In terms of screen shake, there is a small but important distinction to make. There's screen shake where the rendered image moves around and then there's screen shake where the physical camera (such as it is) moves around. Both can look similar but when you physically move the camera, this can have impact on character controls - some things may be using camera space for calculations such as reticle and aiming. Moving the camera can thus be disorienting and frustrating if the camera shake actually moves the reticle off target (or similar effect). Whereas moving the rendered image itself is purely visual and won't impact gameplay itself. Both are useful in different situations as they have different feels to them.
As a counterpoint on camera-motion screen shake: The fact that it affects aim might be a thing you actually want in your design, as it makes it harder to aim (duh) when your character is being shaken by an earthquake or perhaps a boss smashing the ground and trying to destabilize them. Now, there are other ways you could achieve this (as always), but it's something to consider. As always, design works best when you're intentional about how *and why* you do things.
@@LickingFire Oh agreed; that's indeed what I meant. Even within a certain type (shaking from explosions), there is that important design consideration you call out. A player-driven explosion (a thrown grenade) might be best as screen space since it's (presumably) a cool thing the player is doing in the heat of combat. While a thrown enemy grenade may cause game space shaking for that disorienting "I'm in trouble" feel.
Great video as always! The divider in the center of the screen to help with understanding the movement of the camera in relation to the player really helps visualize how the camera is working behind the scenes. Great use of visuals to help with understanding what is going on. Also super happy to see Pyre in this video, it's a hidden gem.
still haven't gotten a chance to play Pyre, but it's sitting in my library waiting. Bastion, Transistor, Pyre and Hades are all fantastic. I can't wait til they get past Hades 2 and show us what's next! (also excited for Hades 2 just love their new stuff so much)
@@LethalLuggage Their first game 'Bastion' was already a massive hit and basically made them the studio to talk about, when talking about indie successes, but you are right Hades has also been massive for them.
The staging of scenes in Gris is fantastic and is a perfect example of how camera work can also be used to emphasize art direction. It's hard to think of a game that handles that interaction as well as Gris does.
One effect that I feel could have been mentioned more is how limiting the speed of the camera can also emphasize aspects of characters or gameplay. A character will feel faster if the camera drags behind them. A perfect example is that Sonic games purposely put a cap on camera speed but not on Sonic, so in certain cases (Sonic 2 Chemical Plant for example) Sonic can outrun the camera entirely. It really makes Sonic feel like 'the fastest thing alive', and it rewards replay to learn when to act in order to maintain that insane speed as long as possible.
I'm not a fan of this personality. I like to be able to actually see what I'm doing. If the character moves to the right side I have no way of seeing what's ahead. I think a better solution is to zoom the camera out as you go faster and faster. Even pulling the character MORE left is helpful and can make tricky fast paced playforming sections much easier and satisfying. But that's my personal taste.
Limiting the visibility is also important to hide secrets. Take the Donkey Kong Country games for instance, the Snes ones have tons of things to discover, as opposed to the zoomed out camera version of the Wii, where the bonus are nothing more than skill challenges...
@@LukieLuke5correct but then you lose the speed feeling. In Sonic and especially in the first generations speed was the trademark of the franchise. With the hardware limitation and most games feeling slow, Sonic speed made the game sell a lot. So not giving the players this anticipation was a good design decision at that time.
@@LukieLuke5 It was a product of it's time. The classic Sonic games weren't meant to be played once. But through trial and error where you can master it and know what is coming up. Which is why the camera is so centered. You are not meant to see what is coming. YOu are supposed to git gud But as players change it's why generations is easier and uses offset camera.
When you talked about tracking two important objects in Pyre, I'm surprised you didn't mention fighting games. That's an entire genre where you have two characters on screen who're equally important (outside some special attacks and entry/ending poses). It's also a genre where positioning and spacing is extremely important, so you need to have a camera that accommodates that.
I guess the video focused on platformers otherwise we'd haeve a pretty big discussion on camera in general, for example it being the "actual toughetst" enemy in fromsoft games. Or how in fighting games as you mentioned camera can have even bigger effects on gameplay when a side switch happens, for example in tekken where an unexcpected side switch has caused a combo drop and a comeback, leading to some of the most memorable moments in tournaments.
@@gjergjipocari8227Those are 3D games, though, which he didn't talk about at all, probably because one more dimension means even more of complex camera mechanic.
Great video, as always, but I think it's important to remember that, in the case of Sonic, making the camera lag behind him was always intentional to give the sense of speed. The camera showed what is in front of Sonic in the original 1993 version of Sonic CD but this feature was removed from the game since the 2011 version. Sonic level design is more similar to a metroidvania than a tradicional platformer, "forwards" can be anywhere, right, left, up or down. If the camera tried to show in front of Sonic like it does with Mario for example it would be a shaky mess
Yeah, I get why some people prefer the Sonic CD camera, but the extra visibility was never worth trading away the sense of speed for me. And I love the design choice to allow Sonic to blast off the edge of the screen as a meta-commentary on his speed - if the game can’t keep up with him, then what can?
Personally, the “sense of speed” gotten from outrunning the camera doesn’t do anything for me. The argument that the path forward is constantly changing in direction in Sonic games does hold a little bit of weight however… …until you consider that there are other platformers that also do this, and find ways with the camera to still have visibility. I think what really matters and what is with sacrifice here is a subjective thing, but when I look at the fact that almost every modern platformer I can think of it that’s worth anything does not have issues like this I can’t help but view Sonic as something clinging to tradition to it’s detriment.
@@technicallyinept2120 which similar platformer have a better camera? The only good games similar to Sonic that comes in my mind are Freedom Planet or Spark the electric jester and both has a camera pretty much identical to Sonic. The camera in Freedom Planet 2 zoom out in some parts but that also happens in 2.5D Sonic. For high speed platformer with open level design I can't think of a better camera. During the 90s there were so many Sonic clones with more traditional camera where it was impossible to see what is going on with blind jumps and enemies coming from nowhere. Putting Sonic in the middle, making the camera move a bit when he jumps and not having enemies on speed section where stuff that always made playing Sonic feels natural. (Yes, Sonic had break this rules before but that's why nobody likes metropolis zone or the Game Great ports of Master System games)
@@GoldenJLR If I recall correctly, FP2’s zooms out dependent on your speed rather than the section of the level you are in. And neither is identical to 2D Sonic. However, the former is similar to 2.5D Sonic as you said, but for a number of reasons 2.5D Sonic just is not what I think of when it comes to matters like this as far “what Sonic does.” Sorry about that. Many, MANY metroidvania’s have you doing more than just heading right to progress to a level, but going in all types of directions. Same with Megaman. Same with Celeste. This is what I was actually talking about when I said that other games are able to do what Sonic does without sacrificing visibility. Not deliver the sense of speed without sacrificing visibility. But now that I think of it, I actually can think of a few platformers that still deliver that. Besides FP already mentioned, the first one that comes to mind is, believe it or not, a mobile game called Skyturns. There are actually a number of things I would love for Sonic to learn from that game. Fancy Pants Adventure also comes to mind. Also a sort of Mirrors Edge fangame made by the same developer.
