Morten Vassvik
@vassvik.bsky.social
Simulation and rendering nerd. Co-founder and CTO @JangaFX
Working on EmberGen and more.
Discord: vassvik @[email protected]
Working on EmberGen and more.
Discord: vassvik @[email protected]
Tested various version in my shader, interesting how it becomes fully pathological when simply adding a `const`: gist.github.com/vassvik/621c...
October 20, 2025 at 12:31 PM
Tested various version in my shader, interesting how it becomes fully pathological when simply adding a `const`: gist.github.com/vassvik/621c...
TFW it works, and the gains are massive
Colorado by number of splines that pass the culling.
Colors are (R, G, B, Y, M, C, W) = (1, 2, 3, 4 ,5, 6, >=7), respectively.
First image uses the outer circle, second image uses the purple region.
Colorado by number of splines that pass the culling.
Colors are (R, G, B, Y, M, C, W) = (1, 2, 3, 4 ,5, 6, >=7), respectively.
First image uses the outer circle, second image uses the purple region.
October 19, 2025 at 8:22 PM
TFW it works, and the gains are massive
Colorado by number of splines that pass the culling.
Colors are (R, G, B, Y, M, C, W) = (1, 2, 3, 4 ,5, 6, >=7), respectively.
First image uses the outer circle, second image uses the purple region.
Colorado by number of splines that pass the culling.
Colors are (R, G, B, Y, M, C, W) = (1, 2, 3, 4 ,5, 6, >=7), respectively.
First image uses the outer circle, second image uses the purple region.
One more, tried to tune it so that temperature diffusion propagates the ignition
October 19, 2025 at 12:18 AM
One more, tried to tune it so that temperature diffusion propagates the ignition
Combustion tests with the new ground, 640M active voxels
October 19, 2025 at 12:16 AM
Combustion tests with the new ground, 640M active voxels
I guess it's the second formulation here that makes it equivalent to GJK?
October 17, 2025 at 9:21 AM
I guess it's the second formulation here that makes it equivalent to GJK?
Simplified construction in 2D with @desmos.com link: www.desmos.com/geometry/qme...
Ordering is a bit wonky, because I wanted the colors to compose in a certain way, but it works because Desmos isn't evaluating the expressions in order.
Ordering is a bit wonky, because I wanted the colors to compose in a certain way, but it works because Desmos isn't evaluating the expressions in order.
October 17, 2025 at 9:11 AM
Simplified construction in 2D with @desmos.com link: www.desmos.com/geometry/qme...
Ordering is a bit wonky, because I wanted the colors to compose in a certain way, but it works because Desmos isn't evaluating the expressions in order.
Ordering is a bit wonky, because I wanted the colors to compose in a certain way, but it works because Desmos isn't evaluating the expressions in order.
Spent some time trying to do conservative culling of geometry for distance function evaluation. Turns out I ended up reinventing something like GJK (which I really have no clue about or experience with) with exact bounds.
October 16, 2025 at 8:22 PM
Spent some time trying to do conservative culling of geometry for distance function evaluation. Turns out I ended up reinventing something like GJK (which I really have no clue about or experience with) with exact bounds.
Fixed a bug and tried beefing up the voxel count: 1.275B active voxels, 3072x3136x1312 bounding box.
Realtime capture in the viewport on 4090.
The projection struggles a bit on this scale, so I brute forced it a bit, so it's a bit slower than it has to be. FMG to initialize should help a lot here.
Realtime capture in the viewport on 4090.
The projection struggles a bit on this scale, so I brute forced it a bit, so it's a bit slower than it has to be. FMG to initialize should help a lot here.
October 14, 2025 at 10:15 AM
Fixed a bug and tried beefing up the voxel count: 1.275B active voxels, 3072x3136x1312 bounding box.
Realtime capture in the viewport on 4090.
The projection struggles a bit on this scale, so I brute forced it a bit, so it's a bit slower than it has to be. FMG to initialize should help a lot here.
Realtime capture in the viewport on 4090.
The projection struggles a bit on this scale, so I brute forced it a bit, so it's a bit slower than it has to be. FMG to initialize should help a lot here.
Finally adding ground collisions
October 12, 2025 at 8:57 PM
Finally adding ground collisions
Few stills
1664x1568x576 bounding box, 536M active voxels. Roughly 65% sparsity.
1664x1568x576 bounding box, 536M active voxels. Roughly 65% sparsity.
September 27, 2025 at 10:03 PM
Few stills
1664x1568x576 bounding box, 536M active voxels. Roughly 65% sparsity.
1664x1568x576 bounding box, 536M active voxels. Roughly 65% sparsity.
Dimmer RGB colors
September 27, 2025 at 9:49 PM
Dimmer RGB colors
Bright Red-Blue-Yellow primaries
September 27, 2025 at 9:49 PM
Bright Red-Blue-Yellow primaries
Some tests I did with hacking in mixbox (github.com/scrtwpns/mix...) support today
Left: RGB blending
Right: Mixbox blending
Interestingly I there's a lot to take from this for smoke scattering, which generally has an absorption component, so subtractive color mixing makes sense I think?
Left: RGB blending
Right: Mixbox blending
Interestingly I there's a lot to take from this for smoke scattering, which generally has an absorption component, so subtractive color mixing makes sense I think?
September 27, 2025 at 9:49 PM
Some tests I did with hacking in mixbox (github.com/scrtwpns/mix...) support today
Left: RGB blending
Right: Mixbox blending
Interestingly I there's a lot to take from this for smoke scattering, which generally has an absorption component, so subtractive color mixing makes sense I think?
Left: RGB blending
Right: Mixbox blending
Interestingly I there's a lot to take from this for smoke scattering, which generally has an absorption component, so subtractive color mixing makes sense I think?
A cool experiment would actually be to have these, but view them under different colored light
September 23, 2025 at 9:44 PM
A cool experiment would actually be to have these, but view them under different colored light
My obsession this last week: What do you mean I can't actually reliably compute the result of this trivial integral?
September 8, 2025 at 7:43 PM
My obsession this last week: What do you mean I can't actually reliably compute the result of this trivial integral?
Bonus: It's easy to add a small adjustment that explicitly elongates the cells by some constant factor as well without changing any of the math.
In fact It's always possible to map from one variant to another by appropriately adjusting the scaling constant
In fact It's always possible to map from one variant to another by appropriately adjusting the scaling constant
August 27, 2025 at 8:46 PM
Bonus: It's easy to add a small adjustment that explicitly elongates the cells by some constant factor as well without changing any of the math.
In fact It's always possible to map from one variant to another by appropriately adjusting the scaling constant
In fact It's always possible to map from one variant to another by appropriately adjusting the scaling constant