#Stellar Plasma-Assisted Rocket Kinetics (SPARK)
1 messages · Page 2 of 1
I mean, either
if there's a way to set up the shader so that the plumes are easily customizable, that'd be even better
though obviously the best solution would be something like Waterfall/LFO
Oh no yeah, its just like the paintable shaders
You can make a ton of variables and people play with them
Paintable one has only a few meaningful ones but the current plumes shader has like 20+
And the vollumetric will have even mor
oh nice, can't wait!
I guess I gotta get my modeling game up so that I can make some engines to utilize them 😆
Lmao i mean u can make real plumes or a pack like that
Also, other than methalox and hydrogen, all other engines are pretty simple modelwise
I wouldn't mind trying my hand at something like a linear aerospike

though that'd be a bit of a challenge for the plumes
since there would have to be like 20 tiny plumes that interact nicely to make one large one at the far end
Not really i mean
The 20 small ones are all the same
So they can all have the same plume
And then you focus on the big one
Also the small ones can be just 1 mesh
And the big one like 6/7
How big should a 30kg tank of Xenon be? The "small" radial tanks are supposedly 0.03t = 30 kg. They're way bigger than my little torus.
Ksp is 0.625 from the original
30 kg of dry mass or wet mass?
COPVs like these should generally be relatively light
and you can estimate the density of xenon in the game from the dimensions of the stock tanks and the difference between wet mass and dry mass
Ksp tanks are all wt mass
I mean, yeah, but you can just set the fuel to 0 and see the mass
so I wasn't sure
but yeah, from the JSON file I can see the dry mass of the tank is 3.75 kg
and it holds 30 kg of xenon
That gives me the idea to add the possibility to change/add things to the menu that appears when u hover a part
Maybe that belongs to LOABE
But as a default/“show-that-its-possible” ill add the dry mass to all tanks
sounds like a good idea!
Hey, I'm having issues getting my localization and icon to show up
Also, I see this in the log
Probsbly check the partName if its correct
Oh thats a diffrent icon!
Munix can u link the icon table
?
WTH Clyde you jack ass
WTH is going on? I can't just copy and paste without freeking Cyde getting in the way and being a total jerk
Is anything I'm sending getting through or is freaking Clyde blocking everything now?
LOL bizare!
btw I imported the stock tank model into 3DS Max and estimated the internal volume to be ~0.03 m^3, so with the mass being 30 kg, the density would be around 1000 kg/m^3 (and Wikipedia is telling me that the density of xenon is ~6 kg/m^3 at standard temperature and pressure, and ~2900 kg/m^3 when liquid, so 1000 kg/m^3 is probably a believable value for the compressed gas)
Good to know!
Have I possibly put my part icon in the wrong place? The guide said to put it in the Assets folder, and that's right where the JSON is.
and the dry mass for your tank should probably be slightly larger than the stock 3.75 kg, since it has more surface area
what's the addressable name of it?
Well, that was probably it. I never clicked the addressible check box for it.
It's checked now, but maybe I need to loose the leading Asset?
yeah, and also probably the png extension
since Lux's docs only say "partName_icon"
yeah, you can try this way, and if it doesn't work, you can try again with .png
I'm not sure about this step, since I haven't done this yet myself
oh yeah and the Staging Icon Asset Address should be one from this list: https://wiki.spacewarp.org/wiki/Staging_Icon_Asset_Address
so Staging-Engines/Staging-ICO-Xenons.png in your case
OK, easy fix. But still no joy on the localizations.
did you set the asset type to sprite in Unity?
I did not.
Where is the asset type set in unity?
could you just send the file?
one issue I see is this: Parts/Manufacturer/SPT100,Text,,Stellar Plasma-Assisted Rocket Kinetics, inc.
you have a comma inside the text in the right column
so you need to put the text into quotation marks
since the comma has a special meaning in CSV
(delimiting columns)
not sure if that's the issue, but it's the only thing that I see is wrong
awesome
It looks like some sort of innertube based pool toy, but it works!
Localization is still not working. I'll look in the logs for more clues.
oh I just noticed
like i said earlier, the folder needs to be called "localizations", in plural
there are a lot of places (including Lux's site) that have it incorrectly as "localization"
I though i put it right 😭😭
sadly not lmao
but you're not the first nor the last tbh
SpaceWarp docs said "localization" for so long
Need to make it shorter....
does it fit if you press shift?
if so, that's fine
even the stock part descriptions don't fit on the half-pane layout
I wonder if it actually produces thrust? I mean, sure the plume is in stealtyh mode, but does it actually propel?
It does, though I see some typos. But of course fixing it is dead easy now!
nice!
I think I see how KSP2 is working out fuel consumption rate. You give it Isp and thrust and it sorts out the consumption rate!
yeah, that should be it
So, in that case stealing the number for EC from Dawn was probably the right thing to do
it's really exciting finally seeing some more modded engines (other than SORRY)
And my little engine needs a nearly perfect 5 EC/s!
Should we maybe add a warning??
A warning that there are more engines? For this little thing that I still need to re-model and then actually paint?
I mean, do people actually use the SW docs anyway? lmao
I'd really like to see some sort of visitor statistics
because I think most people make their parts with the help of your videos and guide
I did! I think we should still be coalescing the best guidance we can. There will be mroe and more folks showing up who want to do some parts moding
the SW docs are pretty outdated to be honest
but nobody likes writing documentation

True.
I don't mind structuring my code so that it's as self-explanatory as possible, and adding doc comments to public APIs and all that, but writing out actual documents with descriptions of classes and methods, and guides on how to use them? No thank you 
No i meant in the logs
hm that's a good point
I was thinking the other way around
let's just enable both
Just like, “you might have missed this”
Would that work? Lmao i think it uses cheeses loading
technically we could just add a second call to LoadLocalizationFromFolder
it's kinda nasty though
there definitely should be just one standard way, but I think we chose the wrong one
Lets move to #🔴space-warp
Hey, I'm wondering if my scaling is wrong in unity? Also, I just went to attack an SPT-100 to the bottom of a small Xenon tank and it wanted to by default put it facing the wrong way.
Doe something that should be 0.3125 diameter, on a size small tank, does that not look too small?
That seems the mm to m thing
On unity
On your mesh file
You can set the import scale
small size tank as in the "SM" size?
Either way looks way too small
the engine should fit 4x across the 1.25 (SM) tank
so it seems alright
at least when I measure the diameter of the tank and the diameter of the engine on my screen
the tank is ~4x larger than the engine
which is correct
that does indeed look right
So it's OK
do we have a 0.625 (XS) engine mount?
I think not right?
because that would fit pretty nicely in 4x symmetry
Like a 9 engine into XD
Look! A Mini Tie Fighter!
with some large solar panels
LXS
Like the poodle
heh that's great
It's got everything it needs but some science parts, and maybe a radio...
and you're even correctly using solar panels and ion engines!
Darth Vader would be proud
Also im gonna make a skirt one
I'm gonna take this to space and see if it can actually change it's orbit!