I'm reminded of games where you can pan the camera yourself in some circumstances, like in Sonic the Hedgehog where you can hold up or down to move the camera in that direction to see what's above or below. And then in Sonic 3&K you can use this trick combined with jumping to break the game in all sorts of hilarious ways, though that falls more under glitches.
In Sonic, it takes far too long to look up or down. That, combined with the fact that the ring system nearly, basically makes you invincible, discourages stopping to do this and instead encourages just kind of rushing ahead blindly. Unless you’re trying for some alternative object that requires not taking damage, like holding on to rings for special stage entry.
The thing I am most proud of in my first game is it's camera, glad to see a lot of the little things I used mentioned here! One thing I did that I haven't seen a lot is having the zoom be tied to the character's speed, which allows for more lookahead when going fast (when you most need it), while allowing slower players more precision.
hey I'm sorry I've tried to answer you before but somehow my comment didn't register. I think the max zoom out was 3 times wider than the normal view. The game was quite low in detail so I didn't think about visual density too much, but maybe I should have! One inconvenient it had was diminishing the feeling of speed, 'cos like backgrounds aren't moving as fast if you zoom out, but I still think the tradeof in visibility was well worth it!
Honestly one of the best and most simple use of camera is rain world. The static camera switching between grids makes the game harder and more fun to play
People complain about the static camera but it's actually the correct choice for the gameplay. You always have to be on the lookout to spot any movement or weird pixels that might be an ambush predator and a moving camera makes that a lot harder. Each screen transition becomes a stressful test of reflexes as a predator might be waiting to snatch you up. It's brutal and unfair and how the game is meant to be. Also it makes audio clues more important! There's a mod that makes the camera dynamic and it just made me feel nauseous.
Rain World's camera system is often criticized due to its nature of hiding offscreen dangers, like a lizard which bites you right as the screen changes. In reality, most such deaths are completely avoidable with skill, and those patient enough to not quit after a couple of deaths will be able to learn from their experience and be more careful next time. It also encourages you to pay more attention to smaller environmental details and audio cues. I won't say it's objectively the best decision, but I personally like it and evidently a lot of people also did, so I don't think there's much to be gained from discussing whether or not it's objectively good or bad. Rain World isn't just a game: it's a work of art, and it's not designed to please everyone. That said, I'm not sure if it was really an intentional design choice. The way Rain World's graphic system works is that all screens are pre-rendered and stored as png files, which removes the need for in-game graphics computation. When the camera moves the game just switches to using a different png. The mentioned camera scroll mod stitches together each pre-rendered screen, which means that some depth elements might not look right, and occasionally there'll be a visible seam between two screens. I've heard in some places that graphical limitations played a role in the game's static camera system.
SB Camera Scroll from the steam workshop is a mod that only does one thing: center the camera on the player in multi-screen rooms, which turns the jarring transition into a gentle scroll. It's very interesting how such a small change can affect the gameplay so much, it makes it feel almost too easy.
_Secret of Mana_ is a great example of how a 2D camera can tangibly affect gameplay. - In the original 16-bit versions, the camera does NOT move until you reach about 2-3 tiles from the edge. This means you have almost no "lookahead" when exploring, but it also means the camera has more "dead zone" to remain stationary during battles. The camera is also blocked from moving off the edges of zones, so any time you CAN actually touch the edge of the screen implies it's an exit to another area (or it's an obvious wall). - On the flipside, your party members are not permitted to travel beyond the camera's reach so while they're normally pretty good at tracking behind your character outside of fights, there CAN be times where your characters are (by some circumstance) split so far apart across the screen, not only is the camera essentially pinned down, the characters are too. (Most notably in the boss battle against Dread Slime) In the game's 2014 remake, neither of the above points apply: - The camera is consistently locked onto your character's position at all times with no "dead zone". - Party members _are_ allowed to stray off-camera without problem (the game will teleport them back to your position if necessary). But where it actually impacts gameplay is how the game spawns monsters when their origin point is just off the edge of the screen -- so while there is a limit of 3 monsters at any time, the constantly-moving camera provides more opportunities for offscreen monsters to spawn in after you kill a different one. This makes individual areas feel more monster-dense than in the original.
Because of how often the trick is used to highlight a beautiful view or grand structure, I think that the camera pulling forwards and zooming out before coming to rest on a static shot primes me to feel a sense of awe or reverence.
Your videos are absolute masterclasses. The amount of focused quality information you pack in 10 minutes is really impressive. Thank you so much for what you do !
Cameras are one of those things that seem sorta simple until you start tweaking it. In a game I recently finished, I'd say I spent easily a couple weeks tweaking the camera before I was happy with it.
This video is like a throwback to GMTK ca 2017. The detailed breakdown of a single very practical topic, running through examples in a million games, even half the games he cites were mentioned in his videos from back then
Hit freezes are one of the coolest camera effects in gaming imo. I can‘t believe you didn‘t cite Smash for a top notch implementation, where the amount of frozen frames corresponds to the severity of the impact!
Always a treat to see Ori and the Will of the Wisps featured. I noticed it using not only damping but also framing; there's a lot of areas where my attention was helpfully drawn to something as Ori entered. There's even a spot where the walls are framed close to the edge of the screen, but then busting through a breakable wall suddenly shifts the camera to show you some of what's behind there! It's neat that they didn't let the camera give away the secret area until you had found it yourself. You can also hold up or down to have Ori look up or crouch, manually shifting the camera a bit in that direction.
I wish more games journalists would get into development. Even if you never go back to your magnet game or never take on another project, the insights you've gained and can pass along now that you've had those experiences are so much more nuanced and relevant to the small, detailed problems your audience will encounter. Thank you for doing what you do and communicating it so clearly!
Amazing video! I would have liked to see a segment on room to room cameras, but I loved the way you approached this topic! My favorite camera technique is the special event one, it is extensively used in Mario+rabbids kingdom battle to show key objects. Most people hate this because they feel it's too restricting, but I absolutely LOVE it.
In my experience, just slowly lerping the camera towards the player position solves all the jitteriness issues pretty well (both sideways and jump-wise) but you need a lot more lookahead to stop the player from escaping that comfortable zone where they see more ahead of themselves than behind. Much easier than implementing all five tricks... but you get an endless struggle of balancing two opposing forces.
Fun fact about not having to track jumps due to level layout: the NES had enough video memory to hold an area four times the size of the screen, but for technical reasons, unless the cartridge provided extra hardware, only half of that could actually be used. The game could choose between using the "bottom" half or the "left" half though, and when you use the bottom half, you can scroll sideways smoothly, and with the left half you can scroll up and down. And this configuration can be changed mid-game, which is how Metroid features both vertical and horizontal corridors, but no huge rooms that scroll in both directions.
3:30 I was studying other games' cameras while working on the camera for my own game recently, and I noticed that in Hollow Knight there are a _lot_ of regions where the camera is hardcoded to be locked along the Y-axis to a specific predetermined height and doesn't follow jumps. An easy place to see this is the long hall with all the boss statues in Godhome, where you can watch the camera do some interesting manuevers as you go between the first and second levels of the room.
