I just changed my entire game 

for any other devs out there, another solution is to start at 360p. 2x is 720, 3x is 1080, 4x is 1440 and 6x is 2160

Im a huge fan of Pixelart. But i think you made the right choice here. Your Game never really looked that commited to the Pixelart Style. It was basically an in between your new style and Pixelart. And thats just my 2 cents, but i think that a game that commits to its art style always looks better.

You say that you should have spent more time planning but I would argue you actually did the right thing. If you try to perfect your plan you will be stuck spinning forever trying to solve every possible problem you may encounter. We aren't actually very good at predicting what will actually be a problem and what wont. It is better to just start building something and solve the problems when you better understand them and actually know they will be problems. I'm not saying don't ever plan or try to avoid problems early but no amount of planning will prevent the need to make large shifts mid project.

6:40 My experience from like 20 yrs of IT projects tells me "skipping ahead doing the real solution" will bring you in more trouble than anything else, because you'll never get finished, you'll produce a lot of "the real solutions" that are completely useless, because you didn't / couldn't know what "the real solution" needs to look like etc. It's far more productive to just accept the fact, that you are producing a "for now that works" solution as long as you're willing and able to refactor stuff, when it's not suitable anymore. Understanding when to live with the compromise and when to throw it out and build something better is the real challenge. It's hard and you need a lot of experience to do that and I have rarely met people who are good at it.... so I guess the real challenge is getting good at understanding when to live with the compromise and when to refactor.

Can't believe Mind Over Magnet got a remake before even getting an official release. Bravo, Mark

An important aspect of Balatro's next fest success was that it had demos for multiple next fests, with each one showcasing a different set of mechanics. It gave more time for the game's hype to snowball and allowed a wider player base to give feedback. Given Balatro's success, I'm curious to see if more indie devs will ditch early access and instead opt for next fest demos.

It takes courage to make a big shift like that. It's too easy to become complacent with how things are because of how much work you think it will take to change. I admire your dedication!

Honestly, I've been watching these videos for quite a while and never realized this game used pixel art. I even recall thinking to myself "finally, an indie that's using vector art". So it's nice to see you fully commit to that style, since I think it sets the game apart visually.

Headsup for as to how I found it; it was linked in the end cards of the Developing 12 video, don't know if you want it there already seeing as the video is unlisted but now you know.

As someone who plays a ton of demos (240 last year) and loves puzzle game, my favourite kind of puzzly game demo is a selection of levels that showcase different mechanics instead of a full "arc" with introduction and more advanced puzzle. I'd still want at least one more advanced level so that I know what to expect, but I think showing different mechanics is way more effective. Neon White's demo was one of the best example imo, it had the first world (which is kind of necessary tutorial wise) and then the second world was entirely custom, it even showed late game weapons. Usually you can get away with that since intro levels aren't that difficult. The Court Of Wanderers was another nice example, at the end of the demo there was kind of a walkway above other levels, so you couldn't interact with them but you could see the kind of stuff the game would have. It doesn't have to be that involved, but I think a lot of games forget to add something that tells the player what to expect from the full game. Talos Principle 2 was kind of a bad example though. It had the intro, first world, and then another more advanced world. But the first puzzle of that world expected you to understand a mechanic that was introduced in a world that was skipped. It's nice to show more than just the start, but gotta make sure the player have everything they need in what's included!

Honestly, I think starting with pixel art and then vectorizing it is a really good way to get clear and easy to read graphics.

One thing that I think is worth noting is that while Balatro had a really successful Steam Next Fest, it's not like it sprung up overnight. The dev has been putting out demos since October at least, possibly even earlier than that, and LPers (thinking about Nothernlion) were able to play it almost half a year before it got the official 1.0 release. The previous builds were good, maybe even great versions of the game. But being able to take their time, listen to feedback, and refine what makes it good is a major reason why it's been so well received imo. Next Fest was like the icing on the cake, and while I'm sure it would have done well regardless, the game we got in the weeks after that last demo shows that you can't rush these things. Quality takes time, and since I believe you can only showcase at one Next Fest (unless I'm mistaken?), it's a real matter of getting that timing right.

Note: You may want to run a pixel mapping process (the kind that requires the lens cap to be on) on whatever camera you're using to film yourself. There always seem to be stray bright pixels on the filmed segments.

A pixel is a pixel, you can't say it's only half. ~ TJ "Henry" Magnet

Oh man, using Photoshop’s vector tools was probably the hardest path you could have taken in remaking the graphics. Learning Figma (my preference because you can make components and reuse elements suuuuuper easily!) or Illustrator (not super friendly to exporting final raster graphics) may have served you better since they are vector-based tools and make shapes and booleans much easier to work with. Also, I actually didn’t even notice that the graphics were pixel art previously. Maybe that’s RU-vid compression, but I always assumed the art style was hi res! In any case, great job taking the plunge and remaking the assets! Definitely not an easy task, but crisp graphics are definitely worth it for the final experience.

