“Cinematic Perspective” is a Unity asset that has the objective of providing the developer with a simple, intuitive and fast way to implement basic camera rig movement, focal and perspective automation as it has been extensively used by filmmakers in the industry to provide a more interesting picture composition or provide meaning to a shot.
The natural use case scenario for this asset varies. It can be implemented for cutscenes, it can be used in gameplay (as in classic games like LucasArts “Grim Fandango”), also it can be used as an interesting way to show architectural design in unity, in hybrid solutions, and various other ways. Creative imagination is the limit.
Some Video
The first is a short film using dissolves, transitions, and various flimic strategies with our new 0.92b version of the asset. It’s simply a knight crossing and bridge and getting tired (yes the writer is as lazy as our main character!). Hope you enjoy.
Features
Realistic movie perspectives
Fixed, Steadicam (Follow), LookAt and Dolly Modes
Transitions between cameras
Use one or multiple cameras (one per rig),
Configure your cameras as desired,
Compatible with most effects,
Auto Zooming
Bobbing / Shake Effect (In Steady Mode)
Excellent performance.
Clean C# well documented code.
Version in advanced beta and still taking ideas
for improvement in this realease.
The “Cinematic Perspective” asset is a class that allows the operation of a set of camera rigs. Camera rigs are virtual devices used to position, orientate and move a cameras based on four modes.
The Steadicam mode (or Follow shot or tracking shot) is a specific camera angle in which the subject being filmed is seemingly pursued by the camera. This particular mode will follow the target or subject (performer) in motion. The LookAt mode is a mode where the rig remains in a fixed position, but the camera faces the target and the Zoom (FOV) can be automated or manually adjusted. The Fixed mode is a mode where the rig remains in a fixed position and the camera faces a defined direction. In this mode the Zoom (FOV) can be manually adjusted. The Dolly mode basically moves the camera independent from the target or subject (performer). The camera can face the target, or another target, or simply focus in a particular direction. The movement is in a vector from point a to b and the movement speed can be edited in an animation curve.
Further Improvements
We are trying to improve our asset and will take into consideration suggestions sent to us. Please don’t doubt and contribute any ideas for improvement you may have. FYI some of the future improvements we are considering are add path mode for better controlling the camera. Improving the steadicam mode for object avoidance, adding other type of triggers like cubes, add multiple targets and improving the api for further flexibility.
This method of setting up camera work appears very visual, which should make the process a lot easier especially when it comes to transitioning from one camera to another. Although I just came across this product today while looking for some other items. I have found several good or great products by wandering and have bookmarked your product just now.
So here’s my first question: How might this work or has it been tested with other products such as Pegasus, which is a camera mover, but could just as well be moving say the car or main character about the set? I think it should work if used with a path finding tool.
The connected question is then perhaps even more important, what tool to use to actually ‘capture’ the action seen by the camera setups of Cinematicperspective? Would Cinematicperspective work with Helios? I know that is asking a lot, since I don’t completely understand using Pegasus or Helios yet myself - having just acquired those. However, I see a great potential use for being able to actually see where the cameras will trigger based on the way you have designed Cinematicperspective.
One item that might be interesting for future development would be to have the camera be able to be set with a custom offset - so the camera is not necessarily aimed dead on center to what is being followed - a sort of other perspective drifted slightly one way or other, even perhaps with some lag when the target being followed turns followed by a slight delay in the camera adjustment.
Can’t believe this product seems not yet to have been discovered by others, because your product is a clever way of solving a complex bit of camera work. I’m sure I will be purchasing this or I won’t be able to sleep at night.
Thanks for your remarks. Let me try to address them:
This product has not yet been tested with Pegasus by Us. The whole component can be disabled from code so to interact with other tools. The main idea is that the set will be rigged with various triggering spheres so when the target (the car or main character) hits the trigger, the camera will place itself and act accordingly.
If used with a pathfinding tool, as used in the taxi demo :
)
we just made sure the transverse sections of the graph are included in the sphere.
You are right, the captures in the demo scenes were done using Helios. We found no problems integrating this solution.
Maybe a solution can be to place an invisible transform and set ti as a target and make this transform part of another hierarchy? would this be a solution?. Yours is a good idea, we will add a target offset in a future release and an “aim speed” to simulate this lag effect. This has to be neatly done as the speed and direction of the moving object has to be taken into consideration.
You are very kind. We’re in the process of waiting for a little more exposure by the asset store (our product has not even been shown in the “new in the asset store” section!). I know you’ve purchased it. We are in the process of a new update that includes transitions. This is a component we will be including in this set for free but it will also be sold separately for other customers. Let’s hope this makes it more useful to you!
We’ve submitted version 0.92b. Various improvements are on the way, including transitions (masked and dissolves), performance and funcitonality. We have also tested our new settings, making a short film called “Lazy Knight’s Bridge Adventure” that can be seen in the first post on the page. We will be adding to new assets in the store in the next days. One if them is the separate version of the transitions script already included in this asset. The other is the screen frame to frame recorder used in the short film. Both will be included in Cinematic Perspective as added value and functionality for free, but will also have independent entries in the asset store.
Seen your asset in the store and purchased it. The idea is great. Are you going to include multiple target switching? I’m going to use it to handle conversations between characters and wanted to place cameras so that they are focused from behind the player they are talinkg (over the shoulder) so to give a more movie-like perspective.
Thanks for your purchase Yamizeru !, this is already included in the api. I guess the best would be to once the targe triggers the camera, switch the target as desired with the setTarget method.
I have a cutscene set up with our own custom code, but I really want to add a steadicam effect to the camera. Can I do that with this tool – add a script to the camera that will give it a bit of naturalistic movement, without using all the other features or interfering with our existing camera code?
I sent the following via email to support, but just in case someone else know or …
I found out last night that Camera Follow scripts do not work with VR (Oculus) as the camera needs to be a child of the object it is to follow. I am building a race game and because this is for VR the camera cannot react to each and every bump. I at least need to minimize these issues, which is why I am looking for a steady-cam solution that will work in VR. I checked the documentation and the unity forum and did not find a reference to VR so I thought I’d ask.
So… can I use your solution to follow an object (car) and steady the camera in VR?
Dear James, we have not tested this solution in a VR environment, but let me provide some ideas: Via the api you can translate the position of the object to follow to the cam and update the scene maybe? or create a dummy object that can be the focus of this. We were going to add image stabilization but Unity has not provided us with any exposure whatsoever (probably due to their cinemachine release) and this impacts our future releases. Sorry for this. We will try and work something out. Please mail me and we will see if we can help you.
The cameras are to be handled from the rig controller script. But if you can place a dummy object in that moves as desired, simply feed the object to our script in steady mode. No scripting needed. Remember this must be a separate cam as to ignore other cams you use in your proyect.