Houston, we have a problem...
It fell apart as soon as it hit orbit. Possibly because I lazy orbited my way up there?
Tried it again, and it was stable for a little while, then litterally wiggled itself apart
Did u turn the engine on?
By pure chance it pops me into orbit < 20 Km from the only other craft I have in orbit
I didn't even get the solar panels extended...
I would try just launching it and seeing what happens
I'm gonna try it again
What? And not cheat? How can the maiden voyage not involve cheating?!
WTF!!!!
Have you ever seen this!?
That's a freaking Mainsail on full power at liftoff here
And no what? It levitates?
Seems like
Mods that mess with lift off bug it
Do it manually
Specially if you have modded engine
In between taking those videos for a while it reached a state of equilibrium and was just hovering, then next thing I know the maisail is just gone
Nope same effect without using K2D2
Size and visual size at 1
OK, I was literally copying the Dawn with those, but sure. I'll change em. Why would those make it pogo?
I remember I had something similar happen with my Super Heavy, remember? lmao
it lifted off fine, and then started dropping like a second or two later
Check if the dawn had a reinforced joint module
Some parts do others dont its so wierd
you probably shouldn't be (publicly) sharing this here, even though it's just a single part definition JSON
it still technically counts as redistributing of game assets
thanks!
I'll take a quick look at the joint
It doesnt
not a multijoint
Its so wierd
I guess it doesn't need it because it's so small?
My engine is more powerful than a Mainsail!
Not in a good way , but I can keep that thing chained to the ground it appears
Its more so about how much the part move sideways
Like, gimbaling engines have a lot of torque sideways
My engine has no gimbal. Though it's at least theoretically possible to do thrust vectoring on a HET
Hey I found out what happened to the mainsail... When the first stage runs out of fuel the whole thing falls 20 or 30' to the ground and bang mainsail gone. After that it's just hovering in space exactly like a brick wouldn't
No thrust, just hover...
that's... interesting
It's doing the same thing with a Dawn in place of the SPT-100
So it's not my part
Actually... the Dawn is ... gone
That's the top of the second stage there... No 3rd stage probe at all. It's like it was totally blown away. Missions of voices cried out in pain, and were silenced.
That's gotta hurt!
Can ya see the solar panels? Darth is trapped in the top of the second stage!
Closed the game and reopened to get find this
It's like orbiting venus with the protomolecule down there...
another day, another slay bug ™️
I'm gonna try a fresh clean campaign.
Protomolecule builders have entered the chat...
Nope. A fewsh campaign does not fix it. Completely fresh buiolt from scratch rocket too
is your game just stuck like this now?

you could try disabling your mod and testing without it
but it'd be weird if it managed to screw up the whole game this way
No, it would be powerful!
Well, powerfully weird, but still... powerful!
How can a codless mod have an effect like this?
no idea
@mellow patrol Did you know, it's not possible to diable a codeless mod?
I can do this
and then quick and restart to get this
Seriously. It's not possible to diaable SPARK. You see, SPARK is like a firmware virus... Once it's in, it's in - and you can only remove it by flashing your BepInEx/plugins folder
See?
That's me changing swinfo.json to swinfo.json.bak in the SPARK folder
uhh
@coral orbit
H-huh‽
@narrow gate this is how you setup your csv that goes in localisations.
I thought I checked the disabled mods against the codeless mod id
I am happy to report it is not my mod.
Hmm, It'll likely be fixed in 1.5
1.5 has a complete rewrite of the mod loading stuff
I regret to report I have no clue what causes this as I just got a pogoing launch vehicle with a Dawn on the probe and SPARK not loaded
how
as a last resort I would try to temporarily move out/rename the winhttp.dll file in your KSP 2 folder
Yes! How!
that will disable BepInEx and by extension all mods
to test if it still happens without mods
I'll need to rebuild the dang rocket... I was using Kesa solar panels
It's mods.
I'll work out which one(s)
yeah, best to just do a binary search
Binary searches are hard with mod dependecies.
eh, there aren't that many
My guess is it's Kesa, as that's the only other part mod I've got in this game
It's Kesa - specifically if you've got the DSPF-S1 solar panels.
I was able to launch a rocket with stock solar panels, and Kesa loaded as a mod, but when I reverted and attacked the Kesa panels I was using I see pogoing
interesting
Not sure about other Kesa parts, but those are bonkers
@broken raptor you should probably know about that, the DSPF-S1 solar panels seem to cause this: #1141807690264367135 message
Not just the DSPF-S1's, but also the smaller DFSP-S2... Which is interesting! Why is one a DSPF and the other DFSP? They both should be DFSP - Deployable Flat Solar Panel
The Deployable curved ones do it, too.
And the DSPF-L12 (which might also need to be the DFSP-L12?)
But not the non-deployable ones! At least not the FSP-XS1
Also not the CSP-XS1, or the CSP-SM.
So, it's just the deployable ones that cause the issue.
Made it to space... Finally!
Not sure I'm makeing any thrust at all...
Maybe I am!
Timewarp is, of course, utterly useless.., but it seems that with the mighty SPT-100 turned on and at full thrust near the PE, the AP is in fact rising!
it does indeed seem so
So... I need a plume! And I need to remodel this thing to be lower poly... And I need to do a materials painting... A few things, and then I'll work on the next few engines. Those should go much quicker with the process sorted out.
I may very well put in an Epstein drive just for kicks, but I'd need to figure out how to make a nozzle bell. That, at least, might be able to use an existing LFO plume as a surogate. It would also give me an excellent reason to go back and re-watch the Expanse - you know, for research purposes...
@chilly minnow what does it take to add an engine shroud for when an engine has something attached on the bottom node? It’s just another mesh that get’s switched on, or what?
Also, since my little thruster is apparently working I’d like to work on a plume. Is that something I can configure using LFO?
ME says thrust is 0. Can you try increasing number of decimal digits? Click on the gear icon, click on increase digit buttons. 5 digits are max at the moment
And it shows thrust in kN, which is weird cause it should switch to N for small numbers.
The thrust should be 0.1 kN for that little engine. It's odd since the Ap was definitely rising and not merely oscilating.
Just noticed that dv and TWR are also reporting 0. I could try taking a look at what the game is showing for those numbers, later in the evening.
Here's what I get with a stock Dawn engine on the same craft.
I'm gonna relaunch with my SPT-100 and double check this with my current minimal set of mods
Hang on, what's this!?
I've got an SPT-100 in the stack, but it's not in the stack
It does not. It looks like this
Granted, there's no shroud... But the stack looks OK
Actually, there is a shroud, but it's transparent aluminum...
Also massless aluminum... Very very thin...
However ME is not showing the stage info for Stage 3...
It's nice to come into here and be like "ah yes, i feel your pain."
Its an electric engine w/ no charge unless it has sun right?
In that case ... its default delta v is zero?
no, iirc it says the dV but nothing happens when thorttled
It uses electricity, but it still must use propellant as fuel. There is no going around Newton's 3rd law. Well, unless it's a solar sail... or if it's magic.
