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

VN RPGM Abandoned Saria Reclaimed [v0.17] [Velminth]

2.40 star(s) 162 Votes

Necro434

New Member
Jun 28, 2021
10
6
Would anyone happen to have a link to 0.13
I wanna see how far the version I was originally interested in got
 
Feb 24, 2019
20
32
Anyone with some image gen skills/experience and a 16gb GPU can continue the development of this game. The steps to follow would be:
1. Being familiar with a stable diffusion interface, getting all the core concepts and basics down over a few days (Automatic1111, maybe comfyUI)
2. Making a LoRA from the already existing art of the game
3. Finding a good combination of prompts and LoRAs using a model like hassakuXL/Pony Diffusion V6 from civitai, to generate new poses and pictures
4. Ta-da, with enough time spent prompting you'll have cut-ins and scenes, in a very similar style to the original.
5. It may also be possible to make animations from there, it's slightly more complicated using only free and local tools. But luma can be a great solution for that purpose.

Do note that this will take more work and time than most people believe it does. And you will need a precise idea of what you want to create each time, and knowledge to use all the tools you have.
This way, someone with the time, passion and programming skills can take up the torch completely, while staying true to the original art and spirit of the game. This is slightly advanced but possible and doable.
Just a suggestion and keywords to help anyone possibly interested, but afraid that the drawing style would be a dead end.
 

BigBiggieBig

Member
Jan 6, 2019
330
903
Anyone with some image gen skills/experience and a 16gb GPU can continue the development of this game. The steps to follow would be:
1. Being familiar with a stable diffusion interface, getting all the core concepts and basics down over a few days (Automatic1111, maybe comfyUI)
2. Making a LoRA from the already existing art of the game
3. Finding a good combination of prompts and LoRAs using a model like hassakuXL/Pony Diffusion V6 from civitai, to generate new poses and pictures
People often skip this information, but being able to photoshop/edit/inpaint generations really elevates them to a point where they lose the generic AI look and tend to look like something acceptable for commercial use. A combination of photoshop, inpainting, LoRAs, a VAE, and a proper model (Could be a custom one, as well) would likely result in a style that is difficult to mimic and is also pleasant to look at.

Of course, it's much easier to boot up Pony V6, type 'Big boobs' and ship out whatever pops up without any changes or fixes as is the case with 99% of games tagged with 'AI CG' on this forum. However, that tends to look like shit and doesn't exactly appeal to a vast majority of users.
 

bruhman969

Member
Apr 21, 2020
136
164
Anyone with some image gen skills/experience and a 16gb GPU can continue the development of this game. The steps to follow would be:
1. Being familiar with a stable diffusion interface, getting all the core concepts and basics down over a few days (Automatic1111, maybe comfyUI)
2. Making a LoRA from the already existing art of the game
3. Finding a good combination of prompts and LoRAs using a model like hassakuXL/Pony Diffusion V6 from civitai, to generate new poses and pictures
4. Ta-da, with enough time spent prompting you'll have cut-ins and scenes, in a very similar style to the original.
5. It may also be possible to make animations from there, it's slightly more complicated using only free and local tools. But luma can be a great solution for that purpose.

Do note that this will take more work and time than most people believe it does. And you will need a precise idea of what you want to create each time, and knowledge to use all the tools you have.
This way, someone with the time, passion and programming skills can take up the torch completely, while staying true to the original art and spirit of the game. This is slightly advanced but possible and doable.
Just a suggestion and keywords to help anyone possibly interested, but afraid that the drawing style would be a dead end.
You should make a team.
 

mbRjpZLD

Member
Apr 28, 2017
159
286
Anyone with some image gen skills/experience and a 16gb GPU can continue the development of this game. The steps to follow would be:
1. Being familiar with a stable diffusion interface, getting all the core concepts and basics down over a few days (Automatic1111, maybe comfyUI)
2. Making a LoRA from the already existing art of the game
3. Finding a good combination of prompts and LoRAs using a model like hassakuXL/Pony Diffusion V6 from civitai, to generate new poses and pictures
4. Ta-da, with enough time spent prompting you'll have cut-ins and scenes, in a very similar style to the original.
5. It may also be possible to make animations from there, it's slightly more complicated using only free and local tools. But luma can be a great solution for that purpose.

Do note that this will take more work and time than most people believe it does. And you will need a precise idea of what you want to create each time, and knowledge to use all the tools you have.
This way, someone with the time, passion and programming skills can take up the torch completely, while staying true to the original art and spirit of the game. This is slightly advanced but possible and doable.
Just a suggestion and keywords to help anyone possibly interested, but afraid that the drawing style would be a dead end.
something like this? ;)
 

HabibHanson

Member
Nov 10, 2019
198
215
Wow this remake really gutted everything out of the game. Is there even a way to unlock new cards or level up? Being stuck at 10 willpower with no option to initiate the old cut-ins makes this incredibly boring.
 

semanuel2630

Newbie
Jun 7, 2020
26
9
Hey guys I think maybe the modders could offer some sollution about this Unfinished Project. A consultation for ideas is being executed there on the UnOffical Versions. Maybe some of you with some Ilustr /Arts Knowedge could offer a good suugestion to them.
 
  • Hey there
Reactions: Vermittler
2.40 star(s) 162 Votes