Previous | Next --- Slide 5 of 52
Back to Lecture Thumbnails

Can we say that in the future ray tracing will replace rasterization to produce high-quality images?


Would it be possible to emulate the effects of raytracing through rasterization and by using additional depth buffers for different camera orientations, and using that depth information to somehow draw reflections and shadows?


As the improvement of hardware is really quick, ray tracing could be a better choice to provide users better experience.


Are interreflections not possible to do with ray tracing? Or would it simply require multiple passes?


In the rasterization case, can't we augment the primitive data structure to have shadow information from other primitives?


Oops, ignore the above, just read the asterisk and the notion of "shadow maps".


Is shadow map a kind of texture map?


For the image on the right, one side of the cube looks very bright and reflects everything like a mirror, while the other sides looks much darker. It makes me feel that different sides of the cube were made of different material. Is this some artifacts of the ray tracing?


what is a possible way to do shadow reflection with rasterization


Are there hybrid approaches where rasterizing first can reduce the RT passes needed?


Ray tracing makes sense that it can better represent things like shadows. But the image on the right also seems to have a sense of reflections. Is rendering these reflections independent from just shooting a straight ray or are these rays also able to bounce off of surfaces to help generate the reflections?


The left misses shadow, refraction on edges, and shows less about surface texture


It appears that raytracing would need to actually trace all the pixels to make sure the shadows, reflections, and other things seem correct. How do some software combine rasterization with ray tracing?


what if we add ambient occlusion to the rasterized image? does it make it closer to the ray-tracing?


In high resolution context, will ray tracing outperform rasterization in hardware?


If you were to simulate ray tracing using rasterization, would that end up taking more processing power than ray tracing at this point? Or rather, is the tradeoff between computational cost and end result enough to justify ray tracing (at least in real time mediums such as games)?


Are there any cases in modern graphics that utilize rasterization method over the ray tracing method?


I'm curious about the comparison of speeds of these 2 approaches.


What is a shadow map and how would we create it? Would it ever be more efficient than simply using ray tracing?