You know what you should do in future renditions? Take a picture of your room, an environment, etc, and try to integrate him in the scene to interact with it some how.
You know what you should do in future renditions? Take a picture of your room, an environment, etc,
I can tell...that facial expression tho! XD Is there anyway to take an image and create a sort of map of it so it could automatically interact appropriately with your models?
I can tell...that facial expression tho! XD Is there anyway to take an image and create a sort of ma
mapping a single image is a bit tricky, you need to recreate the whole interactive environement in 3d, like the screen, keyboard and all, so the light can interact with it, if you look closely to the right monitor, you can see the reflection of the orange squirrel, its subtle, but its make the thing much more realistic, its was my first time though, and had some incorrect placement on the left screen, you can tell by the hard cut between the tail on the right screen.
the process is more easy for a video, because some software can recreate the depth information by the movement of the camera. yes, kinda ironic that its would be easier to do an animated video like that than a single image.
mapping a single image is a bit tricky, you need to recreate the whole interactive environement in 3
The way I envision it without a mapped environment is something like photoshop. Such as pose the model, get a screen shot of it rendered, then apply it in another image.
The way I envision it without a mapped environment is something like photoshop. Such as pose the mod
I really don't follow your though, you don't pose model in photoshop, its not a 3d app, you don't take screenshot of a render but the render file itself, and applying it in another image is part of the post-production.
I really don't follow your though, you don't pose model in photoshop, its not a 3d app, you don't ta
Okay I just figured I was out of the context and din't understand the flow of the conversation anymore.
So, when i want to put character into a real environnement there is some step to do, lets assume its for a movie :
1 (Tracking) the first step is to capture the camera movement, using specialised app like boujou or pftrack, this way, you can import the camera into the 3d app, and its will simulate the mouvement. as a by product, the app will produce 3D point from the 2d image, like for exemple where is located a chair or a car, you will use these point to place your character in the correct position.
2 (Modeling) second step is to recreate part of the environnement in 3d, the purpose of doing that is that when we render the final image, we capture some information like reflection, shadow, refraction and other stuff, its also used to match the light better for the character, for exemple : if you shoot a white room with 3 white wall and 1 black wall, you will want that the light that will bounce to the black wall well behave correctly and won't light the character
3 (Rendering )once its done, the third step is to setup the render, here you want to separate lot of thing, each light will have his own render file with specularity, shadow and stuff, also, all 3d piece you will render will output image with uniform color, like plan red, yellow, green, each, for exemple you have 2 character with fur, you will assign blue and green color to the 1 and second character, and red and yellow for the fur of the character, this way, you can manage to do post-process effect, like changing only the fur of the 1 character without altering the skin. the render will also output depth data, which will be used to recreate the rendered image in 3d in NUKE.
from this point, a single frame will have around 10 to 100+ different images (yes, for a single frame), these image are like puzzle, alone, they mean and look like nothing.
4 (COMPOSITION). Time to bring this all together, for exemple on nuke, you will import all these different image and merge them, like layer, all these image will give you a complete control, you can change the light, color, or even shadow without having to re-rendering everything. with the depth data you will be able to setup lot of post-process effect, like smoke, debris, image and lot of different stuff.
5 (Render (again)) the final step is to render the final movie.
this is all from what I have learn over the year, I may have missed lot of stuff, and some information may be not correct, but that how you achieve a
Okay I just figured I was out of the context and din't understand the flow of the conversation anymo
Sorry, but I'm not reading all that. :/ Anyway, Yeah, you miss understood. Granted I guess I could have said to use the rendered file instead, which is what I meant.
I'm also well aware that photoshop isn't a 3d program.
Sorry, but I'm not reading all that. :/ Anyway, Yeah, you miss understood. Granted I guess I could h
An animation would be interesting. If you ever do an animation, don't settle for less. Even if it takes a lot longer, you should strive for something amazing. And I don't mean long. I just mean a very well done animation.
An animation would be interesting. If you ever do an animation, don't settle for less. Even if it ta
I have thinked for doing animation for very long, the problem is the rendering, take into account some variable : you need at least 24 frame for 1 seconde of animation, assuming you try to optimise your scene to the minimum, and putting less resolution, I could achieve a good render of a single image around 15 to 30min, now multiply that time with the number of frame, for an animation of 1 sec, its would require 12hour of rendering, and 5 sec ... 60 hour, even if we halve the number of frame per second, its still a lot of time where : -your cpu take a lot of stress, i7 aren't made for working for such long amount of time, you need specialized cpu like xeon -you cannot use your computer to do something else -if something is wrong with the render, you can loose all the batch.
I'm working on an application that will dispatch that kind of work between different computer automaticaly, once its done, I will possibly use all the computer at my work and reduce the render time by 10.
I have thinked for doing animation for very long, the problem is the rendering, take into account so
pretty simple, with multi million worth of render server, Disney renderfarm are composed of more than 50.000 cpu that can render hundred of frame at the same time
pretty simple, with multi million worth of render server, Disney renderfarm are composed of more tha
that higly depend of what you want to achieve you know, today computer could easily render a full toy story movie, and what make my render long is the fur, if I drop the fur, could do lot of animation, but i like fur!
that higly depend of what you want to achieve you know, today computer could easily render a full to
I thought he was grabbing fire to start with, but then I realised it was a crystal or something it looks very warm though, so I hope he doesn't burn his paws.
I thought he was grabbing fire to start with, but then I realised it was a crystal or something it l