Ah, I thought this was a solar sail mod for a second, thats my bad
I can't help here, sorry. ME isn't doing any calculations for dv and thrust. All that it's doing is displaying data that the game spits out as stage info. I would assume it's something to do with what and how modules are set for the part, but I have little experience with part modules so could be way off.
I assume this is some kind of goof on my part with the Part Core Data or Module_Engine config.
The craft shown above has a battery and solar panels. When it's on the launch pad it would start with full charge and the engine isn't draining any of that when it's in a deactivated state. There's a tank of Xenon in there right above the engine, plus the engine has its own internal tank of Xenon. So there should be no shortage of fuel or EC when on the pad. Here's a quick test with it mated to an even larger Xenon tank (PB-x750 with 0.7t) and a freaking KR4-P3 Nuclear Reactor that's enabled. There should be EC from the launch clamps to, I suspect. In the VAB the stage info is showing blank on Delta V (not 0, blank like it's null).
If I slap a Dwan on the same rig it shows 16,990 m/s delta V
your engine is not consuming any fuel
Another clue is that with the Dawn, the part info shows thrust limiter slider and an independent throttle switch, btu with SPT-100 we don't get that.
"engineType": "Electric",
"propellant": {
"mixtureName": "XenonEC",
these should be your configs on the Module_Engine
XenonEC is the "compound resource" (they call it recipes) of Xenon+EletricCharge
The Dawn JSON had that as "XenonEC"
I copied that, was it wrong?
I'm sure there must be an error in there, and that could be it
oh you also have XenonEC?
then its right
and the engine type is set to electric correct?
can you send me the JSON for your part pls?
remove the container btw
i mean
either keep it and andd Module_ResourceCapacities
or you remove it
but im not sure that engines with containers work
That could be it. I may have added the wrong thing then.
I removed the container and aligned as closely as i could with the Dawn, even adding a fairing, but none of that had any effect. The fairing did not appear when I attached something to the bottom node, and I’m still not seeing any Delta V for the stage.
Hy. back from Holidays. I saw this bug, with all deployable solar panels. Normaly, v1.1.10 (pushed 15/08) solves this problem. Is'nt it ?
I thought I had all my mods updated and did the test on 8/20, so it would have been after your 8/15 release. I’ll check when I get home
He has been helping a lot! I think I got confused again when I re built for the new model. I’m watching some more videos on it now.
I think I'm getting somewhere with the bakes for this thing. Here's the render - it's not pretty
The material preview looks nice IMO - which probably means I'm doing something wrong with the Blender settings for render
Here's the bake for the base color
Here's metallic
Here's AO
Here's paint base (though I've done nothing with it and this is the default out of TexTools) - I'm thinking there may not be aby parts of this model yet that need user painting. If I add some sort of base or standoff that fits it to a 1.25 m base then it would make sense to paint that I think - but that's not in the model yet
Here's the bake for the object normal
This kind of engine may get warm or even hot - but should not ever get so hot that it glows. There would be a lot of current flowing through it at moderately high voltage - so the magnetic windings, which are partly visible could get pretty warm, but I don't think I need to worry about emission. If I did, it would be from the windings... Hmmm... Why not then? Here's an emission bake with only the magnetic windings selected.
Must be doing that wrong... It looks like this in Blender
A dumb question: How do I persuade TexTools to use the base file name I want for baking images? All my images default to "anode_1_<bakename>". I want them to be "spark_x3_<bakename>"
these are looking awesome!
Here's a render with the world light level bumped up to 10
The real life object I'm modeling is this
In real life the magnetic circuit (the metalic looking rings) are a ferromagnetic material like iron or steel, but in the actual thing they found that covering those with a boron nitride paint helped to reduce some wear on parts of the engine. Boron Nitride is the ceramic material used in the discharge channels and is white.
I thought about making those rings out of the same material as the discharge channels, but felt it looks a little bland that way.
@chilly minnow Here are some plume images from the actual X3. They're green when the fuel is Krypton, but blue when it's Xenon.
They all have a tight bright central part that is from the central cathode, and then there is a diffuse glow from the channel that they're energizing. In real life they pick which channel(s) to use based on the trust and Isp you want. For the highest Isp I think it's just the outer channel, and for sure it's just one channel. For max thrust they turn them all one. In the game I'll just turn them all on as that's far simpler (and probably better looking).
Thats way better volumetric yeah
You can still do it with the mesh based tho
The ion beam is actually diverging slightly, but the illuminated part is following the cathod's electron spray so it looks like it's converging slightly
Can I make the there be a strong white (or bluish white) glow appear from the dark black anode ring recessed in the discharge channel?
If there are hotter parts in this they may be the anodes down in the channels and the magnetic windings visible in the perimeter windows. I have no idea why those square windows are there - possibly just for easier inspection of the test asset? In a rill one there would be magnetic windings like those bracketing each discharge channel, so 6 windings for a 3 channel nested thruster like this, but only the outer channel appears to have these windows? No idea why yet.
You can do whatever you want with it 🙏
Literally
Its pretty much all customizeable
So yeah you can do that
OK, a better question then would be "how do I make it look like that?", but I need to get going to work so I'll save that one for later. I need to get the fbx and texture maps into Unity and mess around with the LFO plume tools you made first I think.
Thank you for those BTW!
Well as for mesh based, s sec ill try smth
And for volumetric, it isnt ready yet but ill see if i can make it for you
With the current version
Awesome! These tools you're making are gonna be so useful! Not just for me I'm sure.
Couldn't resist adding a support base to make this truly a 1.25m part
Making the lower rim and this simple support arm (arrayed at 60 deg rotation about the origin) is a breeze, but what about the other support arms needed to make this more like a real truss? I'm not certain how to get them rotated into the right direction... I think maybe a two step rotation. I could duplicate the one arm that's rotated 45 degrees about y and then rotate that dup by something about x?
Yikes! This my array is all messed up for the tilted arm!
This is better!
Hmmm... why is my render looking goofy? Why is the lower ring partly black?
I may need to add some plumbing under this. There is probably a power conditioning unit with some cabling, and there needs to be some plumbing for the propellant going to various points below the anode rings where it would be injected.
I wonder if the support arms should possibly be something other than steel - something not ferromagnetic. There are some nutso magnetic fields in this which might be "why" it's on a standoff - to keep it further away from the rest of the vessel. Fortunately magnetic field strength falls off at an inverse cube instead of just an inverse square...
Looks amazing, cannot wait to use as I have been wanting more ION stuff.
Thanks! It's really pretty fun to make. The blender part takes me back to working in CAD tools, but I've got a long way to go on the texturing part. Fun stuff!
that looks wrong
Oh object normal
yeah youre supposed to bake tangent normal
sorry i should have clarified
Hy munix. Thanks for your support. Could you confirm me that latest update fixes this bug ?
