don't click here

Sonic Level Design

Discussion in 'General Sonic Discussion' started by Sparks, Mar 20, 2012.

  1. Yuzoboy


    After all these years, I didnt know some of these areas existed. Very interesting information, thank you. I hope there are more on the way with the other zone routes analysed. They could become a feature on some of the videos I do.
  2. Sparks


    Sondro Gomez / Kyle & Lucy
    Updated with Sonic 3K, Homing Attack, Y Axis maps, Character Specific Routes, and updates to elbow ramps. I'll be addressing Sonic CD, and The Right Mindset next time. Let me know if I missed anything else.
  3. Tyty


    ↑ & ↓ & ↻
    Something I'd like to point out, would be the application of the homing attack to potential level design, especially if the homing attack is done well. Simply put, areas where the level is designed to have steps (Like this.). If you put something like a motobug on the steps, something already trivial to kill, Sonic could potentially homing attack his way up rather than jumping on each motobug individually. Hell, if you want it to be a bit more challenging, it could be crabmeats or some other enemy that shoots, as long as the player gets a chance to see what they do. The better player would recognize the chain and homing attack up, but it wouldn't really be annoying (especially if it's short. Just enough so Sonic can't jump the whole way at once). It would lend itself to simply using Sonic's skills, as natural as spindashing before a loop.

    Sega could use it well in a 2D setting. They just haven't, because chains in the air is what they've been doing since Sonic Adventure.
  4. Deef


    Awesome topic and awesome OP. I've been hanging out for a good level design discussion since forever; there's so much to learn.

    My position on the subject is one of very little knowledge. When it comes to actual creation I've always felt like I needed classes in it or something heh; like I know I'm missing too many aspects of it to produce anything more than some fast bits, some slow bits, and a few main paths that interweave then wonder it if it works. That really isn't much to say for having direction.

    So I don't think I can contribute a lot, certainly no experience anyway, but I am hooked on the topic.

    However here is one thing I want to say that feels right I think:

    In designing a level you have to stay aware of the compulsion that the player is experiencing. Your level is something you're giving to the player, each step of it. A good example of this is EHZ1. Just after the first bridge there's that brief incline. That was deliberate. Or that devious spring mentioned in the OP, at the end of the act. So you have to have enough inspiration to build a level like this also. If you just want to get it done... well personally I wouldn't be able to make even an average level that way. But perhaps I'm over dramatising it and it is possible to still produce something people find enjoyable by simply ticking all the boxes.

    Boy it's hilarious to look at a map of Marble Zone after a map of Marble Garden Zone.

    Anyway, some direct responses to your OP:

    · The low/average/high examples were really clarifying for me. See, that gives me a better sense of direction already. I've maybe heard that some time before, it doesn't strike me as completely new, but if I have I didn't absorb the points about each path having a role. So that was good to understand. I would like to hear more on Sonic 3's style in this light. I prefer Sonic 3's levels overall but I do miss the styles of things like GHZ, EHZ, ARZ, CasinoNZ as well, where a level is kind of like stacked layers. I think MHZ and ICZ2 are nice in this regard actually.

    · Also looking forward to some Sonic CD discussion. I never know how I feel in those zones. Wacky Workbench is hell though.

    · Personally I find EHZ more enjoyable that CPZ. Rather than argue level preference, I just want to raise 2 points about EHZ that I think matter in level design:
    (1) I think having what feels like access to open sky makes a big difference.
    (2) EHZ's paths are frequently connected, typically in falling from one place to another. CPZ's not nearly as frequently.

    These 2 points combine to create an argument for an open sandboxy kind of feel in Sonic levels. For me, AIZ suffers among the green zones without a free sky, while ARZ feels nice with frequent sight of a way to change path, while I find PPZ, CCZ and StardustSZ to be a bit too overloaded with free access to other places. Anwyay, a big part of an act's fun for me was seeing how much I could twist (or at least think I am twisting) the path designer's original plans. Either by accessing places from other places it seemed I wasn't meant to access them from, or by simply being able to change path on a whim despite it serving no purpose or progress. CPZ lets the player change paths, but it also has very little in the way of feeling like you can do what you want until the designer's is explicit in allowing you. It has a lot of moments of keeping you glued to one path for quite a run.

    Some examples of that former point (stretching the designer's original plan) are: leaping from a mobius strip to the top of the yellow-spring-to-4-coconuts loop in EHZ, getting on top of loops in HTZ, dropping long ways between gondolas in HTZ again, climbing to the top of ICZ with Knuckles (admittedly truly unintentional), getting from the low to the high path in MHZ or just reaching that very high very powerful diagonal red spring in MHZ that sends you soaring over a huge chunk of terrain. In that last example it's just nice to know there are paths you can't see below you that you're freely passing up.

    Edit - Oh, Violet already raised the issue of how branching occurs, so I look forward to more discussion on that. But I'm basically asking about the contrast of obvious branching, to branching options that do not appear to be deliberate, and if anyone else finds signficant value in that. (It would be self-contradicting to try and create branches that genuinely aren't deliberate.)

    · What's a path swapper?
    · I agree completely about the bottomless pit signs.
    · I will argue against red rings in a later post (out of time).

    Thanks for the topic and the work you put into the OP.
  5. XCubed


    Will Someday Own a Rent-A-Center Oldbie
    I didn't understand what a path swapper was either until looking at your posting. It's probably that circle thing in Scrap Brain Zone that shoots you down, or horizontal.
  6. RetroKoH


    Project Sonic 8x16
    Correct me if I'm wrong, but path swappers are what switch the character from reacting to one set of tiles, to reacting to another, and weren't introduced until Sonic 2 (or maybe Sonic CD). They are primarily used for loops, as well as other forms of terrain to give a sense of depth. Such as the tubes at the start of Chemical Plant Zone. Sonic 1 didn't have these, instead using a more primitive method for their loops.
  7. Deef


    If there had been a Sonic CD context to go on, I would have suggested the gimmicks in Stardust Speedway that literally switch paths on and off.

    But you're probably right KoH. Can you say more about Sonic 1's simpler loop method though?
  8. RetroKoH


    Project Sonic 8x16
    I have no clue how it works. I am not as familiar with the original code, as I am still new to hacking. (I prefer to work with GML) what I do know is pretty much what I told you.
  9. Lilly


    United States
    Shang Mu Architect
    This is a really beautiful guide; it's way more informative than most guides that I've studied in the past! I mean, just dang, never thought about changing the slope inclination of entrance and exits of loops, even after playing the 4th gen hits after all this time. I also feel like a major idiot for not taking character abilities into account either, and after adding a Knuckles-only path to a recent level of mine, I have to say it totally took the emptiness out it! Your guide really helped me out, Sparks, and now I don't feel so lost trying to design new levels for my own fan game.

    If there's one thing I might add to this, it would have to be badnik variety and designs, which is really important. The whole game can feel repetitive if some enemies become reoccurring pests, especially if they don't fit the level theme. Sonic 2 (To some extent.) and 3 typically used between three to even five enemies in one act, whereas Sonic 1 reused certain badniks like Crabmeat in nearly every zone; which made me want to go out of my way just to bop that little nuisance every time! =P Heck, there are some badniks that only appear once in a specific area of one act! (Sonic 3 Hydro City comes to mind with its flies.)

    Sonic After the Sequel and SRB2 are a couple of really great fan-made examples of this, because you're constantly encountering new badniks to troll you and get in your way, forcing the player to approach each section of a level in ways they wouldn't have in previous ones. Even something as simple as a tiny robot turtle can affect level design, and its a detail that I believe shouldn't be an afterthought in any stage.
  10. bluejayfusion


    Well done, man!

    Thank you for this insanely comprehensive coverage of Sonic level design. I don't doubt that their are casual gamers who probably don't appreciate how much thought goes into level design... I believe many of said gamers are currently known as Sonic Team / Dimps.

    Kidding aside, I think its sad that recent Sonic games seem to let most of this stuff fall by the wayside. Because the gameplay is so fast, they largely overlook the amazing depth and character that once went into classic Sonic- all the great layered pixel art backgrounds, the multiple paths, the quirky traps and gadgets. It used to fill me with anticipation to play a new Sonic game and wonder what unique zones awaited.

    I find it strange that the first Sonic Advance game that Dimps did had a lot of promise. Its almost like they lost their passion somewhere along the way... If you take Hot Crater Zone or Route 99 and removed the artwork, I doubt anyone could even distinguish them- all running, sparse enemies, no wonderment.
  11. Putmalk



    Since we're on the topic of level design, I thought I'd post this. Do you think this is consistent with the guidelines posted in the OP? And may I say, it truly is a fantastic guide (I see you posted the character specific paths, thank you!)
  12. Deef


    Can't make a judgement on different path layouts when only shown one path, looking at the map would be more useful/interesting.

