|
Rendering
After my lights were set up, I moved onto rendering. As you might have guessed by now, I have chosen VRay to do the job. The reason is that this rendering engine is fairly simple, yet with its help you can produce very realistic results.
Let’s start from the beginning: I always turn on Frame Buffer so I have a lot more options on my rendering window, such as the mouse tracking function. Frame Buffer doesn`t actually affect the render, although it has some nice post production tools. To be honest, I always use Photoshop for post production, but it’s always nice to have some extra tools.
The first important settings are in the Global Switches tab, namely the Default Lights box which is checked out by default. I always uncheck this, because we don`t need the default lights which the software provides. I only want my lights to influence the scene, nothing else, so unchecking this box is very important. I leave the rest of the options on default. At the Image Sampler tab I set Adaptive DMC and Catmull-Rom. Catmull-Rom is my choice since it provides very sharp edges, which is essential in a scene like this, where a lot of cracks, bumps and little details take place. It’s not the fastest Antialiasing filter, but most of the times its worth the wait. The Adaptive DMC image sampler values are raised to 4 (min subdivs) and 6 (max subdivs) to give me a quality image sampling.
The Environment tab is also critical for a good render, since we need skylight. I check all the three boxes since I want my reflections and refractions to be affected by the skylight. I`ve decided to add HDRI maps to all three slots, which turned out to have a positive effect on my final render. Color mapping type was set to HSV Exponential with the Dark Multiplier set to 1.4, Bright M. to 1.0 and Gamma to 1.8. This setup proved out to give me the desired looks of the final picture. Raising the gamma was important so that I have a slightly bright picture to work with in Photoshop. I always render my images too bright so I can play with it in Photoshop any way I want. A bright picture can easily be darkened down to our expectations, whereas a too dark picture looses an immense amount of quality if we are trying to raise its brightness during post production.
Indirect Illumination was achieved with the Irradiance Map and Light Cache engines. Most of the time I work with these two, since they provide a very good rendering result. At final renders the Irradiance Map is set to a Medium preset to get a quality render. HSph. Subdivs are set to 60, Interp. Samples to 40. Light Cache has some pretty high values also, namely Subdivs 2200, Sample size 0,02 and the Pre-filter is also turned on with maximum values of 1000. Once I got my GI settings set up for a high quality render. I needed to adjust the DMC Sampler: Adaptive amount 0,85 ; Noise threshold 0,008 ; Min. samples 10 ; Global subdivs multiplier 8,0. The Global subdivs multiplier is important since if we don`t raise its values there is a pretty good chance that we will get a noisy render. Still images require at least 2.0 or 3.0 avoid noise on our final render.
My last adjustment before rendering was the VRay lights’ properties. I`ve selected all my lights, pressed the right mouse button and chose the V-Ray properties option. In here, the Diffuse subdivs were set to 2000, so I give my lights more quality.
The only thing missing after my final render was done, were some maps which I could use during post production to enhance my pictures’ quality. I went ahead and used VRay Dirt to make an Ambient Occlusion render. Another render which was necessary was the volume light. I`ve added a volume light to the scene, but before that every other light was deleted and all the objects received a full black colored material. Reflection and specular maps were also rendered.
Post Production
Post Production was done with the help of Photoshop CS3. This process includes the adjustment of the pictures’ levels, brightness, contrast, saturation, color balance and adding some maps with different blend modes to highlight some essential details which are not very visible without these maps.
Ambient Occlusion is important since it highlights and sharpens the edges. For example, without AO, the details on the wardrobe’s decorations weren`t very visible. This was very easily corrected with the help of this map. Specular map is required to control the amount of specularity needed on objects, whereas the reflection map controls the actual reflections on the models. I had a bit of trouble with the reflection map, since it produced a huge amount of reflections on the wooden boards above the windows. I had to manually correct this little glitch. I used various blending modes to combine the render and the maps into one final picture. Another thing I did was darkening the pictures borders a little bit, so that the viewers focus is directed to the middle of the picture. It also gives a nice effect to it.
After post production was done, the final picture looked like this:
Feel free to visit my website where you can view my other work as well or e-mail me if you have any questions.
Thank you for reading.
About the Artist
My name is Horvath Elod, 21 years old, currently I am applying to a 3D Animation course at Birmingham City University. I have been familiar with 3D graphics since 2005 and have been learning it since with passion. My studies in this field took me to Hungary where I have finished the IMMI 3DSMax master-course, where I achieved a “Pass with distinction” standard.
|
