I'll add two small notes that kind of got cut in the edit. This video largely covers the problem with skies that you might encounter in almost all platformers. With that said, if you are making a metroidvania or roguelike, you will generally have to make a lot of your game underground for a totally separate reason. That is, you need to make part of the game underground in order to make the map wrap around on itself correctly. In that case you'll get a map like ghost song or aeterna noctis, where you have 2 or 3 areas above ground, but most areas will still be underground. It's a separate point to the video so it got cut, but it still might be a relevant tidbit. The second point I would like to stress, is that if you are as skilled of an artist as the people making Wonder Boy or Ori, then it can make sense to have an "aesthetic scene" that grabs attention. But if you are similarly skilled as me, attempting to make a nice looking horizon might actually turn out worse than just going for the simple "hollow knight approach". It'll only look nice if you are good enough to make it look nice, otherwise it'll both look bad, and play bad.
but it doesnt always have to be underground right it can be a large magical mansion or a spaceship(like metroid does) so its just have to be a big closed off space but yeah if u make it underground u can add things later down the line thats a plus
Yes, underground is generally a good theme to go by. But why shouldn't solo metroidvania devs explore scenarios like mountains, tree top buildings, high-up towers, suspended bridges, or even several "on the ground surface" locations in addition to those? Underwater is also a great theme. There is so much that can be done with parallax layers to make the horizon look good and consistent with the height, even if you're not a skilled artist. But even if you're not making it look decent, I'd argue it's still more interesting than the monotony of making several foggy underground biomes in fear of making anything amateur.
Hard disagree. The "empty" feeling can sometimes be a great way to give players breathing room. A lot of times in games that are meant claustrophobic and intense, it can feel very nice to finally have some open above ground gameplay, even if there isn't much to see. Sometimes a smooth gradient with some nice granular clouds is all you need.
Pacing and giving the player moments of reprieve are where those 'boring' scenes are great. Understanding how these different environments play a role in the "feel" of a game is where the problem comes. Not every scene should be focused on giving the player something to do. Some are to evoke a feeling.
Ye. This might be fair. But I think a portion of my argument is that for many solo devs it is probably the case that you shouldn't show the sky. As in the ROI on it is probably terrible. I agree that the substance of the video is that it is difficult, but can be done at times or with some compromise. But by and large I want the takeaway to somewhat be that you should truly consider if the difficulty makes it worth it. With that said, I kind of agree that the title is a bit sensationalised, but it's quite tricky titling videos overall.
Oh woah, I clicked on this in my recommended feed and did not expect to see my game mentioned! Great video, and definitely a real issue that I did not think of until I encountered it myself. The vertical camera panning issue is a big reason why I kept room sizes pretty small in Islets. When a room is too big, it's hard to keep the parallax from just showing emptiness when you're too high up. But when it's small, the parallax range is smaller, and you can plan for it to show something at pretty much any position.
Ye good point! I am assuming this is a shift you made intentionally after sheepo? Cause when I compare the two I can see that the rooms in sheepo are generally significantly larger and more zoomed out, but also run into the issue of being empty at times (except for clouds). Btw, really like both your games and admire the artstyle a lot, both games manage to do a lot while still being simple. It's kind of problematic to look at games like Ori and blasphemous as examples of what to do at times because they are frankly impossible to emulate as a solo developer. Both your games look awesome but still manage to be fairly economical (time-wise) in their approach.
@@Nonsensical2D Yeah definitely, Sheepo had some issues with visual layout (also seen by the map being "impossible" since rooms that were 1920x1080 showed as the same size as rooms that were 12000x12000px), and decorating the larger rooms became a bit unmanageable. I'll definitely be rethinking the ways I design areas with the next side-scroller to keep decoration less of a chore lol And thank you! Yeah I think it's partly that I just get a bit exhausted working on a project for way too long, and also just that making games pretty quickly makes the odds of finding success a bit better. I kind of follow the 90/10 rule, where the last 10 percent takes as long as the first 90 for most aspects of game development, so for anything that doesn't need to be 100% perfect, I just skip that last 10% of polish and just save it for things that are a bigger focus (player movement, core art, etc). Maybe is a little lazy, but saves a bunch of time and only sacrifices things that would make a negligible difference
It's worth noting that those mountains at 1:56, as demonstrated in the clip are very short and relatively close. If you create background assets with the idea that they are bigger in scale and further away, the parallax effect makes their faked movement closer to the camera's and you get to have an interesting scene for longer, and also gives a sense of grand scale to the environment. Also about Monster boy I think its in general a bad example for this point because of how uniquely it was made. It was new art placed on top of a very old game, so it's not that the level is flat and empty because the art forced the hands of the designers, but the game already had a very flat and simple level, and the artist decided to fill it up with gorgeous environments so the massive empty space felt purposeful.
When it comes to Wonder Boy I know the art was placed on top of an old game, but it felt like talking about that would detract from the video. I don't think the artist did a poor choice in this particular circumstance, but I do think that the level design in wonder boy is incredibly dated, it frankly feels boring to play. And that section is no different, I don't think it holds up gameplay wise. So I took it as an example largely because it showcases the problem that occurs quite well (and I wanted to talk about the cool parallax foreground), the scene looks amazing and also plays really terribly. I think the cases with Ori or Ghost song are better displays of what happens to the level design when you create an open sky. But frankly there are a lot of games you could look to. I just selected a few that I think illustrated the issue decently well.
Unrelated to the video, but I've noticed the line weights in your art are working against you. Your player character has these super thin lines while elements from the environment often have much bolder lines. I'm guessing this is a product of wanting to make your more important elements more detailed and the less important elements more stylized, which makes a lot of sense, but heavy outlines have a strong association with "intractability" (think cuphead), so the end result looks a bit disjointed and ameture as your eyes are being pulled towards the elements in your scene that are the least important. There's many ways to get around this: Making your character designs more simple to work with a thicker line weight, like hollow knight Giving your environment art a thinner line weight Coloring the outlines of the environment art to make it blend in more (I dont think this alone would be enough, but I think it could help Removing the outline of environment assets altogether (would completely change the artstyle)
Thanks for the feedback! The truth is that I make all my scenes separate from my actual game art, they are usually made quickly to try something out or illustrate a concept, so I don't spend time on it, sometimes I use thick line art, sometimes I use thin line art. But they are by and large completely incompatible with the character since I always use the same 2 characters regardless of the scene. So I have actually played around with most of the suggestions you mentioned, you can even see some scenes in this video with thin lineart, others with thick, some with coloured lineart and some with black lineart. This desert scene also looks off in terms of lineart because I've stolen one or two assets from a previous scene whereas the cactus was made for this scene. But I really do agree with you, If I were to release a game, I would have to ensure consistent line size to some extent :)
This problem dates back to 1989 with Shadow of the Beast, where many parts of the game is just the player running around big open horizons, presumably to showcase the parallax scrolling, which was basically a new thing for computer games at the time.
you should have mentioned one thing though , sometimes its actually better than making stuff indoors , depending on the gameplay , most of the examples you showed were platformers that need obstacles or would become boring , but though i have made a prototype for a game where the main point is running from enemies with complex movesets , and some testing showed me that its better for the game to have wide areas with really simple level design , like just plainly walking , so that it gives enemies place to shine , so i guess it'd be better for my game to have large outdoor sections
That might be possible, and it is true that this video largely focuses on platforming in platformers, So puzzle platformers and action platformers would also be slightly less affected.
Rain World gets around most of these problems by just having a fixed camera so that jumping won't ruin the effect. Rain World also manages to make the sky background beautiful because by filling the horizon with easily repeatable good looking things, and tons of background clouds.
It's weird to make this video and never talks about 2D mario games. It's the grandfather of all the games you talk about, look great, and very often show a reasonable amount of sky.
I mentioned this in another comment, I think Mario coasts on nostalgia quite a bit, making a second rate Mario with free floating platforms will generally just look bad (in my experience). I think the game itself looks great, but generally in spite of the background environment and not because of it.
@@Nonsensical2D I would have to strongly disagree. There are lots of nostalgia driven games that haven't had nearly the staying power of mario. Equally Nintendo has maintained a (I think we'll deserved) reputation for making aesthetically appealing games. There may be good reasons that mario is tough for others to emulate, but to me it's a clear exception to this "rule". Maybe examining that more closely it might be and exception that proves the rule.
