Media Project One – Page 5 – Andy Cartwright

November 21, 2017Andrew Cartwright No Comments

Displacement Map Solution

After coming across the displacement map problem, I have been searching for different solutions and found myself looking on lots of forums which specialised in Cinema 4D – to which I found someone else who was experiencing a similar problem to myself on the forum website called Otoy.

Unfortunately, this problem wasn’t resolved and the user actually found a way around the issues with a noise effector inside of C4D as he was trying to create a rough effect to simulate building bricks.

After not being able to find anything useful on the topic that was relevant to me, I went back to the initial tutorial to see if I had missed anything. On the 19 mins 30 secs mark, the tutorial starts to mention about an issue that they were experiencing which was the exact same problem I had – how the displacement map seems disconnected and is described as “coming apart from the seams”. This was then corrected by altering the Mid-Level of the displacement map. However, that is for the Octane Renderer and that feature isn’t available in the stock version of Cinema 4D.

Even though the Mid-Level setting available in the Octane Renderer wouldn’t work for me, I still noticed that the node was connected the overall displacement material so I went an checked my displacement settings but nothing worked. So my next idea was to try an add a deformer, as I knew the object was a cube – I tried adding a bevel deformer to it to see if this would merge the vertices of the displacement map together, which I did but with an interesting effect.

The effect can be seen below in the renders provided, the ‘before’ render shows that the object looked like before the bevel deformer was added to the cube – not how all the displaced shapes rise up quite high and there is a good amount of colour variation which was achieved through the initial creation process.

Before:

After the bevel deformer was added, it created an odd effect to the surface of the object by flattening the protrusions from the displacement map – also this result may have decreased the overall effect I was wanting to achieve, it also also solved the edge vertices problem that can be seen in the first render.

After:

Overall, the bevel deformer has eliminated the holes in the displacement map on the vertices but, it has also changed the overall effect of the displacement map as the protrusions are lower in height and seems like the deformer has decreased the overall strength of the map. From a close perspective, I really like the effect the deformer has given the object as it resembles the surface of a space station or technologic material – however, I still need to the view the effect as a whole on the object as it might not look aesthetically pleasing form a distance. To further develop this I will continue to alter the settings of the bevel deformers to see what results they produce, which I will document with renders.

November 20, 2017Andrew Cartwright No Comments

Displacement Map Problem

After successful adding the displacement map to the cube and adjusting the segments & sub-polygons, I noticed small section on the vertices of the cube where the map didn’t join up correctly – this can be more specifically be seen in the render below where square/rectangular holes are visible.

As I intend to get some close-ups of the cube for the animation, this will need to be fixed. I think that this could either be done within C4D itself through the mapping settings on the material – the way the texture is mapped onto the shape is currently set to UVW Mapping which might have an involvement with the issue. In addition, it could also be done in Illustrator or After Effects by adding a small white boarder to the edges above all the layers – this would mean that all of the boarders would be at the same displacement level as each colour density represents displacement value. I will research into this further and test any potential solutions.

November 18, 2017Andrew Cartwright No Comments

Final Abstract Cube Creation

As the project deadline is coming close, I have started to make the final adjustments to my project. The videos below show my workflow of how I created the abstract cube as well as discussing key elements within the process that I felt needed highlighting – this process required me to gather all my learned knowledge & research about C4D, lighting, mograph and render settings in order to complete the final asset that will be used within my project.

Part 1

As this was a long process, it was split into 2 parts – the video below is part 1 and focuses on the initial section in Illustrator and After Effects. I discuss key elements about the process as well as highlighting some issues.

Part 2

Part 2 focuses on the Cinema 4D section of the project, as I explain some problems and difficulties I encounter. Unfortunately, I forgot to record the first half of the process which I had covered in my testing experiments, however, the point in which the video starts at is the next stage to complete after the testing video, so the whole process of how I created it is documented.

Renders

Here are the different renders I produced when checking effects of the setting alterations – as you can see they start from when the additional faces were added to the main cube and continue onwards to show small differences in the overall design. Within some renders multiple aspects are changed, from colour settings to light colours & positioning to camera positioning.

Overall, the renders produced show a good documentation of testing conducted on the cube. However, there are still key issues that need to be resolved such as the vibrancy – this needs made more vibrant as the cube looks very flat and washed out which I feel wont be eye-catching for the audience. Additionally, I should look into creating a light map or set of small illuminating objects to stick on the cube, this is to make it look like a machine or building which will further help reinforce the man-made notion I suggested about in my planning.

November 16, 2017Andrew Cartwright No Comments

Render Settings

As my animation is coming towards the final stages of completion, I feel the need to research and test out different render settings in order to speed up my workflow. I often find that I spend a long time waiting for my renders to complete after making small changes which has a detrimental effect on my workflow. The video below gave me some useful times on how to speed rendering up:

Optimising Scenes in Cinema 4D from Luxx on Vimeo.

Overall, the tutorial Optimising Scenes in Cinema 4D focused on two crucial elements: 1) Having low render settings for the development stage of the animation and 2) Having high render settings for the final render when the animation is complete.