    Sigh, it was such awesome music heh.
  13. Shadow Wedge

    Shadow Wedge

    Match-3 game
    That spring near the end of act 1 is pretty dickish since there's no reason to believe it would send you straight into spikes and it's impossible to react in time. Other than that it looked pretty good, although a map would be more helpful.
  14. Cyrus


    Sonic Gear 2.5D Unity3D Project
    Nice guide, I'll be using some of those concepts in my project! :)
  15. Shade Vortex

    Shade Vortex

    The Black Vortex Member
    USA, WA.
    Twitch Streams
  16. Rosie


    aka Rosie Member
    I think your hack shows excellently how levels or indeed a whole game based around Knuckles' abilities should be, aside from this:

    Other than that I think you've done excellently and that's exactly the kind of paths that cater to specific character's abilities that I'd like to see in upcoming Sonic Team games.
  17. Sparks


    Sondro Gomez / Kyle & Lucy
    Spreading out Gimmicks added.

    I would have done a lot more, but while trying to update a few images in the first post, I fucked up a level map pretty seriously. It got fixed, but it still tore up my mood for the day.
  18. Knucklez


    I love 2B 'n' ass. Member
    This guide should definately be wiki'd.
  19. DinodudeEpic


    I find this article to be really helpful in thinking up ideas about what Sonic levels should be about.

    Although, I'm wondering if we can have 3D levels with the same aspects. As in, 3D Sonic Levels with multiple paths not only upwards and downwards but left and right too, and it can have them all interconnect. (Just like what the opening post said, but transferred to a 3D plane.)
  20. Palas


    Don't lose your temper so quickly. Member
    I posted the in the SEGA Forums and I believe it could add something to the discussion:

    Sonic is fast.

    Let's just assume that. That's what we want to see in a Sonic game, I suppose. But how do you make the players perceive that? Surely we can just boost him to win, right? It's the sensation of speed that counts.

    Well, everyone disagrees nowadays. There needs to be platforming for a Sonic game to be good. But... why, exactly? Even Ken Balough did a statement saying that the level design in Sonic Generations (or was it Sonic 4?) was good because the 3-layer scheme as there (the higher you go, the more rewards, but the harder it is to stay there). Okay, let's just assume this is good for a gaming experience. But... why?

    Let me give you a morbid analogy: you want someone to like a mask you've made. It's a common, fullface mask made of plaster. Now, you can give it to him/her and s/he may like it and actually wear it. Okay, but some people will say "yeah, this sucks" and throw it on your face. You may come back with a fancy mask with feathers and glitter and sequins and whatnot, but this will always make someone unhappy. And, obviously, you can only give one mask. So you'll never make as much people like the mask you've made as you'd like to. But~! If you burn their faces with a goddamned flamethrower, hell will they value your mask and wear it!

    It's the same with level design. It's about contrast. It's about making then player notice when s/he's being fast and actually feel something about it, be it relief or thrill. That's why "models" of Level Design will always be insufficient.

    Ken Balough said that what made the level design in classic Sonic games awesome was that the higher you went, the more rewards you got, but the harder it as for you to stay there. There is truth in this statement, but at the same time it's a farfetched generalisation. For instance, how can this hypothesis explain the fact that nearly all speedruns are made through the top layer? How would this explain Sonic CD's level design, in which you could even run mindlessly, but you'd end up caught in some kind of trap or a more difficult path, whereas the ones at the top were mostly empty and allowed you to actually run?

    At the same time, we can't make the counter-generalisation and say we're right. This wouldn't explain Green Hill, for example. What I can suggest is that proper level design in Sonic is never about alternative pathways, exploration, shortcuts etc. It's about building an ambient.

    It's about making the level ahead an amusement park. There is a little bit of this, a little bit of that. A little bit of badniks' localisation, a little bit of slopes or something that provides instant, blazing speed. All in shades of gray, so that the player can experience AS MUCH AS POSSIBLE in a playthrough. This will allow the game to be more full and interesting. It's not about leading the way, but to stimulating or refraining certain behaviours: there are may kinds of players and they all have needs, so you better satisfy as many of them as you can.

    So there must be room for speedrunners, for collectors etc. But how exactly do you do that?

    I'll start with a very basic concept, which I'll call Clockwork. It's an ideal level in which you can follow all the way to the end, and then come back to the start through other one. Basically, you can reach anywhere coming from anywhere. Now, this is perfectly harmonic, as all of the player's choices are reversible. This could be the base for all level design, since it ultimately gives utter, absolute freedom to the designer and to the player.

    That's where the designer can start playing around. The deconstruction of a Clockwork is when a designer places a point of no return somewhere on the map, for instance, and starts placing the gimmicks which provide different sensations. This way, you could make it either what Ken said that's how it was or what I said Sonic CD was like. But~! It's important to unite both paradigms.