The curse of spotting a dead pixel in someones camera, you can't unsee

Thanks for the shoutout! Though by: "all the indie games" you must mean: "about 9 games a year," haha. I always love to see new Developing videos pop up. Nice work on the trailer!

Screw it, *unpixels your art*

It's interesting how you've changed your art style as a solution to a tech problem. I guess an advantage of solo development - if there were a team I'm sure everyone would be too scared to even suggest such a thing! Were you sad to change your original vision so much? (I much prefer the new style, by the way)

Waffeling on the least worst compromise instead of just sitting down and doing it right is something I struggled hard with in my own project, as well :D Almost every shortcut I took got called out by my playtesters, so I had to do better. But it's a much better game since I did! I really love your developing series, it has me nodding along a lot 😄 And your vector art looks really nice, too!

For the demo, I think having a good chunk of world 1 levels with a few later game levels also included would do a good job at showing off the variety of the game! It's pretty common for demos to be curated and not just the opening chunk of a game

the dead pixel on his camera continues to drive me nuts :D

Am I the only one who thought the game was nearly done? I'm surprised to hear he still has to make the rest of the game after World 1

taking a shot every time mark's head does a little wiggle when he talks, will update later

I typically watch RU-vid on one of my 1080p monitors, not fullscreen, which ends up being 720p. So it didn't even really look distinctly like pixel art to me unless it was a zoomed in shot. I think you made the right choice. Pixel perfection is hard to do in Unity at the best of times. When you have a fixed perspective on non-fixed-size screens, well you know.

Love this series, Mark! Everytime i see a new episode of "Developing" it cuts to the top of my To-Watch queue :)

I really love seeing this game develop like this and the way you share all the successes and also the trials you go through - it's both kinda sobering for anyone who wants to make their own game, and a nice way to set a realistic expectation of what it takes. Thanks for sharing, and good luck finishing the game up!

5:18 I love how happy you are with the results if the extra work you put in.

Damn it, Mark. I thought my monitor had a dead pixel.

For a demo, I think adding a bunch of scrambled levels would work. A few levels to introduce world 1 mechanics, a level or 2 to test them, then a few levels to introduce world 2 mechanics and a few levels to test them. Maybe throw in a bit of World 3 or 4 with that process, and it could maybe work.

I really appreciate videos like this. There are so many common game development issues that gamedev youtubers don't touch for some reason, and this is one of them. Maybe they think it's "too boring" or something. But you can make anything interesting. And useful!

When doing something like pixel-art, especially in Unity, you have to plan early. Resolution is important, and integer scaling is as well, but you also need to worry about image import settings, shaders, and the peculiarities of the camera (especially if you have cameras as texture generators). When come other big questions like, do you want to render low resolution fonts or use screen resolution fonts to allow better readability while having a pixel art style? Do you want to render the scene in low res and scale it, or scale objects and directly render to screen? There are major technical decisions to be made for what appears to be a purely aesthetic choice.

In my game, I'm making most of the UI elements fully scalable by using shaders. I have a bunch of material functions in Unreal Engine that can do basic shapes (there are a lot of them in the free UI Material Lab sample by Epic if you want to get started) with a bunch of settings (radius, rounded boxes, antialiasing, lines, triangles, circles, etc.). You can combine them to produce any shapes you want to produce full color scalable materials. My game is a tower defense game. So the timeline bar at the top that shows the upcoming enemy groups are all icons drawn with material functions. The only drawback I've seen is when the resolution or scale gets too small, you don't want things to draw less than 1 pixel wide, then it gets blurry. That only happened once with the health bars and I've limited the scale to prevent it. Probably overkill solution, but damn does it look good. And it'll work on any future resolution like 8k, 16k, etc. or any other screen size.

You had me worried there for a minute. The new art style works just fine by me, and now I have another reason to look forward to the Steam Next Fest.

Really good choice! It already had such a flash game vibe that it just works and some of the touches like the blurred backgrounds are very atmospheric.

Kudos on the brave transition! One thing I’ll mention is that without the perfect pixels, you can now do cool shader deformations to your art, or even animations with squash and stretch, and have it look just as crisp as before. (Especially if you’re using solid colors instead of textures!) Just one possibility to consider :)

Watching other game game devlogs that restart from scratch after 1-2 years and throw everything out made me really scared when I saw the thumbnail. That being said the choice you took here was really good and the game will end up benefiting a lot from it. Great vid and great (so far) looking game. Can't wait for it to be out.

I love every time Mark does a developing he's a bit more frazzled from development then previously

The updated art style looks really good. It's really clean and readable (it was before, but now even more so.)

