Star Wars: Republic Commando RTX

[dry shoal]

https://www.moddb.com/games/star-wars-republic-commando/downloads/rtxconf-star-wars-republic-commando

O just saw Kim's reply lol

My mistake, I totally forgot to delete that post. There should only be one rtx.conf file under RC on moddb

[proven vessel]

ok so i can load, but i couild one before so im gonna relauncha nd reload a few times

ill keep track of the falure to success rate

so these tests were done by launching the game through steam, selecting my profile, hitting load save (loading the autosave right after tutorial), hit start (all but 1 fail froze between this step and the last), and see if i can move and shoot once loaded in, then closing the game

and out of 10 tries, 5 were successful and 5 failed

so 50/50

alll fails were freezes then not responding

[proven vessel]

@dry shoal

[dry shoal]

Man that's a lot worse than I was expecting

I wonder if I can get another person to test with you

[proven vessel]

might be a good idea, it could just be my hardware setup

i do ahve a gtx 3070 but you never know if windos is fucked or soemthing

ima go eat dinner ill be back later so i can do any specifc tests you want me to do then

[dry shoal]

Right now my target for stability is being able to get from new game to the end of level 1's outdoor area without crashing

If that's possible then I'm okay for now

[proven vessel]

ill check

attempt one, froze on the "war has come to the galaxy screen"

closing game starting attempt 2

attempt 2 froze on title screen when i tried to click continue (first time it has froze here)

closing game starting attempt 3

atempt 3, crashed on loading screen between tutorial and mission 1

[proven vessel]

starting attempt 4

attempt 4 failed froze when loading new game

closing game and starting attempt 5

attempt 5 fail froze exact same palce as 2

@dry shoal not looking too stable sadly

at least not for my pc

[dry shoal]

I just don't get it :/

@proven vessel can you copy and paste the contents of your rtx.conf file here? It should open right through notepad

[proven vessel]

Ya sure

Is there by any chance a debug mod for this game?

Like a crash reporter

rtx.autoExposure.autoExposureSpeed = 3.5
rtx.autoExposure.evMinValue = 0
rtx.decalTextures = 0x54BFCD9D49337654, 0xBECA2425EF1F5BBB, 0x3FC44F315253E8B6, 0x594636B12468706E, 0xE40984B803BE9058, 0x3E582C1EBD7E0BFA
rtx.autoExposure.exposureAverageMode = 0
rtx.ignoreTextures = 0x57E726B4370E87B2, 0xEC4D0B8122A6240, 0x9C09C32F2373D680, 0xFCDE3F76E68C7867, 0xBA8815B90166E3FC, 0xE2D7DD30F6779ED4, 0xD6C5AEE1D9B10B4C, 0xB8C8BE16567064EB
rtx.dlfg.enable = False
rtx.defaultToAdvancedUI = False
rtx.uiTextures = 0xAE72AF1064D2BE33, 0xA27764F1A2FC1607, 0xBDCCE4FDDB7BA59F, 0xE13FC20B1323FB4E, 0xFA78CB4A485467CE, 0xD24E9A9FFD6EA68D, 0xC614C9FB02514145, 0x918780B90550BF34, 0xB3EC54C5C6CE76C8, 0xDB941FAC3472CE7A, 0x9086511562411A7B, 0xB9B76A1F20CCA027, 0x4DFF06C1D4CA953F, 0x3A09A2043A8825E8, 0xB9023422FE8090E2
rtx.captureInstanceStageName = capture_2024-02-12_23-46-52.usd
rtx.enableCulling = False
rtx.lightmapTextures = 0x4594E67C92D2804B, 0x7AF112D961ABB78B, 0xDC210CD3550E6E96, 0x879AB280E1B697EC, 0x199655BED326910A, 0x8CF716991FBFE282, 0x57DAD63EC4B7CC52, 0x7846B86111C94257, 0x2CB1CF6F176CA28D, 0x147CC0A85C7543A7, 0xEABF8A5DF5BAFAC9, 0xDE5C1A48FDB2143D
rtx.enableReplacementAssets = False
rtx.enableUnorderedEmissiveParticlesInIndirectRays = True
rtx.enableVolumetricLighting = True
rtx.fallbackLightMode = 0
rtx.particleTextures = 0x3EB483F0AC9DC01E, 0x43E178C132238156, 0x22C57702C7C8AA5, 0xDD0F8417CA1496FD, 0x3D9A5E4DA6804BFD, 0x1093FF5E93E2F381
rtx.postfx.enableChromaticAberration = False
rtx.postfx.enableMotionBlur = False
rtx.postfx.enableVignette = False
rtx.skyAutoDetect = 2
rtx.skyBoxTextures = 0x1CAD434472EAA255, 0x73E8703572F6CAC5, 0x8A17DD2C7FEE0296, 0xE8EB996991595527, 0x6C281ED0EA7BA31F, 0xD1B6EAC2B967892, 0x9FA412F955E0B480, 0xAEF5E17DDB73609, 0x10A83D016EAE4FF0, 0xDBBF35FAAB7DAF08, 0xBCE868373ABF666F
rtx.useRayPortalVirtualInstanceMatching = False
rtx.viewModel.scale = 0.01
rtx.zUp = True

[proven vessel]

@dry shoal

[dry shoal]

wait I don't see dlss on

is it not on for you?

[proven vessel]

its set to auto in the rtx menu in gmae

[dry shoal]

hmm

odd

[proven vessel]

maybe its set to ready only hold on

no its not....

it should be writable

ive set the unlbock property, maybe windows security was doing something weird

ill launch the gaem, change my dlss and send the new file

froze when i tried to make a new game but ii cahnged teh settings, ill send the new text

too much text (good sign actually since i didn't have that warning last time) making a copy and changing it to a txt and sending that

@dry shoal

[dry shoal]

U wibder why some settings are disabled

rtx.neeCache.enable = False

I don't even have this I don't think

[proven vessel]

should i enable it

[dry shoal]

uhh

when in doubt dont mess with it

[proven vessel]

k

[dry shoal]

I wonder if I change my graphics settings if that'll add those

maybe the setting is automatic based on your gpu generation

like shader execution reordering is a perf increasing feature but I think it was only added for the 40 series

[proven vessel]

i have a 3070

fuck it i can just diable it if it is worse

ill try enabling rx.neeCache.enable

[dry shoal]

good luck

[proven vessel]

still froze on new game generation

since no improvement im diablibg it again

[dry shoal]

I'll have to figure that out at some point

I think there might be logs?

check for a file called d3d9.log

in the system folder in RC

[proven vessel]

ok

[dry shoal]

that looks mostly okay

is there a ct.log file?

[dry shoal]

uh

I notice it keeps saying it's not finding an rtx.conf file

[proven vessel]

prob the issue

i can bundle my game folder and send it to you if you want

[dry shoal]

no need, it looks like it does load it in another one

this doesn't make a lot of sense to me

I have one theory I can try later, I'll @ you when I get around to testing it

[proven vessel]

k

[proven vessel]

looks good

[dry shoal]

I think I've made the ideal version of the geonosis level

I went back to the untouched original, forced all static meshes to always be LOD0

forced the dynamic actors (couldn't do all of them) to use LOD0 at all times

I added my anchor mesh that encapsulates the player to guarantee it's always rendered

and kept antiportalactors so I can be selective with the meshes I force to always render

[proven vessel]

nice

[limber marsh]

Does this apply to lights?

[dry shoal]

Right now I don't plan on using any light sources from the base game, though that might change

Since the game wasn't designed with such a realistic lighting simulation in mind, there's random light placements that work fine when baked into an old-school lightmap, but look jarring pathtraced

[limber marsh]

Probably the corridors

i am guessing

[dry shoal]

With corridors id rather light them manually to get results like this 🙂 I think it looks pretty good

[limber marsh]

anywho thats neat, I see a lot of people showing off cool stuff but then when lights are turning off every x feet it almost feels theoretical

Could be aiming at a guy with a rifle and his light be off. Hopefully it less of an issue since a lot of those games are big and open.

anywho, yea lookign back at footage their is a surprising amount of stuff they didnt even light up so will be cool to see lights added in where it should be

[dry shoal]

Which footage are you referring to

[dry shoal]

Well fun part about that, I think I found a guaranteed solution!

[dry shoal]

By using a carefully placed 2 triangle mesh in the original level, the capture will have an anchor I can attach all light sources to where necessary

Since the mech will always be rendered as the bounding box will always be within LoS of the player camera, any assets (or lights) attached under that 2 triangle mesh will also always render

AKA, 0 light culling. It'll be as consistent as you would hope

[limber marsh]

I meant relooking back at real gameplay - things that I assumed had lighting properties didnt

memory vs reality

like all those ship lighting elements, bactahealing, eta

[dry shoal]

Ohh I totally misread!

