Jump to content

Developer Insights #11 – Engine Exhaust Visual Effects


Intercept Games

Recommended Posts

7 hours ago, Bej Kerman said:

Because, Nirgal, Take Two doesn't want you knowing about the thing they want you to preorder. Why? Pfft, customers shouldn't know about the thing they want you to spend 60 pounds on before there's a blooming product to be bought. Rather, let's pretend to make consumers know about how the game is coming along by spouting technical java to give the illusion you know more about the game when really you just know more about what the game is going to be running on. I've nothing against the devs for putting the tech stuff out there, I just want some confirmation that what we're seeing isn't just a really advanced version of Waterfall topped with KSPIE/FFT. I just want some gameplay - it can't tarnish the game's rep much to see an in-dev build of the game working. Seeing the game all in its alpha state won't lead anyone to believe it's all shambles, it's just gonna confirm that it's in its alpha state. Besides all that, I'm just a bit miffed that gaming culture has shifted from "here's a 2 hour demo, please pay between 15 and 30 pounds for this game" (which Elliptic's Rodina does, which I respect) to "here's 12 minutes of gameplay accumulated over 2 years, please pay 60 pounds and maybe we'll deliver the wanted product". Now I know there's a lot of trust in Nate to deliver alright, but it'd be blindfully ignorant to dismiss the thought that the game may flop, and either way you shouldn't be preordering!

This seems like an extreme response, given that you literally can't pre-order the game yet. If/when they open pre-orders, if they still haven't shown any gameplay, then this will be a valid complaint.

Link to comment
Share on other sites

16 hours ago, Bej Kerman said:

Because, Nirgal, Take Two doesn't want you knowing about the thing they want you to preorder. Why? Pfft, customers shouldn't know about the thing they want you to spend 60 pounds on before there's a blooming product to be bought. Rather, let's pretend to make consumers know about how the game is coming along by spouting technical java to give the illusion you know more about the game when really you just know more about what the game is going to be running on. I've nothing against the devs for putting the tech stuff out there, I just want some confirmation that what we're seeing isn't just a really advanced version of Waterfall topped with KSPIE/FFT. I just want some gameplay - it can't tarnish the game's rep much to see an in-dev build of the game working. Seeing the game all in its alpha state won't lead anyone to believe it's all shambles, it's just gonna confirm that it's in its alpha state. Besides all that, I'm just a bit miffed that gaming culture has shifted from "here's a 2 hour demo, please pay between 15 and 30 pounds for this game" (which Elliptic's Rodina does, which I respect) to "here's 12 minutes of gameplay accumulated over 2 years, please pay 60 pounds and maybe we'll deliver the wanted product". Now I know there's a lot of trust in Nate to deliver alright, but it'd be blindfully ignorant to dismiss the thought that the game may flop, and either way you shouldn't be preordering!

Imagine people still not understanding that this isn't their marketing plan but just a way for the community management team to keep the overly paranoid forums at bay for some months with some already existing developing assets and minimal work to do for the actual team making the game.

I know 2019 seems to be 10 years ago with all that happened since then, but do you remember how it was when they were actually actively marketing the game? With Nate going to basically every game show, the trailer being served as an ad on every space-related YT channel and reddit community and them flying relevant members of the community to their studio to make a tour, do you remember that?

This isn't an EA indie title that lives on how much and how often the developer writes on the blog or a big release like a new GTA that has more money for marketing than it has for actually developing the game, probably (and for here on it's just my speculation) they don't even have an internal marketing team as such things are managed by Private Division directly and this is what the community management team and the devs can manage on top of their other, way more important, work.

BTW where did you preorder the game? I thought it wasn't up for sale yet, because, you know, if they are actually asking you to preorder the game now then it almost make sense what you're saying.

 

PS It's funny how there are layers between the devs and the parent company of the publisher but when one has to talk positively it's "Nate's game" and when there's to criticize something it's suddenly "Take Two". Sensationalism is trickling down from crappy article titles to the common talk on forums?

Link to comment
Share on other sites

5 minutes ago, HotVector said:

Why does the plume expand exponentially outwards in a vaccum?

Because there is no air pressure to keep it contained. It also depends on the shape of the exhaust nozzle, hence why there are vacuum rated engines and atmospheric rated engines.

Link to comment
Share on other sites

13 minutes ago, HotVector said:

Why does the plume expand exponentially outwards in a vaccum? I do not think that is how rocket exhaust behaves in a vaccum. It should be more akin to the transitional exhaust. Here's a picture of the Falcon 9 first stage at ~40 km

0*wW55xGyMuqwX3KZJ.jpg

The only reason the transitional exhaust bends back in is because at that degree of expansion it is finally at equal pressure, In a vacuum the pressure in the plume indefinitely accelerates the gas outward as there is no equal pressure. 

Edited by mcwaffles2003
Link to comment
Share on other sites

On 9/22/2021 at 8:03 AM, Bej Kerman said:

Because, Nirgal, Take Two doesn't want you knowing about the thing they want you to preorder.

This sounds like cynical nonsense given that there are 0 ways to pre-order the game right now.

Quote

Rather, let's pretend to make consumers know about how the game is coming along by spouting technical java to give the illusion you know more about the game when really you just know more about what the game is going to be running on.

What "illusion" is going on here? We didn't know how exhaust effects were going to work in KSP2 before, and now we do. Just because you don't appreciate technical explanations doesn't mean that they are "technical java" (whatever that means).

Quote

I've nothing against the devs for putting the tech stuff out there, I just want some confirmation that what we're seeing isn't just a really advanced version of Waterfall topped with KSPIE/FFT.

The dev of Waterfall himself gave that confirmation earlier in this thread.

Quote

I just want some gameplay - it can't tarnish the game's rep much to see an in-dev build of the game working. Seeing the game all in its alpha state won't lead anyone to believe it's all shambles, it's just gonna confirm that it's in its alpha state.

Incorrect. I don't know if you were following KSP2 when it was first announced and its Steam page went up, but some people were complaining that the 2019 pre-alpha visualizations were trash and evidence that KSP2 was going to be a failure. 

Quote

Besides all that, I'm just a bit miffed that gaming culture has shifted from "here's a 2 hour demo, please pay between 15 and 30 pounds for this game" (which Elliptic's Rodina does, which I respect) to "here's 12 minutes of gameplay accumulated over 2 years, please pay 60 pounds and maybe we'll deliver the wanted product". 

If the games you're playing have 12 minutes of gameplay and cost 60 pounds, you need to find better games. 

Quote

either way you shouldn't be preordering!

Great advice, and it seems like Take Two agrees with you since, again, there are no ways to pre-order the game right now.

Link to comment
Share on other sites

Hey all, this is Aaron the visual effects artist on the team. I’d like to chime in on the exhaust comments/questions in this thread.

On 9/17/2021 at 1:56 PM, starcaptain said:

Friend of mine is sus that the blendshapes don't conform to real world maths. He sent these citations from Rocket Propulsion Elements by George P. Sutton.

Your friend is technically correct. We simplified for production reasons.

On 9/18/2021 at 4:08 AM, mcwaffles2003 said:

I really like how this looks and it really contests with waterfall. The outer boundary of the plume looks very natural and soft which is really cool to see. A question I have though is, do the mach diamonds really stay in place and expand as ambient pressure goes down? I would have anticipated they would slowly fade away from the nozzle and elongate as the furthest diamonds would each in turn fade away as pressure decreases.

Also, I know this is an edge case.... but what happens when the vessel goes into atmospheres with far greater pressures than that on kerbin? Will the plume compress even more with more diamonds or is the configuration at kerbin sea level a minimum?

I’ll continue to work on the shock diamonds. The motion and coloring need to be adjusted during throttle and atmospheric pressure transitions. Also, we have plans for adding overpressure to the exhaust in the future.

On 9/18/2021 at 4:24 AM, rextable said:

Out of pure morbid curiosity: do you plan to implement engine startup and shutdown transitions/effects?

There would be 'much awesome' to see engine bell giggle/wobble on startup as the exhaust plume stabilizes. Not to mention floaty globules of propellant blobbing around in the vacuum of space upon engine shutdown. Stuff like that.

I'm not too demanding am I?! 

We do have plans for start up / shut down effects for engines. Implementation is TBD.

On 9/19/2021 at 2:29 AM, neistridlar said:

An other minor issue: In the video demonstration, the speed of the low throttle exhaust seems ridiculously slow. I may very well be wrong, but it is my understanding that exhaust flow speed for chemical rockets should always be supersonic, and that exhaust speed is directly linked to specific impulse of the engine. In other words, the exhaust speed should always be pretty darn fast.