Additionally, the frame size can influence the feel of your game. For example, I've seen a few games take inspiration from Alien Soldier's dash mechanic, with which you can quickly zip from one edge of the screen to the other. However, what a lot of those games miss is that Alien Soldier is displayed in a 4:3 ratio, and thus, each time you dash the screen must pan a little behind you to catch up, and it feels like you're super fast. In contrast, those modern games are displayed in 16:9 and the camera doesn't pan when you dash from one corner to the other, so the effect is greatly deminished.
I think you can still replicate that effect in wider screens, you just need to be a bit more creative. For example, Sonic 2,3, and Mania actually put the camera in a brief _look behind_ mode when Spindashing, before the camera runs off to put Sonic in the middle of the screen again. Of course, you need to make sure the player gets some level of protection if you're going to have a Look behind moment. The Spindash gets away with it due to it being one of the safest positions for the player to be in anyway
I'd strongly recommend developers consider the emotional component of the camera as well. In a game like Rain World or Celeste, the camera is fixed, and you traverse screen sized rooms. (Both have some exceptions, but in these the camera will move on one axis, strictly up/down or left/right, and the few rooms in Celeste with free camera movement emotionally convey the tone of being lost in a sea of confusing space; it is done purposefully). The fixed camera not only facilitates tight platforming, but it makes the player feel a certain way. Celeste is about climbing a mountain, and the fixed camera makes you feel small, and makes the environment feel like a character in and of itself. Rain World is about being at the bottom of the food chain, surviving in a world that isnt designed for you. Ironically enough, the one exception where the camera tracks you is a plot pivotal moment where the environment fades away and the focus is entirely on your character. I wont spoil it but its brilliant. There is even a mod for Rain World that makes the camera scroll, and I find it makes the game miserable to watch. To me, it removes the thematic beauty of the environments and changes the feel of the game entirely. If you want a game where the environment is a character, and the story explores the lead characters relationship with their world, consider a fixed camera.
Even among the really good game design channels, most of them miss little things like the unintuitive camera tricks that make for great play. I love how you catch the design tricks that most players (and some designers even) never think about, and lay them out in such a clear and concrete way.
Something that's really important that I'm not seeing here is that the offset of the camera can also give information subconsciously to the player, not only for seeing what they have ahead. If you have the camera move a little bit forward according to the player X velocity, you will get feedback if you bump into a wall when you are looking at something else on the screen (an enemy about to attack, for example). Same thing when aiming: having the player character offset a little bit to one side tells you that you're aiming to the opposite side of the screen, without you actually having to track the cursor.
8:49 Ultrakill does hitstop really good as well, parrying something makes the game and even the music pause while playing the parry sound. it's very satisfying!
wonderful list of things to spice up your cameras! I'd like to note that while this video might focus on 2D cameras pretty much all these same concepts stay true for 3D camera's too, although things get a little more tricky to get right in 3D sometimes, the basics ideas are the same.
Really cool video, i feel like the camera is the most overlooked element by regular players, but also there isn't a lot of resource on it dev-side, so thank you.
Good video! Very detailed explained how camera movement can improve any game. One thing i did a lot in the past, was moving the camera by its velocity to the target player position until the distance between the player are above some threshold. Even that feels much more nicer. Also its nice when the camera does not go outside the level bounds, so i always clip the camera position by half the visible game area - which looks like the camera was stopped naturally.
I decided to go and take a look at Commander Keen after watching this. I went with Episode 6, as that one has a shake effect. One of the enemies has an attack where it hits the ground and will temporarily stun you if you're on the ground at the time. The camera effect is very simple: the view lurches up so that the ground ends up at a certain height from the bottom of the screen, then goes back to normal height when the stun wears off. Simple but effective. Doesn't work if you're at the top of the level as that's as high as the camera can go, so you only get the stun sprite on Keen and ground-hit sound effect. Normal left-right movement has a small dead-zone near the middle of the screen about four times the width of the player character's sprite. Jumps don't cause vertical movement until you land on a platform or grab a ledge. Falling a long distance will keep you about Keen's height from the bottom of the screen. Using the jump cheat (makes you just keep going up as long as you hold down the jump key) will keep you at the top of the screen when rising a long distance. Going back and checking Episode 4 with its Gimmick Level where you're under water, swimming has a very small dead-zone not much bigger than the player's sprite, but it seems to apply in all directions.
I've spent my entire indie game dev career making 3D games, but I'm now making a 2D game with a friend. It's funny because of course I think of the camera movement in a 3D game, but it didn't even cross my mind to not lock it to the character for a platformer. Thanks for pointing this out :)
My favorite camera shennanigan is how in Sonic 3 (& Knuckles) the camera is deliberatly slowed down when the player builds to much speed in certain situations, like boost pipes in Hydorcitiy Zone. It's fun to imply you are so fast that the game can't keep up.
appreciate seeing a bunch of older indie classics as examples here, fun throwbacks!
ปีที่แล้ว
Thanks for making this video! I've reworked camera system in my game, to include moving ahead of the player - thanks to your advice. And now it feels much better!
celeste's camera is very well made i think. it has look ahead in the form of camera offset triggers, that when you go inside of, they offset the camera by a set x and y value. celeste also has built in damping, which has the consequence of not having madeline on screen if going fast enough (and celeste is one of those games where speed is unlimited and purely up to the platforms), but the simple solution to that is to set the camera offset higher. additionally, celeste has camera target triggers, which focus the camera on a certain point and makes it stationary. it's only madeline moving on screen. it's up to the map maker to use them well, but it helps make celeste a lot more fair.
0:40 fun fact: sonic cd, a 30 year old game actually used a method called extended camera. for that method, whenever sonic runs fast, camera gets moved to the front of him to give him more viewable area and time to react to obstacles.
in celeste mapmaking there are many many camera triggers that you can use to control the feel of the level. there are ones where you offset the camera, ones you set a target, ones you set an area the camera can move inside, etc. it's awesome
This is what gamedesign is all about. While developing a game, you're constantly faced with these problems and challenges, but there have been many thousands of people in the last 30 years who already dealt with these and similar problems and challenges, and channels like this one bridge the link between existing, yet hidden knowledge and the fact that we're all born stupid. This is school! and it's cool
I appreciate the note on options for screen shake 😅 sometimes I struggle with this and at that point in the video just the game footage was starting to be a struggle to view, luckily the video didn’t have segments too long of those kinds of shakes or retro camera jerks for me at the moment Appreciate you Mark
I never gave such academic thought to 2D cameras. I mean, intuitively, I knew what was going on, but on a more fundamental level I never really broke it down like that. Excellent video as always.
I recently implemented this in my two player game! The camera defaults to the center point between the two players, with some additional logic. If the players are facing the same direction, I give a look ahead in that direction; otherwise, I keep the camera centered between the two players. The idea is that if the players want to go different directions, the camera doesn't give preference to either player, but if they're looking the same direction, they're on the same page and a lookahead is fine in that case. Just thought I would give this additional point unless anyone else is making a two player game.
