next »

[William Chyr]

Manifold Garden (previously known as Relativity) is a game that re-imagines the laws of the universe.

Rediscover gravity and explore an Escher-esque world of impossible architecture. Witness infinity in first-person and master its rules to solve physics-defying puzzles. Cultivate a garden to open new paths forward, where an eternal expanse awaits.

Out now on:

• Steam
• Epic Games Store
• Nintendo Switch
• PlayStation 4
• Xbox One
• Apple Arcade
• itch.io

Key Features

• An expansive and visually striking world filled with mind-bending puzzles.
• Manipulate gravity to gain new perspectives and walk on any visible surface.
• Explore expansive architectural structures that repeat infinitely.
• Capture striking and beautiful images with Photo Mode.

Manifold Garden has been nominated for several awards, including “Best Debut” from the Game Developers Choice Awards and BAFTA Game Awards. It was selected as one of the New Yorker’s best games of 2019.





EDIT: Below is the original post written when I first started the devlog. I'm keeping it here as it is for archival purposes. Initially, the game was called Relativity.

---

During Thanksgiving of 2012, I started working on a prototype that would eventually become Relativity . Today being the (more or less) one year anniversary of the game's development, I thought it'd be appropriate to celebrate by starting a dev log here.



Relativity  is a first-person exploration adventure puzzle game set in an Escher-esque world with six different gravity fields. By turning on and off different gravity fields, you can walk on any visible surface and view the world from different perspectives.

It's a mix of physics-based spatial puzzles (like in Portal), environmental/observation puzzles (like in Myst), and also more cryptic metapuzzles (like in Fez). M.C Escher's artwork is a huge inspiration for the game, so you can also expect lots of staircases, and plenty of mindbending visuals.











---
BACKGROUND

I'm primarily an installation artist (you can see some of my work here), so I initially approached game development from a contemporary art perspective. However, I had studied physics in school, and in addition to working in research labs, had also worked in an ad agency as well as an interaction design studio. So I had some programming experience, some design experience, some art experience, and wanted to bring everything together.

The first prototype of Relativity was originally meant to be a quick and dirty project to familiarize myself with Unity3D. I wrote the demo over the course of four days, and showed it to my roommate and a few friends. The feedback was pretty encouraging, so I decided to improve upon it and see what would happen.

Below is a screenshot from the first prototype. The mechanics wasn't that great, and I spent way too much time choosing textures. Eventually everything was rewritten, but this is where it all began...



---
ART STYLE

The art style has gone through a number of pretty major changes, and even now, it's still very much a work-in-progress. I'm pretty much working on the game completely by myself (everything except music/sound), so it isn't really feasible for me to spend time making concept art. Instead, what I do is, I'll implement something quickly so that it's functional and I can test out the game mechanics, then I go through many rounds of iterations, each time adding small details trying to make it a little better. It's very much a "form follows function" approach to art.

Here you can see the progression of the art style over the course of 7 months:
 


---
TECHNICAL

I'm using Unity3D to develop the game, and scripting with C#. I'm using ProGrids and ProBuilder for designing levels. Both are fantastic tools, and I highly recommend them.

---
CURRENT STATE OF THE GAME

Right now, I've got the core game mechanics of shifting gravity field written and pretty well polished. Aside from a few minor tweaks, I'd say it's pretty close to what I have in mind.

In the gif below, the player is shifting from the yellow gravity field to the blue gravity filed, and then back again. Notice how the boxes become active and inactive depending on which field you're in.



At this point, I have probably designed close to 80 puzzles. However, of those, only about 40 are good, and of those, only about 14 are well-designed and polished. I still need to do a lot of iterating on puzzles to make sure that they're challenging but not unreasonably difficult. There's about 3 hours of gameplay in the latest version of the game. I'm aiming for about 40-50 puzzles in the final version, with about 6 - 9 hours of gameplay time.

So far, I've done about 40+ playtest sessions, and the feedback has been very positive. After each session, I make sure to fix a few minor things that didn't work, and add a few minor things to improve gameplay. That's pretty much my workflow for development: iterate, playtest, repeat.

---
ABOUT THIS DEV LOG

I'm really looking forward to sharing updates and progress throughout development with you all, and receiving feedback from the community. I spent the last few days reading other dev logs, and have found it incredibly inspiring to see these projects develop and grow. Hopefully one day this log can serve as a resource for other indie developers as well.

I plan on posting here at least once or twice a week. In addition to regular screenshots and updates about new features or puzzles added, I'll also be talking about some of the more technical challenges of development, such as programming custom physics and working with shaders. I'm still learning a lot of this myself, and there are never enough tutorials out there...

BTW, if you do happen to be working with depth and normal textures in Unity3D, I just wrote up this blog post about working with them that you might find useful.

---
MORE INFO

If you're interested in the game, here are some other channels to stay up-to-date:

Website | Twitter | Facebook | IndieDB

---
Anyway, thank you for taking the time to read this first post! I look forward to hearing from you all.

[melos]

hooray on starting the devlog!

---

i've playtested this and it's quite good and challenging!!