    That's where burning the player's face comes: Marble Zone. Notice how Scrap Brain follows Star Light, which follows Labyrinth Zone, which follows Spring Yard, which follows Marble Zone, which follows Green Hill. Notice a pattern? Grosso modo, it's an alternation of slow zones and fast zones. This makes the player feel the speed better, because there is the comparison with the last slow zone. Also, these are a different experience, in which you have to get a hold of yourself. Not surprisingly, Star Light is a fan favourite from Sonic 1, Chemical Plant is a fan favourite from Sonic 2, Ice Cap is a fan favourite from Sonic 3, Stardust Speedway is a fan favourite from Sonic CD etc. It's about the value you give to the speed you achieve. It's not about too much automation and too few platforming, but the degree of contrast or integration between them. And, funnily, there needs to be contrast even between contrast and integration in order to make the level design complete.

    Also, this can happen WITHIN a given stage. Sections of hardcore platforming, sections of pure speed. This is one thing - but, more importantly, you have to connect them all harmoniously, unlike what was done in Sonic Colors. If you could flatten Sonic Colors' level design, generally speaking, it would probably look like a hive or something like that: a lot of alternative pathways, no cohesion between them because they offered basically the same.

    It's important to have in mind that the feeling of progress in terms of challenge is equally important, so that the player knows what's happening in the game and feels compelled to go on. An useful tool are the skill tolls. These are elements that will "evaluate" your skill at that given time and decide whether you "deserve" to stay in that pathway or to go to an emptier or more dangerous one. These are badniks, flying spike balls, you've seen them all. It's really important not to keep on the same element for a number of reasons and, hell, they've done this a lot in the last games. This, along with the possibility to climb up (or fall down) and try again is important. Give me to pathways and they're just two roads I can cross. Give me a toll post on one of them and this will influence my behaviour.

    That's where rewards come. You will try and cross a skill toll if you know you'll be rewarded. And visible rings or item boxes are one of the elements that can influence your behaviour, too. The Great Carol Yas has stated, "it's all about seeing something and figuring out how to get there". And, hey, Carol Yas was the one guy who made Sonic great at the time, alright?

    Another principle also suggested by Carol Yas is that of long-term, middle-term as short-term objectives that must be put in front of the player. The long-term could be finishing the game or the zone, the middle-term could be finishing the stage; but what's of importance here are the short-term objectives.

    If you play Sonic 2 with such high degree of attention as to spot the short-term objectives, you'll actually lose the point. Because they appear as you play, naturally. Here, the point is to make such recognizable and memorable sections that the player will remember what to do about it. But, first, he'll have to figure out. It's like a puzzle, just applied to Sonic. There could even be a few actual puzzles in Sonic if the designers saw fit, but what's important is the dinamism. You can even be stopped, but there is a whole ambient out there scheming against you, and you must use it to reach your objective - sometimes even more than just passing through the obstacles, you have to use them. That's what our beloved momentum is all about: Sonic being the lord of the stage, not the other way around.

    But I did say that the designer is the one making stuff to mock on the player, right? Right. But how do the two statements relate without being contradictory? Well, it's just that the designer shouldn't, at first, push the player towards the rides in the amusement park. It's just the neon light that may or may not attract the player, and each one likes one type of ride.

    Yes, obstacles make platforming, but how? It's not just direct, frontal threatening. If you take a look, you'll notice that most of the obstacles in classic levels don't hit you - it's you who hits them by doing something that the level didn't want you to. Obstacles have a life on their own and it's your fault if you lose your rings. Not theirs. Spikes fly over your head, badniks shoot things because they shoot things etc. This is what you could call diffuse challenge. The danger is out there, so it does some influence on your behaviour.

    Platforming is generally not about something that you want the player to DO, but about something that you DON'T want the player to do. Jump, and you'll most likely get hit. Stand still for too long and see what happens. That's how the designer plays.


    Now, here, let me show you my views in a practical way. I sketched this level a while ago (some... months, maybe).


    I didn't place any badniks or rings for a reason, nor did I specify what each item box does. Now, on to what the objects are:


    Now, the first section I'd like to drag your attention to is this one:


    There are two points here: first, the player should be able to see the end of the stage (the signpost, the huge ring, whatever) from the starting point, but not the item box. I ask you. Will the player want to see what's to the left? Well, some of them may, some may not. But I won't move a finger to call his attention and say that there could be something to the left. Rings, badniks, everything could be a giveaway. So, if the player gets to the end and sees the item box, there will be the sensation that he didn't explore everything - and, better, s/he'll feel stupid about it because it as there all the time. I suppose this should be able to enhance replay value.