Fantastic video! Makes me wonder what is best for my game. It's a isometric management game so it's probably best to let the player take full control, unless I want to highlight something.
I feel like this is one of the more underappreciated parts of games, I'm not a game maker myself but love learning how games are made and this is one of those things where if you did well making your camera and framing shots then your player should never know how much work is put in to making that camera work. good cameras are not noticed but bad cameras are.
One thing I've been doing following a GDC talk on camera stuff is that I have my screen shake work off a trauma scale ranging from 0 to 1, decreasing the value over time. This way i can manage two variables for shake ( intensity and the degradation rate) while also making it easy to correlate actions with an amount of trauma.
I learned a lot about 2D platformer cameras just by building levels in Levelhead. Because the camera is unlocked by default, you have to manually set areas for camera locks to make the gameplay feel smoother, often adjusting the level design to fit.
I'd love a video on making a good 3D Camera! Soulslikes always notoriously have a camera that gets stuck or unable to see your character. So It'd be interesting to see a compilation of techniques used for 3D
I'd really appreciate if you could do a Designing for Disability episode on 3D cameras and motionsickness. There are a lot of games I'm unable to play because the camera makes me extremely nauseated, and a lot of them could be fixed with a few simple changes. I'm sure I'm not the only person like this, and I think someone with your reach could really make a difference. In my experience, the faster a camera moves, the more likely it is to cause motionsickness. Fast-paced first-person games are typically the worst offenders, but third-person games with a camera tightly locked to the player can be just as bad. Examples of games I can't even watch trailer footage of are Valorant, Mirror's Edge, and Spyro the Dragon. The easiest way to help is to just zoom out the camera and widen the FOV. The more on screen at once, the less the camera has to move, and the smaller those movements can be. In a third-person game, you can have an option for manual camera tracking, so the camera only turns when you tell it to, as well as a slider for camera tracking speed. It is also essential that the player can disable motion effects like camera bobbing, motion blur, and dynamic FOV. Changes to the roll of the camera are much more disorienting than changes to pitch and yaw. Last, and I cannot stress this enough, the HUD must not move, as well as any objects fixed to the camera, such as a player's gun or hand. I can accept not playing fast-paced FPS's, but being locked out of games like Outer Wilds because there's no way to turn off HUD rotation (AFAIK) is just depressing. IMO an example of a game with a good (third-person) camera is Xenoblade Chronicles. You have so much control: you can zoom out farther than most games would let you, or zoom in so far it becomes first-person. You can have the camera rotate automatically or manually, and adjust the speed at which it automatically rotates. You can adjust the height the camera sits at relative to your character. There's even a button to re-center the camera if things get weird. Best of all, most of these options can be dynamically changed while you're playing, without having to enter the menu. The more of these options you add, the better, but even just adding one can make the difference in someone being able to enjoy the game you worked so hard on.
That's because games just need to be fun, and having three dimensions isn't a prerequisite for that, despite some people vehemently proclaiming that anything 2D is automatically bad and/or "nostalgia bait" (which, frankly, sounds like a made-up argument - I've never heard of anyone wanting to play a modern 2D game "for the nostalgia"). And obviously, I'm not saying people can't prefer 3D games over 2D games - they're just different genres. Stating "all 2D games are bad" is like stating "all fictional literature is bad".
This is exactly the kind of content I would love to see more of on this channel. Really apprechiate theese "How can you make your game better by making little changes to X" videos.
I can only wonder if this is a prelude to a longer video covering 3D third person camera systems. That's such a daunting topic that it feels like multiple dissertations could be written about.
While you usually want to frame more viewing space above the player, its important to remember that is because most game secrions are designed to have players interact in that direction. Sometimes its also useful to pan the camera downwards to prevent issues like blind jumps, especially if the player needs to go downwards.
Masahiro Sakurai has talked on his channel about the camera in a similar way that you did, his first video even was about the hit stop mechanic that you also mentioned. That said this is a great video!
#1 rule is always think about what the player really wants to see right now, and the #2 rule is don't assume you know exactly what the player wants to see You want your camera to show them what they want to see but some times it can get in the way of them checking something out or making it very difficult to perform some weird unexpected move/trick, so keep it in moderation.
In terms of screen shake, there is a small but important distinction to make. There's screen shake where the rendered image moves around and then there's screen shake where the physical camera (such as it is) moves around. Both can look similar but when you physically move the camera, this can have impact on character controls - some things may be using camera space for calculations such as reticle and aiming. Moving the camera can thus be disorienting and frustrating if the camera shake actually moves the reticle off target (or similar effect). Whereas moving the rendered image itself is purely visual and won't impact gameplay itself. Both are useful in different situations as they have different feels to them.
As a counterpoint on camera-motion screen shake:
The fact that it affects aim might be a thing you actually want in your design, as it makes it harder to aim (duh) when your character is being shaken by an earthquake or perhaps a boss smashing the ground and trying to destabilize them. Now, there are other ways you could achieve this (as always), but it's something to consider. As always, design works best when you're intentional about how *and why* you do things.
@@LickingFire Oh agreed; that's indeed what I meant. Even within a certain type (shaking from explosions), there is that important design consideration you call out. A player-driven explosion (a thrown grenade) might be best as screen space since it's (presumably) a cool thing the player is doing in the heat of combat. While a thrown enemy grenade may cause game space shaking for that disorienting "I'm in trouble" feel.
@@PhotonBeastthe idea of Ralph Wiggum holding an M4, huddling behind a burnt up car and saying "haha, I'm in danger!" is sending me
@@underdweller thank you for sharing this image with the world 😆
There's also parallax backgrounds to take into account.
Great video as always! The divider in the center of the screen to help with understanding the movement of the camera in relation to the player really helps visualize how the camera is working behind the scenes. Great use of visuals to help with understanding what is going on. Also super happy to see Pyre in this video, it's a hidden gem.
Came down to the comment section specifically to shout out Pyre as well! Absolute masterpiece of a game. ❤️
still haven't gotten a chance to play Pyre, but it's sitting in my library waiting. Bastion, Transistor, Pyre and Hades are all fantastic. I can't wait til they get past Hades 2 and show us what's next! (also excited for Hades 2 just love their new stuff so much)
Is it really hidden? Supergiant has got to be one of the most widely-praised indie dev studios out there.
@@oliviapg to be fair they were a lot smaller before Hades. Now they're practically a household name.
@@LethalLuggage Their first game 'Bastion' was already a massive hit and basically made them the studio to talk about, when talking about indie successes, but you are right Hades has also been massive for them.
The staging of scenes in Gris is fantastic and is a perfect example of how camera work can also be used to emphasize art direction. It's hard to think of a game that handles that interaction as well as Gris does.