[William Chyr]

Thanks! Devlogs for anodyne and Even the Ocean were big inspirations.

[DanDecarlo]

Will there be a story too? Not that it's needed, but it can add some emotional involvement too.

[William Chyr]

Yes, there will definitely be a story. I'm just not sure what it will be yet, or how it will be presented.

I did experiment with a few ideas and methods. One of them was using these text cards that would pop up at certain points, with a bit of text for you to read that gave a narrative, but they felt a little bit too intrusive. Also, by having an object narrative didn't feel right with the ambiance of the game.

Realistically, I don't think I will be able to seriously work on the story until I'm much farther along (like in March or April), as there are so many technical challenges to deal with before that.

[William Chyr]

DevLog Update #1 - 11/24/2013

Today was not as productive as I hoped it would be.

I spent the first part of the day fixing up "Cube Dispensers" - these machines that release new cubes (while simultaneously destroying old ones) when you press a button. Because there are six different gravity fields in the game, I ended up having to do everything six times. At this point, I've abstracted the code enough so that level elements are pretty modular and can switch pretty gravities easily, but it's still not super optimal...



Another thing I worked on is "Color Change Beams" - these are beams of light that allow you to change the color of a cube, and therefore the specific gravity field it belongs to. I wanted to create a sweet transition effect where the new material would slowly fade in over the old material. Alas, writing the shader proved a little too complicated, and I settled for a simple instant-change color effect. I will come back to it in a few days.

[OniWorld]

Man, this looks so awesome. If you ever need a play tester feel free to hit me up Following!

[William Chyr]

Man, this looks so awesome. If you ever need a play tester feel free to hit me up Following!

Thanks so much! The next stable build is still quite a bit ways away, just because there's so much new stuff that I'm adding, but I'll for sure keep you posted.

Also, I just checked out the website for Oni World... man, the art looks sweet! The concept art alone would make for a beautiful book.

[OniWorld]

Also, I just checked out the website for Oni World... man, the art looks sweet! The concept art alone would make for a beautiful book.

Thanks alot! Hey, for all you know what beautiful book may actually exist as some kind of... reward tier?

[fuguelike]

Wow great concept and the optics on some of those screens are really striking.

Seems like your various disciplines are meshing well here. Can't wait to try it out.

Also IMO a game like this doesn't need a story since it would just distract from the core gameplay. That doesn't mean it wouldn't be more approachable for some though I guess.

[William Chyr]

Thanks, fuguelike! I'm very happy to hear that.

Regarding the story - you've got a point there. While a puzzle game certainly doesn't need a story, I do think that having some sort of a narrative can serve as a nice incentive to keep players moving forward.

Ideally, there would be a story for those who'd like one, and for those who just want to solve puzzles, they can ignore the story and still be fine. To me, Portal does this perfectly with GlaDOS's voice-over. But because they've already done it, I need to find an alternative approach.

[noumenus]

Wow this game looks like a trip. I really like the flat colors like the orange in the second pic. It seems to represent your style more so than the grid textures.

[William Chyr]

Wow this game looks like a trip. I really like the flat colors like the orange in the second pic. It seems to represent your style more so than the grid textures.

There actually is a grid texture in the orange part, but because it's so far away, it just appears to be a solid block of color. 

To be honest, I'm still on the fence about the grid textures. I agree with you that the flat colors look better. However, because the game involves stacking these cubes in different ways, you really need to have an idea of the distance between objects. The grid becomes really handy in this respect.

As such, it's basically a trade-off between aesthetics and gameplay. However, I think I just haven't found the right solution to the problem yet. I will definitely experiment with different textures and shaders to determine what looks best.

Also - thanks for the feedback!

[William Chyr]

DevLog Update #2 - 11/26/2013

Spent today and yesterday fine-tuning various puzzles that use color-change beams and dispenser cubes. Mostly this involved replacing the old version of the beams with the new ones, and adjusting solutions that were either too obvious or too tedious to execute.

Tomorrow I will start designing different architectural elements and space to populate the game's world. 

Anyway, got some new screenshots to share:

[OniWorld]

Love it!

I'm sure someone else has said it reminds them of portal, but I get a similar vibe from that as I did playing portal all those years ago. Are you aiming for that simplistic art style or are you going to make it a bit more artsy in the future?

[William Chyr]

I'm sure someone else has said it reminds them of portal, but I get a similar vibe from that as I did playing portal all those years ago. Are you aiming for that simplistic art style or are you going to make it a bit more artsy in the future?

Yeah... it's pretty difficult to make a first-person physics-based puzzle game without drawing up comparisons with Portal. I've sort of been trying to crawl out from under their shadow since day one...

With regards to the art style - I had a very different one when I first started. It had a much more realistic feel to it, with wood and concrete textures. Kind of steam punk meets sci fi, with lots of concrete. However, I'm not a very good modeler, and don't have a lot of experience making textures and normal maps. And I feel that if you go for a realistic look, unless you can pull of something on the calibre of AAA games, it's just going to look lame.

