Wednesday, 6 June 2012

At last, my project is finished. This is the final video, with a breakdown at the end

I am no longer a student, and so, this site will no longer be my "teesside research journal", but just my workblog. I will post work in progress and I will be updating as often as I have new projects going on.

Friday, 20 April 2012

online journal, entry 15

Since last time, I have been working on my composition in Nuke, and today I have been rotoscoping for many hours. I am content with the result.

I am almost done with my soda can animation scene, which will appear at the end of my film. I modelled and animated the can i 3Ds max, and added effects and color correction in After Effects.

this is one image of the sequence I rendered from 3Ds max, and next is the final shot I rendered from AE








There's no sound as of now, but that will be added for the final piece ofcourse. I have had a meeting with a great guy called Alex, and he's happy to make music for my project.

things are looking good.

Tuesday, 10 April 2012

nline journal, entry 14

Per Viggo Bergsvik - Final Year Project
Online journal, entry 14


 Okay. A while since I posted last time, as there is much to do these days.

This is what I have been doing since last time.

* rendered everything from 3ds max
* comping in Nuke (ongoing)
*  begun working on the soda can scene, at the end of the film.

Here are some process images from the soda can scene. I used a subsurface scattering material for the glowing metal in the scene to make it look belivable. The metal cools down and the brand "dare" is revealed. I first tried to create the glow effect by blending and animating maps and self illumination, to create the effect, but this way turned out much better.


SSS material
blending maps and using self illumination




Tuesday, 20 March 2012

Online journal, entry 13

Per Viggo Bergsvik - Final Year Project
Online journal, entry 13


Okay, long between posts these days, but here's some updates.0

I've been doing a lot of different things the last weeks. I have been learning Nuke, a compostiting software which is node based. It uses some complicated systems which are hard to get my head around, but I manage, thanks to the book Nuke 101 by Ron Ganbar.  And in case I really get stuck I allways have Adobe After Effects in backhand. My plan is to do the color correction, effects, and compositing in Nuke.

Also, I have been doing a lot of textures, which are going great. I have written specificly about that in my Contemporary studies in games blog.

The lighting part is the best part. My Hdri seems to lit the scene perfectly, and I don't need to worry about it at all. It just works.

Right now I'm just tweaking the scene, adding/adjusting materials as needed to produce the final render. My plan is to use atleast these renderlayers:

- Beauty
- Ambient Occlusion
- Velocity (for motion blur)
- object ID
- shadow drop

To get more control of the final output I might add  more.

Here's some testrenders illustrating the process so far:


Monday, 5 March 2012

Online journal, entry 12

Per Viggo Bergsvik - Final Year Project
Online journal, entry 12

Okay! I'm done with the animation! It took me 4 weeks less than I counted for on my time schedule. I realised that it would maybe take less time than I planned, but this is a massive releaf. I might even get some textures finished. I will let a preview of my composition speak for itself now. The composition itself is not nearly finished. I have only placed the cg on top of the footage to see the whole picture, and acting in it's current state.



Thursday, 23 February 2012

Online journal, entry 11

Per Viggo Bergsvik - Final Year Project
Online journal, entry 10

I have now tracked footage with Boujou, set the camera up in 3ds max, and begun animation. I need to finish about 15 seconds of animation, and that's not so much. My robot rig seems to be working perfectly so far, although I realise that some parts of my robot could be better designed in terms of movement. For example, the elbow joint has rotation points on two different places, which makes it a little wierd to control. It's descision I made when designing the robot, and it's not wrong, but it definately could be better. 

Other issues is the camera Fov (field of view) , which was a little hard to match in 3ds max as the file from Boujou did not contain the right one. I used a 35mm lens when filming, but when I changed the lens in Max to this it looked wrong. I have set up the camera approximately now though, and it works so far. I'm now on blocking animation. I'll post a preview of my blocking so far on the two first shots.

