Hello everyone! It’s been quite awhile since I’ve written an actual ‘devlog’—we’ve still been around, but we’ve been far more active on our various social media platforms and as such, our blog has kinda fallen by the wayside. Hell, the website shows our last blog post in October of last year. Eek—might have to edit the date on that one… heh heh, er… uh.
There is probably no easy way I can describe how much progress we have made during that time, so instead, I’d like to look back at some of our various social media postings and add some context to them. I won’t lie and say that the past nine months have been easy—quite the opposite, in fact, as we’ve dealt with some of the most difficult, random and inexplicable problems we’ve ever encountered. Sure, it makes sense, the further you go, the harder it gets, as is the case with almost everything—but sitting there two years ago, wide-eyed and motivated, I severely underestimated the kinds of shit we would run into.
For starters, this was an interesting little bug we encountered upon trying to transition to a physics-based movement system.
Those are physics, all right! This was mostly a test, but we ended up not going down this route—and no, it wasn’t because of what happened to poor Bodom. (He actually never fell back down to the ground, so that’s kinda an issue)
We decided that having full physics collision and response is overkill, especially when the main character can’t jump and can’t fall off ledges. So we moved away from this little experiment, but it was good for a laugh or two.
While I took on that particular challenge, Andy took on one of his own: our main menu interface. In game development, or at least as it pertains to Bevontule, it usually takes three tries (OK, as a minimum…) to get it ‘right.’ The menu was no exception. After watching many others play through our most recent demo, both in-person and online, we realized that the menu needed a bit of TLC.
UI design is way more difficult than it might appear at first glance, and it is incredibly easy to overengineer or complicate the problem, especially when you are dealing with multiple inputs that can be used simultaneously—mouse, keyboard and controller. But, of course, showing is better than telling:
I should also add that we have been continually improving our graphics, and you’ll notice that in any videos that feature the steppe.
You can see some of this in this next video, which comes on the heel of a few lingering bugfixes to our AI, essentially improving the enemy’s ability to hurt you better. This also was an early foray into video-editing, which’ll come into play later :)
As a further result of our physics experiments, we were also able to improve our movement code to use something called ‘Root Motion.’ Briefly, this allows a character to be translated based on his actual animations. Thus, if a character is walking forward in an animation, he will move forward at exactly the rate as determined by the frame-by-frame positional changes in the animation itself. That’s maybe a little simplified, but it allows for cool stuff like this:
This Hyderangea would not be able to lunge forward in a realistic manner without root motion, as it would have previously moved at a constant velocity forward no matter how it was animating. This also leads to more realistic character movement—no more feet sliding!
I should also note that we began planning out our Greenlight campaign around early March, and with that, came the first real set of visual upgrades (unfortunately, due to time constraints, we weren’t able to get everything into the trailer)
Courtesy of Andy, we investigated several improvements, starting with volumetric lighting. This allowed us to massively enhance the ‘atmosphere’ of many of our scenes, including the ruined town of Molsha:
This also helps to provide the dusty, foggy feel as shown in the Depths beneath the Steppe:
Much of early April was spent finding (or designing) and recording the appropriate footage—we knew that our Greenlight trailer would be broken roughly into three parts: a general breakdown showing high-level, non-combat play and introductions, a large middle section focusing on combat mechanics, and a final, short section focusing on side-quests and exploration.
I won’t say that it’s the best thing ever committed to film, but I am quite proud of what we were able to come up with, having no real experience with video-editing. Luckily, thanks to an amazing free program called DaVinci Resolve, we were able to build the trailer in a reasonable amount of time. I should also mention the other awesome programs that contributed to our Greenlight campaign overall: Handbrake, GIMP, Canva and likely, many others.
We were also able to enlist the help of an amazingly talented musician, Aakaash Rao, who, despite facing very strict time constraints, produced a phenomenal track to accompany the trailer. You can check it out below!
After passing through Greenlight, we were ready to get back to good ole-fashioned development, revamping our interaction system to more robustly handle more complicated situations and quests. Since this was a purely architectural change, there isn’t much to show here, suffice it to say that the effects will be widely felt in future demos and the eventual release.
Oh yeah, we changed our company name too! We went from Liminal LLC to Multithreaded Games LLC. We think it’s a much more fitting and meaningful name :)
On the visual side of things, we were finally ready to test some of our most sweeping graphical improvements. In no particular order, we moved from Forward to Deferred rendering, added screen space shadowing, and added/edited several effects thanks to Unity’s new post-processing stack, including Ambient Occlusion, Motion Blur, Bloom.
We then revisited the world map area—as we’ve mentioned several times, this is almost its own ‘dungeon’, connecting each of the various locations in-game. This isn’t to say, however, that the world map will be pedestrian by any means, and we plan to fill it with many extra locations, treasure, lore, powerful creatures and amazing vistas. Here’s a sample of overworld movement (note that this has been improved even more since April)
These benefits crept into the battle system, allowing us to improve the look and feel of our area indicators. And we just love finding new and exciting uses for seemingly unrelated things! (Just kidding, updating to 5.6 utterly broke them, so they had to be redone :P)
We then began a two-pronged attack (it sounds way more badass when you frame it in military terms) on our skill system, working on the front and back end, respectively. Business in the front, more business in the back, I think, is how the old adage goes.
Anyhow, I focused on general skill improvements, while adding and implementing new passives, active skills and their upgrades. As of Unity 5.5, the underlying Particle framework has been vastly improved, allowing the use of ‘actual’ lighting effects without having to fake it—as such, it was impossible to resist revamping some of these effects :D
You can actually see both improvements in action in the next video, displaying a single skill upgrade for the Blade Toss skill—also note the improved trail renderer/lighting effects on the actual dagger.
However, for every victory, there is an equal and soul-crushingly blistering defeat—it was none other than the erudite philosopher Abraham Lincoln who nonchalantly uttered these words, memorialized for all time: "With great power, comes great responsibility." I now understand clearly the idea that he was trying to convey.
On the other side of the coin, our skill UI was starting to take shape. It was awesome to finally be able to equip, learn and upgrade skills in-game without using the Unity editor itself:
It should also be mentioned that, as of mid-March, we have been working closely with independent contractors to overhaul our main characters, including some that have yet to be introduced:
We’re working closely with some great partners, not only on the visual front, but on the audio side of things as well. More on that later though! :)
We also found time to implement, fix or improve the following, in no particular order:
- Massively enhanced optimization/mesh combining/atlasing/batching of most models, textures and materials
- Loading screens upon scene transitions
- New Bevontule/Multithreaded Games logos
- Saving and loading of a single save slot (for now)
- Playmaker integration (No idea how or why we survived without this for so long)
- We also implemented the mass murder of approximately 250,000 bugs
This pretty much skims over what we’ve been up to for the past nine months, and while we’ve made almost incomprehensible progress, we’re still climbing ever onwards towards the goal. And while we may not yet see the prize, we still have our eyes on it.
Over this time and really, the entirety of the project, Bevontule has morphed into something even crazier than we could have possibly imagined, but many long-fractured and disparate pieces are finally beginning to assemble.
It certainly doesn’t seem like it, but our tiny studio has been at this for around two years, a period of time that seemed inconceivably long and yet has flown by at breakneck speed. If time flies when you’re having fun, surely it soars when you’re having the time of your life building the project of your dreams.
More than two years ago, we decided to take a risk and for the longest time, had no idea whether it would pay off for us. As of now, we have spent tens of thousands of dollars of our own savings, dipped into retirement accounts, and Andy attempted to sell his own body and astonishingly succeeded, despite my best efforts at intervention. It wasn’t that he was unattractive, per se, but he could’ve earned more than a paltry five dollars.
That being said, I can’t state more resolutely how dedicated we are to seeing this game through. While I’ve never been one to rely on others for support, it has also become increasingly obvious that a large fanbase exists for this game and that we may need to enlist this support sooner rather than later.
As such, it is our intention to announce a crowdfunding campaign in the next few months. We aren’t deadset on a specific day, but we anticipate that it will occur in late September or early October. We believe that we have reached the point in our development where we can very confidently point to the thousands upon thousands of hours of work, our large backlog of development updates, five (or more!) demos, improved and final assets, as well as new and interesting character and enemy concepts. If there is a reason why you think we might be unprepared for such a task, we’d love to hear your feedback and ways in which we can further improve! :)
Our focus for the upcoming months consists of a main goal that relates to and empowers everything else: primarily, this is to release our sixth and most-likely final demo. We believe that this release will absolutely speak for itself and we are firm believers in demonstrating Bevontule’s potential outright, rather than expecting our backers to imagine it. We intend to release this demo for free and whether you’re a backer or not, you’ll still be able to play it.
To all of our fans and supporters, we appreciate it immensely! It’s been a long, winding ride and it’s about to get even more interesting.