Finally! I worked around the rendering issues mentioned previously, so I can finally reveal what I've been working on: the Iron Man 'Heartbreaker' suit.

Proof that I actually was working on something. Ha!

I'm quite pleased the way the renders turned out. I've been working on this so long that all I see is what has to be fixed, but seeing some renders and hearing some feedback lets me know I'm on the right track*.

*Keep reading for a great continuation of this train based metaphor.

Interestingly (not my opinion, an actual certified Very Interesting Thing™), the rim lighting on the models is actually from the material, not from physical lights. Partially, that's because I think you get more control, but really it's because I've never been able to get completely satisfactory results from trying to set up rim lighting. This setup uses the normal of the faces to determine whether it should be highlighted or not. Here is a (simplified) screenshot of it:

The normal is manipulated with the 'Normal' node before being sharpened by the colour ramp. This mixes between the glossy base material and the white highlight. To get a really bright highlight I actually use an emission shader, but to make sure it doesn't cast light onto itself I use a 'Light Path' node so only the camera sees the emission, the objects in the scene just see black (the empty 'Shader' input).

This is a simplified setup, so it just shows the right side rim lighting, if you want other sides of the model to be highlighted, duplicate the 'Normal' and 'ColorRamp' nodes, adjust the normal direction and add them to the other rim nodes with a 'MixRGB' node set to 'Add'.

See I told you it was interesting, and yet you (probably) resisted believing me. Hopefully you will trust me a bit more in future. If not, things are going to get pretty embarrassing for you when I continue to show you Interesting Things™. Let's avert this embarrassment by jumping aboard this train of Trust and riding out this analogy right to the end, together.

Now that all that train business is dealt with I can get on with my work, so that by next week I will be able to show off a new screenshot.

I guess that's the end of the line for this post (THE TRAIN FUN NEVER ENDS).


Ray.

Well, I've managed to stick to my idea of doing weekly blog posts, but then again, I'm only a week in, so maybe that's not quite the achievement I thought it was when I started this sentence.

A bug

It's always a bit pretentious to say you've got a big project but can't talk about it, especially when that limitation is self-imposed, but if I'm going to get the most out of this project, it means working out how to reveal it best. Last week I mentioned that perhaps this would be the week I showed it to the world. Or at least, the small segment of the world that is interested in 3D. Interested in 3D and likely to see my post. Interested in 3D, likely to see my post and actually click on it. Interes...well you get the idea, it's niche:

On Monday I set about starting this reveal process by trying to do an up to date render. I hadn't done a render of the project in a while as it had got so large geometry-wise that my GTX 570, with it's 1GB of GDDR5 RAM, kept crashing midway through the render. But silly old me! GPUs are so often lauded as the be all and end all of graphics performance and rendering that I had stupidly forgotten about my CPU. Sure, renders are going to be a lot slower, but they render!

Of course, the new found optimism that that previous exclamation mark hopefully depicted, was short lived. I have a compositing setup in the scene that renders the current camera angle in the current scene and then renders an alternate angle from another scene (with linked geometry) and combines them into one, nicely laid out image. Trouble is, it now gets to the alternate angle scene and crashes....so the rest of the day was spent trying to reduce the file enough to create a bug report, which I did.

Another bug

Tuesday gave the gift of another bug, this time in my own add-on 'Animated Render Border'. I would love to say it was coding skill that allowed me to solve the bug in half an hour, but really, it was panic. There is nothing quite like the rush you get from seeing the words "does not work" in the same context as a product you have released. Well, there probably is, but not in my quiet life.

And the rest of the week

Luckily, for the rest of the week I actually got some good work done, but it depends on the good developers who work on Blender as to when I'll actually be able to show you something. It's going to get pretty awkward on this blog if for the next few weeks I just have to say "Maybe you'll see it next week!" as eventually you'll presume there is no project.*


Ray.


*"Why not just show the angle that does render then?" I could, but I've waited this long to show it to anyone, so I'd rather wait a bit and do it right, than just show half of the model. I have a few ideas how to get around the bug for next week's post anyway.

I haven't posted here much, but I'm hoping to do a weekly post of what work I've been doing each week. Seeing as I haven't done a blog post for a while, I'll cover a few things I've been working on in the previous months. Does that sound fun? I don't know, but you've already loaded the page now, so you may as well read a bit more.


Scene Nodes:


An interesting little experiment in Blender I did a few weeks ago is something I'm calling 'Scene Nodes'. Blender used to have something called the 'Oops Schematic View' in the Outliner, which allowed you to view your 3D scene in a visual, node based way, showing the relationships between your assets. This feature never made it into the revamped Blender 2.5 so I thought I would have a go at recreating it with Python. Partly, that was because I just thought it would be fun (it was 50% fun, 30% googling, 15% deciding what colour each node should be and 5% crying) but also because I thought that if it proved to be popular I might develop another add-on to sell on the Blender Market.

