These last few days we’ve been working some more on a new art direction for Gateway…
No stone was left unturned, and while the results are encouraging, there’s still a long way to go!
We tried to address the fact that the edges are not very visible. For example, on the next (old screenshot), you almost can’t see there’s a ledge behind the ship:
So we started by trying to dirty the edges:
And then we tweaked the effect a bit:
It helps, but it doesn’t fix it completely…
At the same time, we were also working on a new art direction for the in-game models. We started with tree with less noise than the previous one, but keeping the idea of using cubes as a basis:
It’s a bit confusing in terms of visual perception, though… We also tried using more traditional low-polygon models:
Although they are more easy to read visually, they don’t have as much character as the previous ones…
Going back to the edge issue, I decided to build a edge detection filter that detects ledges and other terrain features. This is the result of a Sobel filter applied with some weights to the normal/depth buffer:
I think I’m making a game with these visuals one day, love it!
Combining the game view with the edge algorithm, and dirtying the terrain a bit, we get this:
This is much easier to read and helps the terrain pop-out, so this is probably what we’ll use as a basis for the terrain.
The tree work carried on in parallel, and we tried a tree made out of cubes, but with much less resolution:
This is easier to read, and it’s kind of midway between the low poly and the cube-based geometry that we started from…
I started then to tackle the rest of the vegetation of the game, replacing the vertex-animated billboards for actual geometry for the blades of grass:
Quite happy with the results here, so we’ll probably keep this as well!
Another iteration of the low-poly tree came in as well:
While work continued in trying to find “THE” tree, I’ve moved on to the river flow (which had a lot of fail cases), and trying to work on the acid cloud effect. This is the starting point:
From here, I ran two separate experiments: using untextured quads, and using cubes as a basis. The quads experiment went rather well:
The cube experiments didn’t go so well… It doesn’t look like gas, but at the same time, I think there’s something there that can be explored further:
So, where we are now?
I think we got the terrain, water flow and small vegetation right, but we’re having a big argument if to go with the cube-based trees or the low-poly ones. This will influence the style of everything else, so it’s rather important we get it right. The cube tree has more personality in my opinion, but it might be hard making the game character in that fashion, while the low-poly is cheaper to produce/render, and may help to the readability of the game…
Regarding the acid cloud, I still feel there’s something to be done about the cube-fog, but currently I don’t have many ideas… Maybe something will hit me during the weekend!
I have a video of it in movement (and the water flow system as well):
The quality is not very good, but it’s enough to have a general idea…
So lately we’ve been using our (little) free time to work on some ideas for Gateway’s new art direction, and although nothing is finalized yet, we’re starting to have some ideas on where we want to take this.
We decided to start with the alien biome, since it’s the more… well, alien, and we think it’s a good way to start the work.
So, this is the current look to the game:
There are several big problems there: the color palette is too chaotic, you can’t really tell the difference between high places and low places (for example, check that part behind the ship, it has a drop there you can’t really see), the art styles are inconsistent (low poly models on the crystals and ship, sprites on the enemies and player, voxels for the terrain), and terrible color palette (for example, check the image above at 160 pixels width):
The yellows take over the scene completely, which is silly for an non-interactive object.
So, we started by playing around with the terrain shader to get a more consistent look, starting by aligning the texture to a voxel grid… We imagined at 25cm x 25cm x 25cm voxel size (currently it’s set to 1m x 25cm x 1m). This way, every single voxel has a single color. The results looked better in terms of less image noise:
There’s no ambient occlusion is that image, by the way. We then played with the voxel size a bit more and with the texture itself, lowering contrast, number of colors and saturation in general:
First image has no ambient occlusion, the second does… And we feel this looks way better, even if the texture still needs loads of work.
After a bit more texture tweaks, we decided to try changing the crystals and ship for voxel version of themselves, using Qubicle’s automatic tools to do the conversion, but we’re not happy with the result. To have any decent results, we have to use a small voxel size, which adds to the image noise, and we’re trying to go the other way. We also added a depth of field effect, and the results there were quite good, it really gives a bit more depth to the scene.
So I thought maybe the problem was the automatic conversion of model to voxel, so I asked the artist to convert the alien tree to voxels… After some problems with the 3ds importer I coded, we got the model in the game… And the results were completely underwhelming…
As I expected, to have any decent detail, the voxels have to be very small, which look good in some positions, but in most it just looks noisy…
So we kind of abandoned the idea of using voxel models, and we’re looking for other options (probably low-poly models, which will require some work to get coherent with the voxel terrain).
While thinking on solutions, we decided to do a small detour and work on the acid cloud effect, try to make it more coherent with everything:
Just pixelated everything a bit, really… Still not good, but we have some ideas (that require a lot of coding) we want to try out (namely a particle system that uses cubes instead of billboards).
At this point, we only had two things that we were happy about: the terrain tweaks (the color-per-voxel part of it) and the depth of field effect. And we weren’t any closer to finding an art identity for the game, that would lead to the rest of the visual overhaul.
I then had an idea (in the car drive to work): what about if I added some noise to the terrain vertices, trying to break the rigidness of the voxels, while keeping it in plain view?
We changed the terrain shader to accommodate this, and voilá:
Suddenly, we’re happy! Although it still needs a lot of work, it shows us a possible path to the future visuals, using this kind of “distorted, boxy” feel to it. Even the tree I was completely unhappy with looks a tad better in this context:
Next step is we’re going to take this baseline and work on it some more, try to replace all 3d objects by low-poly models that work nicely with this type of terrain and see where that take us!
A long way to go, but at least we’ve found the entrance, we think!
I know, I know, a long time before some news on Gateway, so here’s a post to rectify that!
So, last time we talked, I’d gone to PAX South… After that, Gateway was a bit on standby since work on Strikers Edge has been cranked up, getting ready for release (should be a June launch date).
Anyway, early April I went to the Game Finance Market in London, trying to pitch Gateway to financiers and publishers…
Overall, it went fairly well… People seemed to like the gameplay, the business plan, the ideas and concept… What people didn’t like was how it looked.
We knew the game had issues with the visuals, mainly a huge lack of consistency: voxel terrain, low-poly models for some stuff, sprites for dynamic things, weird color-schemes, etc. We were kind of expecting that would not be important for the pitch phase…
Guess what? It is…
So now we need to work on the visuals to be able to go to the next step…
The main problem there is that we’re currently extremely busy with Strikers Edge, so I don’t have much time to do the actual “art direction” work that we require…
The other problem is that we’re considering going to Unity for Gateway (makes it easier to find people to work with us, the porting also becomes simpler and there’s a lot of other advantages, even if we lose the massive performance of pure C++), which means we don’t want to invest a lot in developing the current prototype, but we kind of have to, to be able to demonstrate the final art style…
So this is where we are at the moment, trying to balance Strikers Edge (that’s the important work right now) and Gateway development (which needs a lot of work to get the visuals right).
Strikers Edge has just got into submission for PS4, so code is more or less frozen, which means I get a bit more time to work on Gateway, so hopefully I’ll have some new stuff soon!
So last week I was at PAX South…
I was there to show Strikers Edge to the American audience for the first time, and it was very cool… People seem to enjoy it and we never had a moment’s rest!
I also took the opportunity (and since we could spare the space in our booth) to show Gateway as well, and get some more feedback on the game.
Gateway was also nicely received… Got really nice feedback, especially from couples that wanted something they both could play in couch mode!
It was all a bit hectic, since there was only two of us to greet, talk to the visitors and explain stuff, and two games, but overall it went very well!
So now I’m back in Portugal, coding away on Strikers Edge… The last 10% of a game are really the hardest part!
Yesterday was the the prize cerimoney for the Playstation Talents 2016, and Gateway was one of the 10 finalists, along with some others (links in the bottom).
It was a nice ceremony, with improv comedy and a lot of awards… Unfortunately, Gateway didn’t win “Best Game”, but it did win “Best Use of the Playstation Platform”:
I got it from the hands of last year’s winner, the Strikers Edge team (which is super-cool, since I’m working with them!)…
Besides that, it was nominated for most categories, which is very cool!
The winner was VRock, a VR Band Hero from Game Studio 78. As a big Guitar Hero/Band Hero fan, I’m looking forward to see where that goes from here on!
The other winners were:
Best Child’s Game: An Aztec Tale (Cake Collective)
Press Award: Shutix (Indot Game Studio)
Best Art: Hell Keeper (Badaguedes)
Innovation: VRock (Game Studio 78)
All in all, it was a great day, and it was quite cool talking with all the finalists and taking another step into the gamedev adventure in Portugal!
Big thanks to everybody that supported us, to Sony PlayStation Iberia and Portugal, the press, and all the other finalists!
Next week, I’ll be at PAX South, showing Strikers Edge, and hopefully I’ll have some time to show Gateway to the public and publishers, so wish me luck!
