• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

GAF Indie Game Development Thread 2: High Res Work for Low Res Pay

Status
Not open for further replies.
No art progress to show this week but still did something pretty cool today. About half the team, including myself, went to a middle school to talk to the students a little bit about game development, our game, and our company. It was great seeing kids who were pretty interested in game dev and I was surprised to hear that a group of 8th graders were already working on getting a game together. Even had a couple of kids who knew what Unity and Unreal were. Also got a lot of interest in our game too, which was great to see in person.
 

Minamu

Member
This weekend I'm gonna be trying to make some original level designs as a design test for a potential employer :D It's gonna be great fun, I'm sure of it :)
 

Kalentan

Member
Hey, anyone who works with pixel art able to give me some suggestions on this?

So I have this map I just started working on (so subject to change)

e6qc.png

And it's supposed like this:


And I'm not sure if I should just avoid anything from that angle all together or if there is a good way to convey that you're under a roof still on the top picture.

I would greatly appreciate someones feedback/suggestion on this.

But instead of exiting the temple interior from the south you exit from the north.

Edit: Attempted a 'dithering'(?) effect by suggestion on the Game Maker Discord.

 

Minamu

Member
In Unity, is it possible somehow to make the lights NOT go through 1-sided planes? My outer walls only have textures on the inside, so I can see the map from outside while working (and to save on geometry), but that makes it so sunlight goes right through the walls and messes up the shadowing. I've tried changing my objects to two-sided in the shadow setting but it's not really perfect.
 
In Unity, is it possible somehow to make the lights NOT go through 1-sided planes? My outer walls only have textures on the inside, so I can see the map from outside while working (and to save on geometry), but that makes it so sunlight goes right through the walls and messes up the shadowing. I've tried changing my objects to two-sided in the shadow setting but it's not really perfect.

Just curious, how many lights are in your scene and how is it set up? Sounds like just one directional light with an interior but I just want to be sure.

If it is just an interior, you can make a low-poly capsule that surrounds your interior and set it to Shadows Only. Makes it invisible while also blocking light.

One other question, are you using shadows on your sunlight and how strong are they? I've had an issue where I was setting up lights in different rooms of an apartment and if the shadow strength wasn't set to 1, light would 'bleed' through the walls.
 

missile

Member
Not much news from my side, but here is something I want to experiment with a
bit further;

9Sa7c8T.gif


Here I'm actually trying to solve the full (Kajiya) rendering equation using a
quadrature formula instead of any Monte Carlo techniques whatsoever to compute
the reflected light in a given direction given the incoming light over the
hemisphere for each point on a surface. So I wrote a quadrature hemisphere
integrator and plugged it into said equation getting the picture above. For
the sake of simplicity all surfaces are lambertian, no diffuse intereflection
and no distance attenuation, yet. The lighting looks a bit odd. That's because
the hemisphere was approximated with only 16 patches (more would be dead slow
atm, but I'm working on it). Basically I want to use the hemisphere integrator
to check whether some of my to-be-made custom made BRDFs will preserve energy.
Yeah, I couldn't resist plugging the integrator into the rendering equation. ;)
More to come.
 

Minamu

Member
Just curious, how many lights are in your scene and how is it set up? Sounds like just one directional light with an interior but I just want to be sure.

If it is just an interior, you can make a low-poly capsule that surrounds your interior and set it to Shadows Only. Makes it invisible while also blocking light.

One other question, are you using shadows on your sunlight and how strong are they? I've had an issue where I was setting up lights in different rooms of an apartment and if the shadow strength wasn't set to 1, light would 'bleed' through the walls.
I only have the one directional light the scene starts with, yeah. I do have a small exterior courtyard, but most of the map is indoors. The problem looks like this:


The shadows are using standard settings, I've tried changing the bias and normal bias though. The capsule thing kinda works yeah, good idea. A bit of a hassle though :D
 
I only have the one directional light the scene starts with, yeah. I do have a small exterior courtyard, but most of the map is indoors. The problem looks like this:



The shadows are using standard settings, I've tried changing the bias and normal bias though. The capsule thing kinda works yeah, good idea. A bit of a hassle though :D

Yeah, it can be kind of annoying dealing with lighting in Unity. A quicker solution for your interior area could just be a large plane facing towards the sunlight that would cast a shadow over it all, if you don't use any windows. However, if it does get more complex, you are better off custom modeling your shadow casters.