The result above shows all the different objects and materials in my scene and also the relationship between them. 'Cube.001' (orange) is the child of 'Camera.001' (grey)  and also a parent to 4 objects (orange) and a lamp (yellow). I got it working so that not only could you press a button to generate this node setup, but that duplicating nodes would duplicate the actual object, deleting a node would delete an object and changing the connections between objects would change the real object's parenting.
 
 
Before you go rifling through you wallet to find money to throw at the screen* (because you obviously love the sound of it, not because you're misguidedly attempting to throw loose change into my eyes), while it was quite fun to do, I'm not entirely sure that there's enough interest in it for me to make it an add-on. My last add-on wasn't really cost effective, considering the development time vs monetary return. I'm not ruling out making it an add-on, but I don't think it's a 'high-priority' project.

*and people say I'm a fantasist.**
**If anyone of importance is reading this, I'm lying, no one has ever called me a fantasist. It was just a joke, which we (I) here at www.RayMairlot.co.uk deeply regret.


Experiments with Unity:


I've also been making a game with a friend, not in Unity as the title might suggest, but in Flash. Flash might get criticised a lot, but for making a cross-platform game it really makes a lot of sense (to me). However, I'm well aware that several sites are moving away from flash, at least for video delivery, so I thought I would make some investigations into other game engines, just in case the day comes that Flash content is no longer run by web browsers.


I decided that I would try out Unity as it has a free version and I got to grips with one of the tutorials. I wasn't best pleased to find my simple tutorial scene of a ball rolling over a plane, collecting yellow diamonds, came to an eye-watering 170MB. God help me if I try something complex*. Maybe I can cut that size down somehow, but at least it does cross-platform publishing pretty easily.

*Dear God, I know you're busy, but please help me squash these megabytes. If you help me (for free) I'll tell everyone it was you that helped me and I think that will be great exposure for you. Thanks, Ray.


I had originally thought that if I wanted to learn Unity it would be better if I already had a working game in Flash, so that when trying to re-create it in Unity I wouldn't have to be worrying if the logic of the game was correct, only the conversion of AS3 to C# (one of Unity's coding languages). I made a little Tetris game in Flash, but having seen the large file size Unity creates I was put off from actually trying to convert it. I realised that Flash is still the better option for me, but that Unity was a viable alternative if I really needed it.

That's not to say making the Tetris game was a waste of time; it helped me a lot with realising how to implement OOP concepts properly, which I can put to good use in the game I'm making in Flash.


All in all, quite a few experimental projects, but now I'm back to my main one, which I may start to reveal soon. Or not, know one knows. If people say they know, they're lying.


Unless that people is me.


Ray.

A few days ago I released my first add-on on the Blender Market: Animated Render Border. It's an add-on that allows the render border feature of blender - which allows a portion of an image be rendered - to be animated and track objects. This can cut down render time by skipping rendering the background (which might be blank if using render layers) or by rendering only a specific part of an image for a whole animated preview. I won't go too much into the details of it as that's all I seem to have done over the last few days, and frankly, I've bored myself by writing the same blurb on all the various social sites and forums. Instead, indulge yourself by watching the (relatively) short video below which is a full demo of the add-on, and of course, feel free to head over to the Blender Market to buy yourself a copy:

This add-on was originally started in February this year. It was just an experiment that I thought might be useful at some point. I'd seen a few attempts at such an add-on and a few more blog posts lamenting the lack of such features and surprisingly, I had quickly got the core code to work in the viewport. After that, other priorities took hold and I left the add-on, only to be reminded of it recently by a few posts on Blender.StackExchange and by my own attempts to use it, forgetting I had left it unfinished. I had thought at some point I could release the add-on as a bit of publicity for my site, but the fact that some highly skilled python coders seemed to be coming to the conclusion that an add-on like this could be useful, prompted me to finish it before they might be tempted to do the same and 'steal my thunder'.

At that point I hadn't thought about selling it, I just wanted to quickly finish it and release it. I'd thought about it originally, an extra revenue stream is always welcome, though I wasn't sure if the add-on might be too simple for the Blender Market, but all that slipped from my mind until I was a few weeks away from releasing it for free. Reminded of my idea to sell it I did a bit of investigation and applied to be a vendor on the Blender Market by describing the product I would like to sell. My application was approved, which meant I had to fill some of the extra requirements of releasing on the Blender Market. Don't get me wrong, they're perfectly acceptable requirements, ones I would have set myself, just ones I hadn't been thinking of when I thought I'd "quickly finish and release it". I wasn't just finishing up the code, I now had to think about a video demo, documentation, product images and descriptions, future updates and promotion. I was reminded of each one of these just at the point I thought I might be nearing the end. So the project, like all projects, was a bit longer to finish than expected.

