00:00
00:00
Gagangrene
Equine and Eldritch both start with "E"
Profile pic by: https://twitter.com/tater_art

Age 22

None

Colorado

Joined on 10/5/15

Level:
7
Exp Points:
459 / 550
Exp Rank:
> 100,000
Vote Power:
4.90 votes
Art Scouts
1
Rank:
Civilian
Global Rank:
> 100,000
Blams:
0
Saves:
9
B/P Bonus:
0%
Whistle:
Normal
Medals:
100
Supporter:
4y 3m 11d

Gagangrene's News

Posted by Gagangrene - March 17th, 2020


iu_102008_5523965.gif

Let's start with this. it's an idle animation, but there's some things it doesn't do well:

•It doesn't communicate what it is very well; They're 3 times the size of a horse, and 30 like millennia older, so why do they move so effortlessly? He needs to move less.

•The first draft of this guy wasn't quite the right size. He should be about 30% bigger, in pixels.

•The tail wag is actually wrong. The tip leads the middle of it, and it should be the other way around. This is easy to fix, just move the keyframe cycle back.

•With the exception of the tail, all of his body is perfectly in sync. It's not impossible, but it's not often natural either, and is like twinning.

iu_102009_5523965.gif

I fixed it to the best of my ability.

•The first and most obvious change to make was to slow it down. By nature there's no start or end to the animation, no real highs or lows of the motion, so a simple operation like this is more likely to work fine. Mouse over the timeline, hit S, then 2, then enter, then I'm done.

•I detailed scaling in my last post.

•Also, as a positive consequence of scaling it up, the elderly cracks in this stone's body become more pronounced, which is perfect. :D

•The tail wag is also pretty easy, just take the tail end's keyframes and move them back to before the tail mid's keyframes. But, I also moved the controllers a bit so that the tail wagged less. Because it looked a lot more extreme than the rest of the body. I also did this with the head, taking it's keyframes and moving them slightly out of sync to the torso's movements (I Didn't care which way in time I moved them, probably should've).


Ultimately, I've got a way better idle animation.


Tags:

Posted by Gagangrene - March 16th, 2020


So for Filly Astray, I'm making a boss's sprites in Blender. iu_101696_5523965.png

However, boss says this guy should look a little bigger. Scaling up the sprites is a terrible idea and I think we both know that, so I need to scale up the blender model instead.

This should be simple, but it's not. Firstly, I need to mess with the camera's focal length so it can see the entirety of the model, and in consequence, also alter the render settings to keep the resolution proportionally increased.

iu_101694_5523965.png

Divide the old Focal Length by 2 for the new focal length.

iu_101695_5523965.png

Multiply the old Resolution by 2 for the new resolution, and because it's the inverse of what we divided the Focal Length by.

So now the boss should look exactly the same when I render it (until I scale the actual model up) but should have plenty of room to move around in, because the resolution was increased.

iu_101697_5523965.jpg

Now for the hard part. The whole of the boss isn't one object, it's 23 different objects (19 empties, 1 Guide, 2 meshes, and 1 armature) constraining, parenting, and modifying eachother in different ways. I can't scale just the mesh, because it's armature deformations are applied after the scale. I can't scale just the armature, because it uses the Empties' positions as guides for IK, and the pose won't scale properly. And I've already been animating the positions of the empties, so I'd need to rinse and repeat any transforms for the empties across every keyframe to solve that problem.


The solution though, it's so Bullshit but it works!

iu_101698_5523965.png

Parent everything to another empty, scale the new empty, and then don't touch it. It feels wrong, but I'm not using the 3D data anywhere else but Blender, so it should be fine. Uh, don't know what to do if I were to export and bring this into Unity. Anyways this took way too long to type, time to animate it.


Posted by Gagangrene - March 14th, 2020


iu_101248_5523965.jpg

Through several counts of trial and error using UABE, I modded Deimos into Hot Deimos


Tags:

Posted by Gagangrene - March 11th, 2020


So some 40 days ago, me and my team were in a creative jam, and we got overwhelming issues because we used a Git Repository like an SVN repository, and more importantly put .gitignore in a bad place without knowing it was in a bad place.


