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

Anonus2021

Newbie
Oct 8, 2022
18
7
In which update you plan to add new characters? Also since non-furry was the topic, when do you plan to release this as well if we count every image except card ones?
 

NoodleJacuzzi

Active Member
Game Developer
May 20, 2019
727
1,727
I do think the main appeal of pocketmanz is that its variety of species, not sure who is really advocating for the humanization of pokemon when the base game humans would satisfy their needs. As for tarot cards, well I'm not sure of its demand either.

You could have an option at the start of the game to disable tarot and pocketmanz cards for full human experience, with cheats to reopen them. If they want a dedicated full human experience, then your other games are a perfect fit for those people. For now, it's better to flesh out the standard version for expediency and momentum, and put the human version afterwards as this game is far from complete.
Yeah, i don't think you should mess with the tarot and pocketmanz cards, is just too much work for something that is just an extra in the game.
An option for disabling the cards is the best option i think.
Oops, forgot to actually reply to these.

I think you're both right about not worrying about the cards. With them being optional it's not too big of a deal if the non-furry version doesn't affect those. There's actually already a way to disable them: If you choose to discard the first card you find, you won't ever find any more.

I will say that while yes the main game is probably more important to work on than the non-furry version, the longer I put it off, the bigger workload I'll have waiting for me.

Plus if I figure out the best workflow for collecting and converting the furry images to non-furry ones, I can be careful not to do anything that will make the conversion process much harder. For example, when making the pog coins I actually put a bunch of the character's details out of their usual order, and as such my current conversion setup doesn't work with them; I had to rewrite them to non-furry versions manually.

Plus plus, having a non-furry version of v1 will let people know what to expect. I've seen some people on places like Mopoga saying they're hoping the non-furry version won't use shortstack proportions, I don't want to lead them on.

In which update you plan to add new characters? Also since non-furry was the topic, when do you plan to release this as well if we count every image except card ones?
Hmm good question. I haven't decided on the next round of characters. I should probably finish up Sorbet and Riley first as they don't have a lot left, so maybe v3?

For the non-furry images, it's tough, there's a lot of work involved.
  1. I wrote a python script that grabs all of the .txt files I used to generate the game's images, find which character the prompt is referring to, remove all of their tags and replace them with the equivalent human version of the character's keywords. It wasn't perfect, especially because I did a few images differently, so I'm re-generating some as I type this.
  2. From there, I use Easyfluff to generate the human images. Naturally as a furry model there are a lot of errors involved, so I have to fix up any mistakes myself. Small details won't matter because of the next step but things like characters having tails or missing crucial things like Riley's bowtie and etc need to be manually drawn out/in.
  3. Next, I need to re-tag the images again, because the next step is to use img2img and shift all the game's images to the SDXL AutismMix style as that looks a lot better and does better with details like hands. Easyfluff is a very stubborn model so in order to get proportions right I need to insist upon things like characters having big butts. But since autismmix is very finnicky with tags this means that big butt specifically will often generate the character from behind, so that needs to be removed on any front-facing shot.
  4. From there I'll continually generate, check, re-tag, regenerate until I have the full set of all the images needed for v1.
Now once that's done I'll re-name and sort all the images into their proper folders. But I'll also try and use the finished images to train a new model, which should -hopefully- let me skip step 2 for the future, and maybe even skip step 3 as well, as ideally I'll be able to just generate in autismmix from the start.

Here's the progress currently, I'm waiting on re-generating all the total failures from where my keyword conversion script totally failed.

progress.JPG

I also started playing a game called unicorn overlord and that sucked up a bunch of my free time since I got addicted to it, sorry :p
 

ᴏᴄᴛᴏ

Active Member
Jan 22, 2018
538
1,387
Some of the art is reminiscence of
Well that's probably because he generates his pictures with AI, trained on Syuro's art.

In the start-up screen for the game, he dev mentions they trained their own model on Syuro's artwork, and directs people who like the art to support Syuro directly instead of him.
Based dev.

Also, NoodleJacuzzi, is there any chance you'll be animating the sex scenes or is it too much work? I don't know a thing about AI animation, so I don't know how time consuming or hard it is.
 
Last edited:
  • Thinking Face
Reactions: MrStal0ne

NoodleJacuzzi

Active Member
Game Developer
May 20, 2019
727
1,727
Well alright then, lets wait for the further updates
I hope you'll enjoy them!

Progress is going well, I've finished all the SD1.5 generations and now I'm doing SDXL conversions.
progress2.JPG

Due to the differences in how the models are tagged its a lot of throwing things at the wall to see what works.

You don't have permission to view the spoiler content. Log in or register now.

I plan on doing image sets in chunks here. The actual images needed laid out are:

Logbook images
Gathering encounter images
Pog coins
Magazines
Angelica's scenes
Bluebell's scenes
Cayenne's scenes
Sorbet's scenes
And Riley's scenes

I can't help but think I'm missing something... Oh well.