There is also a bit more pressure when releasing a paid add-on. Getting it functional wasn't good enough, I now had to look at it from the perspective that someone would be buying this and expecting it to work in a professional way. People might be using this in their workflows. What if it crashed and they lost their work? Suddenly, things that I thought were fine for a free release stuck out as dysfunctional. Things that I knew had to be in place for tracking to work had to be checked for and the user warned if they were missing. Every circumstance a user might get into needed to be accounted for. The add-on requires a camera present in the scene. What if they delete the camera? What if they manually turn off settings the add-on has turned on so it can work? What if they're idiots start pouring pasta sauce over their keyboards because I've accidentally pasted a lasagna recipe into the documentation!?*

The days after the release were spent answering questions from all the various places I'd posted it, solving bugs and talking about future features. I hadn't quite anticipated the amount of attention it would need after the release, though I was aware that part of the Blender Market commitment is that authors provide support for products. Luckily, it seems to have settled down now and I have received a modest, but pleasing amount of sales so far. This coming week I hope to be able to return to my modelling projects and improve the add-on steadily over the next few weeks and months.


Ray.

*I will not be held liable for any damages incurred by pasta sauce or any other ingredients which when combined could constitute a 'lasagna'.

You really do start to realise how big a vfx project is once you get to the compositing.
Take shadows for example. This became (and continues to be) perhaps the biggest part of the post processing. I already knew I would have to build any object in 3D if it needed to receive a shadow, but there were several things, having never done a project like this from start to finish, that I didn't realise until I came to the compositing:

Objects...objects everywhere

There were a lot more objects in the scene once I really started to look. Originally I just thought I had to build the main buildings and surfaces. But no, as I looked closer, railings, people, buses, taxis...the more I looked the more I found. Luckily most of this didn't require any new techniques, just more modelling and quite a bit more animation (more on animation in a separate post, later).

Most of the objects can be dealt with quite easily as they're stationary. There was, however, one pedestrian who was walking across the street and was close enough to the camera that the shadow couldn't simply go over them flatly but for realism should match the contours of the body.

I already had a person armature and model from the people I had made to sit on the rollercoaster so it wasn't much of a problem to re-use this model and animate the armature to the person in the footage. It was a bit tedious, and if you saw the 3D view from any other perspective than the camera you might think this person didn't have a lot of bone structure, with knees and elbows bending in directions that would make you wince. Nevertheless, from the camera view the little 3D figure appears to cross the road matching his real-world counterpart fairly well with shadows falling over his body instead of being flat, like some of the more distant people can afford to have.

Foreground people

If mid-ground people weren't annoying enough to recreate, then boy was I in for a treat when I noticed all the fore-ground people, with more detail than my generic 3D human could handle.

When we originally filmed, it took several takes to find one that didn't have cyclists or buses going right through the middle of the shot, but even so, it was impossible to film on a busy street without there being some pedestrians and vehicles in the way.

Any person that appears in front of the shadow means that a part of the shadow has to be removed. The only way to really deal with that was with masking, so I had to painstakingly mask the 3 or 4 people walking across the shot frame by frame. Frame. By. Frame. This was not enjoyable even though it was a fairly short sequence. But, once complete a simple 'Math' node could add the white mask back over the top of the shadow pass, cancelling out the shadow in one fell swoop.

Existing Shadows

These two previous points are annoying but at least I was already aware of those issues even if I wasn't completely aware of the scale they would need to be dealt with. Which left the detail I hadn't been aware of:

The footage has existing shadows. This sounds obvious. And it is! Of course there are shadows in the footage, things in the real world cast shadows. But what was less obvious is that I would somehow have to stop the CG shadows from overlapping them. If they overlapped there would be a dark patch, you can't cast a shadow onto something that's already in shadow without it getting darker than it should be, so the shadows have to very precisely stop at the exact point the real shadows begin.

This started to worry me slightly, I'd missed something which seemed obvious and I wasn't immediately sure how to solve this...

I realised that unless I wanted to try and roto (mask) the shadows very precisely I would have to attempt to use a technique similar to how I'd extracted the ripples from the river. My thought was that I could create a rough mask around the area in the footage with the existing shadow and then isolate the darkest parts of that area, which should be the shadows, with colour correction. For the most part this does seem to work, there's still a bit of fine tuning to do at the exact join between real and CG, but the vector blur helps to, well, blur that line.

A lot of work for just one part of the compositing and there's still lots more to do. Aside from compositing though, putting the people on the rollercoaster was a fun little exercise so I'll try and cover that next time.


Ray.