I don't know why two-sided wouldn't work in this situation, but I haven't used that too often to know why.
 

Minamu

Member
Yeah, it can be kind of annoying dealing with lighting in Unity. A quicker solution for your interior area could just be a large plane facing towards the sunlight that would cast a shadow over it all, if you don't use any windows. However, if it does get more complex, you are better off custom modeling your shadow casters.

I don't know why two-sided wouldn't work in this situation, but I haven't used that too often to know why.
I have a seamless indoor/outdoor scene, and some windows, unfortunately. I think I either have to go back to having solid walls or live with this issue.
 

JackelZXA

Member
Hey, congrats for doing the test all the way too! And who knows - I only got their message a few days after they said on twitter their search was over!

Since you're asking for critiques, here are my thoughts, though keep in mind that's really more my subjective opinion and I'd be curious to hear what an actual seasoned animator has to say (Noogy, where art thou?)

I see what you were going for with a roundhouse kick, but I feel the windup lacks in hip motion to prepare the swing (you basically use your first three frames to slowly raise the right leg while keeping the hips fixed before snapping to a new hip motion on the fourth - I'd probably have recoiled the hip further back on 1-2, raise the leg and trail the hip to show weight on 3-4 then use 5 to do a smear/transition/fast snap movement before the key). As it is the windup feels a bit slow and the resulting kick lacks a bit in impact, as if she were caressing with the kick rather than hitting with it. The followup and return to idle goes a bit fast for me, and I think it could have use 3 to 4 more frames to make it cleaner if you're not smearing it (and even then) - in fact, I had considered a roundhouse myself, and thought the imposed 12 frame limit was a bit short for it, which is why I went for a more straightforward straight kick. Still, for a roundhouse, I think you spend a bit too much time with still hips on 7-8, as I feel the whole body should follow the kick to accentuate the impact. Basically, I'd think the kick is only the result of the body swirling, so you would have the torso/hips moving first, then have the leg kind of trailing behind (though with a bit more force than that :-D), whereas your kick is driving the whole motion and the body doesn't react to it until 8 (arguably 9), leaving you only four frames for a return to idle from a back view, which is... not a lot. You even use the 12th frame as a very small transition before the idle, when I think you might need something a lot more expressive due to how few frames you can spend on that, maybe an exaggerated crouching motion to show the effect of her getting back on the ground before standing up again. I'd suggest going much faster on 7--9, and reduce that motion to maybe a single smearing frame (you'd have your "pre-impact" on 5, clear hit on 6, then "trail" on 7), use 8-9 for a very quick turnaround, then 10 slight squat, 11 squat trail/weight recovery, and 12 stand up/return to idle (and use the belt trails to do a whole clean circle arc on 8-12 to compensate for the snaps). Of course, that sounds nice in theory, but I'd be hard pressed to actually do it :-D

I don't have much to add to the other damage+fall motion as I think yours is better looking than mine :-D

Hope you'll get the call!

Cool to see how two gaffers applied for the same job.

Yeah thanks for the feedback. Those are good suggestions and if I had more time to work on it I'd totally do another version of it trying out all that you suggested. I didnt get a reply all week so I'm betting I didn't get it, but thanks for the kudos. (Kind of a bummer but at least i got some good animation work i can add to my portfolio) I'm basically all-consumed with nier auto-tomato when I'm not doing my day job, so most of my other stuff is kind of on hold until i get those endings!
 
I'm a developer looking for a 2d pixel artist/animator to team with. I have an idea for a game that I would like to develop in Unity and bring to PC and consoles. The best way that I can describe it (in a sentence) is Minecraft meets Towerfall. Cooperative PvE. The design itself encourages multiple replay. I'm also open to input on the design. We would do profit sharing arrangement, but I have money to invest in getting us started and purchasing dev kits if necessary. PM me if interested.
 

Minamu

Member
TextMesh Pro is pretty great and I'm glad Unity bought it out
I might've done it wrong, but after installing the free version, I couldn't manage to create a new assset from a previously existing font file. The generation button just won't do anything :/ So I gave up.

Edit: Completely unrelated question: I have this really old UDK project on my disk, since 2012, and I have no idea how to "install" it, since it comes from another system. I should have all assets and code files etc, it's a complete project, and it should compile perfectly once I get it up and running. But I haven't been able to figure out where each dev folder should go so UDK finds them :/ Also, if I install the very latest/last UDK version, is it likely that my older code will work? I don't remember the exact UDK version installed at school 5 years ago :lol Can anyone help with this?
 

missile

Member
Guys, I hope you don't mind me spoiling all your hires images! xD

Some more progress on fully integrating the rendering equation;

qCKtKOD.gif


The resolution needs to be kept so low to keep it animated at all. xD
It's dead slow (not optimized the slightest), but it gives at least *some*
impression of the illumination.


Edit: Finally some hires renderings. That's how it looks if rendered
out;

S4w8mWt.png

HI (hemisphere integrator, quadrature midpoint)

tIL3dAA.png

MC (uniform random samples over the hemisphere)

GNomJX2.png

difference

For the first picture the HI integrator partitions the hemisphere into 40.000
patches producing a noise free image as it should. For comparison I took my MC
integrator and have spent 40.000 random samples per pixel to match the HI one.
As it seems, the MC integrator (random samples) matches the HI visually. The
two pictures are almost identical minus the massive anti-aliasing inherent to
MC integration. However, more sophisticated scenes will show a greater
difference esp. if the light becomes of higher frequency over the hemisphere.
Similar to MC importance sampling, I will try to track these higher
frequencies with HI by either tessellating the patches as needed or use a
higher quadrature formula (or both) depending on some error estimator, but may
likewise lower the amount patches for the secondary++ bounces to keep the
computational cost within limits. Will see. The slight banding you see in the
pinkish area is due to my current tone-mapper in use (it cuts off a little bit
hard).
 
Anyone with Unity 2D experience know if there's a simple thing I'm doing wrong here. I'm having a hard time getting a rigidbody2d to work flawlessly. While basic movement and jumping works fine, I find if I jump from a high place, I tend to randomly fall into my floors.

Like, 70% of the time it's ok, but far too frequently even smaller jumps just make me pass right into the floor and come to a stop deep inside of it. I have other issues where I'll jump into a wall and it'll be fine, then perform a wall jump off of it into another wall and just go too far into the wall and get stuck.

Really love a lot of Unity's UI, but working with this stuff is making it super tempting to just go back to Game Maker where my code worked great lol.
 

LordRaptor

Member
Anyone with Unity 2D experience know if there's a simple thing I'm doing wrong here. I'm having a hard time getting a rigidbody2d to work flawlessly. While basic movement and jumping works fine, I find if I jump from a high place, I tend to randomly fall into my floors.

You need to increase your physics timestep to more accurately deal with fast moving collisions, and / or probably want to set your 'moving' rigidbody properties to continuous instead of discreet and never sleep and make sure you're flagging stuff like scenery as static.

Honestly, you are probably better off not using the built in 2D physics for something like a platformer and to use raycasts instead - which I'm pretty sure is how GameMaker does it natively.

e: Also if you're familiar at all with the 3D character controller, 5.6 (which should be out soon) is adding a new 2D character controller which will give you nicer fine tuning of where collisions are occurring like the 3D character controller does, where you can get what pats of the controller are colliding, and with what, so it will be easier to determine things like if touching a wall but not grounded
 

mooooose

Member
Hey does anyone have experience using UNet or doing any network multiplayer programming? I have a game that is heavily physics/latency based, so I've been researching for awhile and ready to start.

It looks like I need to use an authoritative server and client side prediction to achieve what I need.

https://www.assetstore.unity3d.com/en/#!/content/14984

This asset package seems great, it's based off of Steam's documented approaches to network programming. But it seems to be a wrapper for UNet, so I'm skeptical.

What is it like using UNet? Is it a good library? What does it cost and how does it scale?

edit - I should preface my game is built in Unity but it's still simple enough I'd be willing to switch engines to one with a better stack for network games.
 

Pehesse

Member
About to launch on Steam in 5 minutes. Freaking out a bit. The moment it hits I'm about to be swamped by either support requests or extreme sadness.

Hope all went well for you!

Guys, I hope you don't mind me spoiling all your hires images! xD

Hey, as far as I'm concerned, your "low res" stuff is more fascinating to look at than plenty of high res out there - though I still won't pretend I understand any of the Science behind any of it :-D
 

missile

Member
... Hey, as far as I'm concerned, your "low res" stuff is more fascinating to look at than plenty of high res out there - though I still won't pretend I understand any of the Science behind any of it :-D
Me neither. I do what I can't. Am just throwing some math stuff around
and see what sticks. xD
 
So I have been experimenting with 2D physics on my platformer game, and as I found out, using dynamic rigidbody for the player character just doesn't work the way I would have wanted:

It handles slopes very poorly with the player skipping up and down at a reduced speed

Jumping cuts horizontal movement speed and is committed (no midair 180 turns nor speed changes). Realistic, maybe, but not the intent.

Even regular movement visibly stutters and looks like the game wasn't keeping up in performance.

So I suppose I might want to use a kinematic based rigidbody or drop it entirely, putting full control (and responsibility)

Thinking of a plan to tackle it, but if I got anything wrong feel free to remind me.

I think I am probably going to rely a lot of ray casts to ensure proper blocking every time the game logic updates. Perhaps one at the very top and another at the very bottom of the box collider towards the direction the player is facing and two pairs from the sides towards up and down... well at least I'll need to have enough so that the player doesn't clip at least.

Probably need to keep track of if the player was grounded or not? Though I think the downward ray casting likely covers it.
 

neko.works

Member
Your game looks amazing!


OMG, this is awesome. I want to play! Take my money, please

Thanks everyone :3

I'm currently considering Steam's Early Access for Light Fairytale. Basically, it would be 50% of the content available 6 months before the full launch.

Any of you guys have an experience with EA? Would you recommend it?

I think that EA would not only give me more funds to complete the game, but also valuable early feedback, and sales expectations for the full game.

Thoughts?
 

LordRaptor

Member
Even regular movement visibly stutters and looks like the game wasn't keeping up in performance.

Not to clickbait with One Weird Trick That Fixes Your Stuttering Platform Game, but you are putting your movement stuff into FixedUpdate() and not Update(), right?

The best results for anything using the built in physics system is getinputs in update, and movement code (including any camera follow code) in FixedUpdate().

But as I said above to superNESjoe, if I was building an 'old school' 2D type game, I wouldn't be using Unity physics at all (well, except raycasts which are technically part of the physics system) because none of those classic 2D titles used "physics systems" per se, and real world physics are actually pretty boring

Probably need to keep track of if the player was grounded or not? Though I think the downward ray casting likely covers it.

A grounded check is useful for all sorts of stuff, like resetting double jumps or having air-control while jumping at lower values than regular grounded movement

I'm currently considering Steam's Early Access for Light Fairytale. Basically, it would be 50% of the content available 6 months before the full launch.

...

Thoughts?

Your game looks really nice, and I assume you have some established goodwill from fans of your previous title - the bike racing game, sorry, can't remember what its called - that should show your willingness to add features as requested by fans, so you might be in a good position for EA.

Having said that - and I have no personal experience with EA as a qualifier - I have heard very mixed results about EA, as many steam users explicitly opt-out of seeing early access titles from their recommendations, and you 'waste' some of your discoverability slots as a result.

To my mind the best kind of title for Early Access is multiplayer based, because you need that constant balance - test - iterate loop, and it helps build good word of mouth that you as a developer are listening and can be trusted to take something enjoyable and keep it enjoyable.
 
Now that Rex is done, there's the dawning realization that it's going to need things like a trailer. The free video editing software out there is not great (currently struggling through OpenShot), and I have no idea how to even begin highlighting the game in something like a trailer. The way things are looking, the end result may just be a basic collection of gameplay clips with no embellishment, and hoping people get what the game's going for from that.
 
Any tips or suggestions regarding optimization for consoles? I've release two games on console with Game Maker Studio, and am currently using Unity to prototype my next game. I see a commonly expressed comment/concern that Unity games just don't carry over to console very well, and that's my main target in terms of platforms.

Any forward suggestions on that?
 

neko.works

Member
Your game looks really nice, and I assume you have some established goodwill from fans of your previous title - the bike racing game, sorry, can't remember what its called - that should show your willingness to add features as requested by fans, so you might be in a good position for EA.

Thanks! The racing game is called Super Night Riders by the way :3

Having said that - and I have no personal experience with EA as a qualifier - I have heard very mixed results about EA, as many steam users explicitly opt-out of seeing early access titles from their recommendations, and you 'waste' some of your discoverability slots as a result.

I haven't checked yet, but does switching from Early Access to full release make the game appear on the new releases list again? With the same visibility rounds? Or all the visibility happen when launching the EA only? That would be a deal breaker :(
 

Kamaki

Member
Very stylish! I'll give it a download on release.

I rarely understand what you post, I mean that in the best way possible, but I can certainly appreciate the work. Always excited to see lighting in action, I'm particularly interested in the performance.

About to launch on Steam in 5 minutes. Freaking out a bit. The moment it hits I'm about to be swamped by either support requests or extreme sadness.
I've watched a few videos of people playing this recently, looks neato. Congrats on the finishing and releasing of a game! Reviews are looking good.

Thanks everyone :3

I'm currently considering Steam's Early Access for Light Fairytale. Basically, it would be 50% of the content available 6 months before the full launch.

Any of you guys have an experience with EA? Would you recommend it?

I think that EA would not only give me more funds to complete the game, but also valuable early feedback, and sales expectations for the full game.

Thoughts?
If you're looking for feedback then I say go for it, if you're looking for a cash injection then maybe not; you run into other issues. What if it doesn't sell well, obviously fingers crossed it sells very well, are you able to justify finishing the story then?
You hear it a lot; something only getting one launch on Steam, whether it's early access or not and I guess it's worth considering what time is right to "launch".
Game looks great by the way, always nice to see your twitter posts on the game.

I've been working on a game, An Aztec Tale, and I've got to say it's starting to wear me down. Hard to go at something for so long and feel like we're not making much headway. I'm sure you can relate! I always tell myself I would post here more often and never do, this time I'm forcing myself to! Sharing is caring after all and it'll probably help with my feelings towards the project.

Saying all of that we did take part in the Playstation Talents Premios contest for Playstation Portugal and came away with a trophy so that was neato. That was the end of January and we since decided to rework the whole thing, we deemed the game wasn't fun and all that. We were aiming to get a demo out for the end of March but that's looking like a bit of a dream at this point especially with EGX Rezzed starting up tomorrow.
Hoping to court a publisher of some kind, if that's even a thing, so we're polishing it right up.

DRgU8rK.png

aEW5TAK.png
 

DemonNite

Member
I've been working on a game, An Aztec Tale, and I've got to say it's starting to wear me down. Hard to go at something for so long and feel like we're not making much headway. I'm sure you can relate! I always tell myself I would post here more often and never do, this time I'm forcing myself to! Sharing is caring after all and it'll probably help with my feelings towards the project.

Saying all of that we did take part in the Playstation Talents Premios contest for Playstation Portugal and came away with a trophy so that was neato. That was the end of January and we since decided to rework the whole thing, we deemed the game wasn't fun and all that. We were aiming to get a demo out for the end of March but that's looking like a bit of a dream at this point especially with EGX Rezzed starting up tomorrow.
Hoping to court a publisher of some kind, if that's even a thing, so we're polishing it right up.

DRgU8rK.png

aEW5TAK.png

reminds me of Mega Lo Mania! :D looks good
 

kkiablo

Neo Member
Here's our game Protoform. We've been working on this for a little over half a year now. It's a 3D stealth puzzle game.

2017033001.jpg

2017033002.jpg

2017033006.jpg

2017033007.jpg


There's been some internal debate as to whether this game's controls are too complicated and the puzzle designs too hard. So we thought to just throw the game out there and let everyone try it and give us some feedback.

Here's the demo download link: http://www.indiedb.com/games/protoform/downloads/protoform
 

DemonNite

Member
Finally been working in the main dungeons again in my game but before that I did some tidying up on the Butcher guy. Mainly added a Beast mode phase that actually doubles him up in size... he wasn't big enough previously :)


However, the biggest improvement across the whole game is adding Inverse Kinematics Look At to all the NPC characters.

RedLinearGoosefish.gif


I think this simple addition has made a HUGE difference in giving these characters a bit more life.

Full scene with new Beast mode phase and IK Look At implemented.

https://www.youtube.com/watch?v=h5FOsmtIc-I
 
Heh, now movement in my platformer is bugged but I see signs of functionality. Saw and fixed a bunch of math and logic errors and I am crossing my finfers that the changes work. Need to copy back the scripts tonight. Not having access to Unity during most of your day kind of slows you down.

By last night, the main character still can't jump, but he falls fine. A bit fast though, so there's still room for improvement.

For detecting whether a sloped segment is walkable, I think it should be fine to attempt two linecasts with a known distance (like a linecast that starts 1 unit taller and 1 unit farther in range) and compare the angle? Or is there a faster way? Also probably would need a method to make the character walk smoothly down a slope too. (Game does not have slopes steeper than 45 degrees for now.)

I think I probably should also put in code to push the character up so that his feet isn't merged into the ground... 😅 sometimes that might happen and I think I will need it once sloped movement is in.
 

LordRaptor

Member
Heh, now movement in my platformer is bugged but I see signs of functionality.

I checked my bookmarks folder, because I'm sure I read a really good tutorial on rolling your own raycast based 2D physics, and I found this link which you'll probably find helpful (and its part of a series with multiple entries)
 

Kamaki

Member
reminds me of Mega Lo Mania! :D looks good
Never heard of it before, looking it up this seems like a rad game. I imagine the similarities end at looking at units run around though.

There's been some internal debate as to whether this game's controls are too complicated and the puzzle designs too hard. So we thought to just throw the game out there and let everyone try it and give us some feedback.
Thought I would give this a go and offer some feedback. Visually I think it looks good; the character feels physical, touching walls and deliberate way she can drop off a ledge turn around an grab a cube.
Sadly I didn't get to see any of the cooler looking environments you posted. I was completely stumped.
JjbLwlw.jpg

Got as far as this room, so probably the first real puzzle. It looks like you need to get the first cube through the first gate to unlock the door but I honestly couldn't work out how.
My greatest frustration came from the turrets, since there's not visual indicator for their hearing range, or perhaps more correctly no indication of how much noise you're making, I found myself repeatedly thinking I was safe only for them to laser me to death.

Let me know how to get past this room and I'll give it another go!
 

RanciscOS

Neo Member
Hi everyone!

I'm currently starting a small game studio and our target is to release a videogame for mobile devices, at least for know since we consider it to be cheaper. However, we have little knowledge about how the pipeline used for videogames. We know what we need, sort of, like 2D artists, animators and a programmer too of course, but we are lost at where to begin.

Does anyone know a website or book to give us some idea? Any advice would help us a lot too!
 
Well, here's a big one. Decided to pass on building an Active-Abilities system for the moment in favor of working on the AI side of things until they're caught up.

Hotline Miami-ish camera, basic AI, and Noise (No actual audio yet)

Pardon the excess usage of annotations stating the obvious. I just wanted to be able to explain what's exactly going on to any old layman who doesn't understand a thing about games and get some last mileage out of them before Google murders them next month in favor of "Cards" bullshit that isn't even a quarter useful in the name of High Overlord Mobile... Y'know, rather than just making annotations work on mobile.

Just wish my progress wasn't so slow right now, but in between working at a "proper" job full-time and still learning the ins and outs of GML, the actual time I've been able to dedicate to programming has been tight.

AI gets a bit daft around the lower left corners of walls right now. Anyone familiar with the subject have any advice/pointers concerning setting up pathfinding and such for NPCs in GameMaker?
 
Status
Not open for further replies.
Top Bottom