Some images will require special attention. Off the top of my head I know that Jasper and Garnet standing next to each other for their intro will require using regional prompting.
The magazines in particular will be larger time sinks than the rest, due to needing custom covers.

I don't understand why your even doing a non-furry version. How will the plot even work?
Because I said I would!

Specifically, when development of the game was just starting, the game's artstyle looked like this:

84811-3975219980-_lyco_Syuro v4-000040_0.9_, syuro, anthro, solo, butt focus, rear view, looki...png 152610-3558208117-_lora_Syuro v4-000040_0.9_, syuro, solo, anthro, mayor, domestic dog, blue b...png 154010-3659407272-_lora_Syuro v4-000040_0.9_, syuro, anthro, chest focus, upper body, cow milf...png 212283-2916715591-_lora_Syuro v4-000040_0.9_, syuro, solo, anthro, lying, on back, sadogato, d...png

I was limited to fluffyrock at the time, a much more cartoony style. I knew it would divisive, especially since this is my first game with a furry artstyle, so to encourage my fanbase to stick with me on the new project I promised a non-furry version, and no matter how long something takes, if I promise to do something, it'll get done.

... Eventually.

Maybe it isn't necessary, but I think my work ethic is the reason some people follow me.

Plus, a lot of the techniques I'm using to both mechanically allow for non-furry art in the game, as well as the techniques I'm using to generate the art, will be usable in my future projects.

Syrup town only exists because of the groundwork I laid making Hentai University, which itself is built on improvements of human alteration app. I'll take the skills and things I made making Syrup Town and use them to improve Hentai University and any other games I make in the future.

Probably a handful of rewriting the intro.
No the plan is for the non-furry version to be a cosmetic change, strictly for the people who don't want to look at furry art as they play. If I were to do a rework, I would say it's a town of shortstacks, or gnomes, and try to replace any mention of animal folk with those.

It's a lot more than just the intro, the main character is a fluff-huffing maniac and I use that as a crutch when I get writer's block, as well as tails and fur being mentioned in many scenes.

Well that's probably because he generates his pictures with AI, trained on Syuro's art.


Based dev.

Also, NoodleJacuzzi, is there any chance you'll be animating the sex scenes or is it too much work? I don't know a thing about AI animation, so I don't know how time consuming or hard it is.
Oh I didn't notice that. It looks more like Dagasi than syuro to me, imo.

I tried out some AI animation things last year and they didn't look great, but that's probably on me since zoom and swaying effects were all I could do. They didn't seem worth it for the extra work and size increase.

It feels like there should be an easy way to get a thrusting animation working, but it's beyond my skill level. Maybe by laying out all my workflows like this, someone who does have the knowhow to do AI animation will make a similar game with that included?

All my games are totally open, everyone's free to take whatever they'd like. If someone could make this game for example better than I could, I'd happily encourage them to and I'd go work on the way too many other ideas I have.
 
Last edited:

HAAAAAAAAAAAA

Newbie
Jul 3, 2021
22
57
I hope you'll enjoy them!

Progress is going well, I've finished all the SD1.5 generations and now I'm doing SDXL conversions.
View attachment 3783400

Due to the differences in how the models are tagged its a lot of throwing things at the wall to see what works.

You don't have permission to view the spoiler content. Log in or register now.

I plan on doing image sets in chunks here. The actual images needed laid out are:

Logbook images
Gathering encounter images
Pog coins
Magazines
Angelica's scenes
Bluebell's scenes
Cayenne's scenes
Sorbet's scenes
And Riley's scenes

I can't help but think I'm missing something... Oh well.

Some images will require special attention. Off the top of my head I know that Jasper and Garnet standing next to each other for their intro will require using regional prompting.
The magazines in particular will be larger time sinks than the rest, due to needing custom covers.



Because I said I would!

Specifically, when development of the game was just starting, the game's artstyle looked like this:

View attachment 3783424 View attachment 3783425 View attachment 3783426 View attachment 3783427

I was limited to fluffyrock at the time, a much more cartoony style. I knew it would divisive, especially since this is my first game with a furry artstyle, so to encourage my fanbase to stick with me on the new project I promised a non-furry version, and no matter how long something takes, if I promise to do something, it'll get done.

... Eventually.

Maybe it isn't necessary, but I think my work ethic is the reason some people follow me.

Plus, a lot of the techniques I'm using to both mechanically allow for non-furry art in the game, as well as the techniques I'm using to generate the art, will be usable in my future projects.

Syrup town only exists because of the groundwork I laid making Hentai University, which itself is built on improvements of human alteration app. I'll take the skills and things I made making Syrup Town and use them to improve Hentai University and any other games I make in the future.



No the plan is for the non-furry version to be a cosmetic change, strictly for the people who don't want to look at furry art as they play. If I were to do a rework, I would say it's a town of shortstacks, or gnomes, and try to replace any mention of animal folk with those.

It's a lot more than just the intro, the main character is a fluff-huffing maniac and I use that as a crutch when I get writer's block, as well as tails and fur being mentioned in many scenes.



Oh I didn't notice that. It looks more like Dagasi than syuro to me, imo.

I tried out some AI animation things last year and they didn't look great, but that's probably on me since zoom and swaying effects were all I could do. They didn't seem worth it for the extra work and size increase.

It feels like there should be an easy way to get a thrusting animation working, but it's beyond my skill level. Maybe by laying out all my workflows like this, someone who does have the knowhow to do AI animation will make a similar game with that included?

All my games are totally open, everyone's free to take whatever they'd like. If someone could make this game for example better than I could, I'd happily encourage them to and I'd go work on the way too many other ideas I have.
You tried Dragon bones animation at all? I've seen a lot of stuff done with it that looks pretty good.
 

NoodleJacuzzi

Active Member
Game Developer
May 20, 2019
727
1,727
block-syrup-fleshy.jpg

Doing final testing on the non-furry version of v1 here:

PC:

Mobile:

These are public test servers, where I test things before uploading them to the game proper. I don't recommend using this link as your primary method of playing the game as it'll generally be buggy and only updated when I intend to test something.

All in all I made images for the following:
  1. Expressions
  2. Logbook images
  3. Gathering encounter images
  4. Random encounter images (like the fairy)
  5. Pog coins
  6. Magazines (the covers were fun to make)
  7. Angelica's scenes
  8. Bluebell's scenes
  9. Cayenne's scenes
  10. Fox twins scenes
  11. Sorbet's scenes
  12. And Riley's scenes
Holy guacamole that was a lot of images to make! As a reminder the non-furry mode is cosmetic only and does not cover the tarot pocketmanz cards. Still, all of the above came out to over a thousand images of shortstacks.

After testing is finished and I find all the images that need fixing, as well as fixing any code that broke, I'll be able to release. The toggle button only appears when both sets of images are present. My plan is to post links for:
  • Downloadable furry version (large png)
  • Downloadable furry version (smaller webp)
  • Downloadable non-furry version (large png)
  • Downloadable non-furry version (smaller webp)
  • Downloadable two-in-one version (smaller webp)
  • APK furry version (webps)
  • APK non-furry version (webps)
So testing is definitely important, especially as I had to do a lot of sweeping changes to the engine to get the toggle working at all. Let me know what you all think/if you find any bugs. If nothing's super broken I'll release tomorrow.

You tried Dragon bones animation at all? I've seen a lot of stuff done with it that looks pretty good.
Nah I'm not interested in learning traditional animation. I know how to get things animated, but it'd be a lot of redrawing and moving things around. Learning to use CSS is already taking up 90% of my brain capacity already.

I was more hoping for an AI tool or something I could entirely automate and just do corrections for. Traditionally animating the scenes would take too long for my smooth zoomer brain. Thank you for the suggestion though!
 
Last edited:
  • Like
Reactions: xanon515x

Jaxelh

New Member
Dec 3, 2019
4
7
Seeding gen-ai off of one specific artist and then naming the game after that same artist is... certainly a choice.
 

NoodleJacuzzi

Active Member
Game Developer
May 20, 2019
727
1,727


v1-5.png

The first non-furry version is officially released! Shortstacks shortstacks shortstacks! Download links for the non-furry versions are in the posts above, the online version has been updated with a toggle system.
  • All costumes and expressions
  • Logbook images
  • Gathering encounter images
  • Random encounter images
  • Pog coins
  • Magazines
  • Angelica's scenes
  • Bluebell's scenes
  • Cayenne's scenes
  • Fox twins scenes
  • Sorbet's scenes
  • And Riley's scenes
I also added a changelog to the title screen (click the version number to see it), and all the v1.2 bugfixes are included.

Saves between the furry and non-furry versions are completely compatible with each other. It's a cosmetic change, strictly for the people who don't want to look at furry art as they play.

I don't have any current plans of doing rewrites, the main character is a fluff-huffing maniac and I use that as a crutch when I get writer's block, and a whole lot of scenes mention tails and fur directly.

As I mention in the post v2 isn't too far behind. If I hadn't gotten so sick I would have probably had it done already, sorry! I've got a lot to do, hardly enough time...

Please let me know if you find any bugs, have any feedback, or if there's some different way I should juggle the furry and non-furry versions of the game.

Seeding gen-ai off of one specific artist and then naming the game after that same artist is... certainly a choice.
It's because I don't want to hide it at all. In the game's intro I directly link Syuro's fanbox as well, telling people to support them instead of me if they like the art, or if they like the game because of the art. And it also isn't just Syuro, many more artists had their works taken to be used to train easyfluff. I'm still trying to make a list of all of them, ideally in a way that shows how many images of each artist were used to train the model. I know fluffyrock had a list like that.

I just like making games, all of my stuff is distributed freely. Also, I mentioned it before in the thread but I tried to have the AI disclaimer put above the overview but it was moved below and into a spoiler, I assume due to formatting rules.
 
4.20 star(s) 17 Votes