4.10 star(s) 154 Votes

ikaruga

Well-Known Member
Modder
Donor
Respected User
Aug 11, 2016
1,192
2,512
I can't save clothing, I would like to save a nude on, or a "no underwear on", but following the in game help, i need to put on camera then middle mouse button, but it removes model, no save outfit ...
 

TaelF

Newbie
Jan 29, 2018
39
42
if the game werent so big on the interaction system thats too clunky to use with mouse & keyboard. maybe as a VR game, but not this.
Game actually isn't as bad if you use the Better Experience improvements, altho that mod has been buggy for me and the new controls always end up breaking, but it kind of plays like Surgeon Simulator, in the sense that its pretty confusing until you get the hang of it, with the default controls I will never understand how to rotate the hand to fap :HideThePain:
 

Hryme

Member
Aug 15, 2016
270
228
Any pros here have a quick idea on how to rate okay. I don't know what I am doing wrong but in my second generation I often get girls that doesn't look good.

I try to vote something I like at 10 for every category in the first week but somehow the girls still spawn ugly.

Is it needed to rate every girl in everything?

I hoped it was possible to just rate what you liked on each girl at 10 and just leave the rest at 0.
 
Jan 7, 2022
362
469
Any pros here have a quick idea on how to rate okay. I don't know what I am doing wrong but in my second generation I often get girls that doesn't look good.

I try to vote something I like at 10 for every category in the first week but somehow the girls still spawn ugly.

Is it needed to rate every girl in everything?

I hoped it was possible to just rate what you liked on each girl at 10 and just leave the rest at 0.
When you open the rating screen, set everything to 5. That does what you think zero does. Five means you don't have an opinion. Then tic the sliders up or down, a little or a lot, depending on what you can observe.
 

Hryme

Member
Aug 15, 2016
270
228
Thanks. That will really help.

Edit: As a user interface improvement the default rating should be set at 5 and not 0 then if the developer reads this.
 
Last edited:

urmum123

Member
Feb 7, 2020
264
126
guys, is it possible to use the monkey editor to make a profile or can you pass a saved character to turn it into a profile or load or load a single interview character with a monkey in game, to be able to score it and have more of them come out? with the monkey editor you can exaggerate even more certain characteristics like the size of the breasts, hips, etc. I would like to create a profile with those exaggerated measures, so that the game starts to create characters related to those measures that are not as an option in the common editor to create profiles.
 
  • Like
Reactions: Vbrz

bloodbeetle

Newbie
May 13, 2018
27
50
guys, is it possible to use the monkey editor to make a profile or can you pass a saved character to turn it into a profile or load or load a single interview character with a monkey in game, to be able to score it and have more of them come out? with the monkey editor you can exaggerate even more certain characteristics like the size of the breasts, hips, etc. I would like to create a profile with those exaggerated measures, so that the game starts to create characters related to those measures that are not as an option in the common editor to create profiles.
just edit the character or clothing on monkey and save it with your phone or camera depending on what you want. i always use it that way
 

urmum123

Member
Feb 7, 2020
264
126
just edit the character or clothing on monkey and save it with your phone or camera depending on what you want. i always use it that way
actually that's what I do, but I would like that little by little the game starts to generate those characters and not that I have to make them manually, it's much more pleasant with the surprise factor.
 
  • Like
Reactions: ricche

arthurthethird

New Member
Feb 25, 2022
5
5
actually that's what I do, but I would like that little by little the game starts to generate those characters and not that I have to make them manually, it's much more pleasant with the surprise factor.
What I did was Editting them using monkey editor then start Rating them based by what I did from the monkey generator and the more higher the rate is the better model you will get after. I've been getting less and less model with a weird ass make up and nylon hair XD
 
  • Like
Reactions: M925Ya

Vbrz

New Member
Sep 7, 2022
12
22
guys, is it possible to use the monkey editor to make a profile or can you pass a saved character to turn it into a profile or load or load a single interview character with a monkey in game, to be able to score it and have more of them come out? with the monkey editor you can exaggerate even more certain characteristics like the size of the breasts, hips, etc. I would like to create a profile with those exaggerated measures, so that the game starts to create characters related to those measures that are not as an option in the common editor to create profiles.
Agree, it seems to be very usefull feature right now. The example of it can be the purpose of creating high-rated model, where I do need several specific options (like legs and arms thickness, tallness etc), and I want the autorating profile to help me with them namely: I dont care about hands/feet size, any will do, but those values will affect the rating. And so, when I meet the girl I like, I would like to update my current autorating profile with all the values from her. It can be done manually, by creating the monkey custom json with all the ratings, screening/papering it somewhere and creating the new profile manually ingame, but it seems quite boring =)