Yeah I'm excited for that stuff too. It'll be a little tricky to pull off though and I expect it won't look like the original effect + dynamic light

I might have to make my own scuffed versions. For example the Geonosis bacta healing station opens up in the middle, but there's a gap inside the casing. If I edit the model to close that gap, I can place a light source inside that casing that will then flood the scene with light when it opens to heal

[proven vessel]

any idea when this is gonna happen

[dry shoal]

I can probably do it today. Basically I'll send you the altered map file + my RemixMesh package. It's a package for the game itself, just has stuff that's helpful for Remix so I called it that. In this case the map file contains one asset from there so it's needed

My theory is that there are slight differences between the 100% vanilla map build and the one I had before. I streamlined the map changes yesterday so it's closer to the original

[proven vessel]

ok

do i need to fresh install?

k

[modern isle]

(is that namatama?)

[proven vessel]

auto - is typing...

im getting ready to read an essay

[dry shoal]

---

Hey guys, cool news:
For context I've been talking with the people working on the Star Wars Republic Commando Remastered mod mainly for technical help on any engine-side tweaks necessary to use Remix to the fullest

But today I got confirmation that once I'm done with this demo, they're down to join forces to see if we can create an RTX Remix adaptation of their vanilla remaster mod ‼️

The biggest impact this has is that it will be significantly faster to make a Remix mod for the entire game, start to finish. It'll still take time but using their assets as a base (mostly AI upscaled, but they're currently replacing environment textures with hand crafted ones) and making adjustments to have PBR will save So. Much. Time. I'm really excited for the possibilities this might bring :) it's not impossible that we would be one of the first full Remix mods out there‼️

[proven vessel]

LETS GOOO

[burnt light]

woah

i think that's the first pin i've seen in #1055020377430048848 that wasn't by me, lmao

anyway, congrats! that's awesome. that modding community seems amazing and i'm glad you can work with them

[dry shoal]

I have the power

(ngl I forgot I was a mod till you said that)

[burnt light]

lol

[dry shoal]

It's a bit messier than that but close enough lol

The SWRC remaster guys are definitely the most friendly I've met in the community

They're very chill, not constantly downplaying ideas

But it works out perfectly since the materials that are being made to replace the AI upscaled textures are actually PBR. Just the guy making them exports as diffuse and normal only since the game engine doesn't do PBR

In other words, his asset has everything it needs for PBR in remix. All it takes is another export

I also agreed to join as another material artist when the time comes so I can contribute to the main mod and the Remix variant at the same time

[celest flume]

can you send me a link to the republic commando discord

did you model this in tris?

please tell me you didnt

[dry shoal]

I did, which was direct advice

from someone in the SWRC modding community iirc

[celest flume]

always model in quads then export to tris

[dry shoal]

https://discord.gg/zCEueRY

[dry shoal]

oh yeah that's what I did

sorry I got stuff mixed up lol

I did it in quads, then when it was time to bring it down in polycount I converted to tries before putting it into UE2

[celest flume]

can I see your wireframe?

[dry shoal]

of the full poly mesh?

[celest flume]

both if you can

[dry shoal]

20k poly <----> 200k poly

[celest flume]

alot of issues with it from what I have seen. is this your first model ever in blender? Anyways your 3 big problems here are 1 Topology this is crucial in any model creation you need to have good topology or your mesh is gonna be a nightmare to work with especially when you are reducing (this would also help fix some of your shading issues ive seen so far) , 2 the armor does not come off as a hard surface, specifically look at the pockets for an example but this applies for the entire model 3, the textures are pretty rough ngl. I would recomend getting substance painter for painting and generating textures. are you just using a base color texture?

[celest flume]

also make sure you models are always in quads in the editor I see your high poly one is but your low poly one isnt

once its lowpoly with good topology and in quads you will be able to easly customize the smoothing groups and make it look 100x better

[dry shoal]

yes lol

first time doing all of this

[celest flume]

well its impressive for your first model

[dry shoal]

and yeah dw, the texture isn't finalized nor does it match this mesh. It was made for a different UV

I think that's why some of the visual issues exist

textures are PBR though, in other shots it's more visible

see

[celest flume]

wait did you just subdivide and smooth the original model?

[dry shoal]

also I'm not gonna lie to you I don't have the faintest clue about topology. It was mentioned a few times as I was making it but no one really went into it

[dry shoal]

just? no. But that was most of it for sure. I did manual tuning via sharpening of edges all across the mesh to tighten the geometry

I did a couple other things in-between but I didn't do a full manual sculpt of any part of the mesh

[celest flume]

ok that explains it

lol

[dry shoal]

the end result looks a good deal different than just the subdivided output

[celest flume]

gta remastered moment'

[dry shoal]

okay it's not that bad lol

[celest flume]

thats what they did

[dry shoal]

I'm proud of it given it's the first time I've done literally anything with a mesh in blender

[dry shoal]

difference is my result looks better :D

[celest flume]

subdivided and smoothed the original models

[dry shoal]

their's looks like they subdivided and called it a day

like there was no other editing lol

[celest flume]

specifically smoothing I think is what they applied to it

[dry shoal]

yeah that sounds right

[celest flume]

Ok so to avoid bad topology I would start from scratch however base it off the original artyle of the game

what did the model look like before you edited it?

[dry shoal]

that's not happening. I don't have the skill level for it and using the original model as a basis saved me a lot of headaches when it came to matching skeletons/vertex groups/etc

all of that was done on it's own

weighting too

[dry shoal]

let me get a comparison, I have it somewhere

original vs high poly!

I think there's a pretty dramatic difference in practice

[dry shoal]

^

[celest flume]

doing it the way your are doing right now will always result in it looking blocky and result in a model with bad toplogy

[dry shoal]

you're telling me it's impossible for me to make it look a little smoother via more editing?

[dry shoal]

those are intended, not perfectly round

[dry shoal]

I'm more than willing to make adjustments that will make it look better in application. But I did imagine I'd start over to replace it down the line

this was a first attempt ever doing something like this

in the 2nd take I'd start similarly but then carve into the high poly mesh to add geometric detail that wasn't there before

[hardy quail]

Dog

Do not subdivied

Just replace that shit wit EA

[dry shoal]

EA?

[dry shoal]

this was a recommended reference another modder gave me to add detail

[celest flume]

adding more detail to a model thats been subdivided to hell is not gonna get you anywhere and will result in a poor quality model with bad topology and smoothing groups

[dry shoal]

but I'll be honest lads, I don't understand the pushback for subdividing. It's not as good as more manual work but for my purposes and skill level I think it turned out shockingly good

[dry shoal]

I did it through a modifier if that makes any difference

not dumb subdivision

it gave very different results to normal subdivision

and was recommended to me by another modeler in the server

[celest flume]

No it doesnt

[dry shoal]

why's that

pls feel free to educate

[celest flume]

again it doesnt fix the fundemntal problems

[dry shoal]

I didn't say it would fix anything

I just wanted to make the general quality better, which it did

if I can edit that 200k result to improve whatever it is you think has issues, then I'm willing to do it

[celest flume]

https://www.youtube.com/watch?v=zSLELihVi6I

[dry shoal]

but I don't have infinite time on my hands given I'm juggling remastering every single asset as well as tinkering with the engine to let me make the most of remix

[celest flume]

https://www.youtube.com/watch?v=rwW6HpOcAuw

[dry shoal]

this seems helpful just off the thumbnail

can I adjust topology in post?

I should also mention this needs to be imported into the engine which has its own headaches attached

that are virtually solved by using the original mesh as a starting point

[hardy quail]

SWBF2 2017 cloke troopee

Subdiv is bas because of unnecessary faces/vertice

And is only meant to be used on models designed for it

Its not made for what you are using it for

[dry shoal]

ohhhhh. Yeah I'm avoiding model rips entirely, as tempting as it is

[hardy quail]

Why

Gl then cuh cuz RTX is gonna look shit with just subdived models

[dry shoal]

copyrighted content, generally frowned upon in the server

[dry shoal]

you're saying good luck to someone who knows quite literally jack shit

[hardy quail]

You are your own person

I used copyrighted shit lmao

[dry shoal]

you are your own person

I don't want to

[hardy quail]

Why is it bad i said good luck!!

Im not forcing you too cuh

[dry shoal]

bc it's the dismissive kind of good luck, not genuine

[hardy quail]

I mean it in both ways in a manner

[dry shoal]

you could be helpful instead of downplaying how decent the final result looked given my experience and that I'm using a texture that isn't even designed for the model, it has a different UV than what the texture was made for and I planned on redoing it for this mesh

[hardy quail]

Ok ive been a bit rude and for that i am sorry, but as someone whos been doin dis shit for years, your method is not good for the intentions

[celest flume]

My issue is your dismissing my critisisms because its "to hard" , learn watch youtube tutorials learning how to model from scratch has never been easier

[dry shoal]

I know

[dry shoal]

I'm not dismissing the criticism, it's valid and objectively you're right

what I'm saying is you're asking too much from someone doing 50 things at once

I'm not gonna put everything into learning modeling for a few weeks to make one good model when it's already at a decent level, when I also am actively learning creating materials, how to use the Remix toolkit, how to make my own emissive maps, how to tune lighting properly, how to edit the original game engine and level files

there's so much bullshit on my plate that I'm happy to take on, but I can't tunnelvision into one thing when I'm doing so many things on my own

you guys are mistaking "I know it's not ideal, but it's what I'm working with given the circumstances" with "uhhh actually I'm right lol subdividing is actually the best way to do this / insert other ignorant stuff here"

[celest flume]

I recomend you stop what your doing right now and start learning the basics of 3d, trying to use remix without that proper knowlege is gonna end up badly

[dry shoal]

really? It's working pretty alright so far for something that isn't final

[celest flume]

Trust me you will look back in a year and think holy shit I thought that looked good? happens to me every year

[dry shoal]

No I won't. Because I know it's not ideal now

again I'm not ignorant or stubbornly naive

I know how limited my skills are, so when I say I'm proud of it I mean I'm proud of it considering I've done all of these things I've never touched before in my life in the span of 2 weeks

when I say "wow I'm pretty happy with it" I mean I know nothing yet I was able to get this result, neat!