This has been a great series, super insightful to anyone who enjoys gaming. Great job as always

well.. this might be... POLARIZING !!!! Hi Mark. New Prince of Persia The Lost Crown Boss Keys? It did some interesting gate keeping on metroidvania skill/map progression.

Great video as always. also I think your camera may be broken or something cause there's constantly a white pixel above your right shoulder

The new art style is lovely and crisp, and matches rather well the graphics of your videos so also good in terms of 'meeting audience expectations' (even if these are very different mediums).

My man, what a great video about your great job! We all wish you the best luck with this game! looking amazing so far!

Watching these devlogs Mark, has been a huge inspiration for me to keep learning GameDev as I plan on keeping up with it. Especially to the point where I'm comfortable with the godot engine usage as my choice. To hopefully make a fun game on my own outside of online courses and use what I learned into action by next year. These keep motivating me. I have an idea already for my first game I wanna create inspired by one of my fave classics growing up in the back of my mind. Once I am finished with the godot course I'm learning from for the engines basics. I can't wait to give it some light in the upcoming future. Thanks for these inspiring devlogs Mark.

So there actually is a solution that could have probably saved you a lot of "redesign everything" headache. You mentioned antialiasing, but there's a sort of twist on this known as a sharp-bilinear shader, which a LOT of emulators use for scaling nowadays where you'll see minimal perceptible blur. The very basic overview is that it scales up to the next highest integer scale and then scales back down using a simple bilinear filter I've used this in a lot of my pixel art projects because you really generally can't see the artifacts unless you actively look for them and it'd probably have worked exceptionally well for your game since your art style doesn't rely heavily on things like bayer-style dithering

Congratulations! You just learned another reason why pixel art is so often created at the lowest practical asset resolution -- it's easier to upscale (by an integer ratio) to fit different display sizes. Option #5 -- basically the opposite of option #4 -- would have been to downscale your level sizes so that the camera can be placed "closer", with more screen pixels per asset pixel. This reduces the overall effect of non-integer scaling (and anti-aliasing), but would DEFINITELY entail re-thinking basically all of your level designs (not to mention the basic parameters of your game physics).

Absolutely love the look of this now. I think your art style works better in vector form than in pixel art :)

Your game looks so great! Congrats for having the motivation to change all of your sprites.

I was like: "YUUSSSSSSS!" when you revealed you were converting the game away from pixel art! Pixel art is great, but I can't tell you how many times I saw an indie game with really cool promotional art, then..."Oh. That's not what the actual game looks like...it's just another pixel art indie game 😴" There are many games coming out that do pixel art well because they have a specific aesthetic they want to reach. One delightful recent example is Arzette and the Jewel of Faramore -- VERY distinctive art style to say the least 🤣On the other hand, many games just look generic, even if the art is well-done. Your vector art is super cute, and seeing them all nice and crisp really helps boost this game's unique vibe!

I'm sure you looked into it, but one way I like to "fix" pixel art when playing retro games is downsampling. You render at an integer scale larger than your monitor and then use bicubic sampling to scale it down to the actual screen size. The end result is a smooth but still crisp image. There used to be a shader for this on RetroArch called "pixellate" too.

Honestly so cool to see you goes from anylizing video game to actually making one. You're so close! Good luck on finishing MoM!

It’s been really interesting going on this journey with you, and I can’t wait to play your game! I’ve got it wishlisted and ready for Next Fest!

New game jam idea: you have to finish your game before the next episode of developing.

Thank you for giving us an update for what's up with the game, and good luck on your quest to get it finished.

Awesome video as always! Always a struggle to get over the sunk cost fallacy in game dev

The long short cut :D

The counter argument is often also true. In fact I'd say usually more often true. Sometimes you just need to ship. Stop futzing with every little imperfect thing, let go, and call it done. Move on to the next thing. By the time you've remade your game 7 times you could have had 7 games and way more experience under your belt.

Always a pleasure to see your journey Mark. Can't wait to get my hands on the end result.

Congratulations on the new art, You added a LOT of value to your project.

Curious, when you redrew the art, is it now in a vector format? Or is it still bitmap/pixels, just at a 4K resolution? (It sounds like it because you used Photoshop's vector tools, but I wasn't sure if you exported it as vectors or as bitmaps)

This was the right solution. It feels like so many indie games do just default to pixel art because it's The Done Thing, regardless of whether it's the best look for their game in particular. So you get games where sprite scaling is all over the map, or where the sprites freely rotate off-grid, or are just not the ideal look for the genre (you mentioned _Balatro_ earlier, and that's a perfect example of all three of those things). Now it looks like a playable version of those Aperture Science training videos or something.

Good luck, Mark! What a journey this has been!

You made the right choice to pivot to this art style. Readability is so important, especially with puzzle games :) Looking forward to the release!