@@owennewburn Do note that I do think the game looks great, I even said as much in the comment. I specifically mentioned that the idea of free floating platforms with a distinct horizon is not something you generally see in modern platformers and rightly so, it generally looks kind of dated. Super Mario looks awesome because of the design and art direction, but I do very much think that the floating platforms is something that is largely maintained because of nostalgia, and not because it is an ingenious solution that no-one else has considered.
@@w1ck3dz0d1ac Ye, I'll probably always avoid showing Nintendo footage on the channel, but I definitely could have mentioned the game (without showing anything).
Man I love these videos man, you always tackle such interesting topics. I encountered same problem when I was making a sidescroller, but wanted the levels to be more open like a metroidvania, but it didnt quite work because how the background worked.
This topic hits home. I'm using asset patching style with the plan of making something like hollow knight but more color and open sky, and I feel that struggle. Not only big assets are difficult to draw, but raise a good point with the horizon line. Vertical scrolling is really difficult to handle with an open sky.
There is a video I would like you to make. My problem is that when I am making a game, I don't like using "placeholder art" because it is not attractive. I want the final art to be in the game while I am making the game, so that the game can look good in my eyes. I don't like ugly graphics in a game. But making the final art will take a lot of time. So I am stuck, because since I have not yet made the final art (which would be attractive), I have no motivation to make the game now. If I use placeholder art, the game might look unattractive to me. I can make my own art, but it will take a lot of time and work. And without the art, I am not motivated to make the game (because it might be ugly to me). And I don't want placeholder art, unless it is attractive. What can I do?
I have been thinking about making a video about that topic, but I sadly don't think there is any perfect solution. You will probably hit that road block somewhere along the way regardless, you can obviously just always make sure to make the art first, but then you might run into the issue where you have great art for a game that you actually don't like.. I don't really like using placeholder art though. I either sit with no art or fill it with art (except for character and animations).
Maybe make some low-res art place holders that look something like what will actually go there. Then when you are done making the game, pretend you are "remastering it", and replace all the low-res stuff with the final art.
This makes sense irl aswell. If you take a puppet and follow it with a static camera angle as the puppet “climbs up a tree” you get the same monotony or lack of visual interest in the background as you get above the horizon level. But if you play with the camera angle you can play with what is in frame and create a sense of dynamic storytelling. This is simply the nature of having a parallaxing flat background without any control over whats in the frame. But I am sure there exists clever ways to solve it, one just have to be aware of it and put on the problem solver hat :) The last example of breaking continuity is one such “sollution”
Ye, I Kind of think that it sucks, it makes me admire Ori even more, because it frankly feels so open and well done, but trying to do the same is just a headache, it frankly feels more sensible to just do what Team Cherry does, otherwise you are just sitting there having to constantly compromise between aesthetics and gameplay.
@@Nonsensical2D It always sadens me how little credit ori gets in many regards. Visually it's perfection, the level design is great and the two complement eachother very well. Nothing feels immersion breaking, because things make sense. Baur's reach is a mountain so it's on top, mouldwood depths is a cavern so it's at the bottom. Inkwater marsh is the forest floor so it sits at the middle. It's awesome how they did this and based on the GDC talks i saw has a lot to do with the macro level design. That being how biomes conect to eachother. I feel like most indie metroidvania's spend too little time going over their macro level design before fully going for the micro (room by room) scale
I'll be honest at first I was like what? I love sky backgrounds in side scrollers. So I wasn't sure about this video. I decided to let you cook and I'm glad I did. I definitely see your point now.
I disagree. The example at 1:40 doesn't really make sense with the mountains disappearing as if they were 10m away. They should remain almost exactly were they are, being as far and as big as they should be. The problem is that for that to work you would probably need assets in proportion to the small assets (i.e. HUGE). But I suppose it could be done somehow with special Unity commands or something.
While I agree that the mountains should be "farther away" and remain more where they are, the problem you run into when doing this is that you need significantly more layers in order to not break the scene (since if the mountains don't move but the rest of the scene does the space between the layers will come apart and it becomes obvious that you are dealing with 2D and not 3D). Not to mention that you still don't actually gain that much height, it still starts looking awkward when you go about half a screen up. I sat with it quite a bit and I personally kind of ran into problems regardless. I very much could make this scene 'better' If I gave it more time, but I don't think the problem with a scene like this would really go away.
@@Nonsensical2D Yes l think you're right, there'll be a technical challenge there, so yes, maybe not show the sky, and try to have some very vague faraway assets.
I've just started learning art, particularly digital art, and I can't express how much your videos have helped me! Thanks so much! Hope to see more from you soon!
I thought of a partial solution(still requires a skilled artist but addresses most of the other issues), have a sky background that can have ground details(like trees, hills, buildings,etc) in repeating pattern close to the ground and similarly have beautiful details(involving say clouds, floating islands, stars if nighttime, sun if daytime,etc) higher up in the sky.
I think there are some great solutions found in the Sonic series. There are a few levels large, tall backgrounds that have a detail at the bottom and sky at the top. They do not adequate "sky" with "horizon" It gives a great sense of verticality and you go up and down throughout the levels, with yes, PLENTY of obstacles. Great examples: Green Hill act 2 (mania) Oil Ocean act 1 (mania) Marble Garden act 2 (s3&K) Great video! always appreciate the practicality of your videos
@@unity.2315 Palarax vertykaly is very small. I try to use it mainly in long corridors and when we go up and down, mask it with trees and other objects.
this wonderboy scene with statue looks awesome, i will be definetely recreating something like that, even to increase game aesthetics in the trailer i think its worth it :D i think its great to introduce an area but its really bad for vertical platformer action
Good video. But I think you can show the sky and the horizon as long as you have interesting background elements close to the player that are part of the level design. stuff like buildings and caves that you enter without changing the scene.
There are tons of 2d platformers that show their horizons. The early Sonic games, every Mario game, Kirby, Donkey Kong Country, etc etc. Not having anything to do in places where you show the sky is a level design problem. Perhaps your foreground elements should fade as you sink below the surface. Don't let them move higher than the midground that the player was just on. And if you're platforming above the ground, have the background match the camera's Y coordinate. You should still see those dunes in the distance as you ascend.
The trick I used to handle open environments in my game were a couple of factors: 1. The planet itself in the game is like one giant gavern with forest or rocky biomes, to technically the world itself is sorta underground so I don't have to worry too much about the horizon unless I want to make it very very important. 2. For outdoor areas above ground, when it comes to platforms like that would normally be placed in midair, I would attach a tree trunk/branches on it and to give it the sense that players are moving from giant tree to giant tree. The only time I don't attach a large branch to a platform is if the assets are completely underground. That way it maintains consistency from room to room while still giving me some freedom to make outdoor scenes for important areas. 🙂 Great video btw!
hi:] unrelated to this video, but in the future would you be interested in doing feedback videos about viewers submitting their game art environments for you to review?
I've thought about it, I've even set up a method to do it. I have been a bit hesitant because I think there is a tricky balance to "reviews/critiques" and I would want to make it valuable. But ye, I'll probably try it out sometime soon!
I'm facing just this problem exactly right now, making my player fly into the sky and showing the land below in the background. The parallax is a killer.
While I do think there's a lot of value in learning how the parallax stuff effects things like the skyline, I feel like that's a "creative problems require creative solutions" type of thing. But maybe that's because I'm a fantasy enjoyer and most fantasy fans absolutely love floating islands, which provides plenty of usages for upper sky level design. Or stuff like birds, clouds, or flying carpets. And even if scifi, I'm used to Sonic running around a large fleet of Eggman battle ships.
Ye I kind of agree in a sense. I kind of like the approach of iconoclasts and Islets to be honest. While the title is a bit harsh, I mainly wanted to highlight the problem, i definitely agree there are workarounds, sometimes the workarounds a worth it, other times not (kind of the depends on the individual dev)
@@Nonsensical2D perhaps future videos on topics like this deserve an extra 5 mins of "examples of workarounds" on how some games deal with such issues. Beyond the examples for the regular essay body of the topic. I just naturally gravitate towards behind the scenes stuff though and always value guides / reminders on how to work through it based on your needs. That next step in the process.
@@amandaslough125 In a sense I want to do that. but the problem is that it tanks the video quite literally. I've tried adding additional context and stuff before but the AVD (retention) goes from 30-40% to 3-5%, basically if people get the gist of the video, they often don't want that additional context. And if you have 5 minutes of stuff that none watches you run the risk of TH-cam not recommending the video. This is a problem that in part is a skill issue, hopefully I can integrate examples and context better in the future and in an interesting way. I'll definitely try. Thanks for the advice.
I don't really thing that makes the difference. It does not matter what you do open or under the principle is actually the same. I think that makes sense with emptiness is that in fact the sky is empty. But it is your game your world, and the sky can be as anyway as you want it to be. I get your point, but thinking like that makes you just locked into the idea that a sky has to be empty! in a game, world of fantasy you decide how empty or not the sky is, as long as it stay consistent with the story you are telling.
Never really thought about this, but makes so much sense now that you call it out! The examples really helped drive home the idea and how other designers have approached the problem
The og super Mario bros showed a lot of sky, had a locked camera and still had plenty of great gameplay, visuals and level design. Even today, it holds up as one of the better platformers out there. Your point is interesting to listen to, but it is very arbitrary and doesn't really hold up when we look at historical examples of great games that are popular to this day.
I have had this comment raised a few times, the original super mario is kind of famous for the bush and the cloud being the same texture in order to save memory. It kind of says something about the lack of background. Mario has great design, but often looks great despite of its backgrounds, not because of it. That is, I dont think a lesser artist can really get away with it to the same extent. There is a better example with rayman origins, which actually does have quite intricate backgrounds, but in rayman, just like in mario, the level design is extremely horizontal. And the problem I describe in the video is more a problem that arises with vertical level design.
The thing is, to some extent the issues I mentioned can be dealt with. Ori is a game with great gameplay and it looks great. You can also maintain a horizontal level design and not encounter too many issues. But by and large dealing with the horizon is difficult (especially when you need vertical scrolling). There is often a compromise you have to make, that many indie developers are less keen on making. As for the mountains, I slapped a shader on them (foliage shader) to try and simulate heat, don't know if it worked though xD
I disagree that its an issue of the sky leading to visually or mechanically unengaging sections. Its mainly. Main thing I can point to is Rain World. The backgrounds uses a lot of repeat assets and just straight up flat colors, but it still is sensible, retains continuity between screens, and when you look up rain world screenshots a vast majority of ones that pop up are the ones that are above ground, not the half of the game that takes place indoors or underground. And you could argue that its because the character isnt centered to the screen normally, but when you enable screen scrolling, the games parallaxing and composition still holds up BEAUTIFULLY. The reason the game can get away with it though is setting. The flat colored backgrounds just come off as fog or some other distant space, the clouds are thick and massive, and the clear horizons that do exist have objects so massively far away that you can only see them do to their sheer scale. The game has a lot of visible sky, but it works so well for the game because the background isnt just a flat blue, its a place. Even if you cant go there even if its a bunch of reused assets and flat colors, it feels like an actual place that exists. That sense of place is what I feel is the key factor, and that is something that is very difficult to design despite the low amount of assets required, and will be much more difficult for some games than others. Nice video though! ^-^ Even if I dont agree with many of the conclusions, you brought up an interesting problem a lot of games have that I never really thought about, and is something good to keep in mind if you someone wants to create memorable enviroments easier. 😊
Thanks for the input and I agree that i should have talked more about this type of workaround. I kind of have a scene that somewhat works with a similar implementation. I also hadnt played rain world when I made this video, but got a somewhat similar comment to yours quite close after release and have since played an hour or two. I will say that it handles the problem quite well, but as you say the fog sort of fakes the feeling of a horizon quite often (I mention a somewhat similar example a bit with ender lillies). In a way I feel like i just communicated some parts poorly. The main point of the video is that having a distinct and natural looking horizon will often create slightly stale gameplay, you can largely deal with this by using it where it makes sense (trading sections or when you know the character is going to walk horizontally), or by making it less distinct (fog) or less natural and nonsensical instead (islets or iconoclasts). All of these things are technically mentioned in the video, just never repeated and thus might get missed. I also wanted to stress that trying to make a distinct horizon might be a mistake for many devs, because of the added difficulty in handling it, but obviously for some situations or settings it just might make sense. Overall, Its a mistake on my part to not make it clearer but hindsight is 20-20.
Thank you for expressing this in clear terms!! My interior areas always come together so much more quickly and with less hassle, and it finally fully makes *sense.*
Haha, I don't normally sit like this, when I film my b-roll I have to ensure that I fit into the frame while avoiding annoying things in the background, so I 'crouch' in order to avoid my head being cut off. I wouldn't say I normally have that good posture, but this looks as awkward as it felt and I wouldn't normally wish it upon myself :D
You don't need gameplay on every square inch of the screen. It is okay to have negative space. I think this video is saying that this is a bad thing without acknowledging the value of all types of space in general.
Ahh, it wasn't my intention to frame it like that. I think the solution Ori uses is basically awesome (in the way it uses the sky sparingly). Frankly most of the solutions that the games in the video use work quite well (Islets and iconoclasts also have quite interesting solutions). My intention was simply to showcase the problem that naturally arises when showing the sky, so that you are aware of it. That way you can approach it carefully and with a solution in mind.
Hmm, I probably could have, but I sadly played it before I started this channel so I don't have that much footage of it. With that said I think it handles it fairly well, if I remember correctly basically every level is completely horizontal which does alleviate the issue to some extent. But I think one thing worth mentioning with both rayman and a game like Ori is that while it does largely manage to show sky and largely manages to have decent level design, but I feel like they do so with quite a bit of difficulty. I think Rayman manages to integrate the setting and the level design in a way that just seems immensely difficult to pull of myself (vines, chains, ruins, trees, they are all used in a way that integrates and make the level design just work, but I don't think it is feasible to spend time achieving those solutions as a solo developer). With that said, now that you mention it I probably should have talked a bit about Rayman, it is an interesting example.
@@Nonsensical2D Along those same lines, you would be remiss not to discuss Mario. 2D Mario games have tons of levels out in the sky and they've worked well for decades. I would certainly recommend you use Mario as a reference if you're ever to revisit this topic.
@@Beakerbite I considered Mario, sonic and the like when making this video but I'm personally not sure its the best example of good looking scenes, the characters and art direction are awesome, but when it comes to the environment it's just free floating platforms that don't relate to the background whatsoever. It can kind of continue doing it because of nostalgia, but it frankly doesn't feel like a good solution for a modern game (unless you truly go for a nostalgia-game like shovel knight) It could be worth revisiting the topic with more examples given, but I feel like being derivative of the Mario solution generally won't work in practice.
Skill issue. Just beacuse it is harder doesnt make it worse. And nothing is wrong with dedicating a section to just be a spectacle. Constant action can be exhausting. It's called pacing, look it up.
I didn't claim there is anything wrong with dedicating a section to just be a spectacle. Frankly I just stated that that is often what happens. Ori is an amazing game and does it incredibly well. I also don't necessarily think harder makes it worse, but for a lot of indie developers it is kind of the case. If it is harder, it takes more time, and time is what indie developers lack. It is a choice that needs to be considered carefully. If you know you can do it well, do it, but Don't do it without acknowledging the difficulties or the drawbacks and knowing that you can overcome them.
@@Nonsensical2DYou literally said the pretty background section of Wonderboy is boring, numerous times. It is meant to be like that. Also, time isn't universally what an indie game developer lacks. You are just projecting your situation as if it is everyone else problem too. Several developers take 7, 10 years to make a game.
Sorry I don't think I understand what you mean, If you mean that I the mountains don't move "enough", then that is because they are technically closer to the camera than they ought to be in "real life". But the reason for this is because if you do place them super far back, then you will need to place an insane amount of layers or get this weird empty effect, and even at that point the parallax still only cover about 1.5 screens, so you don't actually gain that much. I know that it is technically incorrect, but I chose to do it this way for the sake of simplicity. But the parallax is implemented correctly (it's not a script).
