|
Humbug Scoolbus posted:For the maps, midjourney kicks out really good battlemaps actually since v4. Yessssss...... Humbug Scoolbus posted:I listed the prompt I used for that village YEEEEEEEEEEEEEEEEEEEEEEEEEEEES!!!!!!!! Leperflesh posted:Hey guys, I've been sick and basically checked out for a week, but I'm poking around now. How's a new OP going? I'd really love to get the stink of Rutibex off of this thread. I was hoping to get some more feedback and make one more pass at the OP. If anyone can post the thread maybe I will? Anyone strongly want to be the one who posts it?
|
#
?
Jul 4, 2023 23:51
|
|
|
|
| # ? May 2, 2024 07:52 |
|
|
Humbug Scoolbus posted:'How to do AI assisted writing' is what the high school creative writing course I teach morphed into. We're still just making it up, but so far, it's improved my students' critical analysis of writing more than I could have imagined This is absolutely brilliant.
|
|
#
?
Jul 6, 2023 18:50
|
|
|
Ai or ML is here, and it is not going away. Learning how to work with it is a lot more useful than screaming in rage.
|
|
#
?
Jul 6, 2023 20:07
|
|
|
The next version of Stable Diffusion's model, SDXL, has leaked. It's the .9 model so it's not quite there yet but it does have much more "midjourney" like coherence and framing. The exciting thing will be the fine tuning and Control net stuff. In 6 months I expect things to be very good for the casual user.
|
|
#
?
Jul 6, 2023 21:32
|
|
|
Just posting here so I can link said post in another thread and not get probated yet again: Playing around with Foundry as a VN style for narrative games: Elysium in Chicago, 1862:  Price Maxwell has entered to address the coterie! 
|
|
#
?
Jul 10, 2023 18:45
|
|
|
Dang, those look super cool. Not just in a "great detail and lighting" sort of way but nice and consistent composition.
|
|
#
?
Jul 11, 2023 23:32
|
|
|
The Eyes Have It posted:Dang, those look super cool. Not just in a "great detail and lighting" sort of way but nice and consistent composition. Well it's a cutout of the NPC. I'm using the backdrops as backgrounds and then generating portraits to be used as NPC sprites and present the game in a visual novel style.
|
|
#
?
Jul 11, 2023 23:48
|
|
|
That's cool, and I really mean those look good  There's no little off elements or anything that I can tell, either. Nice job!
|
|
#
?
Jul 12, 2023 04:20
|
|
|
Trying out the Claude chatbot. works pretty well...quote:I would like a a world for an RPG. The world is named Hive and it has a technological level of Earth in the 1930s. Here are ten countries and their types of government that on Hive, with a technological level equivalent to Earth in the 1930s: So far it's working pretty good. https://claude.ai/
|
|
#
?
Jul 12, 2023 23:01
|
|
|
Humbug Scoolbus posted:Trying out the Claude chatbot. works pretty well... Iiiinteresting - I wonder what the differences are on the backend between this and vanilla ChatGPT. I'm seeing commentary that Claude's real advantage over ChatGPT 4 is in maintaining lengthy context, like reading entire PDFs and answering questions about them. That's the sort of AI assistant I'd find useful for RPGs: "chatbot, please display every creature that can be conjured with Summon Nature's Ally which has a flying speed of at least 80." ***** Been sitting on some thoughts about AI art in the RPG space for a while now. I finally took the sage advice of KwegiboHB and Megaman's Jockstrap and got Automatic1111 running locally. After getting my bearings, I added ControlNet and some must-have extensions (After Detailer and Inpaint Anything are unbelievably good, jesus), grabbed some checkpoints - and even made my own with checkpoint merge, which was shockingly easy! - and after a couple nights of playing around with it and doing some very basic studying, I'm getting results that are just... Astonishing. The specificity that ControlNet and Inpaint Anything give you is remarkable, and solves my biggest problem with the output I was getting from Midjourney. Human artists can obviously outperform what I'm getting, but - well, storytime. One of the players in my long-running D&D 5e game loves the big bad, the corrupted nature-lich Harrower: conceptually, aesthetically, my portrayal of him, he just digs this weird lich A Lot. It's very flattering! Over the years he's done four or five HeroForge versions of Harrower, and cited this as part of his drive to continually up his game with HeroForge, to the point where he's now showing up on the weekly best-of model list for that site whenever he feels sufficiently inspired. None of them have been quite right though, because HeroForge's toolset is limited, and Harrower looks bizarre. Recently, this player got his first job and decided to celebrate by commissioning some art of Harrower, and asked me for as complete a description as I could provide. I said sure, since it was an interesting exercise, and generated a Google doc as if I was commissioning the art myself, complete with visual references, color palettes, clarifying commentary, etc. Weeks go by, there's back-and-forth with the artist, and finally we get the finished product. And it's just... Wrong. Wrong in basic ways that defy the very clear, several-times repeated and clarified and agreed-upon descriptions of the commission. And it's frustrating, because no matter how many times we try to get the artist to understand what we mean, they either can't comprehend it, or they just refuse to do it that way for reasons of their own. Anyone remember that article or tumblr post or something from years back, where it described someone's trouble getting a piece of art for either MtG or an RPG book? The one where the artist was supposed to make a druid in practical forest garb, and got Big Tiddy Elf GF over and over, no matter how many conversations they had with the artist? Same vibe, but less porny. At some point we both just gave up on getting it right and accepted a subpar result that didn't live up to expectations or, IMO, earn the commission, because my player didn't want to fight with this person and I didn't feel it was my place to do so. After it was over, I remember thinking, "Midjourney can't do Harrower justice either, the visual is too complex - but it would have gotten