Lucas 'Darbeenbo' Annunziata's Workshop Thread
polycounter lvl 13
Thank you for checking out my thread of W.I.P's and finished submissions. I am always looking for critique, even on completed entries.
My Workshop
Views, votes, comments, and pretty pictures
Most recent submission:
What I am currently looking for feedback on:
My Workshop
Views, votes, comments, and pretty pictures
Most recent submission:
What I am currently looking for feedback on:
Replies
http://steamcommunity.com/sharedfiles/filedetails/?id=226232991
________________________________________________________________________
This post is a bit of a post-mortem on the creation of 'Skull Cap of the Betrayed' for Juggernaut. Mostly I want to share my pitfalls throughout this experience so that maybe it will help some others along the way and help myself identify where my weaknesses lie.
Things I should have done:
Research Dota 2 Characters
I am not a Dota 2 player, so I have no clue as to what characters are popularly played (which will have an effect on how many people look at your item). More importantly than that though, I barely glimpsed through the large amount of characters and just chose one that I thought would be easy to make something for. While this was decent reasoning, I didn't even open up the character in game to hear his VO, check out his animations, nothin. I just dove in chomping at the bit because damn it I was going to make an item!
Look at successful and unsuccessful items in the workshop
It took me until the very end of creating this item when I was creating the steam workshop images to even bother looking at what was available for the character already, what had been accepted, and the general likes/dislikes of users voting on the items. While this doesn't HAVE to inform your design/presentation, it is better to be informed and to actively choose to ignore these things, then to just be ignorant. Also having an idea as to how you will market your items is always a good thing. Maybe you'll spend more time on the concept art if you'll use it for marketing purposes.
Concept in 2d before jumping into 3d
I had no idea what I was going to make before I jumped into zbrush and just started sculpting. I didn't gather any reference images, I didn't scribble anything, and I didn't do any paint-overs of my model. and this shows in the final product.
Have an idea of how the color/pattern will look on the model.
Same kind of thing here. General lack of knowing where I was heading lead me to a position where I was trying to retrofit some design onto the model, rather than the model/texture design working in harmony with eachother.
Test the import process before creating the item.
More than anything else, the biggest pain in the ass for me was getting the model imported properly and be skinned. For some reason when I was exporting from maya 2013 using FBX 2013.3 from any of the valve provided maya files, none of my skinning would be present in the fbx files. When I imported the asset it would be sideways and off in the distance. The workarond for this was to export the rig to a new file, then skin the object, then export it. However, the helmet was slightly off. It turns out the the valve provided download of Juggernaut was slightly different the one in game so I used Dota 2 CS: GO SDK Model Viewer Guide to get the source models and Mesa to import them into Maya. Overall, I actually spent more time figuring out all my technical issues than I did creating the item.
Test the low poly in game before baking to get a better idea silhouette, size, and scale.
I had already unwrapped and baked my low poly, and had a base texture pass before I even previewed the mesh in game. This was a poor decision because had I done this, I would have had a better chance to adjust my proportions.
For how little thought and foresight I put into this asset, I'm pleased with certain aspects of it and it has been a good learning experience.
A few things that I'm glad I did:
Helpful things I've found on the way:
Maya Users:
froTools (Toolset for working faster)
froXnormal (Quick xNormal exporter)
froRetopo (Simple tool for doing retopology)
Mesa (xport and import 3d models for the Valve Source Engine)
Nice documentation of your learning process! I'm sure lots of others will learn from it.
I like the model but feel it needs some color separation between the horns and skull.
In game side/top shots it just bleeds into one shape. Some gradient (dark at base of horns) fading to light eeither on horns or skull would help a lo9t I think/.
Research Dota 2 Characters
After finding a character that inspired me, I hpped in game with him, checked out his animations, spells and VO to get more of a feeling for him. From there I read his little paragraph of lore for more inspiration.
Look at successful and unsuccessful items in the workshop
I went ahead and researched what successful sets items were available for him already in game, and what kind of other items are still pending for him in workshop purgatory. Here is Lich's original item set and the two that are in game.
Lich has got this sexy blue color scheme with punches of red that I really enjoy, and his lore is reflected well with his items like the rune covered chains. I really dig the icicles material in the Eldritch Ice set and definitely want to do something similar in my set.
Before I even started planning out a concept, I gathered the technical specs for lich to keep in mind.
Neck
LoD0 Triangle Limit: 1250
LoD1 Triangle Limit: 900
Texture Size: 256H x 256W
Arms
LoD0 Triangle Limit: 400
LoD1 Triangle Limit: 300
Texture Size: 256H x 256W
Back
LoD0 Triangle Limit: 550
LoD1 Triangle Limit: 450
Texture Size: 256H x 256W
Belt
LoD0 Triangle Limit: 1150
LoD1 Triangle Limit: 550
Texture Size: 256H x 256W
The neck and belt had higher than expected poly limits, so this excites me. Looking at the source files, I noticed that there are a few areas of the lich that absolutely need to be covered when designing items, as the underlying texture is meant to be hidden. Namely the neck, the lower legs, and the forearms.
Concept in 2d before jumping into 3d
Concept art has never been my strong suit, so it is something that I have avoided pretty heavily until....now! It may not be the prettiest, but I've got a much better idea of the direction that I want to head now. I know that I'd like to be pretty damn different than the sets already available, but also to keep in mind their strong aspects.
As far as this set goes, I'd like to play with magical rope as a big element, hence the dangling double-noose around the liches neck. Having some broken cuffs around the wrist will give me the opportunity to get a bit of asymetry in there with the number of links attached, and to bring an element that was originally around the neck to a different part of the body. The wrappings around the forearm will be a relative area of rest since it will mostly be normal detail. Currently, there is also the 3 large cuffs around the waist and lower legs. I am toying with the idea of foregoing chains alltogether for this magical ropey stuff, but I'm afreaid of loosing that nice darker grey in the colour scheme.
I'm not completely sold on the back spikey bits yet, so im going to play with more sillhouets there and testing out different size gradients. Currently they are pretty much all the same size, giving it an arachnid-y look, but I think it can be pushed further.
The demon horns are just for shits, but I do think a small red element on the head could be interesting.
Anyway, this is where I'm at. I'll have some more concepts done up before I progress any further. I really want to nail this down a bit before I open up a 3d package. Any feedback would be greatly appreciated.
3 has the "smoothest" and most appealing curves for me, and 4 is really unique due to the really under exaggerated silhouette. It's almost an understatement that might complement the character nicely. I'd be interested to see how it will look from a game view.
I'm excited to see more progress with the Lich and I definitely think silhouette 1 is your best direction because it's less subtle and will be more pronounced in-game.
I'm thinking the ropes are a little bland at the moment silhouette-wise, so some knots are definitely in order there.
As for exporting, this is for blender users https://developer.valvesoftware.com/wiki/Blender_SMD_Tools and its a 1-click export system.
For your lich set crit - it maybe a little late, but since you're not neck deep into the project yet I'll say it. Perhaps you should swap the shoulders and belt design, chain for shoulders and rope for the belt. I know that the default items are chained shoulders but i think it'll work better with your concept. The noose around his neck seemed out of place to me.
of the silhouettes I liked #5 the most. gave him a really interesting Hindu look
Feels like the lines are competing a bit in the middle section. where you have the downward facing spikes on the back mixing with the long curvy shape of the rope. I think maybe you should just take the kink out of the the bottom spike to keep a consistent shape language in that area. (or use the 5th design with 2 upward pointing shapes)
also if your feeling frisky, you could add some sexy pentagram rope bondage to his chest http://puu.sh/6Xcrp.jpg SFW