For some reason next April fools Day I want GMTK and Tom Scott to switch places and present the other's video
Kinda aged badly
@@bobgold202 how
@@cheeseboi8769tom scott quit or something
One effect that I feel could have been mentioned more is how limiting the speed of the camera can also emphasize aspects of characters or gameplay. A character will feel faster if the camera drags behind them. A perfect example is that Sonic games purposely put a cap on camera speed but not on Sonic, so in certain cases (Sonic 2 Chemical Plant for example) Sonic can outrun the camera entirely. It really makes Sonic feel like 'the fastest thing alive', and it rewards replay to learn when to act in order to maintain that insane speed as long as possible.
Yes, it works in Sonic since it only happens in the linear sections where you don't need to be fully aware of what's going on.
I'm not a fan of this personality. I like to be able to actually see what I'm doing. If the character moves to the right side I have no way of seeing what's ahead. I think a better solution is to zoom the camera out as you go faster and faster. Even pulling the character MORE left is helpful and can make tricky fast paced playforming sections much easier and satisfying. But that's my personal taste.
Limiting the visibility is also important to hide secrets. Take the Donkey Kong Country games for instance, the Snes ones have tons of things to discover, as opposed to the zoomed out camera version of the Wii, where the bonus are nothing more than skill challenges...
@@LukieLuke5correct but then you lose the speed feeling.
In Sonic and especially in the first generations speed was the trademark of the franchise.
With the hardware limitation and most games feeling slow, Sonic speed made the game sell a lot.
So not giving the players this anticipation was a good design decision at that time.
@@LukieLuke5 It was a product of it's time. The classic Sonic games weren't meant to be played once. But through trial and error where you can master it and know what is coming up. Which is why the camera is so centered. You are not meant to see what is coming. YOu are supposed to git gud But as players change it's why generations is easier and uses offset camera.
I literally needed this for my project. You’re a blessing 🙏🏻
Im also working on a 2D platformer
Me too. We were planning to change the camera, this video came in the right time
nice
Same here! What a great resource- so grateful for everything GMTK does! 🙏
When you talked about tracking two important objects in Pyre, I'm surprised you didn't mention fighting games. That's an entire genre where you have two characters on screen who're equally important (outside some special attacks and entry/ending poses). It's also a genre where positioning and spacing is extremely important, so you need to have a camera that accommodates that.
I guess the video focused on platformers otherwise we'd haeve a pretty big discussion on camera in general, for example it being the "actual toughetst" enemy in fromsoft games. Or how in fighting games as you mentioned camera can have even bigger effects on gameplay when a side switch happens, for example in tekken where an unexcpected side switch has caused a combo drop and a comeback, leading to some of the most memorable moments in tournaments.
@@gjergjipocari8227Those are 3D games, though, which he didn't talk about at all, probably because one more dimension means even more of complex camera mechanic.
Great video, as always, but I think it's important to remember that, in the case of Sonic, making the camera lag behind him was always intentional to give the sense of speed. The camera showed what is in front of Sonic in the original 1993 version of Sonic CD but this feature was removed from the game since the 2011 version.
Sonic level design is more similar to a metroidvania than a tradicional platformer, "forwards" can be anywhere, right, left, up or down. If the camera tried to show in front of Sonic like it does with Mario for example it would be a shaky mess
Yeah, I get why some people prefer the Sonic CD camera, but the extra visibility was never worth trading away the sense of speed for me. And I love the design choice to allow Sonic to blast off the edge of the screen as a meta-commentary on his speed - if the game can’t keep up with him, then what can?
Sonic is the most misunderstood game
Personally, the “sense of speed” gotten from outrunning the camera doesn’t do anything for me. The argument that the path forward is constantly changing in direction in Sonic games does hold a little bit of weight however…
…until you consider that there are other platformers that also do this, and find ways with the camera to still have visibility.
I think what really matters and what is with sacrifice here is a subjective thing, but when I look at the fact that almost every modern platformer I can think of it that’s worth anything does not have issues like this I can’t help but view Sonic as something clinging to tradition to it’s detriment.
@@technicallyinept2120 which similar platformer have a better camera? The only good games similar to Sonic that comes in my mind are Freedom Planet or Spark the electric jester and both has a camera pretty much identical to Sonic. The camera in Freedom Planet 2 zoom out in some parts but that also happens in 2.5D Sonic.
For high speed platformer with open level design I can't think of a better camera. During the 90s there were so many Sonic clones with more traditional camera where it was impossible to see what is going on with blind jumps and enemies coming from nowhere. Putting Sonic in the middle, making the camera move a bit when he jumps and not having enemies on speed section where stuff that always made playing Sonic feels natural. (Yes, Sonic had break this rules before but that's why nobody likes metropolis zone or the Game Great ports of Master System games)
@@GoldenJLR If I recall correctly, FP2’s zooms out dependent on your speed rather than the section of the level you are in. And neither is identical to 2D Sonic.
However, the former is similar to 2.5D Sonic as you said, but for a number of reasons 2.5D Sonic just is not what I think of when it comes to matters like this as far “what Sonic does.” Sorry about that.
Many, MANY metroidvania’s have you doing more than just heading right to progress to a level, but going in all types of directions. Same with Megaman. Same with Celeste. This is what I was actually talking about when I said that other games are able to do what Sonic does without sacrificing visibility. Not deliver the sense of speed without sacrificing visibility.
But now that I think of it, I actually can think of a few platformers that still deliver that. Besides FP already mentioned, the first one that comes to mind is, believe it or not, a mobile game called Skyturns. There are actually a number of things I would love for Sonic to learn from that game.
Fancy Pants Adventure also comes to mind. Also a sort of Mirrors Edge fangame made by the same developer.
I'm reminded of games where you can pan the camera yourself in some circumstances, like in Sonic the Hedgehog where you can hold up or down to move the camera in that direction to see what's above or below. And then in Sonic 3&K you can use this trick combined with jumping to break the game in all sorts of hilarious ways, though that falls more under glitches.
level wraps go crazy
In Sonic, it takes far too long to look up or down. That, combined with the fact that the ring system nearly, basically makes you invincible, discourages stopping to do this and instead encourages just kind of rushing ahead blindly.
Unless you’re trying for some alternative object that requires not taking damage, like holding on to rings for special stage entry.
@@technicallyinept2120 Or searching for something specific, such as entrances to those Special Stages.
@@gameman250 Nah, I pretty much never use the camera controls for that purpose. Never need to
Thank you for posting this with subtitles already attached. A lot of channels wait a day to a week before adding them for some reason
The thing I am most proud of in my first game is it's camera, glad to see a lot of the little things I used mentioned here!
One thing I did that I haven't seen a lot is having the zoom be tied to the character's speed, which allows for more lookahead when going fast (when you most need it), while allowing slower players more precision.
that's a great idea! may i ask how much did you zoom out and if it had inconvenients, like modifying the visual density on screen?
hey I'm sorry I've tried to answer you before but somehow my comment didn't register.
I think the max zoom out was 3 times wider than the normal view. The game was quite low in detail so I didn't think about visual density too much, but maybe I should have! One inconvenient it had was diminishing the feeling of speed, 'cos like backgrounds aren't moving as fast if you zoom out, but I still think the tradeof in visibility was well worth it!