I would suggest this can be solved by modders (I'm not that smart) in such a way:
When the new girl arrives, profile [current profile] rating file is automatically created (and placed somewhere, saying, with guest_last_... .json). Then [new] profile is created based on [current]. All the profile features I've marked important (all those features should be stated in some separate file, one set for each of 10 groups) are copied from the [game] profile (created in-game normally and set in-game as a autorating profile). All the minor features are being coping from [current] profile. Then [new] profile should be loaded and applyed in-game.

For inspiration, here is the python code for reading the profile.png options (not perfect =):
Python:
# -*- coding: utf-8 -*-
"""
Created on Fri Apr 28 12:15:48 2023
@author: V
"""
# import os
import pprint

game_save_folder = 'C:\\Users\\V\\Documents\\TValle Games\\Some Modeling Agency\\'  #check this path and insert yours!
game_chars_folder = game_save_folder+ 'Characters\\v08\\'
game_profile_folder = game_save_folder+ 'Profiles\\v08\\'
#filename = game_chars_folder+ 'Antoinette Cody.png'     #check this path and insert yours!
filename = game_profile_folder+ 'SlimBusty_MedEasy.png'   #check this path and insert yours!

# chunk generator
def chunk_iter(data):
    total_length = len(data)
    end = 4

    while(end + 8 < total_length):   
        length = int.from_bytes(data[end + 4: end + 8], 'big')
        begin_chunk_type = end + 8
        begin_chunk_data = begin_chunk_type + 4
        end = begin_chunk_data + length

        yield (data[begin_chunk_type: begin_chunk_data],
               data[begin_chunk_data: end])

with open(filename, 'rb') as fobj:
    data = fobj.read()

# check signature
assert data[:8] == b'\x89\x50\x4E\x47\x0D\x0A\x1A\x0A'
print("-" * 80, "chunks of: %s" % filename, "-"*80, sep='\n')

for chunk_type, chunk_data in chunk_iter(data):
    print("chunk type: %s" % chunk_type.decode())
    if   chunk_type == b'le.C':  #for character data
        charChunkData = chunk_data.decode()
        # charChunkData = charChunkData[:100]
        # print("--chunk data:",  chunk_data.decode())
    if   chunk_type == b'lePo': #for profile
        profileChunkData = chunk_data.decode()
        # print("--chunk data:",  chunk_data.decode())
    if   chunk_type == b'iTXt':
        print("--chunk data:",  chunk_data.decode())
    elif chunk_type == b'tEXt':
        print("--chunk data:", *chunk_data.decode('iso-8859-1').split('\0'))
      

prof = profileChunkData[len("rtraitData"):]  #Me was lazy ;P
import json
prof_d = json.loads(prof)
pprint.pp(prof_d)
You can read the marks (interpretations) in-game with monkey custom mods json (stored in \SMA 0.9.8c\BepInEx\config\Monkey\Mods\myCustom_doityourself.json) like this (dont forget to color and bold the caption for tits size!):
JSON:
    {
        "Name": "<color='#FFFAF0'>Face</color> <color='red'>Anger</color> Gain",
        "Type": "Enum",
        "Level": 3,
        "Path": [
            "<MainFemaleCharacter>/Infos/Interpretaciones",
            "InterpretadorDeFemales",
            "interpretacion/interpretacionDeFacialSkin/difficulty/anoyanceGain"
        ]
    }
(read monkey's manual for details)

Well, it seems to work similar to BH alternative genes, but not touching the girls directly - only the autoratings, which, I belive, can help a lot in testing strategies and speed up the process, almost without cheating =)
Me really want at least profile-dump feature for the current model, please point me in case it is already realized and I've missed it =)
 
Last edited:
  • Like
Reactions: urmum123

Unitedbmx69

Newbie
Aug 30, 2019
26
11
Any pros here have a quick idea on how to rate okay. I don't know what I am doing wrong but in my second generation I often get girls that doesn't look good.

I try to vote something I like at 10 for every category in the first week but somehow the girls still spawn ugly.

Is it needed to rate every girl in everything?