[celest flume]

You should be proud but you have to be willing to improve

[dry shoal]

I am willing to improve

you should see all my experimenting and asking questions in #asset-creation

[celest flume]

You are using copyrighted assets btw

[dry shoal]

I'm going there for help, showing progress, getting feedback, iterating based on their advice that they're adjusting because they're considerate of my lack of experience and know it's one step at a time

[dry shoal]

huh

modifying a mesh from a 20 year old game is not the same as using one as-is from a major company that actually still exists

[celest flume]

modifying it slightly doesnt avoid it being based from a model you have no rights to use as a base. just use Ea models

[dry shoal]

like I guess? but it's semantics

it's not slightly, the end result is much less blobby lol

[celest flume]

Lucasfilm still exists

[dry shoal]

lucasarts does not, nor do they care about the republic commando ip

[celest flume]

Lmao Lucasfilm owns all property of lucasarts

it was a subsidary of lucasfilm

[celest flume]

https://www.nintendo.com/us/store/products/star-wars-republic-commando-switch/

[dry shoal]

an abandoned port they made on a low budget that barely performs

it was a quick buck, not an active venture

[celest flume]

semantics

[dry shoal]

still I don't really understand the point. Are you really suggesting I should just steal another model with 0 changes at all vs using my own imperfect work?

[celest flume]

Most mods do this hate to break it to you

[dry shoal]

ok and

they're made by other people

that's their decision to do so

[olive kayak]

Modding and changing assets based on the original is something that is generally tolerated in a lot of older games. Especially if it stays in the same game.

Ripping something out of a different game (even if the same franchise), created by EA is something that can go wrong very fast.

[celest flume]

Give me an example of it going wrong

[dry shoal]

it's my decision not to, especially since I'm actually pushing myself to do something with 3D modeling vs saying "fuck it, I'm a lazy idiot with no passion so let me do literally nothing and steal assets"

[celest flume]

https://www.nexusmods.com/starwarsbattlefront22017

[burnt light]

i'll just say this: it's heavily discouraged to use existing assets from a game. even if it hasn't gone wrong in your experience, it opens you up to a needless potential legal mess

[olive kayak]

Arma 3 Star Wars Mods who have ripped assets from BF2017 - all of them are usually taken down.

Just because some do it, doesn't mean no action is taken or makes it less problematic.

[celest flume]

he wants to remaster the game but not put in any effort when remastering the model, I only suggested this because he didnt want to make an original model

[dry shoal]

I spent multiple days, many hours each making that imperfect end product

you can say a lot, you can't say I didn't put any effort in

[celest flume]

its what you are saying after

your saying basically I already did it this way so this is the way I am going to do it

for convienance sake

[dry shoal]

I said after that it's not final, and that I knew it was flawed, and that I had reasons to start with the original mesh

[dry shoal]

because this project wouldn't get finished otherwise

[naive robin]

That's what they should do. It's their own thing.

[dry shoal]

it's like you're trying to tell me something I already know and for what? I mean if you want to help out on the project and do the meshes because you know your shit then that's fine, I'd totally accept that assistance

[celest flume]

he wants others to come on to help him, he is making a demo for them

[dry shoal]

but you're looking at someone who's juggling so many new skills being learned bit by bit doing this on his spare time as a passion project

[dry shoal]

lol no that's not what's happening

[dry shoal]

I'm making a general demo for RC done on my own as a demo for what Republic Commando with Remix could be

then I'm switching over to make an adaptation of an existing mod but in Remix

they aren't working with me, I'm working with them

[dry shoal]

everything I said there is accurate to what I just said

[naive robin]

Cool, just let 'em cook then. Ain't like they're trying to become a lead artist at Weta or something.

[dry shoal]

yeah and again, I'm making a demo for this community. It's completely seperate to what those guys are doing and what I'm going to do with them for the Remix adaptation of their mod is more limited in scope

I'm going to be a material artist, not a 3D modeler bc I'm not a 3D modeler

they're replacing AI assets with hand-made ones, and I'd like to help

is there anything objectionable there?

[celest flume]

Lmao good luck doing that withought knowing how to model trim sheets

[dry shoal]

lol I can throw a mesh into Mixer or make a generic environment material in instaMAT

I'm not drawing in paint

[dry shoal]

since they're using many original or only slightly upgraded assets poly wise, it's all there

[celest flume]

do you even understand what a material artist is

[dry shoal]

and I still don't understand, do you think I'm trying to go pro? It's a hobby I just picked up and spent dozens upon dozens of hours on

ok if that's not the correct term I'm sorry. It's not a job so it's not a description I was given by an employer

does that make you happy?

I'm going to work on environment materials for another modding team

I'm not making Battlefront 3 by EA

[naive robin]

How could you not go photoscan those rocks in Arizona

[dry shoal]

that would be cool tbf

[celest flume]

quixel exists but I doubt any megascans would be ideal for remix withought unreals nanite

[dry shoal]

that's a joke about how BFII devs photoscanned a lot of stuff

I'm not going to photoscan stuff

[celest flume]

I know

[dry shoal]

I'm not gonna use megascans

and I'm not aiming for BFII level quality across the board

I'm not a multi billion dollar company

[celest flume]

then its not remastered its republic commando : subdivided

[dry shoal]

it's a mod, it's not being sold. I also didn't make the mod called SWRC Remastered

I don't have a name for this demo project either

[burnt light]

megascans also can't be used outside of UE unless you want to pay

and yeah, i think you're going way too hard on auto here somedude. this is just a passion project more than anything from what i can tell, it's not meant to be ultra professional and all that. he's still learning, most of us are here

[dry shoal]

idk, according to somedude people have to name their mods based on the standards set by full dev teams

really curious about his work btw

I feel like actually skilled modelers don't put down a beginner for being excited

[burnt light]

unfortunately that does happen in a lot of fields

i've had people treat me that way when getting into various different things 😦

it's always a shame, and pretty offputting

[dry shoal]

I'm sure they exist, but I'd be shocked if it wasn't just the dunning kruger effect in action

for reference

[celest flume]

Lmao I put you down because your gonna make a mod that looks like gta remastered definitive edition

[burnt light]

it's possible to give feedback without being a jerk 😛

[celest flume]

I tried that he didn’t listen

[burnt light]

i saw it. it wasn't very nice

[dry shoal]

I listened to everything you said and agreed with all of it

[celest flume]

This started out very civil

[dry shoal]

I disagreed at your expectations

[burnt light]

"hey, i think if you continue down this path it might not look so good, imo you may want to reconsider. doing x might help. great work starting out though"

[dry shoal]

scroll up. How many times do I say I know it's not the best, I know it's flawed, I know it's not the best way

you're telling me nothing I don't already know. And if there was any misconceptions I had I was open to you correcting me

it's negativity for the sake of it

or to boost your own ego I guess

sure, I'm probably much worse at 3D modeling related stuff than you

not a high bar

[celest flume]

I said that before I realized he was subdividing the original mesh

[dry shoal]

I still think the mesh looks a hell of a lot better than the GTA trilogy remaster tho :) won't let you lie to me about that one

[dry shoal]

which I said I knew wasn't ideal or the "proper" way btw

the only thing you actually said was to stop what I was doing and learn everything about 3D modeling or everything about this project is a waste of time

[celest flume]

You will be wasting your time it’s possible to learn how to mod and model at the same time I did it

[dry shoal]

I'm well aware. I've had to learn stuff related to modeling, material making, lighting, and remix itself just to get this far

which is 2 weeks in btw

2 weeks of doing everything involved in making a graphics mod on my own

my experience before this: playing games

that's it

and I know for a fact, I'd bet my fkn life's savings on it, that I've done more in these last two to three weeks than you would have starting from the same point in my shoes

that's why I'm proud of myself, and I won't let you shut that down

[celest flume]

You should be proud of yourself but when we tell you your shooting yourself in the foot by subdividing an existing low poly mesh you should listen and start from scratch

Use the original as reference

Along with movie pics

[dry shoal]

actually I think I'll give you some advice now. You don't teach someone by saying they should be ashamed of the work they've done in 3 weeks because they didn't do exactly what you wanted, when you wanted it, right fucking now. Your help isn't wanted here, so go away. I'll be over here having fun, being hyped about each step along the way. Feel free to look at my progress pics and complain that other people are encouraging me to grow with positivity. It's those kinds of people that make all of this so fun and fulfilling for me. You'll be happier if you drop the "fuck you do what I say" mantra

much love, enjoy your night

[proven vessel]

i dotn know what happened but it seems that their trying to give you advice (i may be wrong) so i dotn see why your so hostile?