I disagree about the HK lazy approach being good, i think it's really REALLY bad. Spamming those props in the background makes the rooms boring, uninspired and unrecognizable. Take a room from fungal wastes, any room and you will see that none are recognizable due to the background art. A HK vet will recognize the platform layout but not the background. And that is the issue, apart from their setpiece rooms like that central statue in city of tears, none of the rooms are recognizable. Compare that to something like blasphemous or Ori where each room is unique or at least much more unique than HK's non-setpiece rooms. I think this spamming assets approach is actually so bad, and i don't like how HK gets a free pass on that. Also yes sky areas are indeed hard to pull of if you just have an open sky. But having something like a mountain range in the back, like for example the messenger or maybe floating ruins (also like the messenger) then things get a lot more interesting fast. Personally i think Ori does it the best (honestly it does everything art related the best). They have narrow fun platforming sections most of the time and only open up for the setpiece screens. Like how they show the wellspring early in will of the wisps and have you actually go there later. By having such a big structure there the first room is interesting, and the outside of the wellspring is diverse because you can use this structure to have sensible platforms and varied visuals. The bottom line in all this is that underground/interior areas are much easier to do. They knew this as early as super metroid and castlevania, both of which have very few exterior open sky rooms. There's only a few metroidvania's that pulled it off well, most chose the easy/lazy way but that's perfectly fine. Being mostly underground is ok if your game can make it work sensibly. HK is a bug game so it makes sense that they live in burrows. It's just the execution with object spam that really doesn't work for me
Hmm, I would say I kind of agree with you and kind of don't. I also very much prefer the approach that Ori has, the original Ori (and the blind forest) is my favourite platformer of all time. With that said I simply don't think it is possible to approach game art as a solo dev the same way. What I like about the HK approach is that you can get decent looking results quickly, which frees up time to spend on set pieces. It definitely looks "worse", but it is such an insane time saver compared to what Ori is doing. Like Ori is a 12 hour long game made with lots of artists, whereas HK is a 40 hour long game made by one artist. This doesn't necessarily mean I think you should go to the extreme end (like fungal wastes), but in terms of solo developer efficiency I think it is an aspect that can be important to adopt, even if it is significantly more boring than Ori (mind you I frankly don't understand how they even managed to do it as well as they did, and I don't think I could emulate it regardless of how hard I tried).
Right on. Yeah. I don't think the sky is the issue, it's the lack of focusing where the action is. I could make the sky a sliver in height, have the character at the top of screen, and fill the screen with dirt. Same issue, the main action is pushed to the edge. So it's not the sky that's the issue, and ground isn't the solution. It's about capturing the focus and actions.
I disagree. One of the real advantages of side scrollers is their ability to be so artistic, picturesque if you will. The one you showed in the desert doesn't prove sky shots don't work, it proves that's an ugly game. Besides, you don't want to show your best shots during fights anyway.
I'm sorry, but this just reads to me like "I am bad at this, some other people are also bad at this, therefore you should avoid it". Some of the examples you showed are actually good at it. I just think you're missing the point of these scenes entirely. I'm also an indie game developer, and throughout my studies I've come across multiple instances of the concept of breathing spaces in games. Those can be the safe rooms in horror games like resident evil, the starting ship area of planet Zebes in Super Metroid. Those moments where there's no much to do, but there's still some effort put into the music, layout and visuals of these places. These are meant to make the player wind down and contemplate their progress, to take a breather between action-heavy sequences and refresh them for the next round. These can vary between short moments on a level such as a vine that leads to a safe cloud with coins in it in a Mario game, to entire map areas such as Kakariko village in Zelda. The choice of contrasting focused, indoors spaces with outdoorsy, landscape backgrounds can also contribute to this feeling of relief we try to evoke in the players.
Yes.. I think frankly this might be a miscommunication on my part (but I also talk about quite a few of these things at the end of the video). "some of the examples you showed are actually good at it." I think all of the examples I showed are good at it, that is why I showed them, but as I say in the video, in each of these examples they use a compromise or place it in a very precise context (such as ghost song and ori trading places). And for some of the games, even if the solution is good, such as Ori (which is probably the most impressive of them all), they might not necessarily be that reproducible by a single artist. I think there is a point to these scenes, but my point is that you need to be extremely careful when using it and use it in the right context. And adding it without consideration for the problems it introduces can negatively affect gameplay. And for quite a lot of games, especially if you want to make your art quicker, just adding fog and suggesting a horizon will be the quicker method of working while still creating nice results.
@@Nonsensical2D I do agree that it should be used carefully and done with a certain level of mastery, but I honestly think your proposed solution is generally bad advice, since it is discouraging experimentation and imposing a constraint that's frankly more a "you" thing than a general rule of sorts. Rather than encouraging better ways of creating simple and effective horizons, your suggested shortcut imposes a claustrophobic constraint that may turn people off from even attempting. This is my main issue with all this: this "it's too hard do something else" attitude this whole video exudes.
Hmm, I guess we might actually disagree on this point then. Which is fine, I do see where you are coming from. I agree that I am discouraging experimentation to some extent, but I feel that by and large this is justified. My overall implied premise to this argument is kind of: 1. You are not a super skilled artist 2. You want your game to look good and sell well 3. You have limited amount of time to make your game. and given those three implied goals/limitations, I think showing a clear and distinct sky will in most circumstances come with a huge cost and won't align well with your situation, the ROI is bad. (or rather you should use a distinct sky where it makes sense and doesn't come with this huge cost). Now if you are an indie team with skilled artists or a AAA title, then by all means, you can make your game even more unique and even cooler, but by that point you frankly already know how you will go about handling it. To me this problem has some similarities to the "open world" dream of many beginners, but to a much smaller extent. Everyone agrees the idea of an open world is cool, just that it perhaps isn't feasible for most people.
@@Nonsensical2D As I said in another comment. I still think making amateurish attempts at environmental variety makes for a far more interesting product than producing a more professional looking monotonous one with the same trick being applied all throughout. But this is quite a subjective matter at this point as different people will prioritize different things both in the making and consumption of media. All that said, I wish you good success in your development, and would suggest you to take caution with discouragement on the internet, since the industry is already plenty oppressing and discouraging for newcomers. Really gets on my nerves specifically because listening to similar takes held me back for quite a while in my drive to study and become better both at coding and making art. Cheers!
The world as we know it scrolls horizontally. If you have a horizon, you are basically somewhat limiting your ability to scroll vertically, because if you do, you will have difficulty adding assets and populating the scene, because it'll be tricky to think of a flying asset that will make sense in the context of your game (thus it can quite quickly result in a rather dull background). You can avoid this by not using vertical scrolling, use a nonsensical background, or utilise scene breaks. Or you do what many games do and only use open horizon areas in locations where you don't utilise much level design anyway (like Ori). I'm not saying it is impossible to deal with, just that it is problematic to deal with, and depending on your game and how much time you have, it could be worth it to avoid the sky altogether (like HK). It has nothing to do with Incorrect values on the Y, even though the scene that I created is technically too close to the foreground to be realistic. The problem with placing it really far into the background is that your horizon starts looking really flat (or you need to add too many background layers), while the problem still kind of remains, you still can't do too much vertical scrolling.
wait but blasphemous 2 has plenty of platforming sections with the sky and clouds and spires in the background... im not making a 2d platformer so idk though. i mean??? idunno
I'll add two small notes that kind of got cut in the edit. This video largely covers the problem with skies that you might encounter in almost all platformers. With that said, if you are making a metroidvania or roguelike, you will generally have to make a lot of your game underground for a totally separate reason. That is, you need to make part of the game underground in order to make the map wrap around on itself correctly. In that case you'll get a map like ghost song or aeterna noctis, where you have 2 or 3 areas above ground, but most areas will still be underground. It's a separate point to the video so it got cut, but it still might be a relevant tidbit.
The second point I would like to stress, is that if you are as skilled of an artist as the people making Wonder Boy or Ori, then it can make sense to have an "aesthetic scene" that grabs attention. But if you are similarly skilled as me, attempting to make a nice looking horizon might actually turn out worse than just going for the simple "hollow knight approach". It'll only look nice if you are good enough to make it look nice, otherwise it'll both look bad, and play bad.
but it doesnt always have to be underground right it can be a large magical mansion or a spaceship(like metroid does) so its just have to be a big closed off space but yeah if u make it underground u can add things later down the line thats a plus
Yes, underground is generally a good theme to go by. But why shouldn't solo metroidvania devs explore scenarios like mountains, tree top buildings, high-up towers, suspended bridges, or even several "on the ground surface" locations in addition to those? Underwater is also a great theme. There is so much that can be done with parallax layers to make the horizon look good and consistent with the height, even if you're not a skilled artist. But even if you're not making it look decent, I'd argue it's still more interesting than the monotony of making several foggy underground biomes in fear of making anything amateur.
