Preparing the scene for the final render and managing my time

Since last week’s post I have been making progress on the animation and prepared the scene for rendering. I will however focus on the scene preparation in this text and on the animation part in a future blog post.

My decision to take a short break from refining the animation in order to prepare and organise the scene for the final render was based on my personal preference and working habits. Therefore, the reason was that it allowed me to better estimate the time I have left to complete my project (or any other task) before the due date. Now, there are exactly two weeks left during which I can further refine my animation because I have already prepared the scene for the final rendering. This not only gave me a great sense of progression, but it also allowed me to better plan my next moves when it comes to the animation. I was also able to estimate the time it will take to render the whole scene that now features ray-traced shadows, a three-point lighting system and panels. As a matter of fact, I have already rendered the whole animation before and it took somewhere between 7 to 10 minutes tops, whereas now it takes almost one minute for a single frame out of 650 to be rendered. Consequently, I have safely assumed I will have to spare a whole day for the final render to complete when the time comes.

Before getting started on the scene preparation part, I’d like to address a camera issue I have found in the meantime. In the last blog post ‘Setting up the camera’ I have linked the camera path (which is a circle shape) to the main controller that’s part of the rig. The issue with that being the controller moves up and down at the same time with the character – in fact, the controller was used to move the model’s body up and down when animating. As the controller moves up and down, the path and consequently the camera follows thus negating the effect of the characters weight displayed in the animation. Because the camera was synchronised with the character’s up and down movement on the Z axis, the model looked as if it was floating when the animation played. To solve the issue, I have created a ‘dummy’ helper object and placed it inside the model. I have constrained its position to the to that of the character’s but only set it to inherit the movement on the X and Y axis but not the Z. The camera path had been linked to the helper afterwards. Now, the dummy travels at the same time with the character without moving up and down and so does the camera.

Moving on to the scene preparation, I must mention that I had absolutely no idea what I should even start with. What I did first was to take another look at different 3D animation demo reels I have found on the infamous media platforms. I have noticed that a lot of them if not most present the animation on a rather simple, plain background that’s usually grey.

So, I have tried at first recreating the background by joining three plane shapes together in order to build something like a corridor or a pass.

However, I quickly came to realise it won’t look that good as I have also noticed that the background is usually gradient and has no hard edges in those demo reels. I have then applied the ‘Turbosmooth’ modifier on the panels to create the seemingly smooth, gradient effect.

But that didn’t work out that well either. Seeing nothing worked, I decided to do some more research when I stumbled upon a video where a teapot model was beautifully rendered on a gradient background like the one I decided to go for.

I have then created a new plane shape and used it as a floor. I have extruded the edges (raised) to turn the plane into a box. The hard edges were chamfered to create that smooth gradient effect.

After I was done with the panel, I have started setting up the three-point lighting system. Following the video, I have created three ‘Target Light’ light sources and placed them accordingly (in a triangle shape, with each of the lights representing a corner). Each of the lights were set to cast ray traced shadows. The one on the right was set to emit a cold temperature light (blue hue) and the one on the left to emit a warm light (red hue). The light spot in the centre (at the top, also called a rim light) is set to emit a normal, white colour temperature light.

After setting up the lighting system, I have set up the render settings. The renderer I have used and will use for the final render is the ‘NVIDIA Mental Ray’, as it seems to net out the best results. It is also especially useful since the shadows are ray-traced. Here is one of the initial renders:

Although I have not decided on it yet, I have added a ‘Sky Light’ source to the scene as it seemed to yield even better results:

Similarly to the camera setup, the light sources had been also linked to the dummy helper in order to move at the same time with the character. In the next blog post I will get back on to the animation process.

References:

Preston, C. (2019). Lighting Render Setup in 3DS Max Tutorial. [online] YouTube. Available at: https://www.youtube.com/watch?v=0FDeexXhrN0&list=PLwnsfil42NA4ezewQGANv1lr_hk-SdMKL&index=18&t=938s [Accessed 15 Apr. 2019].

Tiner, J. (2019). Studio Lighting in Vray - 3DS Max 2014. [online] YouTube. Available at: https://www.youtube.com/watch?v=Vp1qRZ-8wno&t=193s [Accessed 15 Apr. 2019].