I personally didn't experience the bug, but I can definitely try to test it now! And then I'm sure when @narrow gate comes back, he'll be glad to test it for you!
no problem. sorry. Schlosrat will test ti later 🙂
i test on my computer, then no bug. I'm thinking it is solved
I just tested it out and it worked fine for me, just so you know
Thanks munix. That's a good news. 👍👍👍
Hi Jim, I'm running Kesa 1.1.9. I assume this is the same version I had on 8/20 when I first observed the issue. I'll see if I can reproduce this now.
I'll be testing with this mod load - SW has an update, but I know that wasn't out when I first observed the issue.
This is the test vehicle: All stock parts other than the 4x DFSP-S2's.
Confirmed, bouncy bouncy!
As I recall in my previous test, this effect was observed with all of your deployable panels, but not with the non-deployable ones.
Curiously, I see that in CKAN there is a Kesa update! I see 1.1.11 is available, but somehow in SW Mod List Kesa is not reporting as out of date. Maybe that too is fixed in the new version? I'll update rn and re-test.
does it happen w/o the panels?
I've only seen it happen with the Kesa deployable panels. Stock panels and Kesa non-deployable panels don't do this. I didn't test without any panels, but presumably that would be the same as stock.
Testing updated now
Confirmed fixed in 1.1.11!
that's because the mod has no public (as far as I could find) GH repository and no version_check URL in its swinfo.json
it's not really a "bug", it's just a lack of an optional feature
@broken raptor You might want to set up a public GH repo and configure your version_check URL. I just assumed you had that and so didn't realize I was running an out of date version of your mod. There could be others in the same boat.
i'm sorry. this is new for me. I have to find how to do this version_check and link that with the github.
is pretty easy
it's in your swinfo.json file
and you have to link to an up-to-date version of the swinfo.json file that you have somewhere online
most people keep their latest release in their GitHub main branch (and develop in a "dev" branch), and then just link the swinfo.json file from the main branch
for example, UitkForKsp2:
"version_check": "https://raw.githubusercontent.com/jan-bures/UitkForKsp2/main/plugin_template/swinfo.json"
it needs to be the address to the raw file, not an HTML page
then what SpaceWarp does is it downloads this file at startup, checks if the version number on GitHub is the same as the currently installed mod version, and if the online version is newer, it lets the user know they can download an update (using the yellow text like SpaceWarp here: #1141807690264367135 message)
@serene abyss do these bakes look OK?
I did an emission bake to, but it's all black as I have no emissive parts
Are there any other backs I'd be likely to need? You mentioned metallic smoothness, but TexTools just has Metallic. There is also a glossiness and a sheen, and a roughness. Maybe Rougness is another one I need?
Is there a way to make it default to PNG each time so I don't have to switch it from Targa, or should I be preferring Targa?
There's the roughness bake
@chilly minnow I'm trying to follow your guide for applying textures, and I'm a bit confused about which bakes go where. The baking options I ame aware of in Blender using TexTools are these
I've created these bakes so far
I think the _diffuse one may be the right one to use for Albedo, and it seems to be happy there. The _metallic one seems to be quite happy being a Metallic Smoothness map. The _ao one seems to be happy as an Occlusion map, and likewise for the _emission texture. However neither of the ones I've got with "normal" in the name are happy in either the Normals Map or Detail Normals Map.
Of course they are!
you almost 99% will use only the tangent
which u already have
its just on the wrong spot!
Does it go on Detail Mask?
Poo
It doesn't seem to like being there
You did say put the normal_tangent on Normals, right?
Also, I'm unclear on how to apply this material to everything. I unpacked the prefab (to delete some things I don't need and that weren't in the bakes anyway) and so here's what I've got
I've still got the old (doesn't work quite right) SPT100, but I've begun a new part which I've cleverly named spark_x3.
Since it's a model of the X3 Nested Hall Thruster
I tried dragging the materail onto each part and I've got to wonder if the UV map isn't completely hosed or something cause this is what I get
Some parts do look metallic, but most are way to dark, and seom are just way wrong
Thanks! I wasn't sure if it would be safe to do that or not. That did make it better, but it's still not right. There must be more for me to do
id say ur uv unwrap is wrong
I kinda forced it to be what it should be. Not actually using the textures, just the crude built in Unity capability to paint
My UV unwrapping is very possibly wrong. It looked OK, but I honestly put in 0 seams and just used UV Smart Project
I'll take another crack at that tomorrow. The baking wasn't horribly long at 1024 wiith x2 oversample
uv smart project only works with seams afaik
The real trick will be to see if I can figure out where I went wrong with this on SPT100 since that had such goofy results - the game not thinking it was really there as an engine.
I'll put in some seams and then give it another shot.
nope
Well, seams might not be necessary, but they might also help it to lay out more compactly. I can also give quixel another shot. Last time I’m not sure what I did wrong but using the material id map didn’t give the results I’d hoped at all.
Mostly though I’d like to get this to be a functioning engine with a plume, then I can go back and make it prettier
Maybe now Unity will be able to color it right? We'll see!
Nope... I'm still doing something wrong, but I don't know what it is... Some things come trhough OK, and some are just weird. They all got baked one after another using the exact same UV Unwrap, and the FBX file was exported right after that.
I rebuilt the texture in Unity, and applied it to each object by dragging it onto them in the hierarchy
@chilly minnow , @serene abyss do my bakes above look OK? I dialed metallic down to 0 for all of them except the metallic bake where they were set back to what they had been (1.0 all cases except for one 0.9 on the actual metallic parts). I used Metallic 0.5 for the anodes and windings in the emissivity map, but it still came out to be all black.
The weird thing is that some parts look OK while others that are using the exact same materials don't.
is the pink blue and yellow one your paint map?
if so you made it wrong, a paint map if done correctly should look white, teal and blue
No, that one is a material ID map, I'm not actually useing it in Unity but would need it if I want to take it to quixel
ah
does the metallic map have smoothness data in the alpha channel
The normal map has a "fix now" button in the Unity inspector. I just clicked it, but nor real change.
also you need a paint map
The metallic map is straight out of TexTools Metallic bake without any changes. The only thing happening there is that I set my metallic sliders back to what they were
set metallic smoothness map to one
also yeah you’re gonna need a paint map
and also to put smoothness info in the alpha channel of your metallic map
and into the alpha channel of the paint map
the ksp2 shader requires a paint map i think
What if I don't want any parts to be painted? I mean, yeah, I do want that, but what if I'm lazy rn and don't want it rn, can't I skip that?
Oh
So I could just take the all black emission map into photoshop and make an all blue (is that the base, right?)
Yes
In which case I think nothing is paintable, right?
but remember to put smoothness info in the alpha channel of the paint map and metallic map
(this is required)
How do you do this?
i dont know how to in photoshop
but in gimp (maybe this will help since they are similar software) i import my smoothness map, copy it, create a layer mask on the metallic map/paint map, and paste it in
OK, so the metallic image needs to be layered with the paint map, and I'll need to find a blending method in PS that applies this to the alpha channel. Does that sound about right?
metallic and paint map are separate
but both need smoothness in alpha
smoothness map is just inverted roughness
also can be called glossiness map
textools can bake it
Nuts, I forgot to make that one. I bet I have to go back and horse around with my metallic sliders before making it.
glossiness is inverted roughness which is the same as smoothness
I bake that with the metallic sliders all at 0, right?