[img]

[img/]

Thursday, 16 February 2012

Online journal, entry 10

Per Viggo Bergsvik - Final Year Project
Online journal, entry 10

I put together the photos i shot at my filming day today, into a full hdri that I can use for lighting when I put my cg characte into my footage. Here's some process images:


the raw images I took on shooting day:




I merged the images from each side into 3 Hdri "sandwiches" in photoshop, then used Hdri Shop to unwrap them. I then put them into photoshop again to stich them together (and remove myself!) so that they could be used for lighting in my scene.





A testrender that shows some early result. I might upload a videoclip too.





I am now dealing with the tracking of the footage I filmed, so that I can put my character into it. I am using Boujou, Nuke and 3Ds max

Friday, 10 February 2012

Online journal, entry 09

Per Viggo Bergsvik - Final Year Project
Online journal, entry 09

I have made a difficult descission that took a lot of thought. As my project takes too much time as it is now, I need to cut down on anything I can. I will therefore only texture parts of the robot, and begin animation as soon as possible. I will use textures and shaders that resembles the final "perfect" product as much as I can, so that it's easy to replace later on when I have more time.

I have been sick for one week of my project, and I didn't get much done. I will send an email to the administration of the uni and try to get a weeks expansion of my project time. This will surely help me..

BUT

Enough jammering. I have done a lot of work since last time, despite all. I have rigged most of the character, and it is nearly ready for animation. I used CAT (character animation tool, I think!)in 3Ds max for some part of the model, like head, torso, hip and parts of the legs. The rest, like arms, fingers and toes are already ready for animation, as I have linked the geometry up. The reason I use a CAT rig inside the robot is because of the instant IK system I get, and the CAT animation layers, which are useful and easy to use. Not rigging the whole character is a cheap solution, and might not be the best on a longer project with lots of animation, but in my case, The Cg character is on screen for about 15 seconds. It will get the job done. I will post an image of the rig in it's current state.


Friday, 27 January 2012

Online journal, entry 08

Per Viggo Bergsvik - Final Year Project
Online journal, entry 08

It's a while since I posted something here now, So I have a lot to catch up on.

I have been modelling my robot, and I have had an unexpected issue. Modelling for realism takes so much time! Especially when modelling complicated mechanical objects, that needs to be unwrapped and subdivided. I have used 3 weeks more of time on modelling and unwrapping than I accounted for, and I fear that the texturing part will actually take longer. I am supposed to be intrigating my finished cg character into the footage by now. Luckily I have accounted almost 6 weeks on the animation part. If I can somehow press it down to half the time, maybe things will work out. But the way I see it, I might not have time to complete this project in the matter of how I wanted, esspecially with all other modules.

On a note, in my other module, Contemporary studies in games, I have decided to do a research on Modelling, texturing and lighting techniques used for creatind realism, and I will fuse it with my Practical project on thise areas. Because of that, I will not go in to deep detail on how I have done my modelling, texturing and lighting in my practical project report.

Here are some images in my process.






Per Viggo Bergsvik

Wednesday, 11 January 2012

Online journal, entry 07

Per Viggo Bergsvik - Final Year Project
Online journal, entry 07


I put together the footage that I did yesterday, and I am posting it here. On a note, I made two mistakes when I did the filming. The first, was to film in a .mov format that only could be opened if you had the codec (made for mac), I had to convert the raw movie files that I had, which means extra work. The second, and bigger mistake was that my footage were a little underexposed, as you may see. I can fix that in post, but nevertheless, it drains some of the quality of the image.



Per Viggo Bergsvik

Monday, 9 January 2012

Online journal, entry 06

Per Viggo Bergsvik - Final Year Project
Online journal, entry 06


Christmas is over and a half a year has gone. I don't feel like I have done so much progress since last post, Because I have been struggling with the design of my CG robot. The functionality vs design issue is harder to overcome than I thought, I and I thought it would be hard from the beginning. Every little piece of the model needs to work well with other parts, look consistent with other parts, and function with other parts (as long as it is going to show in my video, that is). I am nearly done on the drawingboard, and will post the final blueprint for my CG robot whet it is done.

One other thing I need to do, is to get the footage lined up, and the shots in order so it fits thae animatic. No easy task. It seems that real footage and stopmotion-like animatics are different from one another. I should not be surprised. I will also post what I have set together when it is done.


Per Viggo Bergsvik