The speed of the low throttle exhaust has been sped up. Thanks for pointing that out!

On 9/20/2021 at 1:18 AM, Synonym Toast Crunch said:

One question I have is about how it'll handle multiple engines at once.
https://imgur.com/ykdMwO9
As we see here on the Falcon 9, the outer eight plumes bow out from the center as pressure from each plume forces them outward from the center, with the bowing increasing as the atmospheric pressure decreases. Will this effect happen in KSP 2, or will we see nine engines cones expand out and clip into each other with no effect on each other's direction? Is there a way to set a bounding area around each plume and assign a value that determines how much pressure it creates (that falls off with distance), and use that to determine how much effect that can have on any other plume that enters its bounding area, possibly using a parametric deformer on the plume to deform it in different directions?

Currently the exhaust plumes just overlap. I haven’t thought of them interacting in the way you described. I’ll take note and chat with the team.

I love to hear all the feedback and input regarding the effects. If anyone has additional technical feedback or knowledge to drop, please do. I truly want to be as authentic as possible while also balancing development limitations (tech, time, etc).

Since Jon posted his dev diary, we’ve added some heat refraction while in atmosphere, a bit of lens flare at full throttle and the ability for exhaust to bend during rotation. All of this is WIP of course.

einginerefraction.gif

engineflare.gif

exhaustbend.gif

Edited by Aaron VFX
fixed spacing issues
Link to comment
Share on other sites

2 hours ago, Aaron VFX said:

Since Jon posted his dev diary, we’ve added some heat refraction while in atmosphere, a bit of lens flare at full throttle and the ability for exhaust to bend during rotation. All of this is WIP of course.

einginerefraction.gif

engineflare.gif

That looks real cool. 
It might be able give the sort of “shine” that some plumes have. (ie reflectiveness)
 

Spoiler

image0.jpg

This is probably my favorite plume image ever

Also, is that the Rhino engine?

Link to comment
Share on other sites

13 hours ago, Aaron VFX said:

Hey all, this is Aaron the visual effects artist on the team. I’d like to chime in on the exhaust comments/questions in this thread.

Your friend is technically correct. We simplified for production reasons.

I’ll continue to work on the shock diamonds. The motion and coloring need to be adjusted during throttle and atmospheric pressure transitions. Also, we have plans for adding overpressure to the exhaust in the future.

We do have plans for start up / shut down effects for engines. Implementation is TBD.

The speed of the low throttle exhaust has been sped up. Thanks for pointing that out!

Currently the exhaust plumes just overlap. I haven’t thought of them interacting in the way you described. I’ll take note and chat with the team.

I love to hear all the feedback and input regarding the effects. If anyone has additional technical feedback or knowledge to drop, please do. I truly want to be as authentic as possible while also balancing development limitations (tech, time, etc).

Since Jon posted his dev diary, we’ve added some heat refraction while in atmosphere, a bit of lens flare at full throttle and the ability for exhaust to bend during rotation. All of this is WIP of course.

einginerefraction.gif

engineflare.gif

exhaustbend.gif

Welcome to the forums! :grin:

Glad to see you brought gifts!

Link to comment
Share on other sites

18 hours ago, Aaron VFX said:

I love to hear all the feedback and input regarding the effects.

It looks amazing. If I had to suggest something, I noticed that the sea level exhaust plumes end by simply spreading out into a narrow cone of flow and fading. With irl plumes, at low altitudes you get a mixing regime where the exhaust begins to slow down and becomes highly turbulent as it moves downwards along with air that has been dragged along by the plume. This might be ridiculously hard to model so I don't expect the team to tackle that effect. It might be possible to simply fade the shock diamond effect and attempt to seamlessly blend in a second, more sputter-y and turbulent effect which can then be made to spread out and dim in a realistic-looking fashion, through. How hard that would be is definitely not something I would know.

8Q9svsD.jpeg

 

As for supersonic retropropulsion in an atmosphere, there's also some really distinct effects as the rocket exhaust exits the engine and forms a stagnation front as air slams into it in the opposing direcction.

Flowfield-of-jet-issuing-into-an-opposin

If you wanted to model this, perhaps you could take your transitional exhaust effects, truncate them at a certain distance, and then below the engine add a new geometric surface (like a backwards paraboloid) of similar textures and shaders to represent the exhaust backing up and flowing in the other direction. It could be pretty neat to see some falcon 9-style entry burns being performed in the game.

 

Final note not so related to exhaust plumes, I have some questions about how we'll be flying these engines. I've got a few things that I've experienced through mods, and I have to ask, in order of increasing realism and inconvenience:

  • Ullage thrust to start engines in free fall: I know this game isn't going to be hyper-simulationist and all that, but having to scooch forwards a little on RCS before performing a burn always gave me an increased feeling of immersion in the game. A setting to disable/enable this or possibly only including the feature in "hard mode" would be really cool.
  • Thrust development: I imagine the base "normal mode" of the game will have throttle depth and response like in KSP 1 i.e. infinite and instantaneous,  but again if there were going to be a "hard mode" it would be neat to have thrust come on gradually on ignition or application of throttle, and for there also to be thrust residuals when a given engine is turned off. NERVA engines for instance could have much more residual thrust than chemical engines out of the necessity to run hydrogen over the atomic pile to cool it back down. Throttle depth limits would also add to realism in a hypothetical hard mode.
  • Limited numbers of engine starts: Going full realism overhaul here. Could such a hypothetical difficulty include limited numbers of restarts depending on the engine? This would be one of the most hardcore options of all. Possible inclusion? Honestly I'd be fine to see this one not make it into any hypothetical gameplay scenario.

Again, you guys are pulling off incredible work and I'm definitely excited for the end result no matter what the team's vision for it may be. Keep it up!

Edited by Wubslin
Link to comment
Share on other sites

17 hours ago, Wubslin said:

It looks amazing. If I had to suggest something, I noticed that the sea level exhaust plumes end by simply spreading out into a narrow cone of flow and fading. With irl plumes, at low altitudes you get a mixing regime where the exhaust begins to slow down and becomes highly turbulent as it moves downwards along with air that has been dragged along by the plume. This might be ridiculously hard to model so I don't expect the team to tackle that effect. It might be possible to simply fade the shock diamond effect and attempt to seamlessly blend in a second, more sputter-y and turbulent effect which can then be made to spread out and dim in a realistic-looking fashion, through. How hard that would be is definitely not something I would know.

8Q9svsD.jpeg

I would assume they would do that with particles, as achieving that effect with geometry would be difficult and computationally expensive, but with particles would actually be fairly easy. That would explain why we don't see it in these tests, since particles hadn't been added yet.

Link to comment
Share on other sites

5 hours ago, Synonym Toast Crunch said:

I would assume they would do that with particles, as achieving that effect with geometry would be difficult and computationally expensive, but with particles would actually be fairly easy. That would explain why we don't see it in these tests, since particles hadn't been added yet.

That's a really good point! I'm a complete novice when it comes to video game development and that didn't even occur to me. Yeah, this would be pretty much an ideal use for particle effects. You know how KSP 1 would always give your craft this piddly smoke column in an atmosphere that was made of individual puffs? And how if you got going fast enough or your PC slowed down they would be really obvious? Eliminating that effect and replacing it with a much smaller but more lively turbulent stream of fire that moves with your rocket would not only look better but be way easier on the hardware since you would only need a few examples of the effect at any given time. The coordinate system for this effect can be based on the engine itself. Presto! All you'd need to do after that is disinclude solid rocket boosters and you'd be set in terms of not having to deal with those smoke puff trails. Of course we all know that's not going to happen...

Link to comment
Share on other sites

On 10/7/2021 at 12:47 PM, Wubslin said:

That's a really good point! I'm a complete novice when it comes to video game development and that didn't even occur to me. Yeah, this would be pretty much an ideal use for particle effects. You know how KSP 1 would always give your craft this piddly smoke column in an atmosphere that was made of individual puffs? And how if you got going fast enough or your PC slowed down they would be really obvious? Eliminating that effect and replacing it with a much smaller but more lively turbulent stream of fire that moves with your rocket would not only look better but be way easier on the hardware since you would only need a few examples of the effect at any given time. The coordinate system for this effect can be based on the engine itself. Presto! All you'd need to do after that is disinclude solid rocket boosters and you'd be set in terms of not having to deal with those smoke puff trails. Of course we all know that's not going to happen...

just copy and paste the plume code from KSP 1 :P

Link to comment
Share on other sites

  • 4 weeks later...
  • 3 weeks later...
  • 3 weeks later...
This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...