|
| 1 | +### Mardini 2025 - Hot Air Balloon |
| 2 | + |
| 3 | +Today is day 21 of [Mardini 2025][mardini-2025]. The prompt for today is "FLY", |
| 4 | +in the KineFX category. |
| 5 | + |
| 6 | +I was going to do a Vellum sim of a person falling through the sky with a parachute, |
| 7 | +but 1. I'm not actually sure if Vellum handles drag from air resistance like that, so |
| 8 | +it may not even be technically possible; and 2. I wasn't feeling up to animating a person, |
| 9 | +even a little bit; I could have done a ragdoll but that still requires rigging the person, |
| 10 | +and defining the degrees of movement, and in my experience it doesn't work well unless you |
| 11 | +have everything configured perfectly. |
| 12 | + |
| 13 | +Instead, I made a hot air balloon. Then many. Then I made a terrain for them to fly over. |
| 14 | +Then I made clouds so the balloons aren't the only thing in the sky. Then I played around with |
| 15 | +the Karma Physical Sky LOP to see if I could get a nice looking sunrise behind the hill on the |
| 16 | +right. The answer was no, I couldn't; at least not in the time I had for the project. But in |
| 17 | +the process I created a slap comp in order to apply a glow effect, which had helped but was |
| 18 | +no longer needed; but I wanted to do something with the slap comp, so I then went overboard |
| 19 | +playing around with render vars, AOVs, etc. Ultimately I desaturated everything but the |
| 20 | +balloons in order to emphasize them; added a bit of chromatic abberration in post-processing |
| 21 | +(not quite as good as the lens shader variety because it can't "see" behind rendered objects |
| 22 | +to always give the correct color) and played around with selectively blurring parts of the |
| 23 | +image based on depth, whether or not the pixels show balloons, etc. |
| 24 | + |
| 25 | +Basically the project went down several tangents. I don't necessarily think the result is |
| 26 | +any better for it all, but again, the point was practice, not perfection. Getting the failures |
| 27 | +out of the way now at least lets me learn to do things that might not work here, but may work |
| 28 | +in the future, and at that point I'll already know how to do it. |
| 29 | + |
| 30 | +New stuff: |
| 31 | + |
| 32 | + - Karma Sky Atmosphere LOP again. This is technically in the render, but again, I didn't |
| 33 | + end up using any of the more extreme features, just the physically-based light source and |
| 34 | + volumetric atmosphere. So... half-dipping into this for the second time. I'll have to plan |
| 35 | + out a project that actually focuses on this at some point, I'm barely scratching the surface. |
| 36 | + Maybe something in space. |
| 37 | + - I finally got around to playing with the Cloud Shape Generate SOP + related notes that were |
| 38 | + added in Houdini 20, almost two years ago. I used it for one day during Houdini 2024, then |
| 39 | + never touched them again. They're interesting to use, but problematic because what you see in |
| 40 | + the SOPs context doesn't necessarily have any bearing on what you see in the final render. |
| 41 | + Also, it takes a lot of creative engery to fiddle with shapes in such an abstract way to |
| 42 | + attempt to get a specific look. Finally... they make the render take a _horrifically_ long |
| 43 | + time to complete. (As in... 6-8 hours for a 4K image in this project.) So... great tools, |
| 44 | + but I'll probably stay away in the future unless my project specifically requires clouds. |
| 45 | + - Slap Comp, take 2; including glow, chromatic aberration, HSV adjust to artificially increase |
| 46 | + the saturation of everything, selective blur, etc. More details above. |
| 47 | + |
| 48 | +<!-- My entry post is [here][entry-post]. --> |
| 49 | + |
| 50 | +[mardini-2025]: https://www.sidefx.com/community-main-menu/contests-jams/mardini-2025/ |
| 51 | +<!-- [entry-post]: ... --> |
0 commit comments