Nice
now use that as an alpha channel for the paint map and metallic map
this is the biggest reason i dont use blender for texturing anymore
it just gets so annoying baking things
Also, I made this from the TexTools Base Color bake
No, hold on. The one above is glossiness. The one i'm about to post isn't. 🙂
ah
I thought I'd try it out to see if it was useful in the paint layer process. Probably not, but just messing around
I think one thing I could do is to take the material ID map into PS and do a select and replace based on colors. That might make a paint map moderately easy. I'll think about that.
Also, I could give Quizel another go. It seems like I've got a good UV unwrap and a good material ID map, so theoretically I could apply materials there. Would that get me bakes that have the alpha channel set correctly?
nope
only substance can do that
it has presets for unity shaders and benjee10 made a preset for ksp2 shaders
gotcha, thanks
"How do you convert an image to an alpha channel in Photoshop?
Once you have made a selection in Photoshop, you can save it as an alpha channel by navigating to the Channels panel, usually next to the Layers panel. Click on the Save selection as channel button at the bottom of the panel to create a new channel named Alpha 1 which will display your selection as a grayscale image."
Looks like it should be simple. Bring in two images the same size. Convert one to alpha. Merge down and save.
Gonna try it now with metallic unmolested and glossiness as alpha
I may have done this right, the resulting image doesn't look much different to me. That said I'm having a devil of a time making the stupid paint layer. It's probably easier than I think it is, but my photoshop skills are not the best.
I've got a layer with just red for the accent paint, a layer with green for anthing that get's painted in base or accent, and a layer full of blue. You would think it would be trivial to combine these as they need to be. You'd probably bve right too, but my google-fu is failing me on this and the simple steps are eliuding me.
OK, I think I've got a paint map now. What a PITA.
How's this look?
Maybe?
I think that's got an alpha channel from the glossiness image, and I think I've done the same with the Metallic
Neither one really looks different to me with the alpha channel. Here's Metallic without that
See? Looks the same to me too. Should it look different?
Here's how I've got those maps configured in Unity
No effect. None at all. Do I need to reapply the matierial? I've previously applied it to everything, and some things frankly look horrible - like the texture map is tiles across them. In particular the four hexagonal boxes look terrible, All of the veritcal supports in the base look terrible (though the diagonal ones look OK), The outer magnetic circuit looks terrible, and 2 out of 3 of the ceramic dicharge channels look terrible. It just makes no sense whatsoever. It's not like the problems follow particular materials so far as I can tell - though it isn't entirely random either.
Just looking at my paint map I can see it's wrong... It's fixable, but it's wrong. The copper colored parts from the diffuse map should all be blue - no paint allowed.
That said, I don't really see much point in fixing the paint map. The part of the process where thigns are brought into Unity to make the part works in some ways, but the whole texturing thing is just FUBAR.
What should be simple, it not merely hard - it's impossible.
There's nothing wrong with my UV map. It's clearly OK. There's something completely wrong with what Unity is doing with the information I give it.
it doesnt have an alpha channel
some parts should be transparent and have variable opacity
make sure youre exporting as png with transparency
That may be an issue, but I suspect it's a small one relative to some other much larger issue(s). For example, I get the same absolute crap importing it into Quixel
The same crap on the same parts. Perhaps there's an issue with my settings going into TexTools, because the exact same fbx with the same (and more!) textures give the same crappy result of taking the map and wrapping it like some kind wacky wallpaper all over some parts, but then mapping the right parts to the right faces on others. If it were a problem isolated to just some materials, I'd assume that it's an issue with the materials, but the fact that 2 out of three large ceramic parts that should be white are looking like urban camo instead
hahaha! I've applied them so many times it's not funny!
hm
I've done select all and applied them that way
i got one last thing for you to try
select all your objects
go to edit mode
in face mode (3), go to face > triangulate faces
this will make it harder to edit your model though
you can convert them back to quads though
and reexport the fbx
Where is triangulate faces?
you’re in face select mode right?
Do I need to do a select all first?
yeah
Expected?
OK, re-export and then bake 'em all again?
so no export issues should happen
nope no need for rebake
itll retain the uv map
so you can keep yhe current textures
Yeah
Except that apparently I've not done the alpha channel correctly.
yeah
No clue how to fix that - I followed a guide online for PS, but if that didn't work, then I'm not sure what went wrong.
what are you exporting as?
I'll say this though, making the paint map with PS was a bit harder than it should have been..
make sure its png with transparency enabled
PNG
And hide any other layers
I do a lot of exports from PS in PNG, always with transparency - but I'll double check that
because if theres a solid image behind the paint map layer it will export the unerlying layer’s alpha too
Good to know!
youll know you did it right if your image has varying levels of transparency
some oarts are clearly visible others are almost completely transparent
The same goofy wrap of the texture at some goofy scale and orientation around the same beautiful parts (they're beutiful under that goofed up textures - I swear it!)
The blend is 30 mb, but I can put it on github I think with the proto version of this mod. Which files do you need?
If you're willing to take a look at it I'm more than willing to share it!
@narrow gate i found the problem
you have 2 uv maps, one is the wrong one, the other is correct
So to fix this
go here
and delete UVMap from each object
and reexport
automap is the correct uv
your uv maps shouldl ook like this on all objects
Not all actually
some only have one uv map
if so lkeave them be
the ones with automap and UVMap should have UVMap deleted
that should fix it
i had that too
unity (and a lot of other programs) hate more than 1 uv map
afaik only XBOX360 has native support for more than 2 UV maps
Thank you! Thank you! Thank you! This was driving me bonkers!
Should I go back to the model with the quads? I caved a copy before I dis the triangulat faces.
up to you
It was such an easy fix, I went back to the quads. Thanks!
I have no idea how it got like that, but at least know I know what can cause something like that and where to look for it. If I do see something like that again I'll know right away what I need to do.
Hmmm... What did I do to my part?
It seems to be a fluorescent popsicle now... Possibly grape flavored?
That means there was an error during the part crestion process
Aka error on the json
Or on the modules
Ah, good to know! I'm going to scrub my process. One curious thing is that this affected both the previous part I built (SPT100) which is still in SPARK and the X3 shown protruding from the purple push-up popsicle above. Could an error in one impact the build of the whole assembly?
I'm going to document my process. It's slightly modified from your video (https://www.youtube.com/watch?v=9fQg-oMqcH8) based on some things @eager light has told me but is basically the same. I was going to do that anyway to help me as I make more parts in this parts pack , but I will not be surprised if that helps track down whatever I did wrong last night.
ThunderKit Tutorial: https://www.youtube.com/watch?v=EVAFXPNmBWU
KSP2UnityTools: https://github.com/SpaceWarpDev/KSP2UnityTools/releases/latest
@chilly minnow , @eager light, @serene abyss . How does this look for a parts pack production process?
