Saturday, 23 July 2011

Brief Update, Ivy Generator Experiments

Here's just a brief update on my scene I've been playing about with, making it appear more fuller and atmospheric

I added some bead ornamentation, and also placed the needles in a more ideal arrangement, and tied them together with rope more effeciently. I like the way they represent a Ram skull from a distance, these elements further evoking the sacrificial element.






















The scene however still feels somewhat empty...so after some research, I cam across and Ivy Generator, which does a fantastic job at re-creating a simulation of how ivy would naturally grow, and randomises each time you create the effect.

http://graphics.uni-konstanz.de/~luft/ivy_generator/

However, after several hundred attempts at tweaking and experimenting with the many sliders that effect the growth of the vines, the thickness, leaf to branch ratio, and sizing of the leaves not to mention trying to figure out how to import and export geometry, but I finally managed to reach a result I was after, with help from this blog:

http://cgborremans.blogspot.com/2010/12/ivy-generator-and-fixing-materials.html






























After some tweaking of the new geometry in regards to the lighting arrangement, I began to play with the textures so they would read properly, due to them being alpha's maps, I had some trouble getting them to work, along with some coding of the mtl. file that needed to be imported . The shaders needed their transparency fixed so adjusting the outputs was a must, however after doing so the transparent part of the alpha seemed to shine, for reasons still unknown, but after changing what was original set up for to a Phong shader, changing to a Lambert seemed to fix this.









As you can see the textures seemed to have a saturation problem, and feel quite separate from the scene. In order to add some individualize the work from the generator, and to try and remedy this problem, I began to experiment with different textures, however I tended to struggle creating alpha maps from them as I'm not very familiar with this process, leading to white rings around the texture, and the saturation problem.















However, after deleting the textures from the hypershade, and reapplying the old ones, a strange thing occurred, where the past deleted textures still rendered, ignoring the newly placed textures, and even though this was a completely different scene, and geometry, it wouldn't stop happening.








This has grinded me to a stop for the time being, because Im really perplexed as to what the hell is happening lol. After some reflection, I feel the best next step would be to perhaps create the vines, without the leaves, and make my own geometry and alpha's, and place them scarcely around the vine to achieve the same effect.Once I've figured out the problem, I'll make sure to post a step by step process for future efforts. On top of this, I feel there is a need for more in my scene, as it still feel very empty in a sense.









Monday, 11 July 2011

Recent Updates - Plans For the Summer

So after submitting my e4 project again, I was pretty happy with what I achieved, and learnt....although not all went to plan in the final moments before submission, with all my renders done a simple bit of composting at uni turned into a nightmare, as the computers decided they didn't want my work, and crashed numerous amounts of times, which was so frustrating because I had it all :/

I think with this occasion, I've learnt my lesson not to try and utilize every last minute to make my work better, which is I think my worst habit in this course, instead make sure whatever I have is rendered/composited by at least the day before...

Here is my 'Making Of':

The Sting - Making Of


And some renders from the animation



















That aside, I did mange to get a pass, which was a relief but it clearly isn't good enough for my third year.I'm still not happy with what I've submitted however, and I have grown a fondness to what I've already created, so I intend to improve it somewhat, and jump over and hurdles I faced before submission.

The plan for doing so is:

Re-Texture 'Stingy' & E4 Heart
Animate E4 via Blendshapes
Animate Water via Blendshapes
Re-Render With AO Pass & Masked Water 
Tweak Camera Angles/Movement
Composite and Edit in Post

In terms of working with animation with my actual character. This is an element I feel I could pursue elsewhere for the time being, as this project for me was more about developing skills in environment, lighting, texturing and above all, atmosphere. I really think although this is only for an E4 Sting, that it has more potential to become perhaps an animated short in the right hands, so a revisit might be on the board at some point.

As a result, my aim for this summer after finishing up some of the work on the E4 Sting, is to try and build more skills in maya, as well as speed up my digital painitng skills and pipelines. I also want to develop some knowledge of Mudbox, both texturing and modelling, to see if my pipeline can be improved further.

This is going to be done by revisiting my Transcription Project, and applying game texturing techniques, mixed with my developed skills at present. Next, I want to finally challenge rigging and get that down for third year, as this is a vital skill I've not really had time with. Addressing some of the geometry in my character is a must, as it seems to be quite intensive for a game character. I feel some concept art is in the works for this character also.

Friday, 1 July 2011

'The Sting - Final Updates

Apologies for being so quiet recently, I've just decided to crack on and produce work. I feel I really progressed over the past few weeks and added more to my skill base, Texturing wise, I feel like I cracked how to really paint onto UV's, and avoid seems through texture and bump! This was a huge deal to me, and alot of trial and error took place.

Lighting, this was really a new area to me all together, and was very daunting at first, but I manged after several hours of tackling attributes, to come up with the balance I was happy with, one of the key points being 'Break' & 'Make' light links. This was a huge discovery for me, as it allowed me to single out different light sources and apply them solely to particular geometry. This came as a godsend when trying to figure out how to blend the the backdrop to the water. I went through alot of trial and error in this section also, which is where I spent alot of my time. The lights I found most effective were 'Area', 'Point' & 'Spotlight'.

Here's an example of what I went through:

















I also tried my hand at creating a water texture, which I feel worked pretty well considering I was unfamiliar with the technique. Again, alot of trial and error here, as I utilized the Env Ball Node, which felt slightly unpredictable when trying to find a suitable image to reflect, whilst also playing around with Paint Effects to create more of 'Swamp Atmosphere'.












In regards to the 'Needles' & Thread, I used a glow effect on the both to back them pop more from the scene, and create an eerie effect. I experiment with a Cloud Node for the first time, to see what effects it could create, and also used the same node in a dupe to input into the bump, which was very effective, however trial and error once more. To create the Thread, I used C.V curves and extruded along cylinder face. However to save time, and the intricate curls, I wanted to use the same section, and adapt it to each needle, and the merge with another C.V curve, this caused some problems, but I manged to find a solution by snapping the Curve Vertices's to one another, and reversing the flow of each curve, and then using Attach Curve to finish the job.

In terms of rendering, I was attempting to use and Ambient Occlusion for the first time, and manged to pull it off, however, due to the displacement on the water, it began to cast shadows past the geometry of the water where it rose for the displacement. Keepin this in mind, and after speakin to Alan, I was going to try and mask it through the use of surface shaders, ander rende the water on another layer, however, I ran out of time














For me, the re-take on this project was more of a scene based direction, as opposed to character, because I neglected this previously, but also I havent really made an approach towards this direction on this course yet. I didnt manage to animate the character, but however I hope I portrayed my initial style in a much more effective manner, through my time based media. I feel I've learned a great deal from this re-submission, and acquired many skills I had previous been lacking.