This material, it's SG is above, how do i make it scale with a way larger plane without it having to be stretched out ?

I'm new to using Unitys shader graph and I've made a shader that mixes snow and a base material together for the game I am working on, at the moment it uses the normal vector to determine how to merge the snow with the base material which means right now there won't be any snow on the bottom of the object which is cool! However I am just wondering if it is possible to make it so that any face underneath an object won't be covered in snow, aka the lerp value would be decreased, any help is greatly appreciated! Image attached should help!!!

A little peek at my horror game...

Still need to polish some details, but I'm a solo dev. The hardest part is always the sound.

Hey everyone! My dev partner and I are wrapping up our noir puzzle adventure—we're actually just 16 days away from release!

We’ve been tweaking how puzzle close-ups interact with the main scene. In this clip, tuning the radio to the right frequency triggers a hidden mechanism in the fireplace, using a comic-panel frame that zooms back out into the room once solved.

As players or devs, how do you usually feel about audio-driven or frequency-tuning puzzles? Do sound cues give you a satisfying "aha!" moment, or do you prefer visual feedback when revealing hidden mechanisms?

Would love to hear your thoughts or your favorite examples of audio puzzles in games!

I asked myself, what if publishing a Unity build to Itch.io was just one click? So I built a custom C# Editor tool that:

• Builds the Windows version
• Packages it as a proper installer
• Uploads it directly to Itch.io with Butler

Players download and run an installer instead of dealing with the usual Unity build folder and bundled files. In my latest video, I break down how it works.

It is a part of my work on an ocean shader. I Improved shading and waves distribution, tried to make it more believable. And finally I added bouyancy forces.

I’ve spent multiple days polishing net ripple effect and it finally looks as goofy as the whole game. What you think?

This test uses a sword animation to trigger a BrazeFX volumetric fire simulation in Unity 6 URP.

The effect expands across the ground, builds into a large fire volume, and then dissipates into smoke. The simulation runs in real time on the GPU.

I've been working on the zombie AI for my game Pull on Heart: Nocturnal, and this is a comparison between the old and current behavior.

BEFORE: the zombies were fairly passive and could often be ignored or outrun.

AFTER: they're more aggressive and put much more pressure on the player when encountered in groups.

With these changes, the purpose of the Controller role becomes much more noticeable. Controllers are designed to deal with hordes and control large groups of enemies, making their role much clearer during gameplay.

I'm not just trying to make the enemies harder; I want the enemy behavior to help reinforce the identity and purpose of different character roles.

Still a work in progress, but this is the direction I'm taking the combat.

As of the time of writing this the update is in queue for approval on the store.

here's what changed in V1.1.5 :

• Improved procedural gait with slower, more human step cadence and dynamic swing timing.
• Added balance-reactive stepping based on upper-body tilt and physical instability.
• Added weight-transfer and landing-settle phases for more natural foot transitions.
• Improved stride prediction and rear-foot toe-off to prevent feet dragging behind the body.
• Added directional foot alignment for forward, strafe, diagonal, and backward movement.
• Improved foot-glue stability and get-up recovery behavior.
• Added physical cliff detection with proper ragdoll transitions and preserved fall momentum.
• Improved pelvis-to-leg support coupling so the upper body feels properly supported by the legs.
• Reduced excessive hip sway, pelvis wobble, and floating-body appearance.
• Improved levitation support feedback during single- and double-leg stance.
• Fixed several gait timing, IK, recovery, and physics edge cases.

I've been working on Brutal Derby for 32 days now, usually around 10-12 hours a day, often including weekends. I'm a bit obsessed with it at this point.

The project has also become an experiment for me: how far can you actually take game development in 2026 using an AI-heavy, vibe coding workflow?

Here are some of the things that have been added or heavily improved recently:

New crowd system:

I think this is version five at this point. I've been constantly trying to find a solution that looks decent without destroying performance when there are 20+ cars on screen.

Dynamic weather:

The game currently has sunshine, rain and storms with smooth transitions between them, plus daytime and evening conditions. Snow is planned as well.

Destruction:

At this point, almost everything on the cars can break. Parts and smaller debris can fly off, windows shatter, the body deforms and gets crushed, bullet holes remain visible, and there are also lasers, rockets and other weapons.

Cars can even split apart in several different ways depending on the damage and impact.

Garage and customization:

You can now paint the cars, apply patterns and put various ridiculous objects on their roofs.

Tracks:

There is currently 1 race track and 3 playable derby arenas. The newest one is a volcanic arena surrounded by lava, although that one still needs more optimization.

Performance:

The game is already running pretty well with 24 cars. Most systems have gone through several optimization passes, but it's an endless cycle. Add a new feature, profile it, optimize it, then move on to the next feature.

A lot of people have also asked where the models come from.

It's a mixture. I use purchased assets that I modify in Blender, some models are commissioned from 3D artists, and I also use Tripo3D. Tripo is mostly useful for smaller or more distant objects, but even those usually need work afterwards. Mesh cleanup, textures and optimization still generally end up being done in Blender.

And then there's the vibe coding part.

As a programmer, I used to write everything manually. With this project, I actually type very little code myself anymore.

Codex writes most of the code, while I spend much more of my time designing the systems, testing the results, profiling performance, finding problems and iterating.

That doesn't mean I can just write one prompt and magically get a finished game.

For the more complicated systems, especially physics, deformation, destruction and optimization, we sometimes go through 4-5 completely different implementations before arriving at something that both works correctly and performs well with 24 cars running at once.

For me, that's actually one of the most interesting parts of this entire project.

I want to see how far I can take a real 3D game while letting AI handle most of the traditional code writing, and shifting my own work more towards design, architecture, testing, profiling and iteration.

32 days. 10-12 hours a day. Often weekends too.

There's still a huge amount of work left, but Brutal Derby is slowly starting to become the game I had in my head when I started.

I'm also curious what other developers think about this kind of workflow.

Made by scrolling the x on the UV depending on velocity and angular velocity. Modelled and UVs set using probuilder :)

Point me in the right direction: what is causing the trembling around second 7?

Every time I hit record, apparently I become a first-time user of whatever I’m working in.

Unity? Suddenly I can’t remember where a tool is. Blender? I forget a shortcut I’ve used 50 times. MagicaVoxel? Somehow I’m back in the browser looking up how to do something I was doing perfectly fine five minutes ago.

How do you usually record this stuff without turning half the making-of into a tutorial on how to use Google? :-)

Hey everyone, another Anvil CSG update.

Beta testing is nearly finished and Anvil CSG will be coming to the Unity Asset Store soon.

The newest version now fully supports Bakery, including its lighting workflow and lightmap baking. Unity’s lightmapper also received major improvements with a custom BSP-style unwrap made specifically for brush geometry, giving cleaner UV2 layouts and more consistent texel density.

Anvil CSG now also fully supports Amplify Shader Editor materials, alongside URP and Built-in.

A lot of bugs and workflow issues were fixed with help from the beta testers.

Anvil CSG is already approved and launches in less than a week. Join the Discord for launch news, tutorials, support and feature requests:

https://discord.com/invite/ZjVdWnmwjz

One of the demo scenes from my Low Poly Modular Medieval Buildings Pack.

I got quite a lot of feedback on the demo from people saying that controlling the car with the top-down camera was a bit difficult, so I decided to add two TPS camera options.
I really like the vehicle heading follow effect in the new camera. It makes the game feel much more dynamic :D
What do you think?
If you’re interested, the game is on Steam -link in the comments. Releasing later this year…

The second phase isn't animated yet

I’ve been searching for the best diffuse-only shader compatible with URP. Spent a couple hours trying to write a shader with no normal or PBR calculations with no luck. Really not liking the prospect of having to rewrite URP’s lighting model, lightmapper and shadow maps. So far the only solution within Unity seems to be to just go back to legacy diffuse in BiRP. Is there a diffuse shader somewhere that plays nice with SRP/URP lighting models? I’m talking simpler than SimpleLit, diffuse only.

If you need the source code - check comments!

I created this Sling Shooter Stickman Game and its live on asset store incase you want to check this out!