Oh wait, I left off the whole icon part... I need to add that. I'll update it and post again with that as part of the process.
I dig it. thats well thought out and flows well for a new person who wants to learn to mod. everything you need to start parts modding really
Updated with typos fixed, personal info removed, and icon process added.
It's as much for a personal check list and reminder for me as I go along as anything else.
And on to debugging! For some reason, the old part has the new localization info but has now lost it's icon, and the new part has no localization info and also has no icon. Both should have icons... Time to do some log spelunking!
I wonder if it could be because I was hasty with the new part and haven't configured its Module_Engine fully so it doesn't even have a thrust curve, etc. Can an issue like that cause other parts to fail to load their icon or a particular part to fail to load its localization?
The new part's localization is nearly the same as the old part... No! Wait! It's much to same! I need to put in a unique name for it. I'll try this and see if it likes it better...
Curiously, that did not fix the issue... On to log spelunking!
Clue found! And new step info for process... Must convert PNG to Sprite!
[ERR 11:42:50.371] AssetProvider unable to find asset of type 'UnityEngine.Sprite' and key of 'spark_x3_icon.png'.
UnityEngine.Debug:LogError (object)
KSP.Assets.AssetProvider/<>c__DisplayClass11_01<UnityEngine.Sprite>:<Load>b__1 (UnityEngine.ResourceManagement.AsyncOperations.AsyncOperationHandle1<UnityEngine.Sprite>)
DelegateList1<UnityEngine.ResourceManagement.AsyncOperations.AsyncOperationHandle1<UnityEngine.Sprite>>:Invoke (UnityEngine.ResourceManagement.AsyncOperations.AsyncOperationHandle1<UnityEngine.Sprite>) UnityEngine.ResourceManagement.ResourceManager:Update (single) MonoBehaviourCallbackHooks:Update () [ERR 11:42:50.432] AssetProvider unable to find asset of type 'UnityEngine.Sprite' and key of 'SPT100_icon.png'. UnityEngine.Debug:LogError (object) KSP.Assets.AssetProvider/<>c__DisplayClass11_01<UnityEngine.Sprite>:<Load>b__1 (UnityEngine.ResourceManagement.AsyncOperations.AsyncOperationHandle1<UnityEngine.Sprite>) DelegateList1<UnityEngine.ResourceManagement.AsyncOperations.AsyncOperationHandle1<UnityEngine.Sprite>>:Invoke (UnityEngine.ResourceManagement.AsyncOperations.AsyncOperationHandle1<UnityEngine.Sprite>)
UnityEngine.ResourceManagement.ResourceManager:Update (single)
MonoBehaviourCallbackHooks:Update ()
Icons work!
partName should be the same
theres no more than that
Ahh! That explains the localization glitch, and more more improvement for the guide doc!
@narrow gate is your document meant to be missing an example build/load path for addressables?
Not, it is not meant to be that way - good catch! I meant to fix those but got side tracked.
I was being very hasty last night and cleverly thought to myself: "Self! You don't need to configure all these fiddly Module_Engine parameters just to be able to view the part!" Oh how wrong I was...
I'm getting this error when I try to grab an SPT100 and drag it from the parts picker into the VAB for use (whioch it currently doesn't do and used to do just fine)
[Error : Unity Log] NullReferenceException: Object reference not set to an instance of an object
Stack trace:
KSP.Modules.Module_Engine.SetThrustTransforms () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.Modules.Module_Engine.OnInitialize () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.Sim.Definitions.PartBehaviourModule.Init () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.ObjectAssemblyPart.FinalizeModules (UnityEngine.GameObject newObject, KSP.OAB.IObjectAssemblyAvailablePart availablePart) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.ObjectAssemblyPart.FinalizeLoad (KSP.OAB.ObjectAssemblyBuilderEvents events, KSP.OAB.IObjectAssemblyAvailablePart part) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.ObjectAssemblyPartTracker.ManuallyCreatePart (KSP.OAB.IObjectAssemblyAvailablePart partToCreate, System.Collections.Generic.List1[T] existingModuleData, System.Boolean spawnInHand, System.Boolean isInAssembly, KSP.OAB.IObjectAssemblyPart referenceCurrent, System.Nullable1[T] guid) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.ObjectAssemblyPartTracker.CreatePartAtCursor (KSP.OAB.IObjectAssemblyAvailablePart partToCreate, KSP.OAB.IObjectAssemblyPart& part) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.ObjectAssemblyBuilder+ObjectAssemblyBuilderEventsManager.UIBind_SpawnPart (KSP.OAB.IObjectAssemblyAvailablePart part) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.AssemblyPartsButton.CreatePart () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
UnityEngine.Events.InvokableCall.Invoke () (at <25a85da7c6f04932b86e339dfd12957d>:0)
UnityEngine.Events.UnityEvent.Invoke () (at <25a85da7c6f04932b86e339dfd12957d>:0)
UnityEngine.UI.ButtonExtended.OnPointerClick (UnityEngine.EventSystems.PointerEventData eventData) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
UnityEngine.EventSystems.ExecuteEvents.Execute (UnityEngine.EventSystems.IPointerClickHandler handler, UnityEngine.EventSystems.BaseEventData eventData) (at <56e27d38acd347119ce15709c80233a2>:0)
UnityEngine.EventSystems.ExecuteEvents.Execute[T] (UnityEngine.GameObject target, UnityEngine.EventSystems.BaseEventData eventData, UnityEngine.EventSystems.ExecuteEvents+EventFunction1[T1] functor) (at <56e27d38acd347119ce15709c80233a2>:0) UnityEngine.EventSystems.ExecuteEvents:Execute(GameObject, BaseEventData, EventFunction1)
ModifiedInputModule:ProcessMousePress(MouseButtonEventData, MouseButton)
ModifiedInputModule:ProcessMouseEvent(Int32)
ModifiedInputModule:ProcessMouseEvent()
ModifiedInputModule:Process()
UnityEngine.EventSystems.EventSystem:Update()
I removed both the Gimbal and the Generator from the X3, and the Gimbal, Generator and Faring from the SPT100
oh
I get a similar error now with the X3, but for a while it was letting me drag in a purple pushup pop
Where would those be? Are they part of Throttle VFX Manager?
Do I need a Thrust curve?
I believe the atmosphere curve is controlling Isp
also not that
but
just add a empty game objects
on your engine
and call it
thrustTransform
and the game should pick it
On the root part, or some other point?
Like this?
Yay! Engine!
With like Delta V and everything!
Everything but a plume 😦
I'm gonna launch this sucker and see what it can do!
It will propel me by MAGIC!
I kinda like the look of it, too.. Turned out nice! Will be even nicer with a sparkly plume...
Hang on... I'm getting a bunch or errors in the log... I knew it was too good to be true! Progress at any rate. This is much further than I was previously!