Hard disagree. The "empty" feeling can sometimes be a great way to give players breathing room. A lot of times in games that are meant claustrophobic and intense, it can feel very nice to finally have some open above ground gameplay, even if there isn't much to see. Sometimes a smooth gradient with some nice granular clouds is all you need.
Pacing and giving the player moments of reprieve are where those 'boring' scenes are great.
Understanding how these different environments play a role in the "feel" of a game is where the problem comes. Not every scene should be focused on giving the player something to do. Some are to evoke a feeling.
I understand the point but I think it's less of a "bad" thing but more like a thing that "can be utilized in different situations"
This is titled "why you shouldn't show the sky" when it should be called "why it is difficult to show the sky".
Ye. This might be fair. But I think a portion of my argument is that for many solo devs it is probably the case that you shouldn't show the sky. As in the ROI on it is probably terrible. I agree that the substance of the video is that it is difficult, but can be done at times or with some compromise.
But by and large I want the takeaway to somewhat be that you should truly consider if the difficulty makes it worth it. With that said, I kind of agree that the title is a bit sensationalised, but it's quite tricky titling videos overall.
Oh woah, I clicked on this in my recommended feed and did not expect to see my game mentioned! Great video, and definitely a real issue that I did not think of until I encountered it myself. The vertical camera panning issue is a big reason why I kept room sizes pretty small in Islets. When a room is too big, it's hard to keep the parallax from just showing emptiness when you're too high up. But when it's small, the parallax range is smaller, and you can plan for it to show something at pretty much any position.
Ye good point! I am assuming this is a shift you made intentionally after sheepo? Cause when I compare the two I can see that the rooms in sheepo are generally significantly larger and more zoomed out, but also run into the issue of being empty at times (except for clouds).
Btw, really like both your games and admire the artstyle a lot, both games manage to do a lot while still being simple. It's kind of problematic to look at games like Ori and blasphemous as examples of what to do at times because they are frankly impossible to emulate as a solo developer. Both your games look awesome but still manage to be fairly economical (time-wise) in their approach.
@@Nonsensical2D Yeah definitely, Sheepo had some issues with visual layout (also seen by the map being "impossible" since rooms that were 1920x1080 showed as the same size as rooms that were 12000x12000px), and decorating the larger rooms became a bit unmanageable. I'll definitely be rethinking the ways I design areas with the next side-scroller to keep decoration less of a chore lol
And thank you! Yeah I think it's partly that I just get a bit exhausted working on a project for way too long, and also just that making games pretty quickly makes the odds of finding success a bit better. I kind of follow the 90/10 rule, where the last 10 percent takes as long as the first 90 for most aspects of game development, so for anything that doesn't need to be 100% perfect, I just skip that last 10% of polish and just save it for things that are a bigger focus (player movement, core art, etc). Maybe is a little lazy, but saves a bunch of time and only sacrifices things that would make a negligible difference
Don't mind me, just gonna take notes......👀📝
It's worth noting that those mountains at 1:56, as demonstrated in the clip are very short and relatively close. If you create background assets with the idea that they are bigger in scale and further away, the parallax effect makes their faked movement closer to the camera's and you get to have an interesting scene for longer, and also gives a sense of grand scale to the environment.
Also about Monster boy I think its in general a bad example for this point because of how uniquely it was made. It was new art placed on top of a very old game, so it's not that the level is flat and empty because the art forced the hands of the designers, but the game already had a very flat and simple level, and the artist decided to fill it up with gorgeous environments so the massive empty space felt purposeful.
When it comes to Wonder Boy I know the art was placed on top of an old game, but it felt like talking about that would detract from the video. I don't think the artist did a poor choice in this particular circumstance, but I do think that the level design in wonder boy is incredibly dated, it frankly feels boring to play. And that section is no different, I don't think it holds up gameplay wise. So I took it as an example largely because it showcases the problem that occurs quite well (and I wanted to talk about the cool parallax foreground), the scene looks amazing and also plays really terribly. I think the cases with Ori or Ghost song are better displays of what happens to the level design when you create an open sky. But frankly there are a lot of games you could look to. I just selected a few that I think illustrated the issue decently well.
You could put flying enemies in areas that are mostly sky if you want obstacles to avoid.
i disagree with this assessment completely.
How come?
Leaving my marks here to hear the answer.
@@Nonsensical2D it looks good tbh
@@workinprogress6504Same.
Unrelated to the video, but I've noticed the line weights in your art are working against you. Your player character has these super thin lines while elements from the environment often have much bolder lines. I'm guessing this is a product of wanting to make your more important elements more detailed and the less important elements more stylized, which makes a lot of sense, but heavy outlines have a strong association with "intractability" (think cuphead), so the end result looks a bit disjointed and ameture as your eyes are being pulled towards the elements in your scene that are the least important. There's many ways to get around this:
Making your character designs more simple to work with a thicker line weight, like hollow knight
Giving your environment art a thinner line weight
Coloring the outlines of the environment art to make it blend in more (I dont think this alone would be enough, but I think it could help
Removing the outline of environment assets altogether (would completely change the artstyle)
Thanks for the feedback! The truth is that I make all my scenes separate from my actual game art, they are usually made quickly to try something out or illustrate a concept, so I don't spend time on it, sometimes I use thick line art, sometimes I use thin line art. But they are by and large completely incompatible with the character since I always use the same 2 characters regardless of the scene.
So I have actually played around with most of the suggestions you mentioned, you can even see some scenes in this video with thin lineart, others with thick, some with coloured lineart and some with black lineart. This desert scene also looks off in terms of lineart because I've stolen one or two assets from a previous scene whereas the cactus was made for this scene. But I really do agree with you, If I were to release a game, I would have to ensure consistent line size to some extent :)
This problem dates back to 1989 with Shadow of the Beast, where many parts of the game is just the player running around big open horizons, presumably to showcase the parallax scrolling, which was basically a new thing for computer games at the time.
you should have mentioned one thing though , sometimes its actually better than making stuff indoors , depending on the gameplay , most of the examples you showed were platformers that need obstacles or would become boring , but though i have made a prototype for a game where the main point is running from enemies with complex movesets , and some testing showed me that its better for the game to have wide areas with really simple level design , like just plainly walking , so that it gives enemies place to shine , so i guess it'd be better for my game to have large outdoor sections
That might be possible, and it is true that this video largely focuses on platforming in platformers, So puzzle platformers and action platformers would also be slightly less affected.
Rain World gets around most of these problems by just having a fixed camera so that jumping won't ruin the effect. Rain World also manages to make the sky background beautiful because by filling the horizon with easily repeatable good looking things, and tons of background clouds.
I have been meaning to buy Rain world but haven't yet done so. I really like the aesthetic of it so should probably give it a try :)
@@Nonsensical2D ooo yes. it is well worth your time. hope you like it.
also, take the leg route if you want to experience a truly breathtaking moment.
It's weird to make this video and never talks about 2D mario games. It's the grandfather of all the games you talk about, look great, and very often show a reasonable amount of sky.
I mentioned this in another comment, I think Mario coasts on nostalgia quite a bit, making a second rate Mario with free floating platforms will generally just look bad (in my experience). I think the game itself looks great, but generally in spite of the background environment and not because of it.
@@Nonsensical2D I would have to strongly disagree. There are lots of nostalgia driven games that haven't had nearly the staying power of mario. Equally Nintendo has maintained a (I think we'll deserved) reputation for making aesthetically appealing games. There may be good reasons that mario is tough for others to emulate, but to me it's a clear exception to this "rule". Maybe examining that more closely it might be and exception that proves the rule.
@@owennewburn Do note that I do think the game looks great, I even said as much in the comment. I specifically mentioned that the idea of free floating platforms with a distinct horizon is not something you generally see in modern platformers and rightly so, it generally looks kind of dated. Super Mario looks awesome because of the design and art direction, but I do very much think that the floating platforms is something that is largely maintained because of nostalgia, and not because it is an ingenious solution that no-one else has considered.
I just thought he wanted to keep Nintendo from issuing copyright strikes.
@@w1ck3dz0d1ac Ye, I'll probably always avoid showing Nintendo footage on the channel, but I definitely could have mentioned the game (without showing anything).
Man I love these videos man, you always tackle such interesting topics. I encountered same problem when I was making a sidescroller, but wanted the levels to be more open like a metroidvania, but it didnt quite work because how the background worked.
Interesting art direction principles that don't appear in regular art education.