So the main issue was that We have our Root directory, then within it the .gitignore, and right next to it, the Unity Project root folder. It should've been in the Project directory, not next to it. I figured this out on February 10th, a full week after Creative Jam was over.


A lot of what we did in the repository was push files straight to the master branch, while more than one of us were working on the master branch at a time. What we should've been doing is working on separate branches, and then merging branches when we needed files, just as good practice.


Also, today I learned the big distinction between commits and pushes, or at least I think. Commits are purely local. Pushes are just letting everyone else pull those commits.


Posted by Gagangrene - March 2nd, 2020


So I'm making a lil 1-minute movie. Not gonna go into great detail about what that movie looks like till it's done, but I'm doing a lot of compositing to get it to look the way it does:iu_98194_5523965.jpg

iu_98195_5523965.jpg

iu_98196_5523965.jpg

iu_98197_5523965.jpg

A lot of the heavy lifting is done in the compositor graph to get this animation to look the way it does. And due to the nature of the compositor graph, semi-transparent anything is really hard to do. The solution for me has been to render things I need partially transparent separately, and then mix them on top of a new view layer, which keeps things like outlines intact as seen in the second image when you're seeing through a transparent object. However, this solution pretty much ignores the worldspace of objects to accomplish it's goal, and can cause some very dissonant effects when it's worldspace and screenspace positions don't interact properly. I mean like, if in image 2, the outlined character was in front of the transparent wall in the world, the wall would still be drawn on top of them, and that has the potential to mess with your head.

With all that said, how do I get my character to appear behind the transparent wall at one point in time, and at the front in another? The answer is so deceptively simple:

iu_98198_5523965.pngiu_98199_5523965.png

You can literally keyframe what view layer this node uses over time I have the key. It's not too hard to make sense of once you understand, but if it were your typical video editor, it's basically saying that a video layer is supposed to stay in the same place in the layer order, keep it's blending mode, keep it's location on screen etc. But the layer is keyframed to reference a different clip of footage. I have the render layer node keyframed so that it uses a more holistic view layer when the main character is in front of the wall, and so that the render layer node that previously contained the wall is blank. It works so nicely and simply, and I'm just so amazed at how many ways I overthought doing this before I just tapped "I" while mousing over the node.


Tags:

Posted by Gagangrene - February 26th, 2020


Blender I love you but why did you parent/child relationships of objects to vertices "Hooks," but the inverse "Make Vertex Parent" like those two sound any bit similar


Tags:

Posted by Gagangrene - February 20th, 2020


In a fighting game, you usually expect a small little explosive visual to indicate where a hit lands. Me and my peers are working on a fighting game. Our game didn't have those small explosive visuals, so I opened up Photoshop and spent 8 hours across two days making animations and bringing them into Unity to the most of my ability.

iu_95425_5523965.gif


I didn't really grab any references, I just drew off of ideas in my head. You might frown upon this, but the visual is already abstract, so to reference any existing effect is to reference the method, rather than to think the idea yourself. And at that, I was thinking about Madness: Project Nexus 2, so originality was already diluted.

"Starry"

The first one I did was the starry one; as it explodes, shooting stars come out, not the two that explode into single, perceptively-unified stars. I went straight ahead, starting with frame one, and drawing the animation moving forward. So, the first frame is just a big, unsteady star, each spike pointing in a direction I expect the smaller stars to disperse in. The second frame, the smaller stars are nearly in the place they should end in, instantly. All successive frames are just them moving slightly away from the center and shrinking before vanishing. I also started drawing only white, then manually outlined every frame with my brush's blending mode set to "Behind." (If you never use any other brush or layer blending modes, you can almost treat one layer as two layers by using the behind brush blend mode.)

After about an hour or more, I was finished, and asked a few artists what they thought, and what it looked like to them. The main thing I got from them was that it's "magical," like a wizard beam landing. I didn't have any immediate ideas as to how to do it better, so I just tried something else.

"Cut"

In our fighting game, we have a character that fights with a pair of scissors. I thought maybe I could make an effect specific to that. Again, I took the straight-ahead approach to animating it, with only a notion in my head to guide me. This worked mostly nicely, but the effect wasn't violent and fast enough, resembling a stroke rather than a cut. The fix to this was actually pretty easy; I removed the second frame entirely, so the jump between frame 1 and 2 was more significant, and the general motion was more of what I wanted. However, as I was drawing this, I also thought about the cloth and skin that would separate, and tried to abstract that in jagged waves dispersing from the main element.

"Brute"

After making the cut, I wanted to try again with a generic blunt effect. Same method as the "Starry" effect, except with differently-shaped outward elements and more of them. At the end it seemed I overcompensated. But, I thought there was also something unique, because it felt to communicate this absolute pulverization, so I decided to keep it on the shelf.

"Generic"

I literally just tried again, but to get a more reasonable amount of disperso, and I also chose to make it smaller in case we used "Brute" too, so there was more distinction. I think I accomplished this sufficiently, though a criticism I got was that it's too radially symmetrical, and I don't really know what to say about that till it's in.

"Ashesmack"

<There should be something here, but there isn't>


Tags:

Posted by Gagangrene - February 11th, 2020


I opened up Unreal for the first time for real today. I've been in the editor before to see how models look in it, but haven't used it for anything more focused. I started with the particle editor because that's what a group in my class needed, particles.

In the top right of all the windows, there's a little tutorial button, and I watched through each tutorial on the main editor, particle editor, and in incidence the material editor. The particle tutorial pulled up this premade fire particle, so in my usual fashion, I stopped looking at tutorials and just attempted to dissect the creation to figure out how to use it. I came up with this:

iu_93100_5523965.jpg

It's not useful yet. But, in dissecting the fire's material, I found that the fire used am 8x8 sprite sheet and the T_soft_smoke you can see poking out together, I can use info from the Particle editor in the shader graph, and found out that some of my knowledge from using the AE tool particle world has transferable skills.


Posted by Gagangrene - February 10th, 2020


I made a brick texture again:

iu_92770_5523965.png

I started with a dark, moderately saturated brown base color in a 2048x2048 canvas. I extended the canvas, so I could draw out of the borders. I used Photoshop's guides to keep the horizontal lines mostly consistent. I drew the mortar in a single tone of white, then took the lines crossing the borders and moved them 2048 pixels inward on top of the lines inside the borders, so that my lines wouldn't appear to be cut off when the texture tiled. Then I copy/pasted the pattern to the left and right of itself to see how it would tile there.

To get the multiple brick colors, I made a Hue/Saturation layer, turned up the Lightness, back the hue, and then drew over the mask the Mask. I had my foreground color as white, my background color as black, set my brush's foreground/background color jitter to 100%, and turned off "per tip" so that the jitter was applied per stroke, and then colored in each brick with one stroke each. To get it to loop, I'd only draw over complete brick "cells," use the rectangle marqueee tool, select that part of the mask, and drag it over from one side of the canvas to the next.

I also took the white lines again, and on the same layer, copy/pasted it around in a 3x3 grid around the original, and applied an inner glow layer filter to simulate ambient occlusion. We're doing a 2D game so getting Unity to do that isn't viable.

Anyways this texture sucks, it's both too detailed for the game it was made for, and it also just doesn't look that good. The bricks were never meant to look even or well made at all, but there's some bits that were made poorly in that you can clearly see where I stopped the brush, and the layer filter is highlighting holes in the drawn-on mortar pretty well.

This is pretty much how I do all of my looping textures right now, but I realized over the weekended that I can use Blender to get perfect and more convenient looping through it's 3D texture painting mode and some scaling of a plane's UVs. Gonna have to try that out.


Tags:

1

Posted by Gagangrene - February 5th, 2020


Between 5pm of January 31st and 11am of February 2nd, I participated in the Warren Tech Creative Jam.

Not the global game jam, but intentionally coinciding with it. I still made a game with my friends at WT. <There would be a link here, but it doesn't exist. yet.>


Our team consisted of 4 artists, myself included, 3 designers, and 2 programmers.

