like, it does far more than random noise, it can model various terrain types, and different natural processes like erosion and sediment deposition. With manual shapes, you can define the broad strokes of the land you are making, and let it simulate the details of it being a real place. I think it is a very happy "best of both worlds" when it comes to using computer generated topography. Only problem is it is windows only :P
I found a basic intro to it, I think it'd be really useful for mountain making and such! th-cam.com/video/tTY_vj4UIEo/w-d-xo.html
EDIT: just saw the more recent video, where you mentioned it already, oops!
Great video series :D Why dont you draw in rivers from mountains first? that way you can form valleys with the elevation levels. Or is this something you do automatically?
Thanks for the video! I may be a bit behind, but I was trying to follow your plate tectonics video from a while back and I can't really wrap my head around how to create a map in Gplates with more than two plates? Do they somehow connect at points or are they drawn as completely different shapes? did I miss something in the video?
Line up continents / snap
Could you work on the prime meridian side of the map import it back into g plates and move the prime meridian 180* and export it again to work on the other side of the map?
How does he get the level of detail with his coastlines? Whenever I try the same technique on inkscape, all it comes out with is massive curves, despite removing the smoothing