3D action games with freely rotating cameras need a way to focus on objects of interest, otherwise characters moving around enemies can cause the camera to lose sight of their target. It’s difficult to operate the attack buttons and camera at the same time. It’s also difficult to aim characters facing direction accurately when the camera is at their back, so Zelda: Ocarina of Time came up with the compromise of Z-Targeting, which is known in many games as Lock-on.
There are three types of Lock-On; Hard-Lock, Soft-Lock, and Camera Lock. Hard-Lock is toggled by holding or pressing a button to typically orients the character and camera to face a target. Hard-Lock can be found in the Devil May Cry series, Bayonetta, Dark Souls, Legend of Zelda, and Metroid Prime. It’s notable that in DMC and some parts of Bayonetta, Hard-Lock functions independent of the camera, so you’re not forced to look at the enemies you’re locked onto, which helps prevent multiple enemies from getting lost off camera. Soft-Lock functions automatically by orienting the character to face enemies when they attack at a close range. Soft-Lock allows players to move in any direction unrestricted, but always attacks accurately. Soft-Lock is used with the guns in DMC when you’re not locked on. It can also be found in Metroid: Other M, Metal Gear Rising, and DmC. Camera lock is when the camera is fixed to look in one direction, but the character’s actions are unaffected by the camera. This can be found in Metal Gear Rising, which allows you to lock the camera, but Raiden’s actions are unaffected by where the camera is locked. Raiden still soft-locks onto targets as normal.
So what type of lock-on is best? In my opinion Hard-Lock provides the most benefits, especially when the button for lock-on is held instead of toggled. It’s helpful to have the character face their target, enabling them to strafe. It visibly identifies which enemy is being targeted, as opposed to soft lock which has no visible indicator of who you’ll attack. The player can lock-on and lock off
and manually choose new targets as desired. It also allows the obvious boon of being able to use directional inputs with other buttons, by orienting the character relative to the target. Games without a lock-on in a 3D perspective have no absolute directions because the camera can change orientation. The player can only get relative directions such as in the case of Bayonetta where tapping the directions “forward, forward”, and then the attack button or “back, forward,” and attack uses certain moves like stinger. Bayonetta also had a Hard-Lock in addition to this input method. It can be hard to tap the directions with the correct timing, and even harder to tap those directions the right way to successfully hit the exact enemy you want. With a Hard-Lock, the player can just hold forward and press attack at any time. Dark Souls has a Hard-Lock, but the kick and jump attack commands are designed to work without it. They can be difficult to manually operate as a result, since they’re input similarly to a smash attack in Super Smash Bros rather than the easier DMC method of holding a direction and pressing attack.
The trouble is that once Hard-Lock is engaged, combat becomes about the line between the player and the opponent. It becomes 1 dimensional, literally. The only thing that matters is how far your attacks reach and how far both of you are from one another. Game elements have to be introduced so that there is a value to rotating the camera in 3D space for the player otherwise it stop being 3d in function. If the Hard Lock is perfect the player will always face the player’s opponent and will only need to move to avoid getting cornered. To prevent this, many games lock the aim straight forward or slow the character’s rotational speed or make movement slower. This encourages the player to lock on manually only for precision or special directional inputs, and lock off when they don’t need it.
For this reason there is a draw to combat without lock-on. It’s enjoyable to play Dark Souls without lock-on, and it’s technically more efficient for a lot of tasks as it allows the player to run and dodge in any direction unlike during lock-on. Bloodborne was effective by having the quickstep active in lock-on mode and roll active when not to differentiate the two modes. Non-lock-on combat tends to work better in games with isometric or top-down camera angles like Wonderful 101 since the player can clearly see which way the character is pointed but this will affect level layout. To have complex 3D environments and architecture that players walk up, down, and through some type of camera centering button and/or camera lock function quickly becomes necessary for simple navigation.
This video and others by the same guy can show how useful it is to lock off in Dark Souls.