Getting lots of this in my log
[Error : Unity Log] [General] Object reference not set to an instance of an object
at KSP.OAB.ObjectAssemblyPart.get_SurfaceAttachmentRotationOffset () [0x00000] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.OAB.ObjectAssemblyPlacementTool.PerformSurfaceOffsetAdjustment (KSP.OAB.IObjectAssemblyPart target, KSP.OAB.IObjectAssemblyPart current, UnityEngine.Vector3 hitSpot, UnityEngine.Vector3 hitNormal, System.Boolean snapEnabled) [0x001a7] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.OAB.ObjectAssemblyPlacementTool.SnapSelectionToSurface () [0x00085] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.OAB.ObjectAssemblyPlacementTool.OnUpdate () [0x0027d] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at (wrapper delegate-invoke) <Module>.invoke_void()
at KSP.OAB.ObjectAssemblyBuilder.OnUpdate (System.Single deltaTime) [0x00093] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.Game.GameInstance.Update () [0x000bf] in <57799b60a4cd4df8b3c9aec811d65aed>:0
Makes me think I've botched my node attach...
Uhhh... I put the srfAttach node right smack dab where the "top" node its. Is that, how do you put it... No Beuno?
srf attach is set to stack
set it to surface
so problem is that theres 2 stack nodes in exactly the same spot so the game gets confused probably
I think I see the problem, it was this
Yep!
I was being too hasty! And I should also update the guide to catch this sort of thing...
OK, things seem to be working with the x3, but now the SPT100 doesn't want to show up in game. I'm getting these errors
[Error : Unity Log] NullReferenceException: Object reference not set to an instance of an object
Stack trace:
KSP.OAB.AssemblyCheckEnginesNecessaryFuels.Check (KSP.OAB.OABSessionInformation sessionInfo) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.ObjectAssemblyEngineerReport.UpdateActiveConcerns (KSP.OAB.OABSessionInformation sessionInfo) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.ObjectAssemblyEngineerReport.UpdateReport (KSP.OAB.OABSessionInformation stats) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.OABSessionInformation.UpdateEngineersReport () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.OAB.ObjectAssemblyUIEngineeringReport+<ProcessUpdate>d__19.MoveNext () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
UnityEngine.SetupCoroutine.InvokeMoveNext (System.Collections.IEnumerator enumerator, System.IntPtr returnValueAddress) (at <25a85da7c6f04932b86e339dfd12957d>:0)
SPT100 is the old one.
And it's doing this when I try to pull one from the parts picker in the VAB
It's invisible now!
And giving this error
[Error : Unity Log] [General] Object reference not set to an instance of an object
at KSP.Sim.DeltaV.DeltaVEngineInfo..ctor (KSP.Sim.DeltaV.VesselDeltaVComponent inVesselDeltaV, KSP.Modules.Data_Engine inEngine, KSP.Sim.DeltaV.IDeltaVPart part) [0x0004d] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.Sim.DeltaV.VesselDeltaVComponent.UpdateModuleEngines () [0x00084] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.Sim.DeltaV.VesselDeltaVComponent.RunCalculations () [0x00039] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.Sim.DeltaV.VesselDeltaVComponent.CheckDirtyAndRunCalcs () [0x0004f] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.Sim.DeltaV.VesselDeltaVComponent.OnFixedUpdate (System.Double universalTime, System.Double deltaUniversalTime) [0x00000] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.OAB.ObjectAssemblyPartTracker.UpdateAssemblyDeltaV (System.Double universalTime, System.Double deltaUniversalTime) [0x00024] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.OAB.ObjectAssemblyBuilder.OnUpdate (System.Single deltaTime) [0x000f5] in <57799b60a4cd4df8b3c9aec811d65aed>:0
at KSP.Game.GameInstance.Update () [0x000bf] in <57799b60a4cd4df8b3c9aec811d65aed>:0
Actually, pulling either one gives that error.
Though only the SPT100 has a bonus stealth mode (for when you don't want the competitors to see your engine design until it's too late)
It seems to be producing thrust now.
Perhaps I should watch the 2nd video? I wonder what else I might learn... 😛
Which is probably why I'm having so much trouble...
Were you saying that volumetric plumes were likely to be better for an engine line this? If so, why?
I''m not sure this thing even consumes fuel... It's been burning for a while and has now circularized a lightly less than circular orbit, but the tank is still full.
16 km/s dv
The delta V in the tank is still pegged at 15285, same as when it started.
It's added several m/s to the vessal already, shouldn't the tank be going down correspondingly?
It should, yes
Wait... Doh, I've conflated the altitude readout (m) for velocity (m/s) It may have only imparted a fraction of a m/s, which might be why the tank still seems full. I'll let it just keep burning to see if this changes
It's raised the Pe by a number of meters and is now busy raising the Ap
I started the burn a bit before the old Pe and it circularized the 100x98km orbit nicely. Now I've got a 100.1 x 100.4 and slowly rising
Has dV changed at all?
The first orbital speed I saw (or posted) was 2245.8 m/s from ME)
Now it's 2246.1
So it's raised the velocity by 0.3 m/s?
That's approximately correct given it's a nearly circular orbit. And no, I've got no cheats on - except Lazy Orbit Boosted that I used to get in orbit
Cause I'm lazy that way
Kind of like this engine
Boy engines like this one might just get popular once we can do maneuvers under timewarp.
Hmmm, can you post the generated part json?
certainly!
It's 5 times as powerful as the Dawn
And both are ludicoursly overpowered vs. reality
"propellant": {
"mixtureName": "XenonEC",
"mixtureMultiplier": 1.0,
"ignoreForThrustCurve": false,
"ingredientOverrides": [
{
"name": "ElectricCharge",
"unitsPerRecipeUnit": 2060100.0,
"flowMode": "STAGE_PRIORITY_FLOW"
}
]
},
Whats with the ingredient override?
The real engine it's modeled on would produce maybe 5 N of thrust, not 1 kN
Dawn had it, so I do too.
I think this is what lets it consume EC
In addition to consuming Xenon
"resourceSummary": {
"Consumes": [],
"Generates": [],
"Contains": []
},
@chilly minnow said that this does nothing, but I wonder what happens if you put "XenonEC" in the consumes part of this
I could try that.
Not that I think that is the issue
It likely isn't, probably some messed up parameter on the module
The orbital velocity has gone down, which is just an effect of climing towards the Ap, but still... I think I'd like to let this one run for a few hours and just see what it does to the orbit.
Yeah i mean, it probably uses that for filtering of sorts (like in oab) but functionality i dont think it does
Whats ur isp?
