Howdy! Geril here.
This month we've made some assets for a scene that will feature our previously re-introduced female character. She'll be playing Donkey Kong on a classic Nintendo Game Boy system, and the console is even animated — so you can check out the gameplay.
For the Game Boy asset, I made a high-res Game Boy in Blender, baked it to low-res and Lussy finished its textures in Substance Painter.
The hard part was the animation.
I recorded a bit of gameplay from the classic Game Boy Donkey Kong game, and then exported its frames as images. Then Lussy put them together into one big texture.
In Blender, I made a plane for every frame and then made them move separately by shape keys.
So every frame is a shape key that moves the appropriate plane up a bit inside the Game Boy, to bring it just under the screen's surface.
And the animation is just that. As time passes, every frame goes up a bit, becomes visible, then goes down. I synced it to the sound of the gameplay, so it matches up.
After testing it on Sketchfab, I even gave it some bones, so now it looks like Mr. Invisible is playing the game.
We liked it, so we ended up publishing it even without the character.
You can check it out here:
We'll put together the full scene later. I even made a low-res N64 with a Hori controller for it, and Lussy's in the process of modeling her old childhood CRT TV.
Showing posts with label art. Show all posts
Showing posts with label art. Show all posts
2019/04/30
2018/09/30
On shipping something less than great...
We're not proud of every piece of work we've done.
I guess it's part of being a creator, but sometimes the final product of a project is so... not good, that we'd rather it didn't have our names attached to it.
It's a depressing feeling.
Story time.
We worked on a project for about a year. It was a day job for us, making a historical-educational movie that was rendered in Unreal Engine 4. We created AI for big armies and individual characters. It was kinda hard – most of the small team wasn't really familiar with UE4, and the assets were not optimized for videogame technologies (and some weren't even PBR). But the team decided to go ahead anyways.
We finished the AI and most of the logic by May, but the rest of the team had to work on a lot of other parts of the production as well, and so they had no time to test crowd simulation, combat and such.
Mid-June, they started rendering clips for the final movie, but they had trouble with the untested blueprints (as expected of untested technologies). Since we were already pushing the deadline, instead of properly testing and taking a step back and refining the system, the team rushed work and tried to work around smaller issues. On top of that, we didn't have time to properly train them to use the huge and complicated system they described they wanted, so the majority of the features we implemented went unused.
For example, we created a face customizer, kind of like in The Sims. A character could have a randomized, custom face applied, so the crowd didn't look like an army of clones. But instead of using that, they left the crowd without custom faces, so yeah, it does look like an army of clones. Or, at a lot of parts, blueprints weren't even used, just featureless skeletal meshes (no custom faces, no mimicry, not even eye movement).
We also had to animate a lot of scenes in a few days; scenes that would, under normal circumstances, take a few weeks, maybe a month to set up. So you can imagine what the quality of those animations is like.
When we saw the final version of the movie, we were baffled. We suspected, based on our communications with them, that the final result wasn't going to be spectacular, that the team had to cut corners to meet deadlines, but we never imagined they would ignore features we spent months on developing. It was just sad.
After watching (that was a funny situation, trying to keep a straight face in the office for our coworkers while watching the video), we came home and tried to go back to work but it was just too hard. I mean we spent a year on this project, and it looks and feels like we created nothing. Were we this bad at communicating, or was it just too complicated for others to use? Guess we'll never know. But it just about killed all of our motivation for a few weeks.
We're gonna continue working together with them on other projects – they said the main problem was the miscalculated deadline – and we're gonna use the experience we gained to try to make things better. It's just hard to do work after a result like this.
Especially on our own projects.
We wanted to post another Sketchfab thing this month, but this finished movie just shook us.
We've had almost no free time to work on our own stuff, and all for this...
But we digress.
We actually went back to our old project, Project OLP, to see if what we made back then was any good, or if we just imagined it. And... it still works, and it's a lot more developed than any of our other projects... so maybe we should continue working on it.
We'll decide later. But as for now, it's fun to work on something that we have full quality control over.
Still trying to figure out if there's life on mars.
I guess it's part of being a creator, but sometimes the final product of a project is so... not good, that we'd rather it didn't have our names attached to it.
It's a depressing feeling.
Story time.
We worked on a project for about a year. It was a day job for us, making a historical-educational movie that was rendered in Unreal Engine 4. We created AI for big armies and individual characters. It was kinda hard – most of the small team wasn't really familiar with UE4, and the assets were not optimized for videogame technologies (and some weren't even PBR). But the team decided to go ahead anyways.
We finished the AI and most of the logic by May, but the rest of the team had to work on a lot of other parts of the production as well, and so they had no time to test crowd simulation, combat and such.
Mid-June, they started rendering clips for the final movie, but they had trouble with the untested blueprints (as expected of untested technologies). Since we were already pushing the deadline, instead of properly testing and taking a step back and refining the system, the team rushed work and tried to work around smaller issues. On top of that, we didn't have time to properly train them to use the huge and complicated system they described they wanted, so the majority of the features we implemented went unused.