A Soft-lock camera like in MGR remedies a clumsy camera but when multiple enemies are involved, locking the camera onto one of them can be a death sentence. The MGR camera is so close that enemies frequently get up behind the player and it’s practically biased to force you into corners because it pushes off walls so hard it faces straight into them. I feel like no lock-on combat is more successful in Dark Souls due to better camera design and the attack buttons being on the shoulder buttons, so you don’t need to pull your finger off the right stick to attack.
I want to see future games have characters move more to evade attacks rather than using specific dodge moves to invincibly pass through them, which will mean they need to not use cameras that are incredibly close to the character or low to the ground, so players can see where their character and enemies are in relation to each other. I think Dark Souls is very fun when fighting multiple enemies without using lock-on or when fighting bosses without the use of lock-on, but keeping lock-on as an option makes some actions and situations easier, so it’s more of an optional tool to be utilized rather than something necessary to keep the combat functioning at all (though it probably would be unreasonable to not include it).
Freely moving around enemies has troubles though because when you move away from them, you are no longer facing them. Your attacks will go the wrong way completely. Soft-Lock systems attempt to allow players to move around freely, then attack the enemy accurately regardless of rotation. Unfortunately soft lock is unreliable for targeting, because you can’t tell what you’re going to attack before you hit the button. At worst, Soft Lock ultimately reduces combat to just being about the distance from your opponent in much the same way as too good a Hard-Lock. The other trouble is that in a soft lock system, you can’t just hold a direction and press a button for a command move, because every direction is forward, none are back or sideways. In the future it might be interesting to build an automatic camera AI that is attracted to points of interest like enemies that you’re actively engaged in combat with. This can potentially be annoying to players though because they may not want this functionality, they may misunderstand what the camera prioritizes or actively dislike the camera’s choices of focus.
To build on the intelligent camera idea, my initial thought is that elements should be given a ranking, a priority in the camera’s AI. The player character obviously has absolute priority, so it is always in frame (or of course the camera flat out rotates around the character, so it’s not an issue). Next, a point projected ahead of the player character should receive more and more priority based on how long the character has been heading in that direction (much like how cameras autocorrect to the way you are going in existing games). Enemies should be granted priority based on proximity to you, whether they have caused harm to you, and whether you have recently caused harm to them. Further, elements of the environment can be given priority at the discretion of the designer. The idea is that in combat, the camera will orient to fit all combatants into the scene. When running away from enemies it will orient ahead of you as the priority of the enemies you were just fighting falls off and the priority of the direction ahead of you gets stronger. And when passing through without fighting, the camera priority will already be strong and enemy priority needs to build before they are a significant effect on the camera. Maybe the final effect can be visualized like a heatmap, where the camera points to the area with the most heat? Well, it’s more of a linear algebra thing really.
To break off my camera tangent: Moving around enemies, having them all clearly represented on camera, doing it from a close-to-the-character or behind-the-shoulder perspective most of the time, allowing the character to face towards the enemies without reducing their relationship to a one dimensional one or move without changing facing direction. This is tricky with our current controllers.
The thing is, if it could be pulled of, this would enable a game to have a lot more of moving across a 3D plane, and avoiding 3D enemy attacks, while simultaneously attacking them in 3D. There’s a very high level of depth that could potentially appear in a system that can pull this off. Having something be variable across space/time in a meaningful way is one of the easiest ways of giving a mechanic depth.
Think of a movie, movies can depict complex interactions in space like this very easily. The director can simply choose angles that capture everything relevant at any given moment in the frame. Because movies are not possibility spaces of potential events, choosing the right camera position for any one moment is easy. This is why handing camera control to the player is obviously the easiest solution to most camera problems, but players can’t operate cameras without sacrificing access to the face buttons, which sucks. So a good automatic camera is necessary. The problem is, our intuitive human judgment of the best place to put the camera is very hard to express algorithmically, such that in all possible situations the camera will be in a place that captures everything going on. Furthermore, in a movie shot, the actors can be placed to face any direction and simultaneously move while facing any direction, and rotate arbitrarily as they move. The direction that attacks are oriented is tremendously important to making 3D combat work.
