Good news! Work!

About a month and a half ago i was hired by RetroRonin to help build their wonderful voxelMMO calle Voxelnauts! you can see the project website here: voxelnauts.com

Anyways, this is my day job from now on for i dont know how long, but either way im loving it! here're some of the things i've modeled and animated recently. hope you enjoy it!

Parallax and voxels, would it work?

The sequence of photos above are to show the continual steps and changes that had to be made to make a parallax effect work using the voxel model i'm building (still not complete).

Originally i was going with 3 buildings like in the first image but it definitely didn't feel sufficient since it doesnt occupy enough virtual screenspace. By this (virtual screenspace) i mean the space not only onscreen but out of it. Parallax is important in the way that it gives you the impression the world has depth without there actually being any. It's a simple way to use images instead of full 3d scenes to tell the viewer "hey look! i'm actually behind this object!".

Once i knew that 3 buildings wasn't enough i added a fourth...then a fifth....and i'm still not done. What makes this project even more interesting is the fact that i'm building this 3D world but trying to throw it off as 2D...weird. While i could possibly just go full 3D i think the 2D parallax is an interesting take on the project. This allows me to build a full world but retain the focus of the game on the buildings in the "middle space".

In the final image you can get an idea of what i mean by "middle space". The world is devided and in the game the parallax effect will make it so the streetlights and street are faint and pass over your screen but the buildings will always be the focus (hoping you can visualize what i mean). There's lots i can do to make parallax even more interesting using 3d assets, hopefully i can show you more soon.

SO last night after i made Protoman i decided to continue and make both Zero from Megaman X and Captain Falcon.

Zero was a bit tricky for the top of his helmet since i didn't want to make it feel too fat in the front. Captain falcon was a lot simpler to design and I had more fun making him due to all the falcon punch puns i was making in my head but overall i feel Protoman was my best design from last night. I nailed the shading on that a lot more. Although Captain Falcon is great as well!

I think it's complete now ^_^
I'll probably kick myself later when i notice further mistakes, but hey, it was good practice. I'll likely do more characters from the megaman series since they have some decent designs for head models alone.

So, i made a big fix on this now. I knew the scarf needed fixing since the beginning but i figured i'd leave it till the end at the very least. I went ahead and removed the piece I had and made it from scratch instead. Getting the knot tie in there felt perfect and really makes the model so much better. Last thing left is to finalize the shading work :)

Fixed the helmet shading and the front of the scarf (i think both of those are done). Next is the tail end of the scarf and i believe that'll be it for this.

Working on a model of Protoman. Using the classic design since i've previously done the classic megaman model (i like consistency). This is some pretty good practice for shading since his scarf is another notch up the difficulty chain. I'll need to shade his body and the scarf a bit more, buut i definitely don't want to use more colors.

Toying with the color, saturation and Hue of the deer. I like this a lot more, makes the deer pop! The problem though is that Critical Annihilation has a different tone, one that the deer before would suit a lot more. This was a good test however :)

Made a gorgeous deer for Critical Annihilation :)
The results really shocked me.
The process was focused around the head. Taking my own advice about starting with the smallest element, i decided to make the nose and work from there a chunk at a time. First making the head within a 12^3 block, then the chest (20^3), then the legs (16^3) and finally the body (30^3). Cutting it all down and making sure the pixel art feel to it wasn't overboard since Critical Annihilation increases draw calls if i add too much color.

Using a profile image of a deer helps a lot in figuring out the shading and overall shape. Don't eyeball so much, especially with animals since you can go from making a deer to creating a moose. (1 voxel is enough to throw you off, review constantly)

Since the programmer (Pat) loves to blow things up, i decided to make the insides of the model with guts using the Russian Doll function (guts out the inside and makes it a new model component).

This was fun and pretty informative, shows the improvements I've made in the past months :)

At the end of the night i started to try out Unreal Engine 4 to see how to get it working with my voxel models. I found many ups and downs as i played around with it.

First off in the first image we can see a setup i made using a Turn based game blueprint. The blueprint system made by Unreal is actually really easy to comprehend and is relatively straight forward at the surface. The dialogues make sense and Unreal is nice enough to implement a help system/tutorial that guides you in each new section you encounter. Overall the entrance to Unreal wasn't as overwhelming as i expected.

Once i started the blueprint system i decided to start importing assets. I figured that of all things, importing assets should be known to me first since i create them all the time and this is most likely what i'll run into most often in the future.

The second image shows the 2 assets i made within Qubicle for testing in Unreal. The cherry blossom tree and an octagonal tile. Knowing that the OBJ exporter in Qubicle 2.0 is much more refined i decided to make my assets in Qubicle 1.6 (as i normally do) and then use Qubicle 2.0 as the middle man when transferring assets. Originally i was doing this strictly for the mesh exporting, later i found out i'd be using Q2 for many more reasons.

The Third and Fourth images are what happened when i transfered the models from Qubicle 2.0 to Unreal Engine as OBJ files. As you can see, the Cherry blossom is rotated along the Y axis while the Platform looks completely fine. The cherry blossom tree was my first try at importing assets. Since the tree was a random prop i figured having it rotated like that isn't much of a big deal since i can rotate the object in Unreal no problem. However, this was a no go for the tiled platform. The tiled platform originally imported rotated like the tree and that couldn't be something i would keep or else i'd have to take into consideration the rotation of the object when writing a tiling system (which in this case already exists within the blueprint). What i noticed and what you can see in the fourth photo is that the pivot location of the tiled platform is actually on the left hand side of the model, why is that? Is it ok?

It turns out that the pivot location within Unreal 4 is completely different to Qubicle 2.0, both in its orientation and value. The fifth image is what i had to do to the tile piece in Qubicle 2.0 for it to come out the way it did in Unreal 4. Not only did i have to rotate the model in Qubicle but i had to relocate the pivot point on the model (and in the end it still didn't come out centered in Unreal, c'mon!).

In the end i managed to get both files in and working with the tiling script written within the blueprint. I exported the Qubicle models as OBJ with all export functions on and as a single mesh. When importing into Unreal i had to increase the Unit value to 5 for it to scale as it seems in the last image. (i think it's still a tad too small, will go through further testing).

Unreal Engine 4 was fun to use for the first time. Some tweaks were annoying and now knowing what i had to adjust makes me feel a lot better with the tool. Next would be for me to actually start implementing the functions on the objects and replace the player models with something more fitting to my style. I'd also like to go and try out the animation system considering they support FBX in these newer builds. The lighting system in Unreal is a lot of fun to play with and looks fantastic, even with voxel models at the resolution of the cherry tree! The camera controls are a lot more fluid and usable that the ones from Unity and i feel like i can get a lot more done in here. All in all i like what I see and will continue working with the tool. Maybe i'll eventually be able to make something on my own within unreal, who knows.

Here's a turnable gif of the Biplane that will go in Critical Annihilation. Took roughly 4 hours total to make with a cube count of 326,161. Only 1 more vehicle left and i think i'll be good to go!

Made a group of slimes. Not sure about the shading quite yet since i'd like to see how it looks in-game with the transparency set first. Will also be animating them.
I'll go ahead and use Unreal Engine 4 for testing. If i can setup the right environment to make concept scenes along with the right lighting i'll be really pleased.

I was recently contracted to make some assets for a roguelike dungeon crawler. The budget isn't large so i couldn't go too crazy on the amount of objects i can make but i managed to define a style that suits the character perfectly and is pretty new. I'm very pleased with the aesthetic and hope to refine the style more on other characters/objects. I still have a few more models to make, and some animations as well, so i'll be updating this project soon :)