this result in much less than a month, and for free." Years ago, I backed an RPG at a level high enough to have a piece of character art included in the book, with a brief of: "Make a character who fits in this setting, here's some guidelines, follow those and we'll commission it." The process was almost identical, excepting only that I was less specific in my commission because I wanted to make sure the game's author got a book that looked the way they wanted it. I ended up communicating primarily with the artist, and there was no getting through to them. The piece that ended up in the book is nice, but I can't help feeling frustated by it whenever I come across it. Over the years, I've seen other friends commission art of their characters maybe nine or ten times. It's come out correctly perhaps three times. Every other result has been an expensive, time-consuming compromise that left them only vaguely satisfied. One of my other D&D players has had three separate pieces done of her long-running PC, and only one of them came out right, and it cost her something like $300 for that one piece. My sample size is small, but one in three is what we've got, and that's just not good enough. And that's where I'm at with AI art at this point. Human artists can outperform what my customized install of Stable Diffusion can do, no question. But if I'm going to get a result that is "good enough," I would rather get that "good enough" for free, in under an hour, and possibly in under 5 minutes. It's an unbeatable value proposition, and it's only going to get better. I'm very, very glad I'm getting familiar with this technology now, because it's going to be absolutely everywhere in five years, max.
|
|
#
?
Jul 14, 2023 07:28
|
|
|
Megazver posted:Okay, I guess I am the only one posting stuff here for now. Poor Poppy got her head chomped clean off right on the cusp of leveling to 2. RIP Poppy. I decided to bend what the local druid can do a little and have her Reincarnate Poppy (she's level 4, druids get Reincarnate at level 5) so now Poppy Prim is, uh, Gobby Grim? (We'll see if there's an actual name change.)  Also, I am finally actually playing Vampire for the first time after reading all the books years and years ago. Here's my hot Ventrue American Psycho GF:  Kestral posted:Over the years, I've seen other friends commission art of their characters maybe nine or ten times. It's come out correctly perhaps three times. Every other result has been an expensive, time-consuming compromise that left them only vaguely satisfied. One of my other D&D players has had three separate pieces done of her long-running PC, and only one of them came out right, and it cost her something like $300 for that one piece. My sample size is small, but one in three is what we've got, and that's just not good enough. Hard same. Megazver fucked around with this message at 10:52 on Jul 14, 2023 |
|
#
?
Jul 14, 2023 10:29
|
|
|
I've mentioned it before but I commission a lot of art, and I've probably spent thousands of dollars on concept art. And a fairly large chunk of that money spent ultimately didn't result in a design I was wholly satisfied with. Usually while the result was probably quite good in a vacuum didn't fit the idea I had in mind. Something usually the characters proportion and thus their age would feel off, proportions are my bug bear where it is the most request thing artists I've commissioned give me wildly inconsistent results with. Now, some caveats, sometimes this was to be expected, I often stumble across a amateur artist account on reddit or Twitter and they aren't actively doing communications but I wanna see what they do, so I contact them and ask if they wanna do a commission and sometimes they agree and then sometimes I get something pretty good and sometimes I don't. I expect the fact that this is a slot machine as I'm trying to see if I can find an uncut diamond in the rough. The other thing I think drives most of the problems with clients and why so many of them especially in the lower budget range to try ai is because the act of commissioning is also an art. It takes experience to know what sort of references, what sort of description, and how to layout them in such a way to get what you want. It takes a huge time investment of finding and downloading loads of references to figure out how to convey the kind of composite you want the artist to interpret. This whole process to do correctly and reliably sounds a lot like using the ai doesn't it? Finding thousands of references and then compositing them with someone else's spin and skill. It took me like three years and 50,000$ to learn how to best create reference sheets that are reliable, and even then I still have problems, one of my regular artists has literally forgotten how to draw their older style! All the references in the world can't help me there if their hands don't work the same way as it did a year ago! And this of course doesn't cover the usual pitfalls, of commissioning an artist and them disappearing (happened to me for both small and large 250$+ commissions), and artists who sometimes just refuse to do draw what you want despite clear references and refuse to revise it or give wips. And then there's the accessibility issues if your neuro divergent, some of us are shy and really hate having to ask an artist who while clearly is advertising commissions, aren't displaying their prices. Because if it's wildly out of my range, I will feel bad having to say, "ah sorry, my budget isn't sufficient at this time." I'm definitely not going to try to bargain with them either, that feels very rude. Then sometimes there's artists who have a whole Google form you need to fill out first before they give you an estimate which if its not in your range we'll that's a lot of time spent for nothing when they could've just had a clear price sheet with examples to act as a filter. Basically it's a communication and convenience issue where both clients/commissioners and artists have a lot in general they could do to improve the overall experience so more people are satisfied in order to reduce the appeal of AI. Anyways, I had a busy couple of days, today after work ill do one last pass at the op and then pm leper.
|
|
#
?