To pick an example from a completely different genre, a game like Smash Bros gets away with having characters face away from one another because you have the C-stick that allows you to instantly attack in the opposite direction and you can jump or wavedash which temporarily fix your facing direction one way while moving. Imagine if there were twin stick action games in the vein of twin stick shooters, where you could use a second stick to control your facing direction. This would require automatic cameras, and if those simply locked directly to the enemy then the point would be lost anyway.
Implementing this into 3D space like DMC, Bayonetta, or Dark Souls is a control nightmare. With current controllers, I think you need to make sacrifices in some area or have an automated system handle the camera suitably (still having issues with the attack facing problem), you can’t have it all.
Fixed perspective games like Ys Origins or Oath in Felghana have good combat by means of avoiding all the camera and lock-on problems altogether, but at the cost of simpler combat systems and environments. This affords them the ability to mimic bullet hell games and have amazing weaving through enemy attacks at the cost of being mechanically simple.
Lock-on can be a tool and a hazard. If implemented too perfectly (such as in the 3D Zelda series), then none of the actual animations of attacks matter except their range and startup time. If not implemented at all, then players can quickly lose sight of enemies, especially when they need to press face buttons to attack, or aren’t used to using 3D cameras. It’s up to developers to implement camera and lock-on systems that help players stay on top of targets without trivializing what can potentially be an interesting part of combat. Hopefully future developments and refinement will allow for smoother camera systems and better gameplay in the future. Half the battle of developing a 3d game is framing it.
Thanks for this nice article.
I really prefer and like to play action games and first person shooter games. And Call Of Duty WW2 is one of the most one evergreen shooter game. Thanks for this good article and keep on.
Your style is really unique compared to other people I
have read stuff from. Thanks for posting when you’ve got the opportunity,
Guess I’ll just bookmark this site.
I don’t think your opinions on this have much merit as they come from a limited perspective. That perspective being played action games always on a modern console controller, not considering other types of camera design and clearly not having enough experience to gauge whether some of these ‘issues’ are objective flaws.
Simply locking the camera to face your character’s direction (like a first person shooter.) alleviates nearly all of these issues even on the clumsy controller. It’s almost ridiculously easy to accurately target and move at the same time which is why a camera like this is used for shooters first and third person in general. You can strafe and move and face in any direction you like allowing you to easily maneuver around an enemy while still targeting and being able to attack a single enemy. It also means that it’s the player’s fault if enemies sneak up behind them.
The twin stick design (camera controlled with one, movement with the other.) is definitely the clumsier of the two. As someone who still prefers no lock on system in these types of games one of the big difficulties I find isn’t aiming the character in the correct direction to attack, it’s confusing the two sticks when trying to quickly change direction. To target something that is behind or off to the side of what I was just targeting. Navigation is also generally poor in these types of games, characters don’t move with nearly the same ease or precision to avoid attacks.
Even with the lack of precision in movement and poor aiming. In the few games where enemies don’t fly towards you witcher style (don’t cover 10, 15 metres.) and with enemies with a shorter range, the ai in these games can often be poor enough both in intelligence and freedom of movement that it can be very easy to move in and out and land attacks without getting hit. A lot of the difficulty coming from enemies having attacks that aren’t possible to dodge without some form of invincibility because they will fly across the room targeting you perfectly. Or enemies throwing a lot of attacks with very little time in between them that cause a lot of stagger. It would still be quite easy to hit the latter kind if games like Dark souls didn’t restrict your mobility so heavily.