We had a fraction of the graphics design room to share as a workspace, the room with an inconsistent swath of computers, but also plenty of them, with Wacom Bamboo tablets, and the limited real estate of single screens. Unfortunately, the room was also pretty far "inland" in the landmass that was the building, so there were no windows and it was always hot during the day. How the graphic design class functions in there every day is a mystery. We were "prepared" to manage our game with a git repository, since that would work with the school computers as well as laptops we brought from home. It looked like we were well constructed for success till 5pm.

The theme was "Wolves," announced at 5pm.

I specifically wanted to make a twin stick shooter, even without the shooting part. (We're in a school, we literally can't have any resemblance of guns.) We did make a twin stick shooter. You play as the wolf of Pompeii, who historically died chained to it's house, and kill sulphorpus and ashen reanimated corpses as you try and escape the burning city. The chain to your collar is the melee weapon you control, or that's what it would've been if we got it in properly in time. Ri[.

So: Day 1, the end of January 31st.

CHEWY HIT THE HYPERDRIVE

Day 1 was a trip. Not in any drug related sense, just that we broke a bone as we tried to hit the ground running. The bone was the git repository. It was going so good... for a whole 90 minutes... We decided to do pixel art because one of our designers offered to help with art too, saying he can do pixel art. His art wasn't bad, but his animations were out of step, unsurprisingly. I started with making a 3D wolf model that I would later composite and pretender into sprites, but after the night was over I never touched it, and it was never finished. We did get a wolf sprite though. I went to bed first, and as I closed my eyes, I witnessed the shambling git repository.

Day 2, the optimal day to get things done.

I wake up at Breakfast time, and we just humor each other for 30 minutes as we eat. At the end, I ask the programmer who's up, "how's the repository." To my surprise, he and the rest of the team hate it, despite having worked with an svn before. Something went wrong down the line that we weren't prepared to fix, and we didn't want to spend any more of the remaining 28 hours we had working on it We used one or two flash drives for the remainder of the jam. I spent an extra few minutes to sit around with our late team member while he ate breakfast, because I didn't want him to be lonely, and then found myself working on managing the art assets. I was the only one who knew how to even vaguely set up sprite sheets, animations, animators, and such in Unity, so I spent a lot of the time putting all those together, manually. Not doing art, but managing what was there, and coming in. I was fine with this though, everyone was hard at work. After noon though, someone took out their Nintendo Switch. And this is where jade was begot in my veins.

My team went into Party Mode, dancing to Venga Bus and Caramelldansen, wasting time and only entirely returning to earth and work 20 hours later to clean up. Of course there was always at least one of us working, but I wish they'd try and push themselves to work a bit for this special occasion, and because we're on a clock.

Late in the day, some time between and after dinner, I opened up Photoshop to give our wolf a proper walk animation. Because of the perspective and the resolution of the sprite, as well as my inexperience working from a top-down perspective, I think the animation isn't worth bragging about. But I do think I made it look enough like a walk, or a canter as it's called, to the average eye.

Also around this time, I began emphasizing to start building the game, so we could address the biggest issues at the start of the next morning. But, I didn't quite realize when I had communicated my point to them, and so I briefly repeated and continued, till they expressed their annoyance. I sat down and continued to work on the wolf walking animation, also hit build to see what would happen, saw nothing remarkable because I didn't understand Unity scenes in builds and was the first to sleep again.

day3, welcom to die .

I wake up about 6 hours before the deadline. The room is completely silent, the lights are dim, and the only computers that are on shouldn't be. On the floor, everyone is sleeping, scattered across the floor like the gibs of several exploded tables in Gmod, and posed like the like's ragdolls. I sat down and took the time to continue making the wolf canter animation, specifically making the tail move convincingly. I started by animating a little skeletal guide of parallel lines, swaying back and forth and with the most distal segments lagging three frames behind. Then I painted over half those frames with one color to make a silhouette of the tail, and just flipped the silhouette for the other half. Then I locked the transparency and painted a second tone over every frame. This starkly contrasts with how I animated the legs, which was mostly abuse of selecting and moving. Ultimately, the tail looks the best.