Below you can see the trailers of the other finalists, in case you’re interested… I’m missing a couple of them (VRock and Krazy Train) because I couldn’t find a YouTube trailer for them…
Work has almost stalled to a halt after ComicCon, due to the beta of Strikers Edge and Christmas…
We’ve been experimenting with voxel modelling tools, and the idea of switching the characters from billboards to voxel models instead:
This is still experimental, but the idea was to gives more “volume” to the characters, and the ability of easily customize our characters (and maybe even change weapons, etc). The results are neat so far, but we’ll have to see it in game to really reach some conclusions…
Anyway, here’s the really cool Fun Punch Christmas card:
So Merry Christmas and a great New Year from both Fun Punch Games and Spellcaster Studios! For us, we’re sure it’s going to be awesome!
It was very tiring, but very cool!
The fixes we’ve done on Gateway from the feedback we got at the Lisboa Games Week really paid off, we had almost none negative feedback, which feels great!
This was our display space:
Our master of public communication:
A lot of people, specially on Saturday:
And a group photo with all the portuguese devs that were showing off their games, in front of our small but awesome booth!
A big thank goes out to the organization for making this a reality, Nerd Monkeys for sponsoring the event, all the devs for the support and great conversations all around, and all the visitors that spent some time checking our games out!
We come back really energized to continue work on Gateway, full of new and great ideas, so stay tuned for more updates!
So tomorrow we’re on our way to ComicCon Portugal!
We’re going to show Gateway again, in the indie games area, hopefully get more feedback.
This new version includes a few things people suggested at the Lisboa Games Week, and has some more features and balancing on it, so even if you checked it out on LGW, come take a look at this version to see how we’re progressing!
We’ll be there from Thursday until Sunday, from 10am to 10pm!
So today I took the day off so I could work on Gateway…
Most of the work was small tweaks based on feedback I got at the Lisboa Games Week, to prepare an even better demo for the last round of Playstation Talents and for ComicCon Portugal next week.
I noticed in the demo that most people didn’t catch any of the alloy and health packs that dropped from the enemies that you killed… There’s two possible reasons: one, they don’t see the use (that fits with the alloy, but not much with the health packs), and two, they don’t see them at all because of the noise in the scenery. In case the problem is this last one, I added some code for the items to bob up and down, and for them to have a glowing outline (with a neat shader). They are way more visible now, and on ComicCon I’ll be able to test my theory on this…
I also changed the stun mechanics on the game… It was a bit too overpowered, in my opinion… Now, when you hit an enemy with a weapon that can stun, it will increase a stun-meter, and when that meter reaches the top, it will trigger the actual stun, that will last for a few seconds (with a visible effect so you know it is in effect), and then has a cooldown so you can’t keep stunning indefinitely. I think this works better, looks cleaner and it gives rise to a few more potential mechanics in the future (for example, if two players are blasting the same enemy with a stun gun, it will stun faster). The implementation of this wasn’t as good as I’d like, but for the immediate purpose is good enough.
I’ve also tweaked the bosses a lot… Because of the birth of the game as a single-player game, the bosses were to be “designed” and not procedurally generated… Now they are procedurally generated, even if currently it’s just a glimpse of what we will do in the actual game. They can have different payloads, shields, and mechanics as splitting into more enemies when they are destroyed… When I showed the game at Lisboa Games Week, people would breeze through the boss as if it was just another enemy, you only felt that he was a boss because he was bigger than the rest, had a bit more health and was standing in the middle of a temple… Not anymore!
Tomorrow I’ll also be working on this, mostly just polishing, adding a score system (I think that might work in a demo environment, so people have a feeling of having an objective), and a game timer so that sessions have a fixed time.
And it was great!
I got a lot of useful feedback, a lot of people came to try the game (and thanks everybody that voted for me!), and most people seem to have genuine fun with the demo!
Some photos of the event:
The event itself was very nice, with the usual problems on these sort of things… the section where we had our game for show was really hidden behind the other exhibitors, so we had to go and fetch people to play the game… On Sunday, we could clear some space around the display and we didn’t stop all the day, with constant visitors… According to my math, we showed the game to almost 100 people total!
Special note to the Indie Dome, the indie-only part of the event, which was great, a lot of awesome work on display there, and the part reserved for retro gaming was also pretty nice!
So, what’s next for Gateway?
Well, we’ll have to wait for the results of the Playstation Talents awards, but in the meantime we’re pitching the game to potential publishers, doing small fixes based on the received feedback, and preparing to show the game at GameCon Portugal (from 8th to 11th December)!
We’ll keep you guys updated!