This is a discussion of lock-on styles in 3d action games, not a discussion of camera design more broadly. I have experience playing games where the camera is locked to the character, such as God Hand and God of War 2016 on console, and Gunz: The Duel, Warframe, and Chivalry on PC.
The limitation of games like God Hand and God of War is they have limited turning speed on the characters. They need to have a 180 turn button to accommodate for this. It’s the player’s fault if an offscreen enemy hits them in any game, but having cameras that better frame the action without requiring the player to intrevene is generally a good thing, because it allows players to be more focused on the combat without needing to touch the right stick.
The limitation of PC games is that when the character is tied to the camera, then they can’t have commitment on the direction of their swing, so swings in these games are generally designed to be fast (gunz), or they let you change the direction the attack is going as you do it (chivalry).
The other limitation of these styles of games is that they’re worse at platforming than 3rd person free cam games, and Keyboard and Mouse have less buttons available readily (even if they have more buttons total), so they’re not as suited to action games like Devil May Cry, where you have access to many action buttons at once.
A lot of your comment seems to indicate a lack of experience with console freecam games. (Confusing which stick is which? Are you joking me? Left is always movement, right is always camera, this has been a standard for almost 2 decades) So it’s kind of weird that you’d accuse me of a limited perspective.
Also you’re making generalizations about a lot of games, but you’re not citing exactly which games behave the way you claim. It would be helpful so I can be more sure what you mean.
Also, Dark Souls IS a game where you can avoid enemy attacks without invincibility. There was a guy who did a full-game run of dark souls 3 where he refused to block, parry, or dodge. You can google that.
Your whole comment comes across as “man yells at cloud”. It sounds like you’re saying the camera system of the majority of 3d beat em up games is wrong, including: Bayonetta, Devil May Cry, Ninja Gaiden, Metal Gear Rising, Transformers Devastation, Mitsurugi Kamui Hikae, Dark Siders, Kingdom Hearts, God of War 1-3, etc. And instead they should be like some games you failed to mention, presumably because you’re only really familiar with PC Shooters (mixing up the sticks? are you serious?).
Liftrauser AFAIK averages camera position based on enemies. I never really felt like the camera was responsible for my mistakes. This type of camera works really well for more dynamic bullet hell type games.
Heyyy, Did you write this in 2015? Well, I see a lot of comments from 2019 so I guess I am not so late to reply. It is very interesting all the analysis you have made. You see, I am a Unity Game programmer and I am working on a 3d fighter with melee and ranged combat. I am try to find a good reference for a game like this (better if it has PvP) that uses no lock on system at all. The reason I would like to avoid lock on is because it doesn’t match too well with aiming a ranged weapon, and also because my game pretends to be 4 vs 4. But off course I would also like a 1 vs 1 duel to feel nice. I would also like player to be able to run to the sides and backwards while camera is facing towards, to me, it is nice to be able to see your character from various angles and not only from the back. I also like camera to be centered. I could at some point accept a slight offset but I hate cameras that have a big offset even if it is used to shoot. The best references I’ve found are Infamous Second Son , Ratchet and Clank Up your Arsenal, and Dungeon and Dragons Online.
LikeLiked by 1 person
I did indeed write it in 2015.
It sounds like you want to make a mouselook game with melee and ranged combat. You might want the ranged attacks to be mouselook, but the melee to be freecam. (What infamous does)
I would suggest God of War 2018 as another valuable inspiration (mouselook for both melee and ranged)
Mouselook here means that the camera controls the facing direction of the character. If you want to see around the character, you can make it so they run freely normally, but snap to the camera’s angle when attacking or aiming.
Less useful inspirations might be Gunz: The Duel, Vanquish, or Warframe.
LikeLiked by 1 person
Thank you! Yes, I am doing exactly that now. That is the way Ratchet Up your Arsenal and Deadlocked do it. I dislike the offset in Vanquish. Gunz I think the camera is always behind the character. I will check Warframe more and God of War 2018.
Lovely blog you haave here