This topic hits home. I'm using asset patching style with the plan of making something like hollow knight but more color and open sky, and I feel that struggle. Not only big assets are difficult to draw, but raise a good point with the horizon line. Vertical scrolling is really difficult to handle with an open sky.
There is a video I would like you to make.
My problem is that when I am making a game, I don't like using "placeholder art" because it is not attractive. I want the final art to be in the game while I am making the game, so that the game can look good in my eyes. I don't like ugly graphics in a game.
But making the final art will take a lot of time. So I am stuck, because since I have not yet made the final art (which would be attractive), I have no motivation to make the game now. If I use placeholder art, the game might look unattractive to me.
I can make my own art, but it will take a lot of time and work. And without the art, I am not motivated to make the game (because it might be ugly to me). And I don't want placeholder art, unless it is attractive.
What can I do?
I have been thinking about making a video about that topic, but I sadly don't think there is any perfect solution. You will probably hit that road block somewhere along the way regardless, you can obviously just always make sure to make the art first, but then you might run into the issue where you have great art for a game that you actually don't like.. I don't really like using placeholder art though. I either sit with no art or fill it with art (except for character and animations).
Maybe make some low-res art place holders that look something like what will actually go there. Then when you are done making the game, pretend you are "remastering it", and replace all the low-res stuff with the final art.
Thank you. Nice idea.@@xx_amongus_xx6987
This makes sense irl aswell. If you take a puppet and follow it with a static camera angle as the puppet “climbs up a tree” you get the same monotony or lack of visual interest in the background as you get above the horizon level. But if you play with the camera angle you can play with what is in frame and create a sense of dynamic storytelling. This is simply the nature of having a parallaxing flat background without any control over whats in the frame. But I am sure there exists clever ways to solve it, one just have to be aware of it and put on the problem solver hat :)
The last example of breaking continuity is one such “sollution”
Dude I never thought of it this way but this is a problem that has PLAGUED MY LIFE.
Ye, I Kind of think that it sucks, it makes me admire Ori even more, because it frankly feels so open and well done, but trying to do the same is just a headache, it frankly feels more sensible to just do what Team Cherry does, otherwise you are just sitting there having to constantly compromise between aesthetics and gameplay.
@@Nonsensical2D It always sadens me how little credit ori gets in many regards. Visually it's perfection, the level design is great and the two complement eachother very well. Nothing feels immersion breaking, because things make sense. Baur's reach is a mountain so it's on top, mouldwood depths is a cavern so it's at the bottom. Inkwater marsh is the forest floor so it sits at the middle. It's awesome how they did this and based on the GDC talks i saw has a lot to do with the macro level design. That being how biomes conect to eachother. I feel like most indie metroidvania's spend too little time going over their macro level design before fully going for the micro (room by room) scale
I really like your content, thank you !
Great video. You addressed all my concerns or disagreements by the end of the video. I would advice anyone mad to finish it fully first lol
Dude, your info is so helpful.
I'll be honest at first I was like what? I love sky backgrounds in side scrollers. So I wasn't sure about this video. I decided to let you cook and I'm glad I did. I definitely see your point now.
Never heard this before. Great observation!
I disagree. The example at 1:40 doesn't really make sense with the mountains disappearing as if they were 10m away. They should remain almost exactly were they are, being as far and as big as they should be. The problem is that for that to work you would probably need assets in proportion to the small assets (i.e. HUGE). But I suppose it could be done somehow with special Unity commands or something.
While I agree that the mountains should be "farther away" and remain more where they are, the problem you run into when doing this is that you need significantly more layers in order to not break the scene (since if the mountains don't move but the rest of the scene does the space between the layers will come apart and it becomes obvious that you are dealing with 2D and not 3D). Not to mention that you still don't actually gain that much height, it still starts looking awkward when you go about half a screen up. I sat with it quite a bit and I personally kind of ran into problems regardless. I very much could make this scene 'better' If I gave it more time, but I don't think the problem with a scene like this would really go away.
@@Nonsensical2D Yes l think you're right, there'll be a technical challenge there, so yes, maybe not show the sky, and try to have some very vague faraway assets.
I've just started learning art, particularly digital art, and I can't express how much your videos have helped me! Thanks so much!
Hope to see more from you soon!
Great analysis.
I thought of a partial solution(still requires a skilled artist but addresses most of the other issues), have a sky background that can have ground details(like trees, hills, buildings,etc) in repeating pattern close to the ground and similarly have beautiful details(involving say clouds, floating islands, stars if nighttime, sun if daytime,etc) higher up in the sky.
I think there are some great solutions found in the Sonic series.
There are a few levels large, tall backgrounds that have a detail at the bottom and sky at the top. They do not adequate "sky" with "horizon"
It gives a great sense of verticality and you go up and down throughout the levels, with yes, PLENTY of obstacles.
Great examples:
Green Hill act 2 (mania)
Oil Ocean act 1 (mania)
Marble Garden act 2 (s3&K)
Great video! always appreciate the practicality of your videos
Straight into my level design playlist ❤
I've actually been running into this issue while prototyping some art ideas. Horizons are really hard to deal with design-wise in 2d settings.
It's cool that someone is finally talking about it, I went through this problem myself and it's not easy :D
How did you end up handling it? I got the same problem
@@unity.2315 Palarax vertykaly is very small. I try to use it mainly in long corridors and when we go up and down, mask it with trees and other objects.
this wonderboy scene with statue looks awesome, i will be definetely recreating something like that, even to increase game aesthetics in the trailer i think its worth it :D i think its great to introduce an area but its really bad for vertical platformer action
Good video.
But I think you can show the sky and the horizon as long as you have interesting background elements close to the player that are part of the level design.
stuff like buildings and caves that you enter without changing the scene.
really interesting.. i´ll have that in mind. Thx you
Interesting. Never have thought about it. Thanks for sahring😁
There are tons of 2d platformers that show their horizons. The early Sonic games, every Mario game, Kirby, Donkey Kong Country, etc etc.
Not having anything to do in places where you show the sky is a level design problem.
Perhaps your foreground elements should fade as you sink below the surface. Don't let them move higher than the midground that the player was just on.
And if you're platforming above the ground, have the background match the camera's Y coordinate. You should still see those dunes in the distance as you ascend.
Did we forget about 2D mario games? They have amazing looking skies.
Or you can treat the sky like a dungeon and use elements as clouds and flying objects to create an enclosing impression or to avoid continuity.
Interesting points :D
The trick I used to handle open environments in my game were a couple of factors:
1. The planet itself in the game is like one giant gavern with forest or rocky biomes, to technically the world itself is sorta underground so I don't have to worry too much about the horizon unless I want to make it very very important.
2. For outdoor areas above ground, when it comes to platforms like that would normally be placed in midair, I would attach a tree trunk/branches on it and to give it the sense that players are moving from giant tree to giant tree. The only time I don't attach a large branch to a platform is if the assets are completely underground.
That way it maintains consistency from room to room while still giving me some freedom to make outdoor scenes for important areas. 🙂
Great video btw!
hi:] unrelated to this video, but in the future would you be interested in doing feedback videos about viewers submitting their game art environments for you to review?
I've thought about it, I've even set up a method to do it. I have been a bit hesitant because I think there is a tricky balance to "reviews/critiques" and I would want to make it valuable. But ye, I'll probably try it out sometime soon!
You could do something like Parallax, or what Terraria does.
I'm facing just this problem exactly right now, making my player fly into the sky and showing the land below in the background. The parallax is a killer.
nice tips!
While I do think there's a lot of value in learning how the parallax stuff effects things like the skyline, I feel like that's a "creative problems require creative solutions" type of thing.
But maybe that's because I'm a fantasy enjoyer and most fantasy fans absolutely love floating islands, which provides plenty of usages for upper sky level design. Or stuff like birds, clouds, or flying carpets. And even if scifi, I'm used to Sonic running around a large fleet of Eggman battle ships.
Ye I kind of agree in a sense. I kind of like the approach of iconoclasts and Islets to be honest. While the title is a bit harsh, I mainly wanted to highlight the problem, i definitely agree there are workarounds, sometimes the workarounds a worth it, other times not (kind of the depends on the individual dev)
@@Nonsensical2D perhaps future videos on topics like this deserve an extra 5 mins of "examples of workarounds" on how some games deal with such issues. Beyond the examples for the regular essay body of the topic. I just naturally gravitate towards behind the scenes stuff though and always value guides / reminders on how to work through it based on your needs. That next step in the process.