Interesting journey! It's on my Steam wishlist now

One thing I would love to see for the future of this series, if there is any chance you can pull this of: Console releases! Once the game is finished it would be sooo interesting to see the intricacies of having to fine-tune it to fit into different ecosystems with all their own rules and quirks. Not sure if this is realistic but I'll just leave this here as an idea haha

Amazing video! really love how detailed your videos are - Especially Boss Keys. Do you have plans to make a Boss Key's video analyzing Tears of the Kingdom's dungeons? I would love to see that. Cheers!

I took the same approach with my game from the start. I love pixel art, but using it is asking for trouble. High resolution, but simple and clean shapes. Works on any screen.

with that title i couldn't not click on this video jajaja

Great to see your progress! How much of your development pipeline is applicable to people without 1.55m subs and 30k wishlists? Is there anything we can take from your experience that doesn't rely on an inexhaustible supply of playtesters?

I like how smooth everything looks now.

Good info. Appreciate you sharing this.

This reminds of the quote 'A designer knows he has achieved perfection not when there is nothing left to add but when there is nothing left to take away". Video game development is a long and iterative process, patience and experimentation is key!

There is a fifth option to your pixel problem. There is something called interpelation. It adds a very slight blur to the pixels where it'll make the uneven scaling look great. It still looks sharp and you wont get uneven pixel shimmering. If you have watched mylifeingaming's channel, they explain it better. Retro pixel art games are infamous for uneven pixels in 1080p. The retrotink 5x, and some hdmi mods for older consoles have interpelation

Well shit you did it man. You're finishing your first game! Excellent stuff, good job!

You’re insane and I’ve got so much respect for your process, both in making the game and in documenting and explaining it. Thanks for letting us be part of it :)

Im beginning to have Wintergatan's marble machine vibes about this game

"It just works." omg, amazing reference 😂

It looks amazing! Also, you've got a really solid art style, reminds me of ghost trick, specially comparing the ds sprites to the remake, and your sprites to the new version

Sometimes there is an easier solution in there than biting the bullet on a lot of work that seems like the “right” solution. I recommend, in this scenario, doing the common exercise of “time boxing” - carving out a set amount of time (a box) for yourself to try to find a different creative way to solve the problem. If by the end of that time box you haven’t figured something else out that works, you bite the bullet on the solution you know works. This is useful also for experimenting with new features, and deciding whether to spend the time fleshing them out or just cutting them if they’re just not working.

Alternatively you could also have downscaled (or, well, re authored the assets at half the size) everything to 360p instead of 720p instead. That would have fit any size screen

It's very cool to see a take on easy to make art style that is not pixel art. Don't get me wrong, I do love pixel art and love making it but it tends to give less appeal to your game as it just another indie pixel art game.

Awesome ! For the demo I suggest to introduce the beginning of some mechanics of one or two world and at the end a hard level to have a good thinking about it and have a challenge ! When I saw the title I thought : He just scrapped everything again ! But I am happily suprised you did not ! LETS GOO🥳🥳🥳🥳

Watching your into with the dots moving slightly on my phone trips me out. I lose focus and my brain just gets confused. It's actually really cool 😅

It's awfully quiet in here 🤔

Option 5: Lock the screen resolution.

Brilliant to see how it's going! I really like the new art and think I may use a similar approach to update my own art.

All focus on finishing the game! so there was not enough to spare for the video shots :P just teasing, great video and effort Mark. Keep at it!

that release date is pretty close to... you should enter mind over magnet into the GMTK gamejam

Someone got here while the video was unlisted 💀💀💀

I'm using a method that I rarely ever see used. Basically, it's a shader that does filtering, but with a nonlinear, sharper curve. So it's like the best of both worlds. It doesn't look very blurry, and you can scale it any size you want and it still looks kinda retro. Love it!

Change the camera to output to a render texture then use that and scale as you please. This way you get pixel perfect at any resolution without much effort. It’s really easy. Also your main camera has a dead pixel still. It’s been in a lot of your videos.

It baffles me that he didn't optimize the game for 1080p from the start. I thought that was common sense.

I liked the pixel art better

You could put a setting to make all the art pixel art, with a warning that it may not look good on all moniters, maybe specifying the 1880p moniter. That way, the player could choose either. Of course, you could still set this new art as the default, but having a pixel option would be nice.

I love hearing about your game dev process ❤

The artstyle change made the game not pixel art anymore, it just looks like a washed mobile game

I find it interesting that making a game work for Steam Deck is now a core concern of developers from your commentary. Really shows you how much it's pushed Handheld PC gaming.

New style is much better! Great work!

For the demo I would say to have 3 or 4 custom puzzles that show combinations of all of the mechanics of the game. Or at least half of them. That way people can have an indication of the game, but not have the entire game pre-played before actually getting to the game.