[dry shoal]

if you start from the beginning you'll see it

I gave this person an ungodly amount of patience

[proven vessel]

wheres the start exactly?

[dry shoal]

this

so somehow my new geonosis map is in a big box

[proven vessel]

ok stopped reading about 3/4 of the way down, and near the end their total dicks, but at the start i think their intentions were good, but when you gave a good reason as to why your not doing what their saying they got pissy, i do agree you cant tunnel vision when working on 50 different things @dry shoal and for what its worth from someone with no modeling skills, they look good

[proven vessel]

and what i do know a bit about is copywrite of old games and modding old games, your remastering the old models is fine

[dry shoal]

someone's intentions can be good and they can still be unnecessarily rude

I tried to be respectful hoping they'd show some respect back

they never gave it

so the only thing that's left is to move on bc they're doing the opposite of helping me

but going back to mod progress

[proven vessel]

thats right

[dry shoal]

right now I'm trying to use a script another modder made to pull the original assets instead of having to rely on the captured meshes which I'm having trouble with

[proven vessel]

also the bas game looks ebtter then gta remaster

[dry shoal]

it's possible I'm doing something wrong but this bypasses that, or at least it should

[dry shoal]

lol thank you

I look at my model and see the flaws for sure, but I look at the models in GTA trilogy remastered and I see

I just see

nothing else

[proven vessel]

i see shit in tthe form of a model

[dry shoal]

lol

I'm glad my clone mesh didn't turn out like that kind of stuff

I'm gonna see if I can get a good capture and anchor necessary meshes to get to a good starting point

[proven vessel]

well if you need any testing done, you know who to @

[dry shoal]

I reset the project file entirely so I only have captures related to what I hope will be the final level file I'd give with the mod anyways

like before I have all the assets I've used so far on hand so I can just ingest them again when I'm ready to

[proven vessel]

sounds good, although most of it goes over my head, talk about codign i cna follow, talk about models and i get lost...

[dry shoal]

I get lost when people who know their stuff talk about both

[proven vessel]

lol

[dry shoal]

I was in the #asset-creation channel one day in my little corner like " I did something" then 5 minutes later the smart people start talking gibberish

I think I threw in a message like "yes I know exactly what's being discussed here "

[proven vessel]

that happens sometiems for sure....

[celest flume]

i did

you literally did the same exact thing they did

[dry shoal]

thanks to kamil, we got colored sphere lights attached to the lasers!

it was only possible through changing the source code lol

of remix. It's a hardcoded feature that can be exploited if you're crazy enough

check it out! this is a fresh game install so no new assets of any kind

I only swapped in my altered geonosis level file so the save wouldn't crash on every load

[modern isle]

do you understand how absolutely condescending you sound? I've been a technical artist for over a decade working with many different engines and wearing many hats over the years with teams of very varying skill.
I have never seen this kind of behavior actually helping anyone.
By your account people shouldn't even bother trying unless it's a pristine result?
Auto is just starting out and they've already come a long way.

[modern isle]

when replacing textures in a game that doesn't have trimsheets, you don't need to learn how to make trimsheets, right? 🙂

[dry shoal]

Thanks to some people who worked with me through questions about retopology, I'm gonna try it out and add it as part of my model remastering workflow

It'll roughly consist of modifier subdividing, editing the sharpness of edges to reach a desired look instead of something that's just rounded out a bunch, then doing retopology

That should be relatively efficient, lead to good results for my skill level, avoid headaches of brand new skeletal meshes, and so on

shout-out to @wind briar and @naive robin . These guys have been instrumental to even getting this far. Confidence is going up and I think I can do better work

Without them unironically this project likely wouldn't exist right now

[modern isle]

do keep in mind why you're doing the retopology to begin with, you can then take the low poly mesh and bake the high poly details onto it, your normal map is pretty much done already.

That's how you get the maximum out of the model you made

[dry shoal]

Ah yeah that's right. So then what's the right way to go about it? Do you mean I should start with minor edits to make the mesh a little more structured in terms of face distribution, and use that as the source to bake normals onto the proper game ready asset?

[modern isle]

a high poly mesh contains all the small details that you don't want on your low poly mesh as geometry

so if there's a scratch or dent on the armor, that's often actually a modelled feature on the high poly model, obviously you don't want to model that into the low poly. So the high poly contains the extra bells and whistles that then during texturing get transferred onto the low poly as a texture.

[dry shoal]

Hmm I guess I knew that part but my mesh doesn't really have any microdetail

It's the original asset at a higher fidelity but I wouldn't call it more detailed

Though I did say the next time I do work on that mesh that I'd try carving out details manually so

Maybe this is the opportunity to do so?

[modern isle]

maybe, just please don't think about throwing booleans at everything

[dry shoal]

Aren't booleans true and false

What's it mean in the context of modeling

[modern isle]

I'm gonna show you something dangerous

because you'll think "OMG this is great!"

but reckless use of booleans is a very deep pit to fall into

[dry shoal]

Are you kidding? Thinking this and then having it go horribly is the best part!

||don't worry I'm joking||

Wait a sec omg this looks great ‼️

LOL

It does look tempting but I don't have use for it anyways

I'd probably do it with some fancy tools to mess with the polys directly

That probably aren't actually fancy but they're fancy to my dumb ass

There's a decent amount of microdetail inherent to the clone designs that could be capitalized on with normals even if I don't go into real microdetail

I don't trust myself to sculpt little nicks and scratchs. I could always use normals from my texturing to do that procedurally anyways right

[naive robin]

Okay, now show us the wireframe

[dry shoal]

Is the topology all fucked if you do something like that?

That's what would make sense to me at least

[naive robin]

Oh. Like you wouldn't believe

[proven vessel]

nice!

[modern isle]

🥲

[sinful echo]

Outstanding.

[dry shoal]

made improvements to the high poly clone mesh, weird shading issues are largely gone

definitely looks cleaner

I've had enough of blender for a few days so tomorrow I'm gonna look into retopography. Depending on the advice from people smarter than me, my current idea is to retopo the high poly mesh so it's less horrid to work with (so trying to keep as much of the geomtry as possible). Then I'm gonna use that mesh to make a lesser version to be implemented into the engine. If I'm successful baking normals then the mesh that goes into UE2 should look noticeably better than the one I have now

a retopo will also make it significantly easier for me to make a viable UV map

it's definitely exhausing when you're doing things in the least efficient way possible just bc you don't fully know the tools at your disposal, but as long as I get a decent end result and I learn a bit in the process, then it's worth the excessive amount of time put in

I also want to look into batch AI upscaling + add AI generated PBR to the original materials to at least have as placeholders until I get nicer materials in there

I think my plan will be AI upscaling a handful of mesh textures since I couldn't redraw them. I'm gonna use custom made PBR materials for the environment though which will definitely be the most noticeable part

eventually I want to make a higher poly version of the rifle also with custom PBR but that's gonna be a pain in the ass, I can feel it :)

honestly I might be able to make a higher poly mesh then just bake the normals into the original mesh. Dunno how viable that is but it would save me some work for sure

unlike the clone model the guns don't really have noticeable poly edges

[limber marsh]

On the subject of model/asset taking.

Their are brands not to be fucked with

Nintendo / Disney [Star Wars]

Are literally at top of the list

You do you - but history speaks for itself

[dry shoal]

yeah I get you. idk for me even if I could get away with it, there's some pride involved in being able to do something yourself. The skeletal meshes won't be from scratch, but I do spend a lot of time on them and tune them myself. Still learning so I can only manage so much, but I think people would generally prefer just a better looking version of the original with a new PBR asset anyways

[limber marsh]

If anything it would have to be a seperate mod

ontop

like asset replacement of assets in your mod

something "distant"

Like I can think of 10 mods cancelled/deleted off top of my head over last decade that were star wars related

Star Wars is a multibillion business just in gaming. People are out of their mind if they think a remaster with assets stolen will get away with it

Especially since Republic Commando has a "remaster"

They can argue the stolen assets and how it hurts them financially

My 2 cents is as mod grows and stuff is shown people take it seriously, which makes others want to be involved

As you've said your one person. Do what your comfortable and spend your time wisely.

https://www.youtube.com/watch?v=CL5i33CTld8

I am sure this got posted X times in past

But are these asset grabs?

[dry shoal]

actually hasn't been posted here, but I did see a gif on the steam community lol

I don't think those are asset rips. But it looks pre-rendered

but man literally all those assets would be awesome to have

[limber marsh]

🤷🏻‍♂️ Part of my reason to think you should just build up what you can until people come along.

People do it in free time

https://www.youtube.com/watch?v=G0u_q5SlwHU

this was really cool to watch btw

[dry shoal]

yeah absolutely

this would be interesting to look into later, I might see if I can shoot them a message and talk about parts of my project

it's not high priority atm though, and the most practical way to get a fully playable game will be joining the RC Remastered team to work with them on a Remix version

the difference in quality will be fairly significant, but it'll still be improved over the base game in terms of assets + ofc pathtracing

only time will tell how it goes though

[dry shoal]

Before I mentioned in the future I'm looking to join the Republic Commando Remastered team in some capacity. Well today I got a hint at how that might turn out. These shots show one of the new textures that was made for RC Remastered recently. I got the project files, exported at 4K with PBR for remix and boom

the scene is super dim so I'm gonna change that but I'm very impressed

[limber marsh]

Have theyu said anythign to you?

[dry shoal]

different group than the video you linked

[limber marsh]

no, i know

[dry shoal]

yeah I've been talking to them a good bit

[limber marsh]

i actually tried reaching out to them when remix was in beta

they werent interested at time

[dry shoal]

one of them was really interested in Remix but they all have AMD GPUs so

and mostly RX6000 AMD GPUs

so couldn't use the editor and the runtime performs very badly

they're down for me to make a Remix adaptation of their mod as long as there's not a huge ask on their side. But since they're already making PBR textures and just not exporting them as such, there's no extra work that needs to be done there either

they have the assets, level files, all that good stuff

[limber marsh]

nice two birds and all that

[dry shoal]

exactly

and if I work on assets from time to time naturally I'll be sharing it with them too

we both benefit

[dry shoal]

oh but they're also willing to try and help me with where I need to look to do certain engine level stuff which is great too

[limber marsh]

is their any issues with walls/indoors stuff?

[dry shoal]

one of them knows how to disable the default lighting effect when shooting the gun. Normally it just kinda makes the ground brighter but this causes issues with Remix

[burnt light]

that's quite the modding setup 😛

glad that worked out

[limber marsh]

I remember it was on shaders

[dry shoal]

yeah those guys are great. Lots of potential

[dry shoal]

so far shaders have been largely not a concern

at least as of the current remix runtimes and whatnot

[limber marsh]

remix advanced since i've looked at it

and they were simple shaders

except the projection

[dry shoal]

that might change later as new stuff is introduced in the game but for now I'm working on the first level still

[limber marsh]

projectors are a problem

[dry shoal]

yeah the shaders it does use I've seen work fine

[dry shoal]

how so?

[limber marsh]

really?

how is it going to introduce light?

emit

[dry shoal]

ah you're referring to the fancy effects stuff

yeah, unfortunately not possible

as of now at least

[limber marsh]

you can tell projector to render

err force

but it wont effect lighting afaik

cause its all shader related

[dry shoal]

I still don't quite understand. Can you give an example of projector stuff that might not work right? I've mainly just seen bullet decals but there's very likely more that I just haven't noticed

[limber marsh]

https://www.youtube.com/watch?v=a1uOMGeO9lc&t=298s

maybe i am behind times

[dry shoal]

oh the hologram is one of them?

the hologram works actually

[limber marsh]

oh wow, surprising

[dry shoal]

well, I do have an issue atm with my new clone mesh causing a tpose in the hologram but it does work. Also a point light can be attached to the projector to emit light

a cyclinder light going up should work if I can do that properly

[limber marsh]

nifty

probably nothing shader related is an issue then

cause that was most advanced i saw and it was pretty shader limited

[formal cypress]

how would the night vision work with remix?

[dry shoal]

so when I tested it earlier the default behavior makes it spawn a really intense flashlight

like a proper light source

[formal cypress]

sounds about right

[dry shoal]

now, how I can tune that to work well when it's generated by the game?

no idea

that's a problem for future auto

[formal cypress]

is there a way to adjust exposure?

[dry shoal]

it was really cool using no flashlight then using the lasers for light

yes there is, I should probably look into tuning that more down the line

[formal cypress]

yeah maybe have a dim light, with high exposure when it turns on

[wind briar]

@dry shoalhttps://twitter.com/EricaDahlia/status/1758896019654222322

Erica Dahlia (@EricaDahlia) on X

I believe the highpoly is finished. Any feedback is absolutely welcome!
#wip #3d #3dart #StarWars

Twitter · Feb 17, 2024, 4:47 PM

Weird timing

Someone I know just posted this X)

[dry shoal]

I appreciate the thought but I need phase 1 clones specifically

these ones

though this is a pretty good reference lol

[wind briar]

Gosh darnit how many freakin' versions of these things do they need

[dry shoal]

currently experiencing the lovely joys of retopo

[wind briar]

Looking like a solid start!

You probably don't need to make it that dense if this is for the HP though

Because the SUBD will add the density anyway

And making it slightly lighter will make authoring easier

[dry shoal]

hp?

[wind briar]

Ah sorry, HP = Highpoly

[dry shoal]

also I don't actually have a subd modifier rn, the guy in the tutorial removed it though maybe I should use one

[wind briar]

Don't know how Blender handles it but I typically tend to have 1 or 2 subd itterations on when building like this for a HP

[dry shoal]

simple subd?

[wind briar]

So that I can actually see the end result

[dry shoal]

there's also catmull-clark which is what I used originally

ig it's an algorithmic type of subd

[wind briar]

Yeah but I mean more in terms of previewing it

Like in max, when I enable the SubD I can colour its wireframe in one colour, and keep the low poly in another colour

[dry shoal]

added a simple subd

[wind briar]

So that I can easily see both versions whilst authoring

[dry shoal]

oh you mean the mesh behind it

[wind briar]

Nah

[dry shoal]

[wind briar]

I mean the pre-subd version

Like the one you actually author

Important to preview that stuff

Because I can tell you right away, this area won't behave the way you think it will

[dry shoal]

[wind briar]

AND I WOULD STAY TO HELP

BUT

It's almost 3 am

[dry shoal]

no you're good

[wind briar]

And my brain is friiiiiiiiiiiiiiiiied

[dry shoal]

I'm just feeling it out. If it's fkd then so be it, at least I tried

[wind briar]

Hang on

[dry shoal]

(that's my cope answer)

[wind briar]

Shitty mouse photoshop job lol

Something along these lines should be enough to get you what you need after it's been subdivided

[dry shoal]

so I should be using fewer vertices or

[wind briar]

Basically it's a good idea to use less, not that you have to be conserned about polycounts BUT

The more verts you have in there, the easier it becomes to end up with what we call "Blobby" geometry

[dry shoal]

also the problem with the bottom is it extends a bit

[dry shoal]

I'll start over then

[wind briar]

Lots of small points where things could go wrong

If even one of those verts is off for example, it can look like a lump

Whereas if you let the SubD handle the adding of those points, it will do it in a perfectly smooth way

[dry shoal]

yeah I didn't want to use this many edges but the verts wouldn't connect n whatnot

originally I was using fewer verts in general

[wind briar]

And

Around areas like this

You need

Something like this

In order to make the shape hard

Because if you subdivide it the way it is right now

It won't create a corner

But rather a smoothed out shape like this red line

[modern isle]

Am I weird for naming my files _hi instead of _high? Gotta save those characters

[wind briar]

I mean, I name mine _HP and _LP

Sooooooooooooo

Yes, you are definitely werid

And should probably feel bad

[modern isle]

At least I don't call it _lo

[wind briar]

I would genuinely slap you if you did

Nawh like, whatever naming schemes work for ya'll, roll widdit

As long as your use BP_ we're cool

[dry shoal]

this is my naming scheme

[wind briar]

I'm officially banning you from the server

https://tenor.com/view/ban-hammer-cinzou-ban-banido-look-gif-17786909

[dry shoal]

[wind briar]

"HIGHERPOLY"

WHAT IS THAT

[modern isle]

Actually since the model is a skeletal mesh and will be imported into the engine instead of remix, there are limitations to consider. Just worth pointing out that UE2 has some hard limits that have to be considered for this approach.

[dry shoal]

actually UE2 doesn't really have hard limits, at least not in the way ue1 did

[wind briar]

I was under the impression they're just building a HP for baking purposes

[dry shoal]

it can certainly cause instability though

[wind briar]

Whilst the Lowpoly will be a different mesh?

[dry shoal]

and yes I'm trying to do a high poly retopo first

[modern isle]

bluePrint_BP-PLAYERSPAWN_01-final.uasset

[wind briar]

Man, and here I was about to go to sleep

But now I'm fuming with angor

[modern isle]

Haha

Nah for real I had to create default naming conventions for our teams, BP, SM, Sk, MF, MI, etc.

[modern isle]

You are still very much beholden to the 32 bit address limit, but also the engine might handle skinning very poorly. You can make half life 2 crawl within seconds if you know how since it's alpha blending is really inefficient.

[dry shoal]

I was able to get a 20k working before. I can't imagine I'd need any more tris than that

and since I'm only replacing the skeletal meshes I'm relatively safe I think

but I could be totally off the mark

[modern isle]

When in doubt, err on the safe side.