I hoped it was possible to just rate what you liked on each girl at 10 and just leave the rest at 0.
For me works next:
1. on the first iteration try to not decline everything. Really ugly things rate as 0-2,
2. something bad but no absolutely ugly rate as 3-4
3. things that are pretty ok but still not good I rate as 4-6
4. really nice parts i rate as 7-8
5. beautiful parts i rate as 9 and only sometimes 10 ( more often9, i think it makes better variations in future ) PS: such parts appear for me only at lvl 2-3
6. yes, you need to rate everything for better results
7. try to pay your attention not only to the separate parts ( mouth, nose, etc ) but also face and head form - it is very important
8. the first level girls are mostly ugly and it needed nearly 15-20 min to rate than all
9. 2nd lvl girls will be much better, try to correct their parts and except some bad things
10. on lvl 3-4 you will receive pretty good results
11. on lvl 2-3 i also trying to correct not only face and main body, but also breasts, butt, pus ( needed more time to undress her and rate it )

here is one of the lvl 3 girl's face, i belive it not bad results
 
Last edited:
  • Like
Reactions: Buraki2

Unitedbmx69

Newbie
Aug 30, 2019
26
11
can i somehow delete groups on higher lvls ? when characters 8-10lvl+ the game becomes unplayable for me because of the huge amount of different groups with different lvls
 

urmum123

Member
Feb 7, 2020
264
126
Agree, it seems to be very usefull feature right now. The example of it can be the purpose of creating high-rated model, where I do need several specific options (like legs and arms thickness, tallness etc), and I want the autorating profile to help me with them namely: I dont care about hands/feet size, any will do, but those values will affect the rating. And so, when I meet the girl I like, I would like to update my current autorating profile with all the values from her. It can be done manually, by creating the monkey custom json with all the ratings, screening/papering it somewhere and creating the new profile manually ingame, but it seems quite boring =)

I would suggest this can be solved by modders (I'm not that smart) in such a way:
When the new girl arrives, profile [current profile] rating file is automatically created (and placed somewhere, saying, with guest_last_... .json). Then [new] profile is created based on [current]. All the profile features I've marked important (all those features should be stated in some separate file, one set for each of 10 groups) are copied from the [game] profile (created in-game normally and set in-game as a autorating profile). All the minor features are being coping from [current] profile. Then [new] profile should be loaded and applyed in-game.

For inspiration, here is the python code for reading the profile.png options (not perfect =):
Python:
# -*- coding: utf-8 -*-
"""
Created on Fri Apr 28 12:15:48 2023
@author: V
"""
# import os
import pprint

game_save_folder = 'C:\\Users\\V\\Documents\\TValle Games\\Some Modeling Agency\\'  #check this path and insert yours!
game_chars_folder = game_save_folder+ 'Characters\\v08\\'
game_profile_folder = game_save_folder+ 'Profiles\\v08\\'
#filename = game_chars_folder+ 'Antoinette Cody.png'     #check this path and insert yours!
filename = game_profile_folder+ 'SlimBusty_MedEasy.png'   #check this path and insert yours!

# chunk generator
def chunk_iter(data):
    total_length = len(data)
    end = 4

    while(end + 8 < total_length):  
        length = int.from_bytes(data[end + 4: end + 8], 'big')
        begin_chunk_type = end + 8
        begin_chunk_data = begin_chunk_type + 4
        end = begin_chunk_data + length

        yield (data[begin_chunk_type: begin_chunk_data],
               data[begin_chunk_data: end])

with open(filename, 'rb') as fobj:
    data = fobj.read()

# check signature
assert data[:8] == b'\x89\x50\x4E\x47\x0D\x0A\x1A\x0A'
print("-" * 80, "chunks of: %s" % filename, "-"*80, sep='\n')

for chunk_type, chunk_data in chunk_iter(data):
    print("chunk type: %s" % chunk_type.decode())
    if   chunk_type == b'le.C':  #for character data
        charChunkData = chunk_data.decode()
        # charChunkData = charChunkData[:100]
        # print("--chunk data:",  chunk_data.decode())
    if   chunk_type == b'lePo': #for profile
        profileChunkData = chunk_data.decode()
        # print("--chunk data:",  chunk_data.decode())
    if   chunk_type == b'iTXt':
        print("--chunk data:",  chunk_data.decode())
    elif chunk_type == b'tEXt':
        print("--chunk data:", *chunk_data.decode('iso-8859-1').split('\0'))
     

prof = profileChunkData[len("rtraitData"):]  #Me was lazy ;P
import json
prof_d = json.loads(prof)
pprint.pp(prof_d)
You can read the marks (interpretations) in-game with monkey custom mods json (stored in \SMA 0.9.8c\BepInEx\config\Monkey\Mods\myCustom_doityourself.json) like this (dont forget to color and bold the caption for tits size!):
JSON:
    {
        "Name": "<color='#FFFAF0'>Face</color> <color='red'>Anger</color> Gain",
        "Type": "Enum",
        "Level": 3,
        "Path": [
            "<MainFemaleCharacter>/Infos/Interpretaciones",
            "InterpretadorDeFemales",
            "interpretacion/interpretacionDeFacialSkin/difficulty/anoyanceGain"
        ]
    }
(read monkey's manual for details)

Well, it seems to work similar to BH alternative genes, but not touching the girls directly - only the autoratings, which, I belive, can help a lot in testing strategies and speed up the process, almost without cheating =)
Me really want at least profile-dump feature for the current model, please point me in case it is already realized and I've missed it =)
I don't know if it's because of my English or my lack of understanding about programming but I don't know what this is XD. What I understand is that you have done something very useful, is that what I mentioned about putting in profile, characters from the single interview? if so, what should I do to add it to the game? I apologize for my absolute ignorance haha
 

Buraki2

Newbie
Jul 1, 2017
42
109
For me works next:
1. on the first iteration try to not decline everything. Really ugly things rate as 0-2,
2. something bad but no absolutely ugly rate as 3-4
3. things that are pretty ok but still not good I rate as 4-6
4. really nice parts i rate as 7-8
5. beautiful parts i rate as 9 and only sometimes 10 ( more often9, i think it makes better variations in future ) PS: such parts appear for me only at lvl 2-3
6. yes, you need to rate everything for better results
7. try to pay your attention not only to the separate parts ( mouth, nose, etc ) but also face and head form - it is very important
8. the first level girls are mostly ugly and it needed nearly 15-20 min to rate than all
9. 2nd lvl girls will be much better, try to correct their parts and except some bad things
10. on lvl 3-4 you will receive pretty good results
11. on lvl 2-3 i also trying to correct not only face and main body, but also breasts, butt, pus ( needed more time to undress her and rate it )

here is one of the lvl 3 girl's face, i belive it not bad results
Thank you, that's great advice...I'm using your method and the results are great...There's a lot to be gained from rarely using a 10 rating...that little bit of randomness is great. Evaluating behavior is more difficult... here I rely on auto rate... When I rate myself, the results are strange.
 

Thora

Member
Jan 2, 2018
343
1,049
... I would suggest this can be solved by modders (I'm not that smart) in such a way: ...
Monkey is currently in support mode, and I don't plan to add new features short term.

I don't want to discourage other modders, but when I looked at the profile system way back it was a mess. To this day, it is still buggy and I wouldn't go anywhere near this until the core issues have been addresssed.

What you are asking makes sense, but it gets complicated. The problem is not decoding the file or saving it; that is easy to solve.

The core issue is the dev decided to use a completely different classification system for profiles. There are no direct and deterministic transform between the gene system and the profile system.

I'll try to keep this non technical. Let's take race as an example.

Genes is what you save so that's the building block. There are no 'race' gene. The profile system has a hard coded representation of what an Asian looks like (ranges of values for skin tone, nose, mouth, eyes, and a ton more). To know 'how Asian' your model is, the game runs a bunch of very specific computation that transforms genes into an 'evaluation' model. Then this is compared to the hard coded model and scored using average Euclidian distance.

There are several problems with this. First there is no current way to take what you have as genes and make it the target. The target is computed and derived from code, it is not direct. Your perfect 'Eurasian' model would require defining the proper targets, computations and values.

Second, you cannot spawn your perfect Eurasian model because it would require writing the reverse transform setting the genes. There is actually some attempt to solve this in the code, but I think the dev stopped it when he realized how difficult that was.

This is not an unsolvable problem. With enough time and effort I'm sure something can be done that will 'mostly' work. But it is not as trivial and simple as people think. At least for me.

If I ever attack this problem, my approach would be to nuke this entire profile/rating system from orbit first. I believe Q somewhat went this road with BE, with the maintenance nightmare to pay for. No thank you :)

-T-
 

urmum123

Member
Feb 7, 2020
264
126
why every time i try to modify my profile in full mode, the game automatically sets everything to very low and a being from the abyss appears? is there any way to avoid it?
 
4.10 star(s) 154 Votes