For example, we created a face customizer, kind of like in The Sims. A character could have a randomized, custom face applied, so the crowd didn't look like an army of clones. But instead of using that, they left the crowd without custom faces, so yeah, it does look like an army of clones. Or, at a lot of parts, blueprints weren't even used, just featureless skeletal meshes (no custom faces, no mimicry, not even eye movement).
We also had to animate a lot of scenes in a few days; scenes that would, under normal circumstances, take a few weeks, maybe a month to set up. So you can imagine what the quality of those animations is like.
When we saw the final version of the movie, we were baffled. We suspected, based on our communications with them, that the final result wasn't going to be spectacular, that the team had to cut corners to meet deadlines, but we never imagined they would ignore features we spent months on developing. It was just sad.
After watching (that was a funny situation, trying to keep a straight face in the office for our coworkers while watching the video), we came home and tried to go back to work but it was just too hard. I mean we spent a year on this project, and it looks and feels like we created nothing. Were we this bad at communicating, or was it just too complicated for others to use? Guess we'll never know. But it just about killed all of our motivation for a few weeks.
We're gonna continue working together with them on other projects – they said the main problem was the miscalculated deadline – and we're gonna use the experience we gained to try to make things better. It's just hard to do work after a result like this.
Especially on our own projects.
We wanted to post another Sketchfab thing this month, but this finished movie just shook us.
We've had almost no free time to work on our own stuff, and all for this...
But we digress.
We actually went back to our old project, Project OLP, to see if what we made back then was any good, or if we just imagined it. And... it still works, and it's a lot more developed than any of our other projects... so maybe we should continue working on it.
We'll decide later. But as for now, it's fun to work on something that we have full quality control over.
Still trying to figure out if there's life on mars.
2016/11/28
3D Yuumei Fanart: Frey Underground!
Hey!
We've had some free time this month, and decided to mix things up a bit. This is completely unrelated to any of our previous stuff - it's a fanart of Yuumei's Underground picture.
The whole thing started when we realized that our portfolio was a little... one-sided. Not everyone likes the OLP stuff and style, but until now, that was the only thing we could showcase as a team. Separately we've worked in many different styles, but together we've mostly worked in one. So we've decided to try out different styles and subjects for practice, experimentation, and for our portfolio. This was also a challenge for us, to see how far we can get in about a week and how efficient we are. We've probably went a little overboard, but oh well.
Frey, the character is fully rigged and has a morph target set for emotions and complete lip-syncing. Basically we've created a game-ready character as fast as we could.
We've planned to include a violin playing and a random goofy animation as well, but exporting into Sketchfab-friendly FBX made our work hard. The process was very tedious and took up almost a whole additional day. (We're now convinced that Blender hates us) We can't include those animations in the same scene we've uploaded, and we thought making a new scene just for that would be overkill. Maybe we'll upload it someday, if people are interested.
We've decided on this particular piece because I [Lussy] am a long time fan of Yuumei and thought that shiny PBR lights would go well with her art style. It was refreshing to work based on someone else's "concept art". Can't wait for Fisheye Placebo to continue!
Here are some WIP pics:
We hope you liked it! We were trying hard not to butcher the scene. Hope we didn't.
We've had some free time this month, and decided to mix things up a bit. This is completely unrelated to any of our previous stuff - it's a fanart of Yuumei's Underground picture.
The whole thing started when we realized that our portfolio was a little... one-sided. Not everyone likes the OLP stuff and style, but until now, that was the only thing we could showcase as a team. Separately we've worked in many different styles, but together we've mostly worked in one. So we've decided to try out different styles and subjects for practice, experimentation, and for our portfolio. This was also a challenge for us, to see how far we can get in about a week and how efficient we are. We've probably went a little overboard, but oh well.
Frey, the character is fully rigged and has a morph target set for emotions and complete lip-syncing. Basically we've created a game-ready character as fast as we could.
We've planned to include a violin playing and a random goofy animation as well, but exporting into Sketchfab-friendly FBX made our work hard. The process was very tedious and took up almost a whole additional day. (We're now convinced that Blender hates us) We can't include those animations in the same scene we've uploaded, and we thought making a new scene just for that would be overkill. Maybe we'll upload it someday, if people are interested.
We've decided on this particular piece because I [Lussy] am a long time fan of Yuumei and thought that shiny PBR lights would go well with her art style. It was refreshing to work based on someone else's "concept art". Can't wait for Fisheye Placebo to continue!
Here are some WIP pics:
 |
| Frey isn't happy with his incomplete hair |
 |
| Attack of the placeholder normalmaps |
 |
| Frey's violin in Sketchfab's editor |
Some of Frey's test expressions and lip-sync
We love Sketchfab, and it's a shame that it isn't more popular. Whenever we show Sketchfab links or embeds to someone, we have to explain that it isn't a picture or a video, and that the camera can be moved. We always have to either explain the exact controls and settings (for example, this model always loads in SD quality and you have to manually set it to HD) in a wall of text, or leave out the instructions and hope very strongly that people will try clicking on the model. We went with the latter this time.We hope you liked it! We were trying hard not to butcher the scene. Hope we didn't.
Subscribe to:
Posts (Atom)