[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Settings Home
/m/ - Mecha

Thread archived.
You cannot reply anymore.

File: project nimbus.jpg (44 KB, 620x296)
44 KB
I've been shamelessly enjoying this literal Gundam-meets-Ace Combat mashup that's on Steam.

The controls are a bit so so (honestly speaking I seem to mostly be busy managing weapons while tilting my controller in one direction), but it's surprisingly fun.

The story is also pretty fun once you get into the swing of things.
Mirai is for
I tried it but couldn't get into it. I feel like it lacks any polish whatsoever with things as simple as movement feeling very rough. I guess it made me appreciate that creating a game that seems as simple as Ace Combat might actually be a lot more difficult than it looks.
File: 1510599451854.jpg (795 KB, 3500x2644)
795 KB
795 KB JPG

Once I got practiced at actually controlling it I thought it got pretty addictive. It has a nice game flow.

I also like that it's a mecha game both has the mecha action and a proper mecha story (with the character drama inside the cockpits). You don't really get that, besides with Front Mission, Zone of the Enders, Titanfall 2 and Ace Combat (which is basically mecha).
The game felt really raw, even more raw than the trashy War Tech Fighters.
Most of the missions are too easy once you understand the combat mechanics.
The rest are almost impossible because the level/mission designers are retarded.
The graphics are not that good (I don't really care about the graphics, I would be okay with black textures and neon bright polygon edges for 60 fps) but the optimization is some lazy shit.
>inb4 your pc is shit

There's no customization, but the game gives you some mechs to pilot so its not that disappointing.
The gameplay is okay. Mostly it feels like a failed clone of Strike Suit Zero's Strike Suit Mode.
>S37 Terminator "Kasnaya Kometa"
>Andei Dmitriv
>look guys it's a popular Char expy meme alright? sorry we can't fucking get the Russian names correctly though lmao
TL:DR It's an okay game worth pirating.

How would you configure a game like Armored Core 4 on PC, anyway? In terms of controls, I mean.

I'm curious if I ever manage to get one of my "Holy fuck what have you done with the budget" game idea off the ground. In my opinion the most important part of the game is the controls - gameplay, plot, graphics, optimization, other content come second. It's probably an influence I got from playing shmups.
This game is half-finished.
90% of the weapons are fucking useless.
When an enemy goes into pursuit mode, you can fire everything you have at him for minutes and not hit him even once.
Lag out of the ass in the fortress stage.
Animu girl feels out of place.
That fucking walking stage with the girl panting.
Debug mode accessible from the start.
All levels accessible from the start.
What the fuck those options even do.
The developer's "patches" were just the whole game and all its files compiled again.


I liked cockpit mode, though.
Mirai a shit.
I liked it at first, but after a few stages, I realized the devs either don't know anything about balancing, or thought making every weapon underpowered as shit was really cool or something. I mean, the Railgun is the only weapon worth anything, and once you get funnels/particle cannons, you may use those too. But everything else is fucking useless, even missiles.
My main gripe with project Nimbus is that you generally have no idea what the fuck is hitting you, and once you get into a furball, you're gonna get raped from all directions at once.

And what the fuck are those bosses. I swear two of them are fucking impossible without god mode. I mean, I actually got them down to 10% health twice, but that was after a good 5~6 minutes of continually strafing shit and concentrating on not dying instead of hitting them back.

Someone said it above, but Mirai a shit. Yuliana the best.
>How would you configure a game like Armored Core 4 on PC, anyway? In terms of controls, I mean.
Plug in a Gamepad.
Just do what 99% of Japanese devs do when porting their games to PC these days. Tell keyboard users to fuck off.
Wasn't this game supposed to have been ported to PS4 by now?
>project circle strafing
It wouldn't be so bad if the momentum gained from boosters wasn't so huge (so good luck coming back from a maneuver quickly), and the mech didn't spend booster energy like crazy.

They've released it as Project Nimbus: Code Mirai.
File: 1512617774744.jpg (142 KB, 1600x1245)
142 KB
142 KB JPG

I dunno, I think the missiles are pretty good. It's only the decoys and the auto-missiles which I feel don't really give much feedback. Did you remember that the game has on demand bullet time? Because I didn't at first. Once I realized that was a thing (they sort of forgot to add it to the tutorial I think), the game sort of because pseudo-turn based and a lot less of a clusterfuck.


If project Nimbus teaches you anything, it's that easily accessible bullet time mode is the perfect utility for switching weapons.

So basically, if you have a lot of stuff to control, adding a bullet time mode to allow the player to pick his options correctly might be a good idea, without needing to add a pause function that sort of disrupts the gameflow (because you're still receiving feedback from the game)
Enjoyable but buggy. Considering I picked it up for...$15? I think I got my money's worth.

Action is good but needs balancing. Why do I have this stubby MG when I've got a railgun of DOOOM? What do the particle shields do.

The story was campy as all hell but it had some hard hitting points to it.

>Mirai: I'll avoid shooting the cockpit
>Andrei: BS you're treating war as a videogame
It's 2018 Anon, most PC gamers have gamepads and most console gamepads work with PCs.

I think particle shields are just fancy dummies.

I think they should have made the MG a little bit more accurate. It's optimal range seemed to be what, 800 or less? Sort of hard to hit anything at that range, and it does open you up for retaliation. At that range it did out-dps the railgun though.

The gen-4 were all a bit too samey. They should have promoted the zeta-expy to a main suit instead.
I have found that you can hit warships with damned near anything at range, though.

Also, the melee is just bad.
>They've released it as Project Nimbus: Code Mirai

Supposedly it was supposed to release in November, but I can't find it on the PS store, did they delay the western release?
I saw it on the Japanese store when I was looking for the Toaru Majtusu no Virtual On demo a couple days ago.
I'm kind of actually experimenting with different ways of doing this in my own title right now in Unreal (which is literally all placeholder graphics and gameplay, though I am gearing up late next year to find capitol and I'm already talking to other artists, given I mainly do environments not characters) and thus far what I'm realizing there are three 'main' ways of doing things:

1. Direct 1:1 mouse/camera control, with the robot chasing rotation to match, like Warthunder's Instructor mode. I quite like this. In English, you're rotating the camera and the robot revolves to zero the delta rotation to the camera's position.

2. The cursor being a relative position on screen that acts as a kind of joystick cursor -- so if you push to the right, the robot will turn right until the cursor returns to the middle. Seems like a really great idea until you realize it would need a really reliable deadzone so there isn't slight accidental rotation

3. 1:1 Headlook control to the mouse, with dampening to reduce the maximum speed. The problem here was that mouse turns get done in strokes, not in smooth sequences, which means you're actually calculating an average rotational dampening speed, and trying not to flatten out the top which would remove responsiveness in a big way. This also creates enormous problems if G-forces are involved in anyway and my game kind of does involve G-forces and high alpha turns, given its practically a macross simulator at this point.

One thing I did notice is by making some of the movement semi-automatic and doing a lot of you can automate tasks players don't want to do because they're too much of a headache. As players get better though, you hint that if they do all this stuff manually, the robot performs better; that they should know which to use when like the boost is higher. I cut down autobalance and autoboost by something like 75% doing this. Performance + Skill > Ease of use, but the early help is very important.
VR for head tracking in cockpit view, knuckle controllers to manipulate the various switches and such OR have a dual joystick and footpedal mode for the truly hardcore.
>it's that easily accessible bullet time mode is the perfect utility for switching weapons.
Doesn't bullet time make everything slower, even switching weapons?

> It's optimal range seemed to be what, 800 or less?
I think it was 500.
>>16259810 VR for head tracking in cockpit view, knuckle controllers to manipulate the various switches and such OR have a dual joystick and footpedal mode for the truly hardcore.

None of this adds anything to gameplay and serves only to slow the player down in most cases. I'm far more interested in using android and iOS devices as multifunction displays at this point, given the sheer uselessness of VR given the state of how modern games handle physics due to the limitations of single floating point precision which while fine from a distance, makes everything up close look like cartoon physics with objects passing through eachother.
I mean, for a while I did debate it but I don't have a computer up to spec to even play a VR game and the moment you're using VR for switches, you're missing the real point of it, which should be to drive the limbs of the robot in a WALDO arrangement.

The cockpit arrangement in my game doesn't have a seat or room for switches beyond the hand-controllers anyway. In terms of layout, think something like hulkbuster. You're not going to have a big hollow space for a seat there: Its just a mount-on visor, hand-switches and semi-automatic control -- which makes sense, given a robot would be responsible for its own movements and you'd just imply the direction you want to go in and what your target or object of interest is, just like a fly-by-wire system.

I mean, the idea of moving your body 1:1 with a robot is pretty dumb, given the two of you aren't even on the same scale mechanically or balance so most of your own choice of balancing action would topple the robot and the compromises you'd be making to make that work would overcomplicate the system enormously.

One thing I am looking at out of the box though is support for a Tobii Eye 4C at some point because reticles chasing vision look badass and I love the idea of using that as a designator for anti-missile defense or something, as seen in Macross.
I don't know if this is what you're talking about, but I agree that the movement interface should be separate from weapons. No point trying to overcomplicate things by having the player push buttons in a VR cockpit while you can just have a common joystick (or a custom one) with real buttons.
Yeah, but there's whats called 'time to think' where making tactical decisions requires either not being in direct manual control of a vehicle or being in cover for a short time to make a viable tactical decision.

When you introduce the chance for someone to ponder their situation and mentally regroup, it means their next five or six decisions are generally smarter.

The trick to getting good at anything which requires realtime response is bringing the time to make that understanding down and chopping it up into tiny bits and doing it in rapid intervals with very limited information at a time instead of doing a total mental sitrep that might take 3 or 4 seconds. The intervals should be during time where there is minimal action, such as being in transit between two locations but this is kind of a high level skill -- so by in effect making a player able to make the decision as they see fit in their own time you're allowing them to act as an advanced player by shifting the goalposts because a game is a simulation and doesn't need to follow the rules which govern reality.
No no, you're absolutely right. See, at the end of the day, you're following a fool's game by trying to use VR to simulate a cockpit with switches and stuff in any situation demanding high speed skilled response on par with a conventional player because the extra fidelity actually sets you back with the extra information your brain has to process.

One thing enormously ignored by game designers is that triggers are analogue devices that have three potential modifier-states (0%, middle, 100%) and that the 100% can be performed in an order (LMAX + RMAX, RMAX + LMAX, ONLY LMAX, ONLY RMAX, NEITHER) just for the 100% states, with the release states being null. By pairing modifier states with actions with the face-buttons, D-pad or sticks you could say give a fighter manual thrust-vectoring pretty easily depending on the order they're pressed in, or do trim -- with both resetting to their neutral states on release. I'm already doing this in my fighter-mode and I'm using a toggle for an autobalancer system in a softer mode to drive my Gerwalk mode and I'm having some pretty fantastic successes.

I think by contextualizing secondary control around these options, you expand the number of buttons on a controller by something like 3 or 4 fold quite comfortably and by using an Ocarina-Of-Time shift-prompt, you can make it fairly easy to know what the buttons do simply through repetition and guided exploration.

I'm actually having a harder time making it feel like a simulator using a mouse and keyboard than I am with a controller.
File: 1438462300298 - Copy.jpg (37 KB, 500x471)
37 KB

I'm not sure if bullet time slows down switching weapons. That part felt pretty instantaneous to me.

It doesn't matter though since the bullet time physically gives you more time to actually punch in the right prompts on the controller, which matters a lot in boss battles where you are being harried by those fucking funnels all the time.

I also used bullet time a lot just to look around myself, since it didn't affect the camera speed or the speed to switch target. I would also always briefly pause when closing in on a target to give me time to check my spacing before I opened fire with machine guns.

Actually trying to shoot something in bullet time was a waste of meter though, which made it sort of different from how most games do bullet time.
I prefer Armored Core 4, but Project Nimbus was kinda fun. Definitely wasn't as trash as Front Mission Evolved.
Why not focus on a HOTAS joystick? That plus eye tracking would make for some insane maneuverability. Just because a game is on PC doesn't mean that a keyboard and mouse is the best way to control it, so don't feel constrained to extensively support the control method.
It's true that the HOTAS and head-tracking crowd is usually willing to spend money on games, but it's a very very niche market. It's true that a hardcore mecha simulation could definitely finds its customers.
Shit, I'm still hunting down a copy of Steel Battalion. I'm going to have to spend close to $500 on a single game and the proprietary controller. Honestly, I'm not even that upset to be spending it since the game is super fun and the controller looks amazing.
With a bit of fiddling you can use the controller for other games as well, so it's not completely wasted.
File: muh desk.png (3.51 MB, 2048x1536)
3.51 MB
3.51 MB PNG
I used to have a Thrustmaster HOTAS Cougar. I might add a profile but to be entirely honest, I don't think the movements of human hands using the wrist and elbow can mechanically keep up with what the thumbs are capable of, even if the thumbs are less accurate.

You're fetishizing a particular style of play because you think the controls for a fighter would be optimal for a giant robot, and to be entirely honest, they wouldn't.

The eye tracking isn't going to alter mobility -- it would create motion sickness: Its going to be for special targetting and semi-automatic maneuvering (eg, if you're going to slam into a wall and you're less than 30 degrees off say an gap, the robot will correct and try to fly through the hole automatically for you using spline navigation) -- or for special emphasis of targetting when attacking an opponent. Its honestly just a niche feature I'm going to implement a a bonus, really -- not a core gameplay feature.

People have controllers. People generally do not have hotas gear. Implementing it isn't hard, its just time consuming and it means if my game is built around controller and mouse pacing, joystick users are going to cry when they can't make the shots they want or align themselves quickly enough. The answer to that is more specialized automation.

Again, most of the reasons I'm not implementing things aren't because I don't want to (I'd love to) but because I don't actually have the physical tools to do it.

My machine is currently an fx4600 shitbox with a GTX 750 and a Pentium 4, a chromebook and a 2011 Macbook Pro. I did have an i5 but it died and I'm kind of saving up to move rn.
So are you making a Macross style game? You mention gerwalk mode. Also, first or third person?
Both; I'm working for ways to emphasize that first is through some kind of visor and is direct light between you and what is happening around you; and that third-person is a computer-generated image based on heuristics and such.

There are going to be situations where you need to use your own eyes, because the AI of your robot is fooled - or where jamming is so intense that you only see sporatic blurs in 3rd person.

On the other hand, 3rd person is eventually (I've got most of the logic done) going to be capable of pulling back into a map view with some command and control tools akin to an RTS for some stages of the game.

I'm hoping to incorporate this "don't trust what you see" dynamic into a horror section of the game -- and for a sufficient ECM to totally knock out 3rd person for a few seconds akin to whiteout.

As it stands, I'm working on an emissions profile system, where your ability to see something as visible or accurate is based on a radiance value based on your calorific expenditure, available light and any modifiers (a raytrace which detects cover for example adding a -% to visibility or one which contradicts adding a positive). This is all likely to be late-game stuff after I've got the basic gameplay down and the AI working, but with luck it won't be like ARMA's baffling Terminator AI.
And it started as a Starfox fangame but the other folks dropped out so I kept developing it. I found Gerwalk was fairly fun and I made a robot mode that plays a lot like ACFA, with a lockon system like ZOE when closer for CQC -- with some proper flight behavior.

I literally bitched and moaned on this board for years that a good transforming robot sim didn't exist -- where all inertia would instantly stop for some bizzare reason when you transformed, that you'd be world-locked by axis, et, etc...

Some time last January, I started teaching myself basic scripting, got off my ass and started trying to do something about it. I'm very hesitant to show anything because I have a lot of odd features here and there that are half-finished but no gameloop or "winning conditions" yet to show them off.

desu I don't expect to be able to show you anything meaningful for at least another year, but by then it'll be really interesting and not just a bunch of youtube videos of "here is the autobalancer, here is the targetting system, here is how quickboost is handled, dynamic axis, lockon, CQC camera, docking, space movement, climbing sheer surfaces, thrust-vectoring", etc, etc which are all subject to change anyway.
I liked it until it randomly switched focus onto anime girls fighting each other out of nowhere
Fascinating. Any art assets? Need a tester? I've been involved with several games over the years, mostly as a writer and tester, but never anything SFW.
Story is all over the place.
>test this unit
>ok, now escape from this place (you're playing as another character, btw)
>now play as this completely unrelated character in a grunt suit
>activate strike freedom
>let's find peace and understandinguuuuuu
>don't aim at their cockpits
>*women screaming*
>blah blah blah peace blah mirai no tameni
>oh no moeblob 1 was shot down
>[moeblob pants nervously for 5 minutes]
>hardass russian moeblob gets upgrade
>both moeblobs join forces

what the fuck even happened to that guy/girl who escaped from the ship on the first stages
I'm a published author and for the longest time made my income entirely from Amazon book sales so in terms of writers I could go for a sounding board but we're mostly solid.

We have placeholders at the moment -- very very early versions of assets that are awful polycount, not rigged, etc, etc -- I'm mainly focused right now on making it play well.

I have a youtube account with some of the shitty testing videos. Nothing firm or cohesive -- I mainly use it to diagnose problems and get feedback on features. Like I said, its not pretty but I'm getting my ground-work together and discovering my gameloop.

Rotation is currently giving me problems which is to be expected. I've also got a tumblr which is mainly my friends making fun of me with fanart or screenshots and discussions about the project. Its honestly just a place to dump stuff and not one of those muhsocialjustice deals.


Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.