Love how this is educational while also showcasing games that do a good job.
Honestly one of the best and most simple use of camera is rain world. The static camera switching between grids makes the game harder and more fun to play
thou it's annoying when you have to jump between platforms that are close but in different screens. Luckily, this doesn't happen very often.
People complain about the static camera but it's actually the correct choice for the gameplay. You always have to be on the lookout to spot any movement or weird pixels that might be an ambush predator and a moving camera makes that a lot harder. Each screen transition becomes a stressful test of reflexes as a predator might be waiting to snatch you up. It's brutal and unfair and how the game is meant to be. Also it makes audio clues more important!
There's a mod that makes the camera dynamic and it just made me feel nauseous.
Oooh, what an interesting point! I haven't considered it this way yet, but you're right, it does add a lot to the experience.
Rain World's camera system is often criticized due to its nature of hiding offscreen dangers, like a lizard which bites you right as the screen changes. In reality, most such deaths are completely avoidable with skill, and those patient enough to not quit after a couple of deaths will be able to learn from their experience and be more careful next time. It also encourages you to pay more attention to smaller environmental details and audio cues. I won't say it's objectively the best decision, but I personally like it and evidently a lot of people also did, so I don't think there's much to be gained from discussing whether or not it's objectively good or bad. Rain World isn't just a game: it's a work of art, and it's not designed to please everyone.
That said, I'm not sure if it was really an intentional design choice. The way Rain World's graphic system works is that all screens are pre-rendered and stored as png files, which removes the need for in-game graphics computation. When the camera moves the game just switches to using a different png. The mentioned camera scroll mod stitches together each pre-rendered screen, which means that some depth elements might not look right, and occasionally there'll be a visible seam between two screens. I've heard in some places that graphical limitations played a role in the game's static camera system.
SB Camera Scroll from the steam workshop is a mod that only does one thing: center the camera on the player in multi-screen rooms, which turns the jarring transition into a gentle scroll. It's very interesting how such a small change can affect the gameplay so much, it makes it feel almost too easy.
Just a minute in and I'm absolutely loving this video. Could use more of these deeper breakdowns.
i loved Gris's world/background so much i'm SO glad the camera worked as it did because there really was something to apreciate in the bg at all times
_Secret of Mana_ is a great example of how a 2D camera can tangibly affect gameplay.
- In the original 16-bit versions, the camera does NOT move until you reach about 2-3 tiles from the edge. This means you have almost no "lookahead" when exploring, but it also means the camera has more "dead zone" to remain stationary during battles. The camera is also blocked from moving off the edges of zones, so any time you CAN actually touch the edge of the screen implies it's an exit to another area (or it's an obvious wall).
- On the flipside, your party members are not permitted to travel beyond the camera's reach so while they're normally pretty good at tracking behind your character outside of fights, there CAN be times where your characters are (by some circumstance) split so far apart across the screen, not only is the camera essentially pinned down, the characters are too. (Most notably in the boss battle against Dread Slime)
In the game's 2014 remake, neither of the above points apply:
- The camera is consistently locked onto your character's position at all times with no "dead zone".
- Party members _are_ allowed to stray off-camera without problem (the game will teleport them back to your position if necessary).
But where it actually impacts gameplay is how the game spawns monsters when their origin point is just off the edge of the screen -- so while there is a limit of 3 monsters at any time, the constantly-moving camera provides more opportunities for offscreen monsters to spawn in after you kill a different one. This makes individual areas feel more monster-dense than in the original.
Because of how often the trick is used to highlight a beautiful view or grand structure, I think that the camera pulling forwards and zooming out before coming to rest on a static shot primes me to feel a sense of awe or reverence.
Omg, that's a me! 0:58 🙀
Thanks for the informative video and for including Super Catboy! ❤
Your videos are absolute masterclasses. The amount of focused quality information you pack in 10 minutes is really impressive. Thank you so much for what you do !
Cameras are one of those things that seem sorta simple until you start tweaking it. In a game I recently finished, I'd say I spent easily a couple weeks tweaking the camera before I was happy with it.
This video is like a throwback to GMTK ca 2017. The detailed breakdown of a single very practical topic, running through examples in a million games, even half the games he cites were mentioned in his videos from back then
Hit freezes are one of the coolest camera effects in gaming imo. I can‘t believe you didn‘t cite Smash for a top notch implementation, where the amount of frozen frames corresponds to the severity of the impact!
Always a treat to see Ori and the Will of the Wisps featured. I noticed it using not only damping but also framing; there's a lot of areas where my attention was helpfully drawn to something as Ori entered. There's even a spot where the walls are framed close to the edge of the screen, but then busting through a breakable wall suddenly shifts the camera to show you some of what's behind there! It's neat that they didn't let the camera give away the secret area until you had found it yourself. You can also hold up or down to have Ori look up or crouch, manually shifting the camera a bit in that direction.
I just added one of the LookAhead options to my 2D shooter and it feels so much better now! Thanks for the video, I had not thought of these things.
I wish more games journalists would get into development. Even if you never go back to your magnet game or never take on another project, the insights you've gained and can pass along now that you've had those experiences are so much more nuanced and relevant to the small, detailed problems your audience will encounter. Thank you for doing what you do and communicating it so clearly!
Amazing video! I would have liked to see a segment on room to room cameras, but I loved the way you approached this topic! My favorite camera technique is the special event one, it is extensively used in Mario+rabbids kingdom battle to show key objects. Most people hate this because they feel it's too restricting, but I absolutely LOVE it.
Thank you for not saying “dampening”. Appreciated more than you would believe.
In my experience, just slowly lerping the camera towards the player position solves all the jitteriness issues pretty well (both sideways and jump-wise) but you need a lot more lookahead to stop the player from escaping that comfortable zone where they see more ahead of themselves than behind. Much easier than implementing all five tricks... but you get an endless struggle of balancing two opposing forces.
Did you make this comment before watching the whole video? lol
Fun fact about not having to track jumps due to level layout: the NES had enough video memory to hold an area four times the size of the screen, but for technical reasons, unless the cartridge provided extra hardware, only half of that could actually be used. The game could choose between using the "bottom" half or the "left" half though, and when you use the bottom half, you can scroll sideways smoothly, and with the left half you can scroll up and down. And this configuration can be changed mid-game, which is how Metroid features both vertical and horizontal corridors, but no huge rooms that scroll in both directions.
3:30 I was studying other games' cameras while working on the camera for my own game recently, and I noticed that in Hollow Knight there are a _lot_ of regions where the camera is hardcoded to be locked along the Y-axis to a specific predetermined height and doesn't follow jumps. An easy place to see this is the long hall with all the boss statues in Godhome, where you can watch the camera do some interesting manuevers as you go between the first and second levels of the room.
Additionally, the frame size can influence the feel of your game. For example, I've seen a few games take inspiration from Alien Soldier's dash mechanic, with which you can quickly zip from one edge of the screen to the other. However, what a lot of those games miss is that Alien Soldier is displayed in a 4:3 ratio, and thus, each time you dash the screen must pan a little behind you to catch up, and it feels like you're super fast. In contrast, those modern games are displayed in 16:9 and the camera doesn't pan when you dash from one corner to the other, so the effect is greatly deminished.