[wind briar]

Ah, a fellow UE documentation reader I see

https://tenor.com/view/predator-arnold-schwarzenegger-hand-shake-arms-gif-3468629

[dry shoal]

currently working off your reference

[dry shoal]

Copy pasting what I accidently put in #asset-creation

Well after multiple hours all I got wass

https://media.discordapp.net/attachments/1196836285545971712/1208602625722032128/image.png?ex=65e3e1e8&is=65d16ce8&hm=cd737e4ce0cf4f9246e4d1a06782f1f0a494e8df6986874ec6b61404447e33ee&

THIS

This part isn't even done but
I think I'm understanding the mindset behind how to distribute vertices, correct me if I'm wrong
the edges aren't all angled perfectly yet but more vertices where there's curvature >few to none in-between?
Well not none but yk
like I think I'll add one more edge loop in the middle to account for the subtle curvature and let normals n whatnot do the rest
Maybe I'm overthinking it through
calling it a night

[proven vessel]

@dry shoal just downgraded my windows form 11 to 10 so im gonna unable to test new configs for a few days while i get my computer operational again

[dry shoal]

That works out. I'm at an interesting point where I might be able to make some big progress on the visual side of things, so it'll be a bit until I'm back on making configs

[dry shoal]

thanks to a modder working on SWRC Remastered, I was able to use a script to export all game assets from the game files and have them auto converted to .fbx for use in blender and/or Remix

[dry shoal]

^so the same guy who made the script suggested it might be possible to automate a very tedious process with skeletal meshes

basically the way umodel works (program for viewing/extracting UE game assets) makes it so you can't really get everything in one go

animations can be ignored. The problem is the Armature. The GLTF mesh includes the original mesh normals while the PSK mesh does not. To get what I believe would be the ideal setup, I need to cmbine the mesh of the GLTF with the .PSK armature

this also has a scaling step inbetween since they import into blender very differently

but I also could be worrying about a non-issue. @wind briar since I'm trying to make higher quality versions of the original meshes, how important is having the original normals vs no smoothing at all

are those normals utilized in the subd process / other things I might use

[wind briar]

You don't have to care about the original normals at all

Since you're making a new thing and baking new ones

[dry shoal]

well shit that saves some work

I didn't know if any of the algorithms used might reference normals

feels like it would for subd but maybe not

[modern isle]

You don't have to worry about the old normals at all. You're modifying and changing the normals anyway and in the end import a different model with different textures, whatever normals existed are pretty much irrelevant.

[dry shoal]

this is awesome to know, thanks. Did you see my retopo changes above? I wonder if those look a lot better than before

[dry shoal]

.

[modern isle]

Better than before, I'd probably make a line through the middle, but the additional lines near the edges help to flatten the normals (don't worry too much about that concept yet, we'll get to vertex normals and normal angle at some point) and tighten up the smoothing

[dry shoal]

got some pretty sweet news. I got in contact of the producer behind this render and he's going to share a few assets from the render that he himself worked on!!!!

[limber marsh]

its a good slice of game to so maybe can show it off as proof of concept depending on how much

err slice of intro

[dry shoal]

implementing it will take time. The assets he has on hand are a mix of things where some can be replaced through Remix (aka use the full quality mesh from the render), some need to be replaced at least partially in-engine. So those meshes will have to be optimized

[limber marsh]

@dry shoal I know I linked this once before but just wanted to see since it may of appeared to be same video

https://www.youtube.com/watch?v=G0u_q5SlwHU&t=0s

its not completely relevant to remix but it shows alot

of explaining how scene was lit

[dry shoal]

yeah I did see!

I showed the guy some preview shots of my Remix progress and he thinks there's potential, but there was a period where I got pretty busy so I'm waiting on him for whenever he's ready

I don't want to spoil what he agreed to share with me :) rather keep it a surprise

[limber marsh]

oh...

that explains alot

Layout Artist at Blur Studio

the guy who did the lighting

Blur is nuts

[dry shoal]

yep, super talented people worked on that

guy I'm talking to did some of the 3D art

aka exactly what I need

[dry shoal]

retopo progress

I'm tired boss

I want to get to the fun stuff

fuck the topology

[naive robin]

Take a break, switch to something else

[dry shoal]

way ahead of you friend

[naive robin]

You make learn the best when passionate, so if you're not vibin it, just switch to something you want to do.

[dry shoal]

restarted the Remix project file hopefully for the final time. I know a lot more now

once again shoutout to runestorm's material. subtle POM thrown in there for good measure

I've been tunnelvisioning hard on the clone mesh. I think I need to put that on hold so I can actually make some progress

I forced all static meshes in the level to always display at full quality which works well for captures. dynamic meshes are still affected by LOD despite my efforts tinkering with the engine. I left in antiportal actors to save on performance where needed, and the current plan is to use my anchor mesh only as needed

if I notice something popping out of existance, that mesh is getting anchored

but otherwise I want to take advantage of the optimization systems bc why not

the trusty 1 tri anchor

the original level is very much indirectly lit. Still undecided if I wanna change that drastically. if this angle stays (indirect) then I need PBR on those cliff/arena meshes stat. Looks awful with the plastic sheen

ok i forgot u can change roughness values

already looks much better, but still not nearly where it should be

now that I have some experience extracting level meshes from the game, I can get a true 1:1 mesh of the ground. If I play my cards right, I can import it into instamat and make a nice custom texture instead of relying on repeating UVs :)

like being able to hand paint where certain mixes of materials are for instance

in this very shot you can see on the ground that two materials are supposed to blend. I could recreate that with very high quality and better

(might need to steal Rune's experience for help on that one )

exciting times ahead!

fuck the clone model !!!! it's time for the big cool stuff

clone will come eventually tho dw

night folks

[dry shoal]

thanks to Remix's mesh captures, I was able to bypass one of the UE2 limitations of BSP geometry not being an extractable mesh

I'm gonna prioritize using original extracted assets bc no better place than the source. But for levels Remix's captures are very helpful in this regard

the topology is once again awful, but because it's all squares it can be extremely easily retopo'd. I'd also make my own UV map from advice I get from the modeler people here lol

[proven vessel]

👍

[modern isle]

are you sure about what you just said?

literally 2 clicks in the editor to export the entire level + props

[dry shoal]

Not possible with RC specifically, according to other modders

[modern isle]

and you tried?

[modern isle]

I unfortunately don't have a working unrealED build for SWRC, the one shipped with the game crashes on launch

ah there we go, had to lower bumpmapping quality.... it was set to low in-game but it probably didn't save correctly

[modern isle]

well, for SWRC it's a bit more involved but I got it mostly working

all geo and light positions exported

¯_(ツ)_/¯

the magic of .t3d

need to fix up the static meshes or textures being included but I'm sure I can get that working too

[modern isle]

some meshes import now

once I got a full scene working we in biz, lads.

all the BSP imports correctly already, so that's the important thing done

[modern isle]

welp, ran into a regression, but the BSP is a clean export now... no meshes though

[dry shoal]

Wow this is news to me lol

[dry shoal]

Yeah there's a community fix that also includes a more stable Unrealed build

Fixes the bump mapping too

[modern isle]

I got all BSP geometry, and I got the individual meshes. Just not combined, but I do have a full .t3d with the proper references, just need to fix up some paths for the meshes and they should import correctly

[dry shoal]

Yeah that BSP geo should be quite helpful

I have the meshes all exported but the textures are more complicated. They combine textures in the base game for a blend but I haven't been able to cleanly get a final combined texture got a mesh

[modern isle]

are they using vertex color for the blend?

[dry shoal]

I imagine yes but I'm not 100% sure

There's multiple textures involved, the final combined texture isn't just one thing that applies to the whole mesh so it gets complicated

(which I think is due to using vertex colors?)

[modern isle]

went through the arena_hr folder and didn't find any meshes that really made use of vertex color, I wonder if UED supported in-editor painting of vertex colors

bruh

it's literally just a mesh lol

I was already a bit suspicious because the floor was flat BSP and that doesn't support any vertex colors or anything

[dry shoal]

The gravel patch is a mesh above the BSP geometry. It's used for terrain blending with the BSP surface texture

If you were to select all matching static meshes in Unrealed you'll see only a few chunks of the ground are actually highlighted

For Remix I'm probably just gonna delete those meshes entirely since I'll be able to do something similar within one texturemap

[modern isle]

how you're wording it is wrong. It is a transparent mesh that gives additional variation due to it being overlaid, but there is no terrain blending of any kind going on

[modern isle]

I would advise against that

[dry shoal]

Interesting. So how would something like that be tagged in the Remix runtime? I could never get it looking right, just looks like a solid mesh with its own distinct texture

[dry shoal]

Wouldn't it just be more complicated

[modern isle]

it's literally just a transparent model. Like a pane of glass

[modern isle]

why would it, you can replace this flat transparent model with a highly detailed pile of gravel :)

[dry shoal]

Wait is that the original mesh?

[dry shoal]

So you'd suggest doing an asset replacement that intersects with the ground using its own materials for more variety

[modern isle]

yes

saves you all that headache that you're about to bring upon yourself

[dry shoal]

That makes complete sense and I didn't really think about it for some reason lol

[modern isle]

and if you want more variation you can just add another dirt pile model and use the anchor to position it in the world

[dry shoal]

Yeah that's a good point

The thing I'm running into though is the blend on the textures that I replaced with your asset. You can see in the editor that there's burn/scorch marks on parts of meshes

Not quite sure how to replicate something like that without making a new texturemap for the mesh

[dry shoal]

Wait that's genius

Did you import the mesh into instaMAT, make a material for it then bake a new mesh?

[modern isle]

I exported the original to get the shape and dimensions, quadratized then added a subdiv modifier, made a heightmap in instamat that I could use for a displacement modifier, added a vertex map so that the outer edge is lowered a few centimeters into the ground so that it never juts out.

[dry shoal]

What's the general process for making height maps in instaMAT? Can it be done easily based on the mesh itself or something

Or is there a lot of manual stuff involved

[modern isle]

a heightmap can be just a noise texture if you want

[dry shoal]

The mesh snapping to the new shape at 0:39 feels like the "rest of the fucking owl" meme

[modern isle]

I just added in a flat plane at 39

[dry shoal]

I'm guessing this is C4D? In that case I might not be able to get away with just using instaMAT for it

[modern isle]

you can do it in instamat too from what I've seen in the node graph, I just know how to do it in C4D in a few minutes :P

[dry shoal]

A flat plane based on the heightmap?

[modern isle]

no, the grey thing is literally just a flat floor

[dry shoal]

Oh LOL

[modern isle]

I already applied the displacement before with the modifier, it's just a very simple "move this up based on heightmap" thing

that's the heightmap

[dry shoal]

So you're applying the displacement using the heightmap as a reference right

[naive robin]

Noise generators > pick a noise > blur it a bit > heightmap

[dry shoal]

And then you're showing what the mesh would look like at that point when intersecting with the ground

[dry shoal]

I starting to sense a pattern where there are easy solutions to things that seem like they would be complicated

I think it stems from "idk what I'm doing so I don't fk with that" but then I find I totally can do it

[naive robin]

The more you know what and how to use the tools you got the faster things go.

[dry shoal]

Yeah like I'll see Rune do some cool shit in a few minutes when getting a similar result would take me hours simply because I don't know what's available that will do the seemingly hard stuff for me

[dry shoal]

HOLY SHIT I COULD MAKE ROCKS DOING THIS

[naive robin]

Tfw discord mobile iOS doesn’t have webm support

[dry shoal]

Does for me

[modern isle]

tfw apple doesn't have webm support

[dry shoal]

Android users on top

[dry shoal]

But yeah I didn't realize how versatile this stuff is

Like in the span of that short video I thought of numerous ways I could apply something like that

It could be making mud, a pile of gravel, rocky mesh surfaces, mountains, etc etc

[naive robin]

There’s lots of videos of artists walking through their workflow, different from tutorials, that I find helpful for finding out new ways of doing things.

Workshops I think they’re called

[dry shoal]

I'll have to check a couple but in my experience the problem is I'm doing oddly specific things that don't benefit from what someone might be doing in those videos

Maybe a select few bits would be applicable but those I'm mostly already aware of

I will say generally speaking though it feels like each day I ask questions I get answers/new understandings of things that makes me feel like I can be more ambitious/do things I thought I couldn't on my own

For example, getting outside assets for large canyon walls/huge boulders that surround the playable level. Using some of the stuff here I could totally make my own versions of those assets but with geometric detail that isn't too insane while also being a huge jump over the original asset

Add a new mesh, scale it up, subdivide, use a noise map for displacement across the surface

I can then import that mesh into instaMAT to make textures based on that displacement/gradients/curvatures/all the above so it's actually unique

[wind briar]

For cliffs and stuff, there's a very neat and simple workflow that I use all the time

So I sculpt some tileable rocky things and bake them down right, this is just to give me max control but you can use noise generators or heightmaps from resources online etc

Blockout your general cliff as a big blob, then subdivide the hell out of it

[modern isle]

instamat can do it too

[wind briar]

After that, use your heightmap to displace the model until it looks good

And THEN

We use a sneaky cheat way that you shouldn't rely on much, but works fine for rocks called "Decimation"

Decimation modifiers will analyze the mesh and automatically nuke down the polygon count to whatever you specify whilst still retaining the shapes as best as it can

[modern isle]

Also known as a Polygon Reduction, except instead of being fully naive in it's approach it first combines polygons that have the lowest relative angle to each other. Meaning if something is flat, those get merged before anything gets touched that changes the shape

[dry shoal]

Subdivide it using those rocky surfaces you made beforehand? Is that a thing?

[modern isle]

the rocks are baked... into a .... repeat after me...

heightmap

[dry shoal]

OHHHHHHHHH

So traggey essentially hand makes heightmaps

[modern isle]

yeah, just like how I did it in instamat with the tech wall, it can be modelled too

[dry shoal]

Wait this looks great

That would make my job so much easier

So if I'm understanding right, it's applying the heightmap over the whole mesh, then using that as a reference to algorithmically add vertices and move them around to form a more variable surface height mesh. Then it optimizes it and bakes mesh normals for you? @modern isle

[modern isle]

it's not adding verticies in this case

all it's doing is taking the texel at the vertex location and moving it based on the value/brightness of the heightmap at this point

what I'm doing is before applying the displacement I subdivide it, then after displacing I decimate to get back to a more reasonable polygon count

[dry shoal]

Ah so by default it'll just move the vertices it can to the best of it's ability

Subdividing = more vertices to work with, therefore more accurate displacement

?

[modern isle]

correct

[dry shoal]

[dry shoal]

Today's plan: make progress on the clone retopo, potentially export the BSP geometry to a mesh ( @modern isle how'd you do that btw ), start work on materials to be used in the environment

Bonus if I'm feeling extra up to it: improve some level geometry. More specifically, the roof above the first underground area in the level. At the moment it's just a flat mesh that shares textures with other meshes which causes problems

And rune while you're here, if I use a heightmap for displacement on a mesh then when I add a material to it with the same height mesh, would leaving out the displacement within Remix provide the same/better visual result but without POM's limitations?

[modern isle]

• Open the Republic Commando Unreal Editor, open a map and then go to File -> Export and select the .t3d file format. Save it somewhere accessible.
• Download Unreal Engine 4.27 or whatever version works for you.
• Create an EMPTY project for UE4 and name it UnrealTournament (important) and place it in a folder that is named "RepublicCommandoUE4" or whatever you want to name it so that the filepath is this \RepublicCommandoUE4\UnrealTournament\UnrealTournament.uproject
• Use UT4X Converter https://github.com/xtremexp/UT4X-Converter to convert the UE2 style t3d to UE4 compliant t3d. In the settings for UT2004 select Republic Commando path and in the UT4 Editor folder put your \RepublicCommandoUE4 folder (not UnrealTournament)
• File -> Unreal Tournament 2004 -> Convert map to UE4 and pick your .t3d you exported from the RC editor, leave all the other settings as is and hit convert
• Once it's done it will open a folder with your converted .t3d file, open your UE4 project and create a new EMPTY level. Once you're in that empty scene, open up the .t3d file with a text editor and select all and copy to clipboard, click into the viewport and paste.
• If you want SOME assets to import, go into the staticmesh folder and drag the .obj files in the restrictedassets/wip/mapname folder in the content browser. Many will error out since they're not correctly converted.

[dry shoal]

Thanks a ton! I actually know a guy who's been working on a Remix-like system for porting entire old unreal games to UE5. As a part of that he has tons of t3d files on hand, so I might be able to get the meshes from him directly

[limber marsh]

deus ex guy?

https://twitter.com/DXU241

DXU24 (@DXU241) on X

Game Developer. Bringing Unreal Engine 1 to Unreal Engine 5 at runtime.

https://t.co/TH1hJNJx5A
https://t.co/Ud7G3P42rE

Twitter

if not - kind of interested

[dry shoal]

Different guy. Seems like the one you linked is exclusively doing UE1

The guy I've been talking to lately is trying to make a system that will port UE1, 2, and 3 games to UE5

So converting game assets, importing entire levels, and potentially even scripting? I'd have to double check on that bc that's kind of insane

[limber marsh]

🤷🏻‍♂️ Seems insane. Cool if possible. XXXX% more applicable then remix tbh

[dry shoal]

He feels exactly the same way

[limber marsh]

Does he have anything published? Twitter / youtube / etc?

