Tenuous Atmospheres: How Will They Work...?

Photoshop let you add a nice gradient, which I'd love to be able to do via a shader, but it's not trivial. There is no atmosphere shader, just the skybox shader which is designed for space, so there's no easy way to orient a gradient with the horizon / sun (this would need to be done in the game engine). I could fake it for a still shot, but it would "fall apart" as soon as I start flying around.

But then again, I might be able to come up with something kinda clever - I haven't given up hope just yet! In the meantime, your effort was not in vain :D
Out of curiosity, you manipulated the skybox shader and replaced the black (sky) with blue? (and probably made the stars dimmer?)
 
Out of curiosity, you manipulated the skybox shader and replaced the black (sky) with blue? (and probably made the stars dimmer?)
Among other things, yes. If you go out into space, it's still sky blue. There is no true "sky" shader, just the space skybox shader, thus this limitation.
 
Among other things, yes. If you go out into space, it's still sky blue. There is no true "sky" shader, just the space skybox shader, thus this limitation.
I see. And this skybox shader (I saw the shader text you posted in the other thread, bunch of xyzr stuff) does not have / cant get any info about where you are (e.g. height above surface, distance from center of planet, or whatever)? So there is no variable to use to calculate some gradient or dynamic behavior?

(You need the source code and cant reverse engineer it?)

Again just fmi/curiosity:)
 
I see. And this skybox shader (I saw the shader text you posted in the other thread, bunch of xyzr stuff) does not have / cant get any info about where you are (e.g. height above surface, distance from center of planet, or whatever)? So there is no variable to use to calculate some gradient or dynamic behavior?

(You need the source code and cant reverse engineer it?)

Again just fmi/curiosity:)
These are pixel shaders, so most places you see XYZW it's actually RGBA (which I use in my custom code) - both mean the same thing programmatically in a shader. Any place where it actually is a true XYZ coordinate, it's a coordinate in screen space, not model space. So for example, I edited the vertex shader (which deals with geometry rather than pixel color) to get rid of that awful "SUPERCRUISE ACTIVE" sign that floats above my nose in VR. I did this by throwing out any vertices in that quadrant of the screen. But as soon as I moved my head (head tracking), it all fell apart because that text moved to a different part of the screen.

I would need access to variables inside the actual game engine, and these are locked up tight. Reverse-engineering the actual ED game code would be both a nightmare and against the ToS. In fact, it might well be considered illegal here in America thanks to the dumb DMCA. Shaders, on the other hand, are "open", very much like HTML or Java, which allows one shader to work on a variety of different graphic cards.

Now there are some tricks I might be able to employ to do some simple things, like make the sky grow darker as the sun goes down. This requires sharing data from one shader (scene lighting) with another shader (skybox). This is doable in theory, but I'm still working out how 3DMigoto does this. The hardest part of all this is that 3DMigoto's documentation is rubbish, so there's been lots of trial-and-error.
 
These are pixel shaders, so most places you see XYZW it's actually RGBA (which I use in my custom code) - both mean the same thing programmatically in a shader. Any place where it actually is a true XYZ coordinate, it's a coordinate in screen space, not model space. So for example, I edited the vertex shader (which deals with geometry rather than pixel color) to get rid of that awful "SUPERCRUISE ACTIVE" sign that floats above my nose in VR. I did this by throwing out any vertices in that quadrant of the screen. But as soon as I moved my head (head tracking), it all fell apart because that text moved to a different part of the screen.

I would need access to variables inside the actual game engine, and these are locked up tight. Reverse-engineering the actual ED game code would be both a nightmare and against the ToS. In fact, it might well be considered illegal here in America thanks to the dumb DMCA. Shaders, on the other hand, are "open", very much like HTML or Java, which allows one shader to work on a variety of different graphic cards.

Now there are some tricks I might be able to employ to do some simple things, like make the sky grow darker as the sun goes down. This requires sharing data from one shader (scene lighting) with another shader (skybox). This is doable in theory, but I'm still working out how 3DMigoto does this. The hardest part of all this is that 3DMigoto's documentation is rubbish, so there's been lots of trial-and-error.
Thanks for the explanation. Much clearer for me now.

fyi, on a sidenote, my ED VR runs on a cloud Shado
 
Last edited:
Atmo4.jpg


Yeah, this is pretty lame, but it's "something". I'm hitting a wall of what I can do with the knowledge I have (specifically, what data Frontier is passing into its shaders). I think I'll at least wait until we get a better answer regarding if atmospheric flight will support VR or not before continuing with these specific experiments.
 
Yeah, this is pretty lame, but it's "something". I'm hitting a wall of what I can do with the knowledge I have (specifically, what data Frontier is passing into its shaders). I think I'll at least wait until we get a better answer regarding if atmospheric flight will support VR or not before continuing with these specific experiments.

Hey, Zak put his stern face on about that:

"For now best to assume all Odyssey content is not VR compatible."

That means you have to keep going ;)
 
View attachment 180489

Yeah, this is pretty lame, but it's "something". I'm hitting a wall of what I can do with the knowledge I have (specifically, what data Frontier is passing into its shaders). I think I'll at least wait until we get a better answer regarding if atmospheric flight will support VR or not before continuing with these specific experiments.
That's pretty great work - managing to gradient that edge would be super hard I would guess, from what knowledge I have of shaders - been watching your stuff and it's super interesting, was wondering if you might get to somewhere like this.
 
Just so I'm not off topic in my own thread...

@Straha Yeagar - Have you had any joy with your digging into thin / untagged / thick atmospheres? :)


Well that's rotten news... Makes no sense to me unless Frontier is truly abandoning VR completely, which I guess won't surprise me. Sad, sad, sad, sad...

Ach, they could well still get to it post launch, I don't think it means a definite abandonment. (The Legs aspect is almost certainly a ton more work than the existing VR, to get it to match current market expectations etc).

If Odyssey sells well, or the VR market opens further for them (via PS5VR or whatever), I could see them putting the hours in.

In the meantime though, come support the pitch for an experimental branch in my signature ;)
 
Ach, they could well still get to it post launch, I don't think it means a definite abandonment. (The Legs aspect is almost certainly a ton more work than the existing VR, to get it to match current market expectations etc).
I can live without VR space legs, but being able to fly over an atmospheric world in VR is just a no-brainer. Having done so in a "pseudo-atmo" setting, I can tell you it's something we don't want to miss out on!
 
I can live without VR space legs, but being able to fly over an atmospheric world in VR is just a no-brainer. Having done so in a "pseudo-atmo" setting, I can tell you it's something we don't want to miss out on!

Sure I get that. (Even though I do want VR Legs gameplay ;))

The idea of the experimental branch is that it would probably allow Ship & SRV VR access from the off. With the Legs aspects being more likely to be extra rustic. (But forgivably so, because it's a 'test branch' etc).
 
Last edited:
Well that's rotten news... Makes no sense to me unless Frontier is truly abandoning VR completely, which I guess won't surprise me. Sad, sad, sad, sad...
This is expectation management. It says nothing about the plans or state of development other than there may be news about it in the future. Part of the CM job.
 
Weighed against Frontier's track record, it does not leave me hopeful.
Well the new guy seems alright, and has demonstrated tangible difference and upgrade from the crap we usually get.. at least as far as communication goes. Given that we owe them a chance at least. Having said its also fact that frontier aren't the developers you're looking for, they've had a chance to make good for many years now and simply decided not to, so no expectations. Just an open mind.

EDIT: If someone has a broken leg or arm they have a valid excuse.. this just isn't the case.
 
Last edited:
Haaaaang on, there just might be a wee giveaway here about just how thick a "tenuous" atmosphere can be...

Wildly speculative hats on...




If you zoom right in to this piece of concept art, to the top left monitor above the window, there's what appears to be the status readout for an airlock.

It includes the text "Target ATM: 755 MMHG".

755 MMHG is about 1 Atmosphere. So what you'd expect for the interior pressure of a surface base.

But it also includes two percentage indicators. One at 71% and one at 21%.

It's impossible to say for sure, as the icons for these values are not immediately meaningful to me... maybe they're just filler the artist plopped in to pad out the display... or that the values themselves are plucked out of thin air for this concept... or this is just an early mockup and the in-game values have changed since this stage of development...

But it seems (maybe?) implicit that 71 could be the percentage of "Target ATM" currently measured inside the airlock, as it climbs towards the ideal 755 MMHG...

And 21% would be the baseline, ambient, exterior pressure

So if that (very reaching) assumption is correct, it might mean atmospheres up to about one quarter the thickness (or at least pressure) of Earth's.
 
Haaaaang on, there just might be a wee giveaway here about just how thick a "tenuous" atmosphere can be...

Wildly speculative hats on...




If you zoom right in to this piece of concept art, to the top left monitor above the window, there's what appears to be the status readout for an airlock.

It includes the text "Target ATM: 755 MMHG".

755 MMHG is about 1 Atmosphere. So what you'd expect for the interior pressure of a surface base.

But it also includes two percentage indicators. One at 71% and one at 21%.

It's impossible to say for sure, as the icons for these values are not immediately meaningful to me... maybe they're just filler the artist plopped in to pad out the display... or that the values themselves are plucked out of thin air for this concept... or this is just an early mockup and the in-game values have changed since this stage of development...

But it seems (maybe?) implicit that 71 could be the percentage of "Target ATM" currently measured inside the airlock, as it climbs towards the ideal 755 MMHG...

And 21% would be the baseline, ambient, exterior pressure

So if that (very reaching) assumption is correct, it might mean atmospheres up to about one quarter the thickness (or at least pressure) of Earth's.

Hah nice spot 😄

I’d be wary of reading too much into concept art, as you say, but it’s certainly a fun theory :)

Assuming they’ve done their due diligence on the numbers, another possible reading of it is that the target pressure is just for the base internals though right?

The 71% & 21% also seem to have an N and an O symbol respectively. Poss just readings of the main gases, Nitrogen & Oxygen? (Which would be in a similar ballpark to Earth. Probably breathable I’d guess?)
 
Hah nice spot 😄

I’d be wary of reading too much into concept art, as you say, but it’s certainly a fun theory :)

Assuming they’ve done their due diligence on the numbers, another possible reading of it is that the target pressure is just for the base internals though right?

The 71% & 21% also seem to have an N and an O symbol respectively. Poss just readings of the main gases, Nitrogen & Oxygen? (Which would be in a similar ballpark to Earth. Probably breathable I’d guess?)
I think you're right actually. I didn't read that first symbol as an "O", but it could well be. Earth's atmosphere is indeed 21% oxygen, 78% nitrogen. So it's not far off that.
 
Top Bottom