I think you can still replicate that effect in wider screens, you just need to be a bit more creative. For example, Sonic 2,3, and Mania actually put the camera in a brief _look behind_ mode when Spindashing, before the camera runs off to put Sonic in the middle of the screen again.
Of course, you need to make sure the player gets some level of protection if you're going to have a Look behind moment. The Spindash gets away with it due to it being one of the safest positions for the player to be in anyway
I'd strongly recommend developers consider the emotional component of the camera as well. In a game like Rain World or Celeste, the camera is fixed, and you traverse screen sized rooms. (Both have some exceptions, but in these the camera will move on one axis, strictly up/down or left/right, and the few rooms in Celeste with free camera movement emotionally convey the tone of being lost in a sea of confusing space; it is done purposefully). The fixed camera not only facilitates tight platforming, but it makes the player feel a certain way. Celeste is about climbing a mountain, and the fixed camera makes you feel small, and makes the environment feel like a character in and of itself. Rain World is about being at the bottom of the food chain, surviving in a world that isnt designed for you. Ironically enough, the one exception where the camera tracks you is a plot pivotal moment where the environment fades away and the focus is entirely on your character. I wont spoil it but its brilliant.
There is even a mod for Rain World that makes the camera scroll, and I find it makes the game miserable to watch. To me, it removes the thematic beauty of the environments and changes the feel of the game entirely.
If you want a game where the environment is a character, and the story explores the lead characters relationship with their world, consider a fixed camera.
You don't know how happy I was to see Sonic on one of your thumbnails, great video too!
Even among the really good game design channels, most of them miss little things like the unintuitive camera tricks that make for great play. I love how you catch the design tricks that most players (and some designers even) never think about, and lay them out in such a clear and concrete way.
I really enjoy the great number of examples you include in these videos. Good Work!
you always enlighten me on how much really goes on in making a game and how many choices a game maker has to give their game a specific feel!
Something that's really important that I'm not seeing here is that the offset of the camera can also give information subconsciously to the player, not only for seeing what they have ahead. If you have the camera move a little bit forward according to the player X velocity, you will get feedback if you bump into a wall when you are looking at something else on the screen (an enemy about to attack, for example). Same thing when aiming: having the player character offset a little bit to one side tells you that you're aiming to the opposite side of the screen, without you actually having to track the cursor.
8:49 Ultrakill does hitstop really good as well, parrying something makes the game and even the music pause while playing the parry sound.
it's very satisfying!
wonderful list of things to spice up your cameras! I'd like to note that while this video might focus on 2D cameras pretty much all these same concepts stay true for 3D camera's too, although things get a little more tricky to get right in 3D sometimes, the basics ideas are the same.
Really cool video, i feel like the camera is the most overlooked element by regular players, but also there isn't a lot of resource on it dev-side, so thank you.
It's definitely one of those "if you do it right, (probably) no one will notice, but if you mess it up, everyone will" things
Good video! Very detailed explained how camera movement can improve any game.
One thing i did a lot in the past, was moving the camera by its velocity to the target player position until the distance between the player are above some threshold.
Even that feels much more nicer. Also its nice when the camera does not go outside the level bounds, so i always clip the camera position by half the visible game area - which looks like the camera was stopped naturally.
This is incredibly helpful, thank you! We'll definitely apply some more of these tricks in our game, you have great timing :)
I really love these in depth looks at decisions I wasn't aware developers have to make
Wow I’d never even thought how much the camera movement affects the feel of 2D games! I’ll definitely notice it now. Illuminating video as always.
i'm glad that this video mentioned your Platformer Toolkit because that is great content that more people should know about
I decided to go and take a look at Commander Keen after watching this. I went with Episode 6, as that one has a shake effect. One of the enemies has an attack where it hits the ground and will temporarily stun you if you're on the ground at the time. The camera effect is very simple: the view lurches up so that the ground ends up at a certain height from the bottom of the screen, then goes back to normal height when the stun wears off. Simple but effective. Doesn't work if you're at the top of the level as that's as high as the camera can go, so you only get the stun sprite on Keen and ground-hit sound effect.
Normal left-right movement has a small dead-zone near the middle of the screen about four times the width of the player character's sprite. Jumps don't cause vertical movement until you land on a platform or grab a ledge. Falling a long distance will keep you about Keen's height from the bottom of the screen. Using the jump cheat (makes you just keep going up as long as you hold down the jump key) will keep you at the top of the screen when rising a long distance. Going back and checking Episode 4 with its Gimmick Level where you're under water, swimming has a very small dead-zone not much bigger than the player's sprite, but it seems to apply in all directions.
I've spent my entire indie game dev career making 3D games, but I'm now making a 2D game with a friend. It's funny because of course I think of the camera movement in a 3D game, but it didn't even cross my mind to not lock it to the character for a platformer. Thanks for pointing this out :)
Some good advice! Eager to see how the camera feels in your game.
My favorite camera shennanigan is how in Sonic 3 (& Knuckles) the camera is deliberatly slowed down when the player builds to much speed in certain situations, like boost pipes in Hydorcitiy Zone. It's fun to imply you are so fast that the game can't keep up.
appreciate seeing a bunch of older indie classics as examples here, fun throwbacks!
Thanks for making this video!
I've reworked camera system in my game, to include moving ahead of the player - thanks to your advice. And now it feels much better!
celeste's camera is very well made i think. it has look ahead in the form of camera offset triggers, that when you go inside of, they offset the camera by a set x and y value. celeste also has built in damping, which has the consequence of not having madeline on screen if going fast enough (and celeste is one of those games where speed is unlimited and purely up to the platforms), but the simple solution to that is to set the camera offset higher. additionally, celeste has camera target triggers, which focus the camera on a certain point and makes it stationary. it's only madeline moving on screen. it's up to the map maker to use them well, but it helps make celeste a lot more fair.
0:40 fun fact: sonic cd, a 30 year old game actually used a method called extended camera. for that method, whenever sonic runs fast, camera gets moved to the front of him to give him more viewable area and time to react to obstacles.
Literally the GMTK episode I needed at the moment. Took a break from the game I'm making to watch this and review what I'm doing XD tHank you!
6:09 great of you to show Gravity Circuit! It's the best action platformer I've played in years.
in celeste mapmaking there are many many camera triggers that you can use to control the feel of the level. there are ones where you offset the camera, ones you set a target, ones you set an area the camera can move inside, etc. it's awesome
This is what gamedesign is all about. While developing a game, you're constantly faced with these problems and challenges, but there have been many thousands of people in the last 30 years who already dealt with these and similar problems and challenges, and channels like this one bridge the link between existing, yet hidden knowledge and the fact that we're all born stupid. This is school! and it's cool
I appreciate the note on options for screen shake 😅 sometimes I struggle with this and at that point in the video just the game footage was starting to be a struggle to view, luckily the video didn’t have segments too long of those kinds of shakes or retro camera jerks for me at the moment
Appreciate you Mark
This video and comments are a wealth of knowledge. I have nothing to add to the conversation. I learnt a lot, so thanks everyone
I never gave such academic thought to 2D cameras. I mean, intuitively, I knew what was going on, but on a more fundamental level I never really broke it down like that. Excellent video as always.