Low Render Settings:

Within the tutorial it went through each individual setting, basically turning ever setting down a couple units so that the overall speed of the render is decreased. All of the settings were specific to the physical render engine which I intended to use anyway – the most significant alteration of the renderer was the Sample Quality, as the setting was changed to medium, this drastically changed the values within the setting which was, at first concerning as I thought it would provide enough detail. However, I later found out that this would be more than sufficient enough for development stages of project as it would allow for the render times to drop drastically which would further allow me to make alterations much faster.

Example render using the low settings:

High Render Setting:

After I had used the low render setting to make the desired alterations to the animation, I had to figure out a higher setting for the final render which involved turning everything up. As I was planning on using the physical renderer, I decided to turn the default Sampling Quality setting from medium in the low setting to high – this further bumped up the units to produce more detail and most notably decreased the error threshold to 1%. After conducting a test render, I felt that I could have changed the Sampling Subdivisions up a couple of units more to produce more detail which would be desirable at the cost of a bit extra time, which I feel it is worth.

Example render using the high settings:

Conclusion

Using the low settings will produced really fast renders which will in turn speed up my workflow as I will be able to make small changes quickly to my animation, however, this speed will come at a sacrifice as the quality wont be as good. Oppositely, using the high settings will produce really high quality renders which will be very beneficial when coming into after effects as it’ll look very aesthetically pleasing, however, this quality will also come at a sacrifice as the time it will take to render each time will be very long. The image below shows two test renders I conducted with the different render settings. Overall, I feel that both images look identical on screen which is very surprising, however, the low render setting took far less time in comparison to the high render setting – approximately 5 mins 10 sec less. The low setting would be far more beneficial for me to use in the development stage of the animation as it will allow for render times to decrease, which in turn will increased my workflow allowing me to change and tweak settings fast. After the development stage has been completed, I can then switch to the high render setting which will give me better detail ready to export to After Effects.

November 15, 2017Andrew Cartwright No Comments

Height Map Research & Testing

A continuation of the displacement creation inside of C4D, this time I tested a couple different ways in which I could make the map more defined on the cube as which I scrolled in the edges were slightly bevelled. The video below shows my workflow:

Within this time-lapse video I was trying to tackle my problem with the displacement map – on certain areas of the map I experienced organic looking protrusions coming from the object and I didn’t know why it was happening. In addition, I also saw the some of the edges seemed very bevelled even when the segments were turned all the way up – soon after the software locked up and froze due to all of the RAM being used up on the computer. As this was either to do with the displacement map itself or the amount of segments on the object, I set about experimenting with all of the different settings.

Firstly, I tried to turn up/down the amount of segments within the cube itself to see if this had any effect. I found that the less number of segments the object had, the less flexibility the face of the cube had to deform to the displacement map, this often resulted in a less detailed map.

Secondly, I tried to alter the height and strength settings of the map but this didn’t really benefit the clarity of the protrusions in any way – When the height settings were increased, the physical length in which the protrusions came out to were increased as well and the this was the same for the strength setting too. Changing these settings at times did help occasionally, however, this wasn’t very beneficial as the whole aesthetic of the cube had changed to something I didn’t want.

Lastly, I tried to add Sub-Polygon Displacement to the cube which was under the displacement tab within the material editor. With this particular project I found that it drastically slowed down my computer and wouldn’t even allow me to render as it produced an error message: ‘Run out of application memory’ and prevented the scene to be rendered any further. I created another project file and loaded up the same displacement map (enabling the Sub-Polygon Displacement) and found that even with 1 segment for the X, Y & Z settings, the displacement map worked.

*Sub-Polygon Displacement – “In principle, sub-polygon displacement (referred to as SPD in the following text) is similar to displacement: an object is deformed during rendering based on a texture’s grayscale palette or, in some modes, color palette. The trick lies in the object’s own internal, adjustable and relatively high subdivision, which allows for very detailed structures without having to actually subdivide the object permanently. In many cases, it is impossible to achieve the same degree of detail through modeling because too much memory would be required.”

Sub-Polygon Displacement disabled

Sub-Polygon Displacement enabled

Map Resulting Geometry

All of the renders produced had 1 segment for the X, Y & Z settings – found that this was much faster to render with my computer and I also feel like the edges of each protrusion are less bevelled than before.

Conclusion

After experimenting with the sub-polygon displacement, I have come to the conclusion that the displacement map is far too detailed for what I need, meaning that there are too many fine details in the map that the displacement cannot displace – this is evident in the physical renders that are produced as you can see some of the shapes created in the map itself aren’t very geometric in shape and are more organic looking. To overcome this issue, I will create another displacement map that is more simplistic, this will allow for the displacement to produced protrusions which are more defined which could also be developed by experiment with colour in it. Additionally, Im looking at adding extra elements to the cube that might cover some of the sides – examples of this could be rectangular faces in the centre of the cube to add some diversity to the shape.