CREATE YOUR AI CUM SLUT ON CANDY.AI TRY FOR FREE
x

Fan Art Being A DIK: Fan Art

5.00 star(s) 2 Votes
Discussion Reviews (2)
Pillock

Pillock

Member
Nov 23, 2019
485
7,386
Instancing Optimization
Memory v Speed
For those that don't know about the benefit of using memory regarding Instancing Optimization. Here are some quick examples.
10 instances of Quinn set on Memory.
10 instances of Quinn set on Speed. (It calculates each instance as if it was a G8F figure)
100 instances of Quinn set on Memory.
I wouldn't even bother trying 100 instances using speed. lol
Memory 10 instances.jpg Speed 10 instances.jpg Memory 100 instances.jpg
Now this is a dumb example, but think of a bar scene with hundreds of glasses, a dozen bar stools all those tables and chairs etc or a forest, if all the multiple items were instances, it might allow for an extra character in the scene. It could mean the difference between your scene rendering or not.

A quick note this only helps with rendering and Vram, scene navigation will still behave normally, the more there is on screen the slower your system will become.
 
Last edited:
  • Like
Reactions: g1nk, Tjure, The Glorious LIME and 1 other person
Night Hacker

Night Hacker

Forum Fanatic
Jul 3, 2021
4,580
22,328
m4dsk1llz said:
Just for your own piece of mind, the 'eye' problems are almost entirely DAZ fault. They inverted the direction the eyes moves from G3 to G8. Actually inverted is not the right term, they basically rotated the eyeballs on the Z axis 180° when creating G8. I don't know if that was a fuck up, a correction, or just DAZ's way to mess with creators.

If you are doing a conversion from G3 to G8 and also align nodes, the alignment is preserved during the conversion, so your G8 eyes now move in an unexpected manner. The key in creating morphs is to not align them if converting from one gen to another, then the orientation of the new morphs inherit the orientation of the base character (G8 or V8 or M8)
Click to expand...
That and certain eye morphs messed up the blinking. I ended up having to unload ALL my G8F morphs and back them up. Now when I load in a character, if one of the morphs is missing, I add back in what is needed, noting the ones that are what I call "parasite morphs" that attach to everything. So G8F eyes are working again. That was my main problem. I haven't done a lot of conversions from G3 to G8, but I'll keep your advice in mind. Thanks.
 
Night Hacker

Night Hacker

Forum Fanatic
Jul 3, 2021
4,580
22,328
Pillock said:
Instancing Optimization
Memory v Speed
For those that don't know about the benefit of using memory regarding Instancing Optimization. Here are some quick examples.
10 instances of Quinn set on Memory.
10 instances of Quinn set on Speed. (It calculates each instance as if it was a G8F figure)
100 instances of Quinn set on Memory.
I wouldn't even bother trying 100 instances using speed. lol
View attachment 2679949 View attachment 2679950 View attachment 2679951
Now this is a dumb example, but think of a bar scene with hundreds of glasses, a dozen bar stools all those tables and chairs etc or a forest, if all the multiple items were instances, it might allow for an extra character in the scene. It could mean the difference between your scene rendering or not.

A quick note this only helps with rendering and Vram, scene navigation will still behave normally, the more there is on screen the slower your system will become.
Click to expand...
I actually programmed instancing from scratch in my own application. What it does is that it loads the one object into memory, and then renders the same object over and over again without needing to take up more memory for multiple objects. This not only results in not wasting memory, but man is it EVER fast, at least in my own programming. I created a little program with a planet and 10,000 instances of an asteroid creating Saturn like rings with them you could fly though and it was really fast. I could easily do WAY more than that. Love instances. In my own program I had different shapes and sizes for my asteroids, but the objects and textures could not be changed, but the shape could (stretching, resizing etc). Sadly, I don't think you can resize them in DAZ, I think all instances are identical. There's no reason why they have to be though, from a programming perspective.

Clipboard01.jpg
 
  • Like
Reactions: GeekyLion
Pillock

Pillock

Member
Nov 23, 2019
485
7,386
Night Hacker said:
I actually programmed instancing from scratch in my own application. What it does is that it loads the one object into memory, and then renders the same object over and over again without needing to take up more memory for multiple objects. This not only results in not wasting memory, but man is it EVER fast, at least in my own programming. I created a little program with a planet and 10,000 instances of an asteroid creating Saturn like rings with them you could fly though and it was really fast. I could easily do WAY more than that. Love instances. In my own program I had different shapes and sizes for my asteroids, but the objects and textures could not be changed, but the shape could (stretching, resizing etc). Sadly, I don't think you can resize them in DAZ, I think all instances are identical. There's no reason why they have to be though, from a programming perspective.

View attachment 2680351
Click to expand...
The instances can be scaled and rotated in Daz, just not retextured, also if you edit the original mesh changes are made live to the instances. One thing though the instances are invisible when it comes to Dforce, you have to use the original mesh or an actual copy for it to work, A great use for it is background city scapes, load up one of the Urban Sprawl blocks and create an instance or four of the entire block. Awesome stuff.
lilyinstances.jpg
 
  • Like
Reactions: Night Hacker
scooman

scooman

Member
Oct 8, 2022
263
2,231
Night Hacker said:
Well, while playing around with Sage's hair, I realized all her G3F assets were available as G8F assets. Not quite sure why I went only with G3F... probably due to issues I had with G8F female eye problems which turned me off of them at the time, not sure now... but I went ahead and manually recreated Sage as a Genesis 8 Female. I didn't use a script to convert, but recreated her like I create all my assets and I wanted to make certain she looked the same, and she does. I noticed that I somehow missed the fact that Aurora comes with a belly button ring, so I didn't need an extra asset for that (somehow I missed that with G3F). It's been a while since I worked on the G3F model and I learned a lot since then. ANYHOW, without further adieu, here's the G8F version. The render is of the G8F model, again, no script conversions here...

Sage (G8F) Assets
Body:
You must be registered to see the links

Skin:
You must be registered to see the links

Alina Hair:
You must be registered to see the links
(loose hair)
Leony Hair:
You must be registered to see the links
(ponytail hair)
Chunky Pigtails:
You must be registered to see the links
(latest)
Genitals:
You must be registered to see the links


View attachment 2678254

As always, my Sage (G8F) Scene Subset requiring the above assets (which you probably already have)...
View attachment 2678256

I also updated my post with my list of asset links and added this.
Click to expand...
As always... WONDERFULL work Night Hacker. Thank you
 
M

MetaJosh

New Member
Oct 14, 2017
6
1
Does anyone have a stable diffusion model for the characters from this game? I saw one on CivitAi, but didn't download it and can't find it anymore.
 
You must log in or register to reply here.
Top Bottom
5.00 star(s) 2 Votes