Comparing Godot to Unreal, my man, get me some of that weed you smoke
Unity VaM 2 [v1.1.1 Beta] [Meshed VR]
- Thread starter Crazygamer43
- Start date
Tha D Wizzy
Newbie
- Mar 16, 2022
- 84
- 138
Only project GWN looks even close to what VaM is on terms of desired market. Mind you VaM isn't a "game", its mainly a real time DaZ poser. That's what most of us creators use it for anyway.
- Sep 8, 2019
- 196
- 498
You should maybe learn to read, I clearly said "there isn't any good competitors right now".Comparing Godot to Unreal, my man, get me some of that weed you smoke
Reading is a very useful skill, believe me.
Because the long-term idea would be to move from static scenes to creating live, real-time interactions in virtual environments with 3D characters, streaming live similar to Chaturbate but with full 3D VR immersion, everything built in 3D. With VaM2 this still isn’t very practical, especially when it comes to real-time facial and body animation of DAZ3D characters. Meanwhile, with MetaHuman almost everything is ready to take adult entertainment to the next level. Imagine fully 3D-created virtual models that could either be trained with ChatGPT or directly animated by a real woman using mocap devices to stream and interact live. This could generate a huge amount of money if that space opens up.
Considering how early the project is "not practical" may not be the right word, just "not possible yet because there's no mechanics". Doing that in Unity with the current tool set is actually very practical and viable, considering what people already did in VaM1. I recall that there was a creator doing live streaming with VAM with a complex setup, streamdeck and all. For VAM1 the biggest issues are performance and fidelity, and having tried myself to do realtime mocap and facial animation i know it's possible and not that hard, with the possibility of getting good visual results (mechanically), just not performant enough to do that in real time with good enough quality. With VAM2 those two main limitations will go away (i really hope so), so i see a great path for what you described in there when the main features are implemented.
It's really hard to visualize that if you haven't dealt with VAM2's AddonKit, and it's understandable because there's nothing showing off any complex stuff yet. But even if DAZ stuff isn't good enough, a great selling point is the possibility to import any model you want in VAM2, you can literally go and do that right now with the SDK if you want to learn and get some results. It's just a matter of who will do it first and show what's possible.
People are waiting for features (plugins and hair, at least) and official support for that on the hub for publishing, but i saw people doing many experiments. I did some furry stuff and ported a (almost) full VAM1 scene over by rebuilding the assets and exporting animations. Mocap is from Virtamateur. I, at least, am excited about all the custom stuff you can import but most creators will be using the standard DAZ figures for morphs and all.
You don't have permission to view the spoiler content.
Log in or register now.
The main limitation of VAM2 will be the Daz3d characters, since the facial rigging and morphs are horrendous, and the facial wrinkles are practically nonexistent as far as I know. It would be necessary to recreate this entire wrinkle system again, and this is not an easy model to do well and perfectly adapt to all characters, in contrast to metahumans and character creator characters, which have already achieved this.
You must be registered to see the links
It's surprising to see the advances and customizations of metahumans in body and facial shapes, where very few morphs are missing to have perfect sexual characteristics, and if there's anyone who can create genitals for metahumans, it's meshes.
Look at the quality of metahuman wrinkles.
The customization of arms, legs, hips, torso, etc., now allows for a wide variety of creations and customizations, creating diverse characters faster and faster.
It does look impressive!
A little before VAM2 beta came around i was looking to make a 'animation player' for VR. I looked to Unity and UE, and while i haven't tried UE, from what i read from comparisons between both engines Unreal seems to be very easy to achieve something good fast, but the complexity to make advanced stuff on it is higher than Unity's. For a sandbox game like VAM2 you need to consider how easy will be for you community to do stuff on it, both from code and assets. Photo realism is great but doesn't mean much when you can't do a lot of customizations (from visual assets to code) without being an expert. Unity, or DAZ, will not give you that quality right away like Unreal but it could be achievable with some work.
I read some discussions about HD morphs in DAZ, you can get some very good quality with them. Meshed said the plan is to build a custom (and DAZ independent, since it's currently gated) way to create HD morphs for VAM2, which looks really detailed in DAZ, and Genesis 9 has an even higher polycount. He also talked about wrinkle maps and the integration in his custom skinned mesh renderer, so that level of detail may be achieved in Unity. How good will it be, we'll find out eventually, but i believe we can see that level in quality while enjoying the freedom Unity provides for custom stuff.
Also, might be just my personal taste, but the vast majority of DAZ renders i see always look 'off'. With some light work VAM1 looks way better, and the current Gen8 figure in VAM2 already does look superior in it's unfinished state to most stuff out there. While DAZ alone may not deliver, with some improvements on the egnine side it has a ton of potential.
Dukez
Member
- Dec 19, 2020
- 448
- 1,602
Honestly someone could clone VAM into UE5 and have a far superior product. I would venture that it would even release before VaM2 ever does so it would render VaM2 obsolete on arrival.
Tha D Wizzy
Newbie
- Mar 16, 2022
- 84
- 138
This blows VaM and DaZ out of the water WTF. It was a dream to have actual facial expressions integrated into the program. The plugin creators do all the work, Meshed owes his success to them!
If we had metahuman figures, potentially we could get 1:1 clones of real people or atleast near 94% way easier, I dont think no daz3d model will ever mimic a real person to this degree unfortunately :'(
BUt obviously using real persons image without consent is bad, only if you're not using their name commercially or defaming them then were good
Is that raytraced lighting, because we still havent seen any sorta raytracing in vam yet, maybe it would look something better.
I like how people keep saying "Just use Metahumans and make clones of real people"
Ok. Do that. Open up UE5 and make a high quality lookalike with Metahumans and report back to the thread.
Like, If UE5 is so easy, and you can just magically turn Metahumans into whoever you want like you can with VaM and Daz3d, why isn't anyone doing it?
Ok. Do that. Open up UE5 and make a high quality lookalike with Metahumans and report back to the thread.
Like, If UE5 is so easy, and you can just magically turn Metahumans into whoever you want like you can with VaM and Daz3d, why isn't anyone doing it?
well duh, you need experience in order to get good results and also it is easier, as in easier to get more details accurate. If you ever created a vam model you should know the topology doesn't work well with certain body shapes like cartoonish game characters or extremely thick people . also attempting to replicate 1:1 face details is almost impossible on a vam model, like what?
and also it is easier, as in easier to get more details accurate. If you ever created a vam model you should know the topology doesn't work well with certain body shapes like cartoonish game characters or extremely thick people . also attempting to replicate 1:1 face details is almost impossible on a vam model, like what?
Kintaroe
Active Member
- Apr 5, 2019
- 750
- 1,992
UE 4 and 5 are a mess. Too many inconsistencies to even think about creating anything with it. Its just too much work for the result u are getting. Ill leave that engine for the big gaming companies
well duh, you need experience in order to get good results
If Metahumans require a high degree of skill to use, and are not particularly user friendly in terms of modification, they are worse than Daz3d.
While on a technical level yes, things like proper face mapping and wrinkling and things are great to have, if you can't get a Metahuman looking like the person you want it to at all, then those models being higher fidelity doesn't mean anything for this sort of game. The entire point is the customization.
If Metahumans could be flipped into a lookalike builder with ease, you'd see lots of clones out there using metahumans. The reality is the only people using Metahumans to match real life people are big budget AAA studios who literally get full 3D scans of actors they convert.
Velomous
Member
- Jan 14, 2024
- 400
- 413
You are correct, but this is a lot harder than you'd think.
Games like VAM are going down an unpaved road, you have to implement a lot of key features from scratch with minimal guidance, and unreal has a lot of issues. Prototyping in unreal with blueprints is fantastic, but you don't want to actually use that in a final game or you get a slow piece of garbage (something that'd make even the original VAM look fast, this is why a lot of unreal indie games and even bigger unreal games have garbage performance, overuse of blueprints, there's a technique to it, where you re-implement things that may be causing slowdowns in c++ and then access the c++ functions through blueprints, that's how you get the best of both worlds, but modern programmers are allergic to optimization so... go figure.)
Just implementing softbody (realistic jiggle simulation) in unreal is an absolute nightmare, but there are a dozen different established ways to do it in unity, this might even be the main reason the vam devs are using unity, me and a couple others have been trying to create good soft body for humanoids in unreal a while back, and we found a lot of partial solutions but nothing truly satisfying. There is exactly 1 game made in unreal i know of that had impressive looking softbody sim (better than vam even).
Also unreal's rendering engine is a total mess and lumen sucks ass, there is still no proper solution for unreal's infamous ghosting that doesn't sacrifice important graphical features for instance. Also the colors, you cannot get unreal to render colors accurately, or at least doing so requires esoteric knowledge, you can look on the epic forums you'll find a bunch of posts asking how to fix unreal rendering colors way off base, but you will not find answers to those questions. It seems like very few people know how to work the rendering engine and the few who do know aren't sharing the knowledge. All these unreal games that are being released they have this problem, the devs making those games just accepted it, just let it be fucked, settled for 'good enough', probably fucked with the material properties to make things look more correct and things of that nature.
It's all doable, truly, and a really exceptional dev, or a team of good ones could do it too (naughty sandbox dev comes to mind), but a product like this is extremely complicated to create. The vam devs may not be very competent, as shown by how dogshit vam's performance and memory optimization is in general, and how long they have taken to make vam2. But they do keep going, keep working, keep progressing, and that's the most important part for a project like this.
The engine isn't really the problem, unity is a pretty decent engine when it comes down to it, i do agree that unreal is better after working with both but... Just moving a game to a different engine won't really matter, it's the devs, not the engine, that make the game. It's the competence of the devs that determines how good everything is, not the engine. And if the engine has a problem, it's up to the devs to fix or work around it in the end too.
Also I laughed when someone was comparing godot to unreal, godot isn't even in the same ballpark as unity and unreal. An advanced 3d game like this would be borderline impossible to make in godot.
Less that and more of studios thinking they can use BP to shove programming work onto designers to save time and costs
Unreal is a fantastic engine if used correctly. The issue is that it's much cheaper to use it incorrectly. Plus the documentation is abysmal dogshit
Last edited: