Ironically, I think a large part of the sense of scale comes from it being a vertical video. Really helps with stacking all these layers of environment on top of each other. Anyway, Rückenfigur (https://en.wikipedia.org/wiki/R%C3%BCckenfigur) style concept art where there's a tiny figure in front of a huge layered environment has been common in games for decades, the difference is that with ai we can create """concepts""" that look nearly like a playable demo.
This game looks cool, but also the trailer just showing 3 static scenes from the same angle for a minute is a sure sign that this is very very early.
In a youtube demo, the author says he has been working on this idea for a while, and was motivated to deliver after the GIF. So, kinda? Thanks for the reference.
I don’t think it’s ai because it correctly resolves both vertical and horizontal parallax with correct perspective lines applied to the scene - something ai still seems to struggle at greatly
I had an idea a little back about using hand-drawn mipmaps (.5x, 1x, 2x, etc) scales for sprites, with some method for nearest neighbor, (e.g. with a 3x scale taking pixels from the 2x and the 4x scale). I imagined I wasn't the first to have that idea, but it was just a passing thought. I wonder if OP is doing anything like that?
If you go frame by frame, the texture becomes more detailed, about 8x as many pixels. It looks like a normal LOD switch as the camera gets close to the wall. It might be a texture without enough in-between scales.
This looks really cool and I will doubtless play, if it ever comes out. Unfortunately I get Hytale vibes from this website. I don't understand the point of announcing such a project, not least with such an elaborate landing page, so early in development.
Agreed. I tend not to announce or even share plans for impending projects until I have something nominally concrete.
Interestingly there was a paper from a few years back where they found that a participant's commitment was found to diminish in the long term if they prematurely shared their goals.
Given the call to action is to wishlist the game or subscribe to their mailing list, I guess they could use the interest to validate their idea or even to try and get funding.
The standard marketing advice in game development is to announce your game and launch your Steam page as early as possible to start collecting wishlists. In fact, considering how much footage and gameplay they are showing off I would say they are announcing it too late. So I don't agree that it's too early in the development compared to other indie titles.
It's to give yourself some form of accountability if you have a strong habit of never completing projects. When you've made a website for it, suddenly it feels like something that you need to work on.
I appreciate the work but as someone who grew up playing these games and quake and blood and all those novalogic titles, it was never about the graphics. They were just fun.
In a sense the pastiche nostalgia reminds me of the 80s retro boom. It is way over the top to the degree that nothing back then looked or sounded like that. It was more low key and subdued.
I wouldn't go so far as to say it was "never" about the graphics. There was definitely a sense of wonder when I first saw Wolfenstein 3D, then Doom, and later Quake. It was the graphics that got you first and made you think, "Oh wow, I can wander around this world?!" Thankfully, they were fun games as well.
I think what's getting people interested in this project is the sense of wonder they get from the visuals. Who knows if they can deliver, but there's something about how this looks that gives me a sense of wonder.
As a child of the 80s, I see it as an homage to King’s Quest and games like that (with a touch of Defender of the Crown), but I suppose the beauty of an alpha release ad is it’s simply a Rorschach blot for us to project whatever we want to see on it.
This is super trippy. I saw the screenshots first and though "Huh, they really nailed that early point-and-click RPG vibe. Neat, but only mildly interesting". Then I clicked on the video and realized it's 3D. "Oh..."
The castle in the distance really gives me some sort of surreal feeling of fleshing out areas that didn't exist in the old games. Back then, the highly degraded road in the distance you understood to just sort of be skipped if you clicked on the castle, whereas with this, that little brown, poorly rendered strip is an actual place you will traverse and have come alive. Neat.
Fun art style and I'm sure there's some interesting tech behind the renderer/shader - to my eye it looks like the pixels don't shift much as the camera pans, which would naturally happen a fair bit if you're simply doing nearest-neighbour sampling. I'd like to see some gameplay though; so far it's just a tech demo, albeit a very pretty one.
A recent game, Eclipsium[0], was a short horror game with a similar esthetic. I really liked it for the simple gameplay and the short (less than 4 hours) completion time.
I like the pixelated look, but that's probably nostalgia.
Presumably this effect was achieved using a shader rather than an old school software renderer like old-school Quake and Half-Life (the last games I remember where you could switch between OpenGL and "software renderer")
Yeah, I don't think this is anything revolutionary. I was thinking they were actually doing something a-la DOOM's pseudo 3D, but no, this is definitely just downscaled 3D.
By temporal I refer to DLSS, FSR, TSR, etc. In this game's case, I imagine it's not as simple as downscaling, there is probably some shader work involved similar to this:
Temporal sampling techniques do (in supported titles) have the option to supersample and generate higher resolution images, but they are most commonly used to efficiently downsample for FPS gains with minimal quality degradation.
From the Wikipedia page for DLSS:
The goal of these technologies is to allow the majority of the graphics pipeline to run at a lower resolution for increased performance, and then infer a higher resolution image from this that approximates the same level of detail as if the image had been rendered at this higher resolution. This allows for higher graphical settings and/or frame rates for a given output resolution, depending on user preference.
The difference is, this is "normal" geometry, with fixed 64x64 textures (or similar) mapped with hard edges (nearest-neighbour filtering, no bilinear/trilinear/blur).
Voxels is quite different: the world is made of cubes of 1 size or various size cubes (any texturing method can apply here, or without, some even use gradient shaders on the cubes).
So you get that pixel art vibe, mapped into the 3D world with smooth movements. The 3D window isn't made of stairs.
In this case, this is done "properly". Because some developers just use normal geometry, normal texturing with bilinear filtering, and slap a post processing shader on the camera view, which makes the view "blocky" without the world actually being blocky - which is cheap, and typically looks cheap.
Why would I want to play a game where the poor get poorer and the rich get richer, trying to get a bit less poor myself? I already do that in the real world.
I believe the concept for this project is an AI-generated pixel art GIF that went very viral on X a few months ago:
https://x.com/de5imulate/status/1947024682118488116
I am not sure this project is capturing the strangely compelling sense of scale shown in that GIF, but it's nice to see someone taking a shot at it.
Ironically, I think a large part of the sense of scale comes from it being a vertical video. Really helps with stacking all these layers of environment on top of each other. Anyway, Rückenfigur (https://en.wikipedia.org/wiki/R%C3%BCckenfigur) style concept art where there's a tiny figure in front of a huge layered environment has been common in games for decades, the difference is that with ai we can create """concepts""" that look nearly like a playable demo.
This game looks cool, but also the trailer just showing 3 static scenes from the same angle for a minute is a sure sign that this is very very early.
In a youtube demo, the author says he has been working on this idea for a while, and was motivated to deliver after the GIF. So, kinda? Thanks for the reference.
https://youtu.be/v9FMp5QgOvY?si=iKXJSNIuLiU5nOnj
Two other replies to this comment seem to have missed the fact that the inspiration for the game is AI-generated, the game itself is not made with AI.
What makes you believe that? The video the author put out doesn’t mention that.
Cool gif nevertheless.
https://x.com/de5imulate/status/1949291379655373061
I don’t think it’s ai because it correctly resolves both vertical and horizontal parallax with correct perspective lines applied to the scene - something ai still seems to struggle at greatly
It does not appear to be AI, and is a passion project by a single developer.
https://www.youtube.com/watch?v=v9FMp5QgOvY
The concept (the X link) is AI-generated with Midjourney, the game isn't.
Makes me think of the voxel displacement renderer ( https://blog.danielschroeder.me/blog/voxel-displacement-mode... ) from a year ago.
I had an idea a little back about using hand-drawn mipmaps (.5x, 1x, 2x, etc) scales for sprites, with some method for nearest neighbor, (e.g. with a 3x scale taking pixels from the 2x and the 4x scale). I imagined I wasn't the first to have that idea, but it was just a passing thought. I wonder if OP is doing anything like that?
Anyways, very neat.
That mipmaps thing is an interesting idea!
I'm sure I wasn't the first to think of it, and I think that's roughly what we're seeing here: https://www.youtube.com/watch?v=v9FMp5QgOvY&t=108s
If you go frame by frame, the texture becomes more detailed, about 8x as many pixels. It looks like a normal LOD switch as the camera gets close to the wall. It might be a texture without enough in-between scales.
This looks really cool and I will doubtless play, if it ever comes out. Unfortunately I get Hytale vibes from this website. I don't understand the point of announcing such a project, not least with such an elaborate landing page, so early in development.
Agreed. I tend not to announce or even share plans for impending projects until I have something nominally concrete.
Interestingly there was a paper from a few years back where they found that a participant's commitment was found to diminish in the long term if they prematurely shared their goals.
https://psycnet.apa.org/record/2009-06873-015
Given the call to action is to wishlist the game or subscribe to their mailing list, I guess they could use the interest to validate their idea or even to try and get funding.
The standard marketing advice in game development is to announce your game and launch your Steam page as early as possible to start collecting wishlists. In fact, considering how much footage and gameplay they are showing off I would say they are announcing it too late. So I don't agree that it's too early in the development compared to other indie titles.
It's to give yourself some form of accountability if you have a strong habit of never completing projects. When you've made a website for it, suddenly it feels like something that you need to work on.
I’m still waiting for Eitr.
I’m sure miegakure is going to come out any day now.
I appreciate the work but as someone who grew up playing these games and quake and blood and all those novalogic titles, it was never about the graphics. They were just fun.