    Second, it's important to start a level showing off your gimmicks (the mini black hole, in this case), just to tell the player how they work without harming him/her. This way the game becomes more intuitive.

    The second important section is this:


    There are three ways from here. To the left, downwards and to the right. If the player takes the left or the bottom path, s/he'll inevitably reach a trigger. If s/he chooses the path to the right, there is a possibility that the first trigger s/he'll face will be the last one... and I don't want that, because it's too threatening to be the first one to meet and I want the player to be prepared by this time. It's obligatory, anyway. So, in theory, only experienced players seeking speedruns should come here (and I'm not even sure this would be the fastest way anyway). That's why I probably should put a skill toll here - the spikes are an example, but there could be a badnik "guarding" this path or, on the other hand, I could attract the player with visible, small rewards on the other paths. This relates with that principle I talked about, in which the paths which are hardest to reach usually become the easiest to cross - it's not always "more rewards, more difficult to stay there".

    Another section:


    I put those two item boxes there to attract the player to the other 4 ones that he can't reach. Some rings may be necessary to keep the interesting alive (like a trail of... I don't know, candy. Didn't a fairy tale use something like this?). I think that's how you stimulate exploration: "holy cow, 4 delicious item boxes... but how do I get there?" If you take a look, you'll realise it's a long way to reach these (more on this later). But~! Some players are more prone and have more iron will than others, but eventually one ill have to look everyhere to find a way. And it's optional, which prevents loads of frustration (even if not completely).

    Now, on the teleporters (required to reach the 4 item boxes placed outside of the maze). I didn't put any order between them. But I did make them visible, yet enigmatic (those two, side by side with an item box in-between). I wonder if this was a good choice. I mean, inevitably the first one will have to be that one, down there. It think it's the only possibility, but I'm not sure. Anyway, Sonic CD did a lot of this: show you something, leave you ondering just hat the heck that might be... and I find it interesting, even though it may look displaced sometimes.

    Moving forward:


    It is possible to blaze through this level. But~! Go ahead. Try it and see what happens. If you're reckless enough or have set a strategy beforehand, you can cross the center of this level and make it in record time. But, in order to do this, you have to avoid dying. Twice. And you'll have to be fast, but I would NEVER move a finger to make you faster. This is important. The reckless have a place in the world, too. Saying "who cares" and throwing yourself forward without fear is a positive trait, right? There should be reward for this sometimes. But there must be danger, too.

    The last section I'd like to talk about is this one:


    There is some precision to accessing the path to the left. If you can't or got attracted to the rewards (the checkpoint post, for example), you're screwed: here's one of the two sections where I applied that principle enunciated by Ken Balough (and by pretty much every fan out there): risk and reward. The risk is enormous: you have 5 seconds to climb a hole lot of platforms and reach the top, otherwie, you're dead. Challenge, people. There must be some life-threatening challenge (which the player is actually aware of) and I haven't seen this for a while.


    Now, is music a part of level design? What about art?

    I may be a little too holistic about level design. For me, structure s important, but one level must correlate with almost everything else in the game in order to be effective. And music, for instance, is very important. The pace of the tune may stimulate you or refrain you... or maybe make some efforts displaced. Let me give you an example.

    Sonic Colors has fast paced music. Amazing music, don't get me wrong, but if you jut TRY to stand still somewhere, the game will make this action seem weird. You gave to be always running in Sonic Colors, and this much I find intriguing. This way, platforming becomes undermined, even if very complex and intelligent (like in Starlight Carnival). Or, at least, it is transformed.

    On the other hand, Sonic CD's ambient, grim music made Metallic Madness just insurmountable, because it felt horribly, horribly wrong to run. Even in the Good Future. These effects must be balanced - how?

    The first thing to have in mind is that the tune must be easy to remember. This helps A LOT in creating a distinctive feeling for the stage. Actually, not having a catchy tune is probably the worst thing that can happen to a stage, because then you'll most likely remember what the stage was like, but not feel attached to it.

    But it must be fitting. Now, Sonic Colors' music WAS fitting in a way. Aquarium Park's tunes are just so Aquarium Park-ish I can't even deal. But~! Let me tell you, Planet Wisp Act 3 is a better tune than Act 1 and 2. Not because I like it more, but because it's also more "neutral" regarding behaviours. It doesn't feel awkward to be slow, but it also isn't weird to blast off.


    I agree with the OP and ith some replies to various degrees, but, still, I think I could only fully express my opinion with a giant wall of text. I'm sorry about that.