You can check on the OAB the fuel usage too
There are gridded Ion engines that can get up to around 20k
What does RealISPValue do?
xD
Calculated at runtime
Its the isp at said thrust level and atm pressure
I think that either the isp is too high
Or smth
Cuz ur consuming tons per hour
xD
"atmosphereCurve": {
"fCurve": {
"keys": [
{
"time": 0.0,
"value": 4200.0,
"inTangent": 0.0,
"outTangent": 0.0,
"inWeight": 0.333333343,
"outWeight": 0.333333343,
"weightedMode": "None",
"tangentMode": 136
},
{
"time": 1.0,
"value": 100.0,
"inTangent": -1273.98926,
"outTangent": -1273.98926,
"inWeight": 0.333333343,
"outWeight": 0.333333343,
"weightedMode": "None",
"tangentMode": 0
},
{
"time": 1.2,
"value": 0.001,
"inTangent": 0.0,
"outTangent": 0.0,
"inWeight": 0.333333343,
"outWeight": 0.333333343,
"weightedMode": "None",
"tangentMode": 136
}
],
"length": 3,
"preWrapMode": "ClampForever",
"postWrapMode": "ClampForever"
},
"_minTime": 3.40282347E+38,
"_maxTime": -3.40282347E+38
},
This is the ISP curve?
Try burning under timewarp for a while schloshrat?
I can't seem to get a good read on the amount of fuel left in the tank, but I thought it was starting around 1 ton?
Yes
Yep, that's the Isp curve
I'm at maximum warp (Warp factor 6 at my altitude) and I've now obrited Kerbin many times. The tank has not depleted
If floating point roundoff prevents insanely high ISP engines ... I will scream
I think its more so
I got a few of these when Lazy orbit put me into orbit
[Error : Unity Log] NullReferenceException: Object reference not set to an instance of an object
Stack trace:
KSP.Modules.Module_Drag.GetNodeOrientationInAttachCoords (KSP.Sim.AttachNodeData attachNode) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.Modules.Module_Drag.SetExposedArea () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.Modules.Module_Drag+<RecalculateExposedAreaCoroutine>d__44.MoveNext () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
UnityEngine.SetupCoroutine.InvokeMoveNext (System.Collections.IEnumerator enumerator, System.IntPtr returnValueAddress) (at <25a85da7c6f04932b86e339dfd12957d>:0)
That it would take years to deplet the tank
Commencing burn test now
Only a 100 hours per one ton
I dont think that 6x warp would be enough xD
yeah, but it'd be enough to see a slight depletion
Maybe not enough for the ui to show?
I may have found the problem
Oh no
I think its the EC override?
It appears that although I commanded the thrust to full I might not have ever actually activated the engine... When I activate this one...
The default XenonEC recipe is 1:1
Yep, it is
So you need the override
I don't have cheat menu installed...
so you might just have been seeing orbital decay?
Yes! Or orbial raising in this case
I'm going to reset it back to high Isp and try it again.
This time with a fully activate engine... Plumes sure are nice to have!
Interesting, the dawn engines max thrust is 0.2
😅
Yeah, 0.2kilo Nutons... That's insanely high for Ion
The shader is pretty much done as you can see on #1085910413239140416
The engine I've modeled IRL produces about 5 N
The Dawn should produce less than 1
But things are beefed up for Kerbals...
kerbals are just great engineers
They truely are!
I mean, if you think about it, they have loads of data from all the explosions
So they know how to not make smth explode
They apparently know how to bend laws of physics, too!
They use the power of the Kraken to do that, why do you think it gets so angry at people
Ahhh! That explains it!
It was the dreaded Engine activation Bug alrightly...
Just look at all it puts out now
If only I could ajust pointing during timewarp...
There is still possibly this bug issue... And I bneed a plume!
[Error : Unity Log] NullReferenceException: Object reference not set to an instance of an object
Stack trace:
KSP.Modules.Module_Drag.GetNodeOrientationInAttachCoords (KSP.Sim.AttachNodeData attachNode) (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.Modules.Module_Drag.SetExposedArea () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
KSP.Modules.Module_Drag+<RecalculateExposedAreaCoroutine>d__44.MoveNext () (at <57799b60a4cd4df8b3c9aec811d65aed>:0)
UnityEngine.SetupCoroutine.InvokeMoveNext (System.Collections.IEnumerator enumerator, System.IntPtr returnValueAddress) (at <25a85da7c6f04932b86e339dfd12957d>:0)
@chilly minnow what do I need to do to make a plume? I mean, really any kind of plume. I don't care how shitty it may look
yeah, you are really missing a plume
Gosh, luxes plumes have come a long way since the first version of SORRY
Heck, I mioght settle for an indicator light...
First version was… we dont talk about them
An Ion engine is kind of like a hyperactive nightlight...
transparent cones
I mean technically a really short really blue plume would work?
Do you have an ion engine shader?
He will soon!
The thing is viewing angles
Bur first I need to learn how to make something with his tools
Meshbased doesnt wrok well with them, they only look good from the side
So they have to be volumentric?
Either that or complex math to know how they look from below
And above
Honestly idek where to start
How hard can it be? Just write the Looks Great equation, set it equal to 1, and integrate over 4 pi sterradians, amiright?
😅 oh if it was that simple
Seriously, though - maybe we're overcomplicating an Ion engine plume. It shouldn't really be evolving and changing over time. It's sort of either on or off
Voluemtric took me 5 days
3 of those of me bashing my head asking why does the dot product not work
When i get home ill build a ion plume and ull see 🙏
Here are some samples, though these are "burning" krypton
Actuially, the blue one in there is Xenon. My bad
But I'm thinking it's just a few diffuse cones extending out with a brighter one dead center
That thing in the middle looks sexy
That's the cathode! It's pretty, but it's not really where the thrust comes from oddly
Yes, in the middle. That's the electron stray from the cathode interacting with the plasma coming our of the annular discharge chambers
So there's this faint spray that comes out from the annualr discharge rings diverging at about 3 degrees or so, and then there's a shorter bright spike right in the center
Well will be hard to describe
Easier done than said?
Yup
Let me try to draw
The idea here is that this profile weill be rotated 360 degrees
To make a cylinder
So the top one would be the ion
And the bottom one a. Normal plume
You can use greys to tell “less plume here”
So the shorter lines at the top are the light emerging from the annular discharge chambers. It will form bright rings. And the the other thing becomes the plume when rotated. I think you may only need to rotate 180 unless you make a half shape.
I think a half shape would be better as it would have to be mirrored to line up properly
Yeah, but works either way
Yep. So what do I need to do!
I need to make that profiled shader but its easy
Since it doesnt have any math
How will the shader respond to increasing the thrust, will it scale out the plume or is it just on or off?
Oh come on! Who uses an Ion engine at 1/4 impulse? It's full power, baby! Burn, baby, burn!
Fair point, but the question still stands
What if I make overpowered magic engine that has a plume similar to an ion engine but at full power it might be way too much?
No, it's a very reasonable question, but (a) I'm a smart ass, and (b) I really don't think many people will want to dial this thrust down.
That said, in about 30 minutes time I' have changed my nice circular orbit in to a reentry trajectory with an apogee at 141km
Wait until I make a very light lander that can land on a celestial body with this engine
Extremeli lighy
I should make an extra line parts mod
10 grama bateries
1kg lander
This engine is 0.35 tons
Very small probe core
If I make it smaller, the thrust is less