In a sense the pastiche nostalgia reminds me of the 80s retro boom. It is way over the top to the degree that nothing back then looked or sounded like that. It was more low key and subdued.
I wouldn't go so far as to say it was "never" about the graphics. There was definitely a sense of wonder when I first saw Wolfenstein 3D, then Doom, and later Quake. It was the graphics that got you first and made you think, "Oh wow, I can wander around this world?!" Thankfully, they were fun games as well.
I think what's getting people interested in this project is the sense of wonder they get from the visuals. Who knows if they can deliver, but there's something about how this looks that gives me a sense of wonder.
As a child of the 80s, I see it as an homage to King’s Quest and games like that (with a touch of Defender of the Crown), but I suppose the beauty of an alpha release ad is it’s simply a Rorschach blot for us to project whatever we want to see on it.
This is super trippy. I saw the screenshots first and though "Huh, they really nailed that early point-and-click RPG vibe. Neat, but only mildly interesting". Then I clicked on the video and realized it's 3D. "Oh..."
The castle in the distance really gives me some sort of surreal feeling of fleshing out areas that didn't exist in the old games. Back then, the highly degraded road in the distance you understood to just sort of be skipped if you clicked on the castle, whereas with this, that little brown, poorly rendered strip is an actual place you will traverse and have come alive. Neat.
Fun art style and I'm sure there's some interesting tech behind the renderer/shader - to my eye it looks like the pixels don't shift much as the camera pans, which would naturally happen a fair bit if you're simply doing nearest-neighbour sampling. I'd like to see some gameplay though; so far it's just a tech demo, albeit a very pretty one.
Looks amazing.
A recent game, Eclipsium[0], was a short horror game with a similar esthetic. I really liked it for the simple gameplay and the short (less than 4 hours) completion time.
I wonder if it can carry a larger scale.
[0] https://store.steampowered.com/app/2419670/Eclipsium/
The creator talks a bit more about it in this video: https://youtu.be/v9FMp5QgOvY
I like the pixelated look, but that's probably nostalgia. Presumably this effect was achieved using a shader rather than an old school software renderer like old-school Quake and Half-Life (the last games I remember where you could switch between OpenGL and "software renderer")
"3D pixel art" = a 3D engine running in a lower resolution? It looks like 320x200
Yeah, I don't think this is anything revolutionary. I was thinking they were actually doing something a-la DOOM's pseudo 3D, but no, this is definitely just downscaled 3D.
Downscaled 3D doesn’t create such a stable pixel structure across movements.
Depends on the technique used. Temporal methods don’t look this way but standard downscaling can.
I don’t really see how. What do you mean by “temporal methods”?
By temporal I refer to DLSS, FSR, TSR, etc. In this game's case, I imagine it's not as simple as downscaling, there is probably some shader work involved similar to this:
https://youtu.be/Cwh_TpouPfQ
DLSS, FSR, TSR are for upscaling, not downscaling, so I’m not sure how they would apply anyway.
The creator hints a bit at things in this video: https://youtu.be/v9FMp5QgOvY?t=2m55s
Temporal sampling techniques do (in supported titles) have the option to supersample and generate higher resolution images, but they are most commonly used to efficiently downsample for FPS gains with minimal quality degradation.
From the Wikipedia page for DLSS:
The goal of these technologies is to allow the majority of the graphics pipeline to run at a lower resolution for increased performance, and then infer a higher resolution image from this that approximates the same level of detail as if the image had been rendered at this higher resolution. This allows for higher graphical settings and/or frame rates for a given output resolution, depending on user preference.
This is everything we imagined point and click games to be as kids.
What’s the difference between this and voxels?
The difference is, this is "normal" geometry, with fixed 64x64 textures (or similar) mapped with hard edges (nearest-neighbour filtering, no bilinear/trilinear/blur).
Voxels is quite different: the world is made of cubes of 1 size or various size cubes (any texturing method can apply here, or without, some even use gradient shaders on the cubes).
So you get that pixel art vibe, mapped into the 3D world with smooth movements. The 3D window isn't made of stairs.
In this case, this is done "properly". Because some developers just use normal geometry, normal texturing with bilinear filtering, and slap a post processing shader on the camera view, which makes the view "blocky" without the world actually being blocky - which is cheap, and typically looks cheap.
this seems to be inspired by the extremely viral ai-generated video:
https://x.com/de5imulate/status/1947024682118488116
Love the name. They should keep it
[dead]
> Wishlist and follow dev updates
How about no? Damn, this has been going on for a while now. I have around 5 titles in my wishlist that are dead by now.
"Early alpha trailer"
For that reason, I'm out.
What separates being out from being in? None of us can play the game yet.
Why would I want to play a game where the poor get poorer and the rich get richer, trying to get a bit less poor myself? I already do that in the real world.