Right now this is not possible.
Anybody see good reasons to do this type operation and wich type, why?
1 Like
What modifiers would you want to add to the camera? I can’t see a bevel mod doing anything.
Right, bevel nothing. But maybe other?
Riging/physic? Why always anchoring to object that have this modificator?
Eg. simulations for Leans (as new feature)?
It’s speculative question and post due feature request:
Camera Lock to collection - not only object
I think the logici behind it is: the geometrical objects do have (geometrical) modifier to …modify… their geometry. And cameras, light don’t need them (then there has to be distinguished between geo-, light-, view-modifier), but you can parent the camera, light to a dummy and do weird things with drivers…
So it is not necessary to introduce a new type of modifier ?
In 3ds Max you can put modifiers on the camera. Using a Lattice ( FFD in Max jargon) was an easy way to simulate tilt and shift.
You can also put bend modifiers on cameras to make the camera rotate. Use cases are rare but I guess it doesn’*t hurt to have it.
1 Like
Use cases are rare but I guess it doesn’*t hurt to have it.
Well, as i said you can parent to anything… And even MAX does things different than others, so any Lightwave or C4D user can say: but wihch this one i can…
So for me the real question is: does it have to be done like in some other program because you can not do it in the this program? (Or didn’t found a maybe even simpler solution with that other tool?)
… but wait… the original question of this thread ask for it (camera modi) but on the other hand @ManBlender ask:
Anybody see good reasons to do this type operation and wich type, why?
My answer: No, parent camera (or light) to any geometrical object and inherit any modification/animation you want.
I’m eager to learn, so tell me what type would be neede to solve waht problem? (I believe the developers also would ask: why should we (do/ work on) this if you can make it with…)
Well, having something like a lattice on a camera makes it possible to animate the shape of the rendered image.
If it is possible to use a bend modifier on a camera you can use a driver to easily link the distortion to some other bend modifier.
As i said this kind of stuff is rarely required but can be super useful in these rare cases.
Obviously it wouldnt have to be done exactly like in program x.
Okay, thinking about this … latticing or bending the camera… this would be a modification of the ordinary camera projection (persp,ortho)… (a dream or LSD sequence??) but wouldn’t it be easier to render the camera the normal way and do a image manipulation in postproduction. You could even try different kinds of effects just by 2D effect… much cheaper on performance and flexibiliy to produce a special look.
What about this type problem?
Is this not good option to use “Array” or something to divide main render image or images from other cameras (something like “middle composition”)?
Then let we imagine that once of camera doing “lag” or other type “delay” on camera as some effect.
Why trying to use (and even implement) a modifier on a camera if someone want different cameras with different perspectives and zooming (animation) ???
Modifier just as tool with new options and features as I described - convert final composition from main camera? Then many combination can be possible:
- mirror for once of all square cameras
- track line for this squares (eg. animate 4 in 1 or 8 in 1)
- manage this squares as object to multiply (array)
- give effects (eg. change square for roundable - some masks [eg. broken])
- lag/delay for separated cameras
- different colors (eg. black and white, gray spectrum, or something like infrared, etc. - for join 2 separated views)
Well, real cameras can be heavily modified, so why not modify 3d cameras? Just look how heavily you can bend, tilt, shift, twist real cameras. In CGI it could be even less limited of course.
to @ManBlender:
- mirror for (which?) square: well simple image mirror
- track line…??? 4/8 in1 : i don’t understand that… multiple cameras? (then my other answer would apply; settup multiple camera)
- manage…array: … ??
- lag/delay for sep…: do this with second cam and follow… mod
- different colors… ?? for a camera?? so differen t image effects.
to @Lumpengnom:
That’s an other question. Doing double camera composition or… what effect would be achieved with that bended setup? What would tilt be? Shift by x or y are camera settings.
I’m not fighting here, but back to the original question:
Add modifier to camera.
What do you want the camera to do which is not possible in blender? Because only that would be a reason to implement it and even any developer for another payed program wouldn’t do an extra, if it is already doable.
- Mirror for cameras view, but not one, but when we have 4cameras in 1camera (like picture in picture)
- Track these 4 camera on some main layer like on 2d viewport (operating this 4squares of cameras picture as output
- yes,
array mortificationto easy multiplication of square outputs on this 2d ascountx, y fit type - lag/delay on this 2d vieport as dashboard (let we imagine more than 1,2 sec delay as 2minutes from animation - don’t think it can be done by clone that following something)
- different colors… yes, but on these 2in1 or 4in1 (soo, I don’t think can be done on separated images)
What do you want the camera to do which is not possible in blender?
The animated pictures in picture as 2in1, 4in1 (separated - like quad view) as some dashboard in 2d layer and 3d layer (eg. like linked post, the 4 CCV cameras in 1 monitor).
I was asking especially why this arrayed n in 1, because in games this is used for simulating the monitors for multiple camera. But for any (short) movie one would render the different view and composite them. The advantage would be that you can change or even add camera. And Blender has a feature (which i didn’t have used yet) Stereoscopy with MultiView. So i don’t get (and i bleiee the developers also not) why this has to be a modifier?? You can do it, so why to pick on camera modifiers ??? (And just repeating demand doesn’t explain anything… ) 
I’m not so advanced around blender and especially games, and never do in UPBGE this case with multiple camera but talking about in parallel way at the same time. It’s something like mirrors in car for me, but mirrors are mirrors.
I not meet the right solution on internet yet around picture in picture.
We talking about some solution, but not confirmed any of this in the needs about what we talking and partially indicating around this topic.
https://docs.blender.org/manual/en/2.80/render/output/multiview/usage.html#viewport-preview
My idea was about some modifier that can divide final composition in way that main picture is build from other as cameras sources. Then these sources can be eg. animated on 2layer. More… as example explode for 1 from 4 cameras (as square - just effect fro this 1 of 4 picture). Trying combine some of existing modifiers to usage for camera.
Repeating case was as example, to mining this idea and other.
Not sure it can be more easy to produce kaleidoscope by modifiers or in other way.
On examples PIP I always meeting the external render as movie, ok - but small modifications around scenes require render separated sources if we going about contex with PIP from 2/3d viewport camersa
It is for tilting the lens against the image plane. Now, Blender can shift. But can it tilt? I actually don’t know. I guess it can be faked with the DOF settings but we have a lot of other “unneeded” real camera settings such as focal length as well.
What tilt does in real cameras is explained here:
Anyway, if you could put a lattice modifer on the camera you could easily use it to tilt and shift and even a lot more which is not possible with real cameras.
You could - for example - shift only one “corner” of the camera.
Or you could twist it.
Or use a noise modifier on the camera to create a weird wobbly distortion of the image.
Give it a cloth sim?
No idea how useful that would be but I can imagine it would be popular among the motion design folks.
I guess most of this stuff can be faked in post but having direct access to things in 3D enables you to directly combine it with other effects, for example, via drivers.
Also you don’t have to render more than you actually need. If you fake a tilt in post, or a noise or a twist you will often have to cut off a part of the image.
Is this worth developer time at the moment? Not in my opinion as I’d rather have performance increase, improved Cycles and a whole bunch of other things but I can see the usefulness of camera modifiers.