Universe Sandbox
Universe Sandbox => Universe Sandbox ² | Discussion => Topic started by: Dan Dixon on August 18, 2015, 08:33:27 PM
-
The Logarithmic Z Buffer Test 4 is our fourth opt-in, experimental build we're releasing to test major graphics changes we've made under the hood.
You must own Universe Sandbox ² to participate. You can buy a Steam code from our website:
http://universesandbox.com/2
To switch to this build:
1. Right-click on Universe Sandbox ² in your Steam Library and select 'Properties'
2. Click on the 'BETAS' tab
3. Change the beta to 'log-z-test'
4. Click the 'CLOSE' button and wait for the version to update
5. Click 'PLAY' and enjoy!
If you've already using the log-z-test build, you will automatically update. Just run it.
You will be asked to run a Performance Test upon launching the build (that will send us 6 screenshots taken during the test). As there are still a number of known issues, you'll likely want to switch back to Alpha 15.2 after testing this build out.
We're changing all the graphics to use logarithmic z-buffers. This website demo does a great job of illustrating what this means:
http://threejs.org/examples/webgl_camera_logarithmicdepthbuffer.html
Let us know if you run into issues... black screens, doesn't run at all, please point these out!
And thanks again for everyone's help!
[Updated to indicate latest test version...]
-
Startup is much smoother than before, no more driver hanging. Supernovae are bugged though. There are wireframe spheres, and the bright light it creates sometimes blinks black on the left side of my screen. I think the blinking has something to do with my graphics card or driver.
Heres a screenshot of the supernova (http://i.imgur.com/ndRHBvol.jpg)
EDIT: Collision glows are much too large. Molten planets look off in some way. almost as if they were pictures edited to make the exposure way too high(http://i.imgur.com/H452BFsl.jpg)
EDIT2: I don't know if this is intentional or not, but planets fall out of orbit, they seem to take multiple passes through each other before merging.
EDIT3: Under the option "ComputationDevice" my cpu and gpu are listed as AMP cpu/gpu instead of AMD. Also, setting this option to AMP CPU or AMP GPU crashes the game with this error "Runtime Error! Program: This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information."
-
Supernovae are bugged though. There are wireframe spheres...
Chad said:
By the way, did we want to leave in the supernova debug spheres?
I can turn them off for the build.
I said:
It's fine for the test.
Perhaps we should have turned them off. :)
They're debug visualizations to show the light edge (white sphere) and debris edge (orange sphere).
Collision glows are much too large. Molten planets look off in some way. almost as if they were pictures edited to make the exposure way too high
Collisions glows... hmm. We'll look into that. How did you generate that screenshot? Is that during a collision?
planets fall out of orbit, they seem to take multiple passes through each other before merging.
ComputationDevice should be set to Managed. Are you changing the ComputationDevice to Native? There's a bug with Native right now.
Under the option "ComputationDevice" my cpu and gpu are listed as AMP cpu/gpu instead of AMD.
AMP CPU or AMP GPU should be hidden. They don't work at all, yet.
-
Just to be clear, it is supposed to say AMP (Accellerated Massive Parallelism) and doesn't have anything to do with AMD.
It's a library we're working on testing that could give us some significant performance boosts on machines that support it, as well as allowing us to leverage GPU computation again.
https://en.wikipedia.org/wiki/C%2B%2B_AMP
-
This seems to run very smoothly, more so than 15.2
Issues:
- Ice and Clouds do not always render on earth
- Planets do not appear molten when hot
- Trying to use either AMP option crashes the game
- Collision glows do not appear with impact craters.
-
The first star was all black,but now work,there are some news,there is water but not cloud
The craters are smoothed,the supernova now are less bug,and look beautifull ;)
Mh yes the collision bright is omnipotent but for big collision is cool :)
-
Collisions glows... hmm. We'll look into that. How did you generate that screenshot? Is that during a collision?
Yep, it was after a fragment passed through the parent planet.
ComputationDevice should be set to Managed. Are you changing the ComputationDevice to Native? There's a bug with Native right now.
Nope, this was on Managed.
Just to be clear, it is supposed to say AMP (Accellerated Massive Parallelism) and doesn't have anything to do with AMD.
Ah, ok. I didn't know that. I thought it was a misspelling, haha.
-
bads:
I opened the "Supernova Type Ia" simulation. Every time a particle falls on the surface of Sirius B, the star gets bigger for a brief instant. There isn't an explosion or something. The star (as well as it's luminous halo) gets twice as big for an instant. I can't take an screenshot of it because it happens so quickly.
There is a serious issue with Earth's polar ice cap when I press 5 to launch it. It's a circle on top of the planet. Also, the temperature is all weird. The planet looks frozen when the avg temp. is 15ºC.
The collision fireballs are awful. So are the supernovae.
The loading screen is extremely laggy. I has never crashed the game, though.
goods:
The main solar system sim. seems to be running a lot more smoothly.
The hot an molten planets seem to look much better and more dynamic.
I'll play some more later...
-
I don't seems to have any problems
The only thing that I found which was a bit weird is when there is a super nova explosion, The whole skybox get's White
I posted a Screenshot with the feedback tool
Some Craters seems to show less lava than they used to. mostly the small one.
I don't recall seeing anything else
Did you upgraded the terrain? It looks better than the last time i've played
-
We just pushed log-z-test 4. If you can... run it one more time, please!
-
it is installing some Microsoft VC Redist thing.
-
Wow, it is definitely better now! There is no longer that Sirius B glitch and the supernovae are ok now. That big blue halo around the supernova seems kind of unrealistic though. Well, that doesn't matter. What matters is that things are much better now.
I'll play some more and see if I spot anything weird going on
-
it is installing some Microsoft VC Redist thing.
It is support library code which is needed to run code built with visual c++. There is now two modes of computation:
Managed mode c# CPU as well as native mode c++ CPU in a dll (for now only working on windows). The latter is generally faster/more effective, but currently have some issues, so it cannot be considered done.
-
Now it's very good :),there aren't many bugs,but and thing maybe,the graphics of craters,they are quite pixel :(
But after this,it's really good!
Good work :)
-
1st screenshot is taken from "Tidally locked Earth climate" scenario. It looks like the water and ice/snow are not rendered unless they are illuminated. The night lights are always on. Also note the overly bright mountain peaks.
2nd screenshot is from "Climate simulation" scenario. There is no snow, the colours are strange, the night light are always on. When I increase the sim speed to >3 months/second the colours become normal but still no snow.
-
Thanks for the report Plutonium, the climate bug you posted with the thin green line should be fixed, as well as the climate not initializing right.