[dry shoal]

No he's very much in active development. I don't think he uploads anywhere but he has sent screenshots of levels fully imported into the UE5 editor, though without their textures

I would invite you to the server he's been most active posting progress in but it's a dev team server, not a public facing one

[naive robin]

👀

[dry shoal]

Cool stuff right? Interested to see where that goes

We had a bit of a conversation about the practicalities of Remix vs a full engine port. While the engine port certainly gives more direct control, I think Remix has a place too just based on the leap in technology alone

But a successful UE5 porting system would definitely be more practical than building a remix mod I'd say

no pathtracing tho

[limber marsh]

Tbh, even the remix team member said same thing at one point

he wanted to build a translation layer for UE5

instead or remix

but i am guessing legality

so it does sound "possible"

[dry shoal]

I don't think that's any hit to Remix though, as Remix can theoretically cover wayyyy more games

[limber marsh]

tbh, not really.

not till they can figure out shaders

UE3?

[dry shoal]

And on a personal note, I think conceptually it's less interesting. Really cool and a huge accomplishment if it's successful. But under the hood Remix is doing more impressive things

[limber marsh]

Thats 100 pc games

all high profile

Bioshock 1/2/3, batman series

[dry shoal]

What? All fixed function games, all mostly fixed function games with shader model 1

[limber marsh]

shader model 1 is such a small amount of games

[dry shoal]

I agree UE3 has a lot of high profile titles but he hasn't even begin to cover that, I imagine it'll be significantly more complex there

[limber marsh]

fixed function DX8/9 is also

also its not "shader model 1" its basically how complicated the shaders are

and how much of game

like FEAR is dx8

and no way is it going to work any time soon

100s/1000s of shaders

I was a 3d vision modder, we basically had to edit shaders

to run 3d vision games

[helixmod]

the amount of games that can run on remixs current specification [if it works 100%] is still pretty low

Like Xbox original is outside remixes capabilities

[dry shoal]

That's why I mentioned theoretically. But I know it'll be a challenge for FEAR for sure

[dry shoal]

Remix is also in an early beta while the tool he's developing to port to UE5 isn't available in any capacity nor has it had any game run yet

I don't want to downplay what he's done so far and the goal he has, but one exists today and the other doesn't

[limber marsh]

that is true that its a pipe dream till it works

[dry shoal]

I truly wish him the best of luck though, it could do amazing things if it was possible

[limber marsh]

like that n64 raytracing thing

[dry shoal]

I think it'd also be awesome to see what a successful RC port would look like

[limber marsh]

Looks amazing

but two years from shown and still in development.

[naive robin]

We tried to make remix work with helix mod but we couldn’t get the loading order to work right :/

[dry shoal]

Yeah unfortunately he's had less time on his hands. Tbh I think making it an N64 plugin never made much sense. I feel like if he could make a standardized system to plug all these decompilation projects into, that would be really cool

[dry shoal]

It's super cool that you worked on this btw. I always would see it on Pcgamingwiki but never looked into it myself

AFAIK there's a lot of passionate people who want the 3Dvision stuff to keep going

[limber marsh]

Been a while since I was involved at one point I was "the active guy" but yea, they are still going. Its cool that its still going and its effecting stuff outside of 3d vision. Went from helixmod->3dmigoto and thats in more stuff then i know. I saw some stuff where it modded resident evil [non 3d] and its still used for texture mods - afaik people are still using it for genshin impact / FF7-R

not to mention people can play stuff in vr now

with it

[dry shoal]

I think I've seen 3dmigoto sometime too

Small world

[dry shoal]

I accomplished very little yesterday, but I did run into an obstacle that I know how to overcome now. I might be able to start making significant progress to remastering the environmental assets

[main turret]

Heyo @dry shoal , quick question but in terms of putting together a playable build, are you only focusing on the first level of the game as of now?

[dry shoal]

For the current project as much as I'd like to push beyond it, if I want to really push the fidelity then only the first level is feasible. It's a largely solo project, and every single component is something I'm doing for the first time ever. Modeling, environment materials, PBR character model textures, and even some base game modding to further enable Remix's capabilities.

However there is a potential plan in place for after the release of this "Republic Commando RTX" demo project. There's another, larger team of modders working on further improving a remaster of their own within the native engine. If all goes to plan, I'll be teaming up with them and leveraging their existing and future assets to make a Republic Commando Remix mod that would cover the entire game. However it'll be less visually ambitious than what I'm trying to accomplish with my demo

In other words, a pathtraced adaptation of an existing mod. Since they're working within the confines of the native engine, they're limited in the asset quality they can output without having instability issues which I've come to learn RC has a reputation for. However for the Remix adaptation of their mod I should be able to do PBR materials across the board, as their current workflow actually involves modern material making and simply exporting to a flat texture + normal map that can be used by the game. If you haven't spotted the ace, that means the work to increase fidelity on Remix specifically is already done despite that not being on their mind when making this stuff. It just needs to have a new export that includes things Remix can use that the native engine can't

I'm kind of on the brink of a major increase in progress. I have most of the understanding needed to push up asset quality, it's a matter of execution and getting assistance with weird unexpected issues along the way (almost entirely just due to lack of experience). I'm currently working on a much improved version of the base game's clone model, with the next character model I'll work on being the battle droids. But I shift between priorities as to not burn myself out on one thing that's taking longer than intended, so it's not a given that the next big step in progress is completing the ideal version of that clone model

Hope that gives some insight :)

[limber marsh]

time is something thats the most important part of everything

the sad truth is that 95% of ambitious stuff never happens.

So they may be reluctant, without saying as much

so if you keep on showing them what it could look like you can win them over.

they been remastering the game for XX years

i know your saying that you can make do with what you get but just saying in case something happens/ it takes time for them to give/issues understanding/etc

[dry shoal]

actually it's basically locked in already :) I asked explicitly, told them the scope of their own involvement I'd need for the adaptation (not very much), and they said they were down

especially since I'll be contributing to assets myself in some capacity

just the Remix version will be higher poly, and the materials will be PBR while the native engine's won't

[limber marsh]

It is kind of crazy that I am already eating my words. Their are engine limitations but I see they got alot of stuff working in Fallout which shouldnt be possible.

[burnt light]

Barnyard was also exactly like this (development only stopped because it broke on my system). compatibility with games this heavy in shader utilization has been great for quite a while now, it's just taking until now for us to pursue them more 🙂

[dry shoal]

Ha, one of those times where we want our skepticism to be wrong huh? I'm interested to see where that goes too. I imagine ENBs will quickly become irrelevant in those communities as more people are able to use Remix at a performance level they're comfortable with

People played Fallout/Skyrim for years using ENBs and mods that pushed the engine in the most inefficient ways, getting like 40-50fps on the most monster GPUs out there. That community would embrace remix more than most I think

[burnt light]

initial reception to the FNV images i shared in the xNVSE server wasn't as good as i hoped. only one person replied and it was pretty disappointing, lol

so i'm not so confident

BUT i think it's pursuing anyway

[limber marsh]

Tbh, people here are often amazed something just works but the truth is a game needs to look worse before it looks better.

like we can imagine what it "could" look like when its done but anyone else just sees something worse

I would say it would be nice to do a spot 100% but i am not sure of what issues game has

[burnt light]

i really feel like FNV is the exception. it's a straight upgrade visually from the base game as far as i'm concerned

so that's why i was especially disappointed

buut in general, i agree

[limber marsh]

That was my life as a 3D gamer/modder

you can see it as better [perhaps it is] but its not enough yet

to change from visual look/style to a diehard fan. X% vs XX% difference

[burnt light]

yeah...

i guess i'm the type of modder to not really care what the fanbase wants and just do what i feel is right for the original game's art style

[dry shoal]

working on doing retopology for the original mesh. This is as far as I'm going today but I'm very proud for being super minimal with it. Only doing enough to match the shape but with superior topology. Hopefully it'll give a strong base to improve the mesh with. Likely with more handcrafting than anything else I've done so far due to how intricate it is

[dry shoal]

my first ever made from scratch* material!!!! Threw it onto the old gun texture to test and yep, now light has a more direct effect on how the material looks. No more light being baked into the texture

in the future I'd like to add emissives to the ammo display counter and the magazine's ammo counter

nothing crazy, just a subtle glow. I have an ID map for the rest of the gun so in theory I could overhaul the textures relatively quick but the UV map of the original mesh sucks and I'd want to just redo it. But if I redo it, it should be on a remastered mesh regardless

otherwise why waste all that effort

gonna do some gaming but I'm slowly getting back into the groove of things. I enjoy making materials (not models, fuck blender)

[dry shoal]

made a second material !

gonna try applying this. Need to check my ID map to see if it covers all it needs

[dry shoal]

who needs an ID map when you can just cover the whole thing bc fk it why not

I think I need to lower the roughness. It's not reflecting enough light

[modern isle]

I think baking some mesh normals will greatly help with the look

oh btw I was bored. In case you need a clone commando visor material. It's set up to blend in the emissive only in crevices so that it gives that nice rim glow

[dry shoal]

I'd love to give it a shot, looks stunning

Idk how you do it, great stuff

[dry shoal]

Certainly. Even looking at the preview in mixer it was clearly very wrong

[modern isle]

step 1: add some noise
step 2: add some more noise
step 3: add some... you guessed it! Grunge

[dry shoal]

Eventually I'll figure this out lol

Idea makes sense, just the execution will take time to nail down