I recently implemented this in my two player game! The camera defaults to the center point between the two players, with some additional logic. If the players are facing the same direction, I give a look ahead in that direction; otherwise, I keep the camera centered between the two players. The idea is that if the players want to go different directions, the camera doesn't give preference to either player, but if they're looking the same direction, they're on the same page and a lookahead is fine in that case.
Just thought I would give this additional point unless anyone else is making a two player game.
Every time gmtk uploads, a rush of dopamine grand slams my brain
0:28 Thought he was talking about cubing for a second
Fantastic video! Makes me wonder what is best for my game. It's a isometric management game so it's probably best to let the player take full control, unless I want to highlight something.
This might be your best video yet! Brilliant work.
Great video! This really helped me think about a camera when making my own game - I love your work, but this is one of your best!
I feel like this is one of the more underappreciated parts of games, I'm not a game maker myself but love learning how games are made and this is one of those things where if you did well making your camera and framing shots then your player should never know how much work is put in to making that camera work. good cameras are not noticed but bad cameras are.
What a treat for indie dev! Best theories for Camera movements
One thing I've been doing following a GDC talk on camera stuff is that I have my screen shake work off a trauma scale ranging from 0 to 1, decreasing the value over time. This way i can manage two variables for shake ( intensity and the degradation rate) while also making it easy to correlate actions with an amount of trauma.
Just like the decay envelope on an oscillator in audio synthesis! I wonder what effects you could pull off if you implemented a full ADSR envelope?
I learned a lot about 2D platformer cameras just by building levels in Levelhead. Because the camera is unlocked by default, you have to manually set areas for camera locks to make the gameplay feel smoother, often adjusting the level design to fit.
Very insightful and presented understandably. Thank you!
I'm so glad Muno was featured here. From the RoA workshop to the world!!
this video had me watching some clips i recorded of games ive played to see how their camera works. neat !
This video is an excellent reminder of all the fantastic 2D games that have came out since the indie Renaissance
I'd love a video on making a good 3D Camera! Soulslikes always notoriously have a camera that gets stuck or unable to see your character. So It'd be interesting to see a compilation of techniques used for 3D
Amazing video! I love how it can be practically applied, and the resources! Mark, you are truly a blessing ❤
Never thought about the camera too much, but this is good to know. Great video!
Love this man's videos! I always wondered what is GMTK top 10 in games of all history so far.
I'd really appreciate if you could do a Designing for Disability episode on 3D cameras and motionsickness. There are a lot of games I'm unable to play because the camera makes me extremely nauseated, and a lot of them could be fixed with a few simple changes. I'm sure I'm not the only person like this, and I think someone with your reach could really make a difference.
In my experience, the faster a camera moves, the more likely it is to cause motionsickness. Fast-paced first-person games are typically the worst offenders, but third-person games with a camera tightly locked to the player can be just as bad. Examples of games I can't even watch trailer footage of are Valorant, Mirror's Edge, and Spyro the Dragon.
The easiest way to help is to just zoom out the camera and widen the FOV. The more on screen at once, the less the camera has to move, and the smaller those movements can be. In a third-person game, you can have an option for manual camera tracking, so the camera only turns when you tell it to, as well as a slider for camera tracking speed.
It is also essential that the player can disable motion effects like camera bobbing, motion blur, and dynamic FOV. Changes to the roll of the camera are much more disorienting than changes to pitch and yaw. Last, and I cannot stress this enough, the HUD must not move, as well as any objects fixed to the camera, such as a player's gun or hand. I can accept not playing fast-paced FPS's, but being locked out of games like Outer Wilds because there's no way to turn off HUD rotation (AFAIK) is just depressing.
IMO an example of a game with a good (third-person) camera is Xenoblade Chronicles. You have so much control: you can zoom out farther than most games would let you, or zoom in so far it becomes first-person. You can have the camera rotate automatically or manually, and adjust the speed at which it automatically rotates. You can adjust the height the camera sits at relative to your character. There's even a button to re-center the camera if things get weird. Best of all, most of these options can be dynamically changed while you're playing, without having to enter the menu.
The more of these options you add, the better, but even just adding one can make the difference in someone being able to enjoy the game you worked so hard on.
Your timing is impeccable.
Camera movement: **exists**
Rain World: Nah, I'm not feeling it
Im amazed how its 2023, yet 2D games can still be hella entertaining in an era of 3D gaming
That's because games just need to be fun, and having three dimensions isn't a prerequisite for that, despite some people vehemently proclaiming that anything 2D is automatically bad and/or "nostalgia bait" (which, frankly, sounds like a made-up argument - I've never heard of anyone wanting to play a modern 2D game "for the nostalgia").
And obviously, I'm not saying people can't prefer 3D games over 2D games - they're just different genres. Stating "all 2D games are bad" is like stating "all fictional literature is bad".
Mark just dropped the line "in the fantasy basketball RPG Pyre" so casually like that isn't an absolutely insane combination of genres
This was good, but now, we need the 3D Camera Tutorial
This is exactly the kind of content I would love to see more of on this channel. Really apprechiate theese "How can you make your game better by making little changes to X" videos.
Glad you talked about 2D Sonic and how hard it is to predict where you're going. I'm always running into walls in those games.
1:18 Cave Story jumpscare ( Seriously, it really caught me off-guard )
I do not make video games, but I love your channel.
You really made it easy to understand, as always.
Fantastic video. I will actively look now for more camera behaviour in general
I can only wonder if this is a prelude to a longer video covering 3D third person camera systems. That's such a daunting topic that it feels like multiple dissertations could be written about.
Videos like this will hopefully spawn a new indie game golden age
As brief as it was, it was great seeing Pyre here. Supergiant's most under appreciated game.
Every time that a notification GMTK popups up, I run to see what is next!
Some of the plosive "P's" in this video were immensively powerful.
One super helpful tip I once heard about screenshake is to add "more than you need but less than you want"
2:26 that was your first swear mark , let history record it
While you usually want to frame more viewing space above the player, its important to remember that is because most game secrions are designed to have players interact in that direction. Sometimes its also useful to pan the camera downwards to prevent issues like blind jumps, especially if the player needs to go downwards.
Masahiro Sakurai has talked on his channel about the camera in a similar way that you did, his first video even was about the hit stop mechanic that you also mentioned. That said this is a great video!
Mathematically, the 3D camera video should be nine times longer 😂
#1 rule is always think about what the player really wants to see right now, and the #2 rule is don't assume you know exactly what the player wants to see
You want your camera to show them what they want to see but some times it can get in the way of them checking something out or making it very difficult to perform some weird unexpected move/trick, so keep it in moderation.