Ultimate space simulation software

 
User avatar
gamadh
Astronaut
Astronaut
Posts: 58
Joined: 03 Jan 2017 09:48
Location: Earth/Brazil

Work progress - 0.9.8.1

18 Sep 2017 08:03

The best is the enemy of the good. And reasoning this way, the app should never be released because there will always be some other feature that can go in if we wait "a bit more"... unless Vladimir can find a quick way to simulate a whole universe within itself (and that means simulating Space Engine too, inception with turtles all the way down).

Vladimir can take the time he wants. But he did ask if he should delay the next version another month to get another feature in. I'm just making my voice heard... although I would have kept quiet if someone hadn't presumed to speak for 
it is for this same reason that the nasa needs to launch its probes after making several upgrades in the equipment otherwise the nasa would be making eternal updates according to the technology of the time in which the probe needs to be launched
In such a hostile universe it is not necessary be the center to be special, just be alive
 
User avatar
SpaceEngineer
Author of SpaceEngine
Author of SpaceEngine
Topic Author
Posts: 1125
Joined: 17 May 2016 22:16
Location: Saint-Petersburg
Contact:

Work progress - 0.9.8.1

18 Sep 2017 08:41

Such a flag will influence oceans as well as the atmosphere. As long as "terra" is no longer necessary a planet with seas, SE could generate or not generate seas as a result of equilibrium calculations. So there must be 2 options in the script: either define both ocean and atmosphere with their composition, or define the initial volatiles and allow SE to perform calculations.
 
User avatar
Sion
Astronaut
Astronaut
Posts: 43
Joined: 04 Nov 2016 21:46

Work progress - 0.9.8.1

19 Sep 2017 08:24

If there is one thing, I say change the temprate parameters. It's awkward when you are finding planets where the temperature is below freezing point, and they have high ESIs, despite being much colder, some planets in the negatives.
 
User avatar
SpaceEngineer
Author of SpaceEngine
Author of SpaceEngine
Topic Author
Posts: 1125
Joined: 17 May 2016 22:16
Location: Saint-Petersburg
Contact:

Work progress - 0.9.8.1

19 Sep 2017 09:42

ESI is calculated using a standard formula. It is a continuous quantity, unlike the temperature classes subdivision.
 
User avatar
FastFourierTransform
Pioneer
Pioneer
Posts: 356
Joined: 17 Nov 2016 15:09

Work progress - 0.9.8.1

19 Sep 2017 13:47

Wow FastFourierTransform, thanks for your kind words! Sorry for a long delay before reply, I was at a vacations :)
Thank you for your detailed responses Vladimir!! And excuse me for the delay.
I am pretty sure that JPL has a data files for hundreds of asteroids which they had modelled together with planets and moons. That data files should be quite large, using them in SE would require implementation of a some disk data streaming mechanism. But I believe that you can't notice a difference in position of some asteroid modelled using DExxx or a simple Kepler orbit. Unless you are planning a space mission, or planning to use SE to make an astronomical observation. I don't think that precise asteroids ephemerides worth the effort.
Ok, I understand. Maybe in the future there could be a system to directly load orbital solutions from JPL to Space Engine for spacific asteroids. SE could be a usefull tool also for planing observations.


Also I'm very happy to hear that you are going to work in rotational models for the planets, also the erratic rotators, and the SPICE integration of spacaft trajectories (that is going to be amazing in so many levels).
The best what SE could have in the end is some analytical, periodic modelling of perturbation of a stable orbits. Some precession, wobbling change of eccentricity and inclination. Just like rotation models.
This is understandable.

I have a few more questions:
1) Precession of Mercury's perihelion: I saw it in SE and looks great, but only occurs from the year 1550 to the 2560. Considering this is a more or less periodic phenomena could you make the timespan for the precession bigger? In that way we could see further evolution of Mercury's orbit. It's a very small angle now and If one has to explain it to people they are not going to clearly see it if it is not with longer duration.

2) When you increase the time-flow speed and end using a timescale of 10[sup]6[/sup], the orbits clearly "jump". For example if you go to Saturn and watch it from 0.5 AU with that time-flow, you can clearly see how for each step the orbit change in a discrete jump. I know this is because for each timestep you compute a new keplerian orbit. But could the timestep be smaller? or could it be interpolated for more points between one and the next orbit render so this effect get's unnoticed?

3) I saw that you have implemented TASS 1.7 for the Saturnian system; the orbital theory that acounts for the perturbations over Mimas, Enceladus, Tethys, Dione, Rhea, Titan, Hyperion and Iapetus. Is there some solution for the rest of the moons of Saturn?
I think is important we can see for example the orbital exchange between Janus and Epimetheus. A very interesting phenomenon that occurs every 4 years. Seen from Epimetheus how the 200 km sized Janus aproaches and then leaves swaping orbits between them it's a dream. In SE for now you just see them in opposite sides of Saturn, never coming close, never swaping orbits :(
 
User avatar
SpaceEngineer
Author of SpaceEngine
Author of SpaceEngine
Topic Author
Posts: 1125
Joined: 17 May 2016 22:16
Location: Saint-Petersburg
Contact:

Work progress - 0.9.8.1

21 Sep 2017 16:37

1) Precession of Mercury's perihelion: I saw it in SE and looks great, but only occurs from the year 1550 to the 2560. Considering this is a more or less periodic phenomena could you make the timespan for the precession bigger? In that way we could see further evolution of Mercury's orbit. It's a very small angle now and If one has to explain it to people they are not going to clearly see it if it is not with longer duration.
JPL ephemeris does not calculate that value directly, so it cannot be extrapolated above the end points of the simulation.
2) When you increase the time-flow speed and end using a timescale of 106, the orbits clearly "jump". For example if you go to Saturn and watch it from 0.5 AU with that time-flow, you can clearly see how for each step the orbit change in a discrete jump. I know this is because for each timestep you compute a new keplerian orbit. But could the timestep be smaller? or could it be interpolated for more points between one and the next orbit render so this effect get's unnoticed?
I don't understand what do you mean. Low resolution of the orbit mesh, where moving through the next vertex make a noticeable line fracture?
3) I saw that you have implemented TASS 1.7 for the Saturnian system; the orbital theory that acounts for the perturbations over Mimas, Enceladus, Tethys, Dione, Rhea, Titan, Hyperion and Iapetus. Is there some solution for the rest of the moons of Saturn?
Yes, and for minor moons of Jupiter and other giants. You have to dig into literature to found them. There are no "collective" theories, just a separate ones for each individual body or small sets of bodies (like trojan moons of Tethys and Dione).
I think is important we can see for example the orbital exchange between Janus and Epimetheus. A very interesting phenomenon that occurs every 4 years. Seen from Epimetheus how the 200 km sized Janus aproaches and then leaves swaping orbits between them it's a dream. In SE for now you just see them in opposite sides of Saturn, never coming close, never swaping orbits
SE needs implementation of a custom numerical/analytical theory of movement of these moons.
 
User avatar
SpaceEngineer
Author of SpaceEngine
Author of SpaceEngine
Topic Author
Posts: 1125
Joined: 17 May 2016 22:16
Location: Saint-Petersburg
Contact:

Work progress - 0.9.8.1

21 Sep 2017 18:44

Experimenting with a detail textures. Here is a 4 "octaves" of a different textures with a different scales. Astronaut model to scale.
scr00839.jpg
scr00842.jpg
scr00843.jpg
Here is a single tiled texture
scr00844.jpg
scr00845.jpg
The problem is what this detail pattern is the same across the whole planet. Planet texture generator must produce some sort of a mask texture, which will define where to place pebbles, where sand, where grass. These textures will be generated in addition to the elevation/normal and color/specular textures, thus increasing memory load by 33%.

Another problem is a noticeable tiling, but there are many methods to fight against it.

It's possible also to implement the Parallax occlusion mapping to add illusion of a volume to stones.
 
User avatar
DoctorOfSpace
Star Engineer
Star Engineer
Posts: 1363
Joined: 22 Aug 2016 15:04
Location: SpaceX Mars Colony
Contact:

Work progress - 0.9.8.1

21 Sep 2017 19:34

Very nice, time for a new texture atlas
CPU: Ryzen 9 5900X @5Ghz - RAM: G.Skill Ripjaws V 64 GB (2 x 32 GB) DDR4-3600 CL18 - GPU: MSI RTX 40​90 GAMING ​TRIO 24G
Quando omni flunkus, moritati
 
User avatar
LiveLife42
Space Tourist
Space Tourist
Posts: 34
Joined: 23 Nov 2016 14:59
Location: Sol System
Contact:

Work progress - 0.9.8.1

21 Sep 2017 21:47

Looks great, can't wait to see the final look.
 
User avatar
TheRedstoneHive
Explorer
Explorer
Posts: 173
Joined: 03 Mar 2017 00:56
Location: Gliese 667Cc
Contact:

Work progress - 0.9.8.1

22 Sep 2017 00:58

I've heard that the Duke nebulae will not be coming in 0.981, and just 3D clouds, is this true?
 
User avatar
Salvo
Pioneer
Pioneer
Posts: 451
Joined: 03 Nov 2016 07:19
Location: Veneto, Italy
Contact:

Work progress - 0.9.8.1

22 Sep 2017 01:34

Another problem is a noticeable tiling, but there are many methods to fight against it.
I know a technique called "Macro Texture Variation", but I never found any raw implementation of it, just on terrain generators of game engines.

I think it consists in randomly rotating and maybe also scaling parts of texture, but I have no idea how it works... It usually also darkens parts of it but this is already done by SE.
The universe is not required to be in perfect harmony with human ambition.

CPU: Intel Core i7 4770 GPU: Sapphire Radeon RX 570 RAM: 8 GBs
 
User avatar
SpaceEngineer
Author of SpaceEngine
Author of SpaceEngine
Topic Author
Posts: 1125
Joined: 17 May 2016 22:16
Location: Saint-Petersburg
Contact:

Work progress - 0.9.8.1

22 Sep 2017 02:19

I think it consists in randomly rotating and maybe also scaling parts of texture, but I have no idea how it works... It usually also darkens parts of it but this is already done by SE.
Yes, SE already uses blending with rotated part of the same texture while generating low-scale patterns (they now should be called mid-scaled).
 
User avatar
ettore_bilbo
Explorer
Explorer
Posts: 204
Joined: 11 Nov 2016 05:33
Location: Italy

Work progress - 0.9.8.1

22 Sep 2017 05:03

Planet texture generator must produce some sort of a mask texture, which will define where to place pebbles, where sand, where grass. These textures will be generated in addition to the elevation/normal and color/specular textures, thus increasing memory load by 33%.
Maybe it's a silly question but, You can't use elevation and color textures themselves like a sort of mask? 
 
User avatar
Fireinthehole
Space Pilot
Space Pilot
Posts: 82
Joined: 02 Nov 2016 15:54
Location: Lysekil, Sweden

Work progress - 0.9.8.1

22 Sep 2017 06:02

I've heard that the Duke nebulae will not be coming in 0.981, and just 3D clouds, is this true?
I believe it's the other way around.
 
User avatar
ZackG
Space Tourist
Space Tourist
Posts: 37
Joined: 11 Dec 2016 13:15
Contact:

Work progress - 0.9.8.1

22 Sep 2017 13:36

I've heard that the Duke nebulae will not be coming in 0.981, and just 3D clouds, is this true?
Where did you hear that from?

Who is online

Users browsing this forum: Google [Bot] and 20 guests