@@amandaslough125 In a sense I want to do that. but the problem is that it tanks the video quite literally. I've tried adding additional context and stuff before but the AVD (retention) goes from 30-40% to 3-5%, basically if people get the gist of the video, they often don't want that additional context. And if you have 5 minutes of stuff that none watches you run the risk of TH-cam not recommending the video. This is a problem that in part is a skill issue, hopefully I can integrate examples and context better in the future and in an interesting way. I'll definitely try. Thanks for the advice.
I: love to draw clouds...
Video: Why You should Avoid Showing the Sky!
Okay :'(
how about gris?
I don't really thing that makes the difference. It does not matter what you do open or under the principle is actually the same. I think that makes sense with emptiness is that in fact the sky is empty. But it is your game your world, and the sky can be as anyway as you want it to be. I get your point, but thinking like that makes you just locked into the idea that a sky has to be empty! in a game, world of fantasy you decide how empty or not the sky is, as long as it stay consistent with the story you are telling.
thanks
Never really thought about this, but makes so much sense now that you call it out! The examples really helped drive home the idea and how other designers have approached the problem
The og super Mario bros showed a lot of sky, had a locked camera and still had plenty of great gameplay, visuals and level design. Even today, it holds up as one of the better platformers out there. Your point is interesting to listen to, but it is very arbitrary and doesn't really hold up when we look at historical examples of great games that are popular to this day.
I have had this comment raised a few times, the original super mario is kind of famous for the bush and the cloud being the same texture in order to save memory. It kind of says something about the lack of background. Mario has great design, but often looks great despite of its backgrounds, not because of it. That is, I dont think a lesser artist can really get away with it to the same extent. There is a better example with rayman origins, which actually does have quite intricate backgrounds, but in rayman, just like in mario, the level design is extremely horizontal. And the problem I describe in the video is more a problem that arises with vertical level design.
How do you spell the name of the game at 07:15 ?
Iconoclasts
@@Nonsensical2DThanks!
My toxic trait is that I want to prove this wrong so bad. Anyways why are your mountains breathing?
The thing is, to some extent the issues I mentioned can be dealt with. Ori is a game with great gameplay and it looks great. You can also maintain a horizontal level design and not encounter too many issues. But by and large dealing with the horizon is difficult (especially when you need vertical scrolling). There is often a compromise you have to make, that many indie developers are less keen on making.
As for the mountains, I slapped a shader on them (foliage shader) to try and simulate heat, don't know if it worked though xD
I disagree that its an issue of the sky leading to visually or mechanically unengaging sections. Its mainly. Main thing I can point to is Rain World. The backgrounds uses a lot of repeat assets and just straight up flat colors, but it still is sensible, retains continuity between screens, and when you look up rain world screenshots a vast majority of ones that pop up are the ones that are above ground, not the half of the game that takes place indoors or underground.
And you could argue that its because the character isnt centered to the screen normally, but when you enable screen scrolling, the games parallaxing and composition still holds up BEAUTIFULLY.
The reason the game can get away with it though is setting. The flat colored backgrounds just come off as fog or some other distant space, the clouds are thick and massive, and the clear horizons that do exist have objects so massively far away that you can only see them do to their sheer scale. The game has a lot of visible sky, but it works so well for the game because the background isnt just a flat blue, its a place. Even if you cant go there even if its a bunch of reused assets and flat colors, it feels like an actual place that exists. That sense of place is what I feel is the key factor, and that is something that is very difficult to design despite the low amount of assets required, and will be much more difficult for some games than others.
Nice video though! ^-^ Even if I dont agree with many of the conclusions, you brought up an interesting problem a lot of games have that I never really thought about, and is something good to keep in mind if you someone wants to create memorable enviroments easier. 😊
Thanks for the input and I agree that i should have talked more about this type of workaround. I kind of have a scene that somewhat works with a similar implementation. I also hadnt played rain world when I made this video, but got a somewhat similar comment to yours quite close after release and have since played an hour or two. I will say that it handles the problem quite well, but as you say the fog sort of fakes the feeling of a horizon quite often (I mention a somewhat similar example a bit with ender lillies). In a way I feel like i just communicated some parts poorly. The main point of the video is that having a distinct and natural looking horizon will often create slightly stale gameplay, you can largely deal with this by using it where it makes sense (trading sections or when you know the character is going to walk horizontally), or by making it less distinct (fog) or less natural and nonsensical instead (islets or iconoclasts). All of these things are technically mentioned in the video, just never repeated and thus might get missed. I also wanted to stress that trying to make a distinct horizon might be a mistake for many devs, because of the added difficulty in handling it, but obviously for some situations or settings it just might make sense. Overall, Its a mistake on my part to not make it clearer but hindsight is 20-20.
Thank you for expressing this in clear terms!! My interior areas always come together so much more quickly and with less hassle, and it finally fully makes *sense.*
Me seen the title while making the world in my game That is a floating islend in the sky: Oh shi*
4:42 Ouch, that posture hurts to just look at 😬 Take care of your shoulders before it's too late dude
Haha, I don't normally sit like this, when I film my b-roll I have to ensure that I fit into the frame while avoiding annoying things in the background, so I 'crouch' in order to avoid my head being cut off. I wouldn't say I normally have that good posture, but this looks as awkward as it felt and I wouldn't normally wish it upon myself :D
What game is this 7:40?
Islets :)
@@Nonsensical2DOh nice
👍👍👍👌👌👌🤟🤟
You don't need gameplay on every square inch of the screen. It is okay to have negative space. I think this video is saying that this is a bad thing without acknowledging the value of all types of space in general.
Ahh, it wasn't my intention to frame it like that. I think the solution Ori uses is basically awesome (in the way it uses the sky sparingly). Frankly most of the solutions that the games in the video use work quite well (Islets and iconoclasts also have quite interesting solutions). My intention was simply to showcase the problem that naturally arises when showing the sky, so that you are aware of it. That way you can approach it carefully and with a solution in mind.
Sonic mania
It’s disappointing that you didn’t mention Rayman Legends.
Hmm, I probably could have, but I sadly played it before I started this channel so I don't have that much footage of it. With that said I think it handles it fairly well, if I remember correctly basically every level is completely horizontal which does alleviate the issue to some extent.
But I think one thing worth mentioning with both rayman and a game like Ori is that while it does largely manage to show sky and largely manages to have decent level design, but I feel like they do so with quite a bit of difficulty. I think Rayman manages to integrate the setting and the level design in a way that just seems immensely difficult to pull of myself (vines, chains, ruins, trees, they are all used in a way that integrates and make the level design just work, but I don't think it is feasible to spend time achieving those solutions as a solo developer). With that said, now that you mention it I probably should have talked a bit about Rayman, it is an interesting example.
Other games to check out that handle sky well are Contra and Battletoads. Maybe even Mario's cloud levels.
@@Nonsensical2D Along those same lines, you would be remiss not to discuss Mario. 2D Mario games have tons of levels out in the sky and they've worked well for decades. I would certainly recommend you use Mario as a reference if you're ever to revisit this topic.
@@Beakerbite I considered Mario, sonic and the like when making this video but I'm personally not sure its the best example of good looking scenes, the characters and art direction are awesome, but when it comes to the environment it's just free floating platforms that don't relate to the background whatsoever. It can kind of continue doing it because of nostalgia, but it frankly doesn't feel like a good solution for a modern game (unless you truly go for a nostalgia-game like shovel knight) It could be worth revisiting the topic with more examples given, but I feel like being derivative of the Mario solution generally won't work in practice.
You guys need more Super Mario world in your game research about sky’.
Every Mario platformer solved this
Skill issue. Just beacuse it is harder doesnt make it worse. And nothing is wrong with dedicating a section to just be a spectacle. Constant action can be exhausting. It's called pacing, look it up.
I didn't claim there is anything wrong with dedicating a section to just be a spectacle. Frankly I just stated that that is often what happens. Ori is an amazing game and does it incredibly well. I also don't necessarily think harder makes it worse, but for a lot of indie developers it is kind of the case. If it is harder, it takes more time, and time is what indie developers lack. It is a choice that needs to be considered carefully. If you know you can do it well, do it, but Don't do it without acknowledging the difficulties or the drawbacks and knowing that you can overcome them.
@@Nonsensical2DYou literally said the pretty background section of Wonderboy is boring, numerous times. It is meant to be like that. Also, time isn't universally what an indie game developer lacks. You are just projecting your situation as if it is everyone else problem too. Several developers take 7, 10 years to make a game.
y axis mod= 0.99
So you actually get this wrong. The current methods you used are all wrong, You need to parallax in all 4 directions not just 2
Sorry I don't think I understand what you mean, If you mean that I the mountains don't move "enough", then that is because they are technically closer to the camera than they ought to be in "real life". But the reason for this is because if you do place them super far back, then you will need to place an insane amount of layers or get this weird empty effect, and even at that point the parallax still only cover about 1.5 screens, so you don't actually gain that much. I know that it is technically incorrect, but I chose to do it this way for the sake of simplicity. But the parallax is implemented correctly (it's not a script).
@@Nonsensical2D Try a 4D method, paralax the ground into the background so it looks like your looking downward the higher you climb, might be fun.
I disagree about the HK lazy approach being good, i think it's really REALLY bad. Spamming those props in the background makes the rooms boring, uninspired and unrecognizable. Take a room from fungal wastes, any room and you will see that none are recognizable due to the background art. A HK vet will recognize the platform layout but not the background. And that is the issue, apart from their setpiece rooms like that central statue in city of tears, none of the rooms are recognizable. Compare that to something like blasphemous or Ori where each room is unique or at least much more unique than HK's non-setpiece rooms. I think this spamming assets approach is actually so bad, and i don't like how HK gets a free pass on that. Also yes sky areas are indeed hard to pull of if you just have an open sky. But having something like a mountain range in the back, like for example the messenger or maybe floating ruins (also like the messenger) then things get a lot more interesting fast. Personally i think Ori does it the best (honestly it does everything art related the best). They have narrow fun platforming sections most of the time and only open up for the setpiece screens. Like how they show the wellspring early in will of the wisps and have you actually go there later. By having such a big structure there the first room is interesting, and the outside of the wellspring is diverse because you can use this structure to have sensible platforms and varied visuals. The bottom line in all this is that underground/interior areas are much easier to do. They knew this as early as super metroid and castlevania, both of which have very few exterior open sky rooms. There's only a few metroidvania's that pulled it off well, most chose the easy/lazy way but that's perfectly fine. Being mostly underground is ok if your game can make it work sensibly. HK is a bug game so it makes sense that they live in burrows. It's just the execution with object spam that really doesn't work for me
Hmm, I would say I kind of agree with you and kind of don't. I also very much prefer the approach that Ori has, the original Ori (and the blind forest) is my favourite platformer of all time. With that said I simply don't think it is possible to approach game art as a solo dev the same way. What I like about the HK approach is that you can get decent looking results quickly, which frees up time to spend on set pieces. It definitely looks "worse", but it is such an insane time saver compared to what Ori is doing. Like Ori is a 12 hour long game made with lots of artists, whereas HK is a 40 hour long game made by one artist. This doesn't necessarily mean I think you should go to the extreme end (like fungal wastes), but in terms of solo developer efficiency I think it is an aspect that can be important to adopt, even if it is significantly more boring than Ori (mind you I frankly don't understand how they even managed to do it as well as they did, and I don't think I could emulate it regardless of how hard I tried).
Right on. Yeah. I don't think the sky is the issue, it's the lack of focusing where the action is.
I could make the sky a sliver in height, have the character at the top of screen, and fill the screen with dirt.
Same issue, the main action is pushed to the edge.
So it's not the sky that's the issue, and ground isn't the solution. It's about capturing the focus and actions.
I'd say skill issue
That's a great perspective. I wonder how couldn't I noticed that yet.
I disagree. One of the real advantages of side scrollers is their ability to be so artistic, picturesque if you will. The one you showed in the desert doesn't prove sky shots don't work, it proves that's an ugly game. Besides, you don't want to show your best shots during fights anyway.
It's looks like Nonsensical statement
I'm sorry, but this just reads to me like "I am bad at this, some other people are also bad at this, therefore you should avoid it". Some of the examples you showed are actually good at it. I just think you're missing the point of these scenes entirely. I'm also an indie game developer, and throughout my studies I've come across multiple instances of the concept of breathing spaces in games. Those can be the safe rooms in horror games like resident evil, the starting ship area of planet Zebes in Super Metroid. Those moments where there's no much to do, but there's still some effort put into the music, layout and visuals of these places. These are meant to make the player wind down and contemplate their progress, to take a breather between action-heavy sequences and refresh them for the next round. These can vary between short moments on a level such as a vine that leads to a safe cloud with coins in it in a Mario game, to entire map areas such as Kakariko village in Zelda. The choice of contrasting focused, indoors spaces with outdoorsy, landscape backgrounds can also contribute to this feeling of relief we try to evoke in the players.
Yes.. I think frankly this might be a miscommunication on my part (but I also talk about quite a few of these things at the end of the video).
"some of the examples you showed are actually good at it." I think all of the examples I showed are good at it, that is why I showed them, but as I say in the video, in each of these examples they use a compromise or place it in a very precise context (such as ghost song and ori trading places). And for some of the games, even if the solution is good, such as Ori (which is probably the most impressive of them all), they might not necessarily be that reproducible by a single artist. I think there is a point to these scenes, but my point is that you need to be extremely careful when using it and use it in the right context. And adding it without consideration for the problems it introduces can negatively affect gameplay. And for quite a lot of games, especially if you want to make your art quicker, just adding fog and suggesting a horizon will be the quicker method of working while still creating nice results.
@@Nonsensical2D I do agree that it should be used carefully and done with a certain level of mastery, but I honestly think your proposed solution is generally bad advice, since it is discouraging experimentation and imposing a constraint that's frankly more a "you" thing than a general rule of sorts. Rather than encouraging better ways of creating simple and effective horizons, your suggested shortcut imposes a claustrophobic constraint that may turn people off from even attempting. This is my main issue with all this: this "it's too hard do something else" attitude this whole video exudes.
Hmm, I guess we might actually disagree on this point then. Which is fine, I do see where you are coming from. I agree that I am discouraging experimentation to some extent, but I feel that by and large this is justified. My overall implied premise to this argument is kind of:
1. You are not a super skilled artist
2. You want your game to look good and sell well
3. You have limited amount of time to make your game.
and given those three implied goals/limitations, I think showing a clear and distinct sky will in most circumstances come with a huge cost and won't align well with your situation, the ROI is bad. (or rather you should use a distinct sky where it makes sense and doesn't come with this huge cost).
Now if you are an indie team with skilled artists or a AAA title, then by all means, you can make your game even more unique and even cooler, but by that point you frankly already know how you will go about handling it. To me this problem has some similarities to the "open world" dream of many beginners, but to a much smaller extent. Everyone agrees the idea of an open world is cool, just that it perhaps isn't feasible for most people.
@@Nonsensical2D As I said in another comment. I still think making amateurish attempts at environmental variety makes for a far more interesting product than producing a more professional looking monotonous one with the same trick being applied all throughout. But this is quite a subjective matter at this point as different people will prioritize different things both in the making and consumption of media. All that said, I wish you good success in your development, and would suggest you to take caution with discouragement on the internet, since the industry is already plenty oppressing and discouraging for newcomers. Really gets on my nerves specifically because listening to similar takes held me back for quite a while in my drive to study and become better both at coding and making art. Cheers!
Nah, still don't get the point. Sounds like complains about parallax layer Y speed incorrect values that are not right for the context.
The world as we know it scrolls horizontally. If you have a horizon, you are basically somewhat limiting your ability to scroll vertically, because if you do, you will have difficulty adding assets and populating the scene, because it'll be tricky to think of a flying asset that will make sense in the context of your game (thus it can quite quickly result in a rather dull background). You can avoid this by not using vertical scrolling, use a nonsensical background, or utilise scene breaks. Or you do what many games do and only use open horizon areas in locations where you don't utilise much level design anyway (like Ori). I'm not saying it is impossible to deal with, just that it is problematic to deal with, and depending on your game and how much time you have, it could be worth it to avoid the sky altogether (like HK).
It has nothing to do with Incorrect values on the Y, even though the scene that I created is technically too close to the foreground to be realistic. The problem with placing it really far into the background is that your horizon starts looking really flat (or you need to add too many background layers), while the problem still kind of remains, you still can't do too much vertical scrolling.
wait but blasphemous 2 has plenty of platforming sections with the sky and clouds and spires in the background... im not making a 2d platformer so idk though.
i mean??? idunno
L take