So I decided to go for a more minimalist approach, focusing first on gameplay and mechanics, and letting the art style evolve from that. In this way, every visual thing you see in the game, whether it's a symbol or a color, serves to provide some sort of information. A box is blue for a specific reason, and a door has a symbol on it for a reason. This is different from the early version where I would have a box with a wood texture on it, but for no particular reason.

When Jonathan Blow was working on Braid, he already had all the puzzles designed before he implemented the art style. He said one advantage with this  way of working was that he could see if the art style was interfering with the puzzles. For example, if the new art style made moving platforms not as visible against the background, he would know that he'd have to change the art.

I'm trying to follow a similar approach - get all the puzzles down, make sure all the visual cues are there. Then, once that's complete, start adding little elements to make everything look better, but always checking to make sure it's not interfering with gameplay. So, the game will definitely have a minimalist art style, but will look much nicer and artsier than what you see right now.

[noumenus]

Wow this game looks like a trip. I really like the flat colors like the orange in the second pic. It seems to represent your style more so than the grid textures.

There actually is a grid texture in the orange part, but because it's so far away, it just appears to be a solid block of color. 

To be honest, I'm still on the fence about the grid textures. I agree with you that the flat colors look better. However, because the game involves stacking these cubes in different ways, you really need to have an idea of the distance between objects. The grid becomes really handy in this respect.

As such, it's basically a trade-off between aesthetics and gameplay. However, I think I just haven't found the right solution to the problem yet. I will definitely experiment with different textures and shaders to determine what looks best.

Also - thanks for the feedback!

Oh I see, that makes sense. Just a thought, but maybe the grid only appears on floors you are standing on and the walls you are looking at? Everything else will switch back to solid colors.

[OniWorld]

Yeah... it's pretty difficult to make a first-person physics-based puzzle game without drawing up comparisons with Portal. I've sort of been trying to crawl out from under their shadow since day one...

With regards to the art style - I had a very different one when I first started. It had a much more realistic feel to it, with wood and concrete textures. Kind of steam punk meets sci fi, with lots of concrete. However, I'm not a very good modeler, and don't have a lot of experience making textures and normal maps. And I feel that if you go for a realistic look, unless you can pull of something on the calibre of AAA games, it's just going to look lame.

So I decided to go for a more minimalist approach, focusing first on gameplay and mechanics, and letting the art style evolve from that. In this way, every visual thing you see in the game, whether it's a symbol or a color, serves to provide some sort of information. A box is blue for a specific reason, and a door has a symbol on it for a reason. This is different from the early version where I would have a box with a wood texture on it, but for no particular reason.

When Jonathan Blow was working on Braid, he already had all the puzzles designed before he implemented the art style. He said one advantage with this  way of working was that he could see if the art style was interfering with the puzzles. For example, if the new art style made moving platforms not as visible against the background, he would know that he'd have to change the art.

I'm trying to follow a similar approach - get all the puzzles down, make sure all the visual cues are there. Then, once that's complete, start adding little elements to make everything look better, but always checking to make sure it's not interfering with gameplay. So, the game will definitely have a minimalist art style, but will look much nicer and artsier than what you see right now.

I agree, I don't really mind that I get the same vibe - as I loved that vibe. But no game wants to be compared to another one with a massive budget, really.

I think that if you can, you should go with a minimal style. Like you said, realistic wouldn't look right - especially for this kind of game. I think if you can implement SSAO and have mainly white for the game levels, but having colours to highlight the important aspects (cubes/lasers etc) it could work really well. It helps the player focus and know what is important in the game, and depending on the setting it could fit in. The best reference I can find is the "fluffy" dimension in quantum conundrum:


I really do think you're progressing right though, don't focus on art - it's not important. A lot can be done with shaders, there is a game on here that is being developed without any textures and the art looks and feels awesome.

[William Chyr]

I think that if you can, you should go with a minimal style. Like you said, realistic wouldn't look right - especially for this kind of game. I think if you can implement SSAO and have mainly white for the game levels, but having colours to highlight the important aspects (cubes/lasers etc) it could work really well. It helps the player focus and know what is important in the game, and depending on the setting it could fit in. The best reference I can find is the "fluffy" dimension in quantum conundrum:

Yes - ambient occlusion is exactly the way I've been thinking of doing it. This image here is actually quite close to what I'm aiming for:



I think it's done in a 3D modeling software though, so getting it to look that nice in the game might be a little tricky. I need to play around with the lightmapping settings in Unity some more.

It's also quite funny you brought up Quantum Conundrum as an example, as the gravity shifting effect I have in my game is directly inspired by their dimension shifting effect (the ring of light passing through the environment and changing the textures). However, I had never thought of the fluffy dimension as being an art style that primarily uses ambient occlusion (I thought Unfinished Swan was the only example). Now that you mention it, it definitely makes sense and is a very good reference, especially seeing as there are still subtle hints of different textures, despite everything looking washed out.

Going to go play some Quantum Conundrum now...

[William Chyr]

DevLog Update #3 - 11/28/2013

Happy Thanksgiving, friends!

I think I’m finally getting the hang of lightmapping in Unity. It really makes a huge difference in the final result. Anyway, here’s a sneak peak at some puzzles and architectural elements I’m currently working on: