What constraints are involved in working within a 3D environment - polygon count, file size and rendering time?
--------------------------------------------------------------------------------------------------------------------------
Polygon Count and File Size
There is quite a lot of restrictions whilst creating 3D graphics, especially when you have to do real-time rendering every time the game camera moves. Even with modern day computers which have evolved and been developed to stop constraints from happening, it is still very common for artist to have limitations on 3D modelling. For example this could include problems for the computer processor or even something simple like Memory space issues due to the amount of memory creating 3D graphics can take up. There is a lot of issues which can happen if the constraints are ignored. Here is a few;
Polygons Vs. Triangles
When using 3D modelling in programs such as Maya, blender, 3DMax etc. The models is usually built up of polygons, for the polygons to create the model they are joined together and this creates a mesh of polygons. However counting the polygons is quite important. The polygon count is basically the total number of polygons in the model and scene. During the modelling the computer processor calculates the position of each vertex and polygon which has been created within the application.
This means that the more polygons there are, the more hard work it is on the processor. This count is a lot more important for assets in games or movies simply because creating a asset will take a lot of polygons to complete.
Depending on how much process power the artist computer has, How much polygons are used and if the polygon count is left unchecked, this could cause problems with the computer processor and cause disruption whilst creating the model. If this is the case then converting the polygons into a triangle could stop the polygon count issue, this is because triangles are the simplest polygons having only three edges and three vertices, making a three sided face. However modelling with triangles is difficult because triangles tend to have a problem with subdividing which can cause a deformation whilst modelling with polygons is a lot more easier. Considering this, It is probably most beneficial to convert the polygons into triangles for the final stage. This will take away alot of pressure from the processor as well as having a more smoother surface for the model. The main different between the polygon count and the triangle count is that game engines only work with triangles, that is why the polygons are all converted into triangles automatically before being exported into a game engine. And secondly a polygon count can be a bit misleading for instance a poly means 'many' so polygon is a many sided shape so a cube could only be 6 polygons but then be 12 triangles. Triangle count is some what more accurate for game designers than polygon count is.
Triangle Count vs. Vertex Count
Vertex count count is the amount of vertex used in the programme. A vertex is each angular point of a polygon. Whilst a triangle count is the count of all the triangles, Just like the polygon count explained above. Vertex count is more important for the performance and memory than the triangle count although some artists still use the triangle count as a performance measurement. This is simply because triangle count and vertex count can be similar if the triangles are joined up to one another for example if two triangles are joined together their will be 4 vertices whilst if 3 triangles are joined together then it will be 5 vertices. However changes to shading, smoothing and material changes from triangle to triangle can cause a break within the models surface which means the vertices will be duplicated at the breaks when rendered so the model can be sent in render able chunks to the graphics card. Using to much smoothing groups and material assignments can also lead to a higher vertex count, this can put to much pressure on the computer processor causing a slower performance and use more memory due to the extra vertices.
Rendering Time
Depending on the size of the 3D model or animation, rendering can vary for different models for example it could take a few seconds or several hours. The more complex and realistic the animations are for example animations which pixar or dream works make. The rendering could take an insane amount of time, Even sometimes up to days however it all depends on how powerful the computer processor speed is for example if the processor is powerful then the faster the rendering will be.
Real time
Real time rendering is used in interactive media for instance in a game animations have to be in real time rendering because as the player interacts with the game the game camera will move, Every time the game camera moves real time rendering will make sure that the game is being rendered. This means creating images fast enough on the computer so the viewer can interact with a virtual environment. Usually games run at around 30 frames per second, the rate at which image are displayed is then measured by the frames per second. However with most modern computers games are able to play at a much higher framerate meaning that games are usually around 30Fps - 60Fps, To be able to make the expectations without any lags the graphics have to be some-what suitable for that game for instance if a game which runs 60FPS with assets including a lot of elements then the computer may not be able to play the game smoothly. Real time rendering is usually used in games, interactive worlds and VRML.
Non Real-time
Non Real- time rendering is sort of the opposite of real time rendering. Real time rendering is rendering every single time the camera is moving although Non-Real time rendering is basically when a animation that doesn't need to be rendered every single time due to the user not effecting the animation. Non real time rendering, Only takes one render instead of a render every FPS like the real-time rendering. This is beneficial because the artist can add as much detail as possible and create a more realistic and detailed animation without having as much pressure on the processor. O artist creating the artist is done creating the animation spending as much as needed setting up the camera, the artist will then finally rendering the animation for it first and final time then it will be imported into a video game. This means that whilst in the game the user can not change the camera angle of the animation due to the artist already rendering the camera for that animation. This basically means that the player will see exactly what has been rendered in the animation.
Links