Jul 14, 2023 14:26
|
|
|
Megazver posted:
Looks rad, how did you make this? Also, how you're playing V5!
|
|
#
?
Jul 14, 2023 14:28
|
|
|
Fuzz posted:Looks rad, how did you make this? It's gonna be V20, not V5. Hopefully I've play some of that one as well, eventually. Initial prompt was "smiling punk latina vampire, clipart, pop art, flat shaded, clean and sharp inking, high contrast chiaroscuro portraits, black palette"    After a couple of Zooms, I photoshopped out the weird necklace and pencilled-in the fangs. A couple more attempts where I tried to get it to draw some fangs:    And a bonus Evil Vampire Toddler that I trolled my GM with: 
|
|
#
?
Jul 14, 2023 15:16
|
|
|
So the AI voice stuff has gotten really good over the last two years or so but its finally started to leak out from commercial companies like elevenlabs into the open source world. 99% of it is all horny porn stuff, as is par for the course, but some real good shitposts are making their way through the cracks. Please enjoy these two weirdly high effort Neco Arc covers. "Without Me" by Eminem, with a boatload of fun art. https://www.youtube.com/watch?v=3ZaWUzrqg8M And "Snake Eater" from MGS3. No art here but, well, just listen all the way through. https://www.youtube.com/watch?v=HamZ8rSsyhI Edit: posted this in the wrong AI thread but eh, will leave my shame intact. Objective Action fucked around with this message at 17:53 on Jul 14, 2023 |
|
#
?
Jul 14, 2023 17:46
|
|
|
Megazver posted:It's gonna be V20, not V5. Hopefully I've play some of that one as well, eventually. These are loving amazing. I'mma have to steal that style. And I've seen fellow players commission RPG character illustrations a grand total of ZERO times in 35 years of gaming, although that doesn't discount that some people do. E: Ugh, sorry for low content post. Here are some D&D character portraits I did with MJ for players who wanted them.  My new Bard  A friend's Inca Style Dark Elf Cleric  Another friend's BadAss barbarian. Doctor Zero fucked around with this message at 18:08 on Jul 14, 2023 |
|
#
?
Jul 14, 2023 17:57
|
|
|
Objective Action posted:So the AI voice stuff has gotten really good over the last two years or so but its finally started to leak out from commercial companies like elevenlabs into the open source world. All the AI threads kind of run together so that's understandable. The AI voice stuff is getting really good right now. The first one I heard that really impressed me was this video of Dagoth Ur from Morrowind saying a Charles Manson courtroom speech. It follows the inflections in Mansons voice but changes all the references to Morrowind stuff. The main difference I hear is just in the overall audio quality of the two voices, with Manson being recorded on older analog equipment. I wonder how good AI is at generating the effect of old recordings yet. https://www.youtube.com/watch?v=vIlXkNswM1Y This is the Death Grips song, it gets loud right after the quote of Manson at the beginning so don't listen to loud. https://www.youtube.com/watch?v=-qxa91EwFco AI voice could be really benificial to DM's though. Most people are bad at accents or only have a couple in their pocket. Being able to give NPC's unique voices that you could tag and return to would go a long way to make NPC's feel more real.
|
|
#
?
Jul 14, 2023 18:58
|
|
|
From a Legend of the Five Rings game I'm in...   
|
|
#
?
Jul 14, 2023 20:03
|
|
|
Megazver posted:It's gonna be V20, not V5. Hopefully I've play some of that one as well, eventually. Hot dang, thanks so much for sharing the prompt details. Some of those terms (especially "flat shaded, clean and sharp inking, high contrast chiaroscuro portraits") look like they might be relevant to a project I'm working on, making images that I ultimately draw out using a plotter. Which is something that works much better with vector art, to which this kind of style is suited. Anyway, this gives me some new thoughts, thanks!
|
|
#
?
Jul 14, 2023 20:22
|
|
|
Doctor Zero posted:These are loving amazing. I'mma have to steal that style. Well yeah, especially in an online game where there's a >75% chance the game will die within 5 sessions or weeks anyway, it would be a frustrating waste of money and just promote people dragging the same loving character into every game solely to get their money's worth on the art.
|
|
#
?
Jul 14, 2023 20:37
|
|
|
Raenir Salazar posted:It takes experience to know what sort of references, what sort of description, and how to layout them in such a way to get what you want. It takes a huge time investment of finding and downloading loads of references to figure out how to convey the kind of composite you want the artist to interpret. This whole process to do correctly and reliably sounds a lot like using the ai doesn't it? I think you've hit the nail on the head in more ways than one. For example, the three-page nonsense low-effort "backstory" mentioned earlier in the thread in a way is in many ways the_problem_with_ai.txt but the thing is that it probably also could be a perfectly fine little backstory with just a better prompt. Like if I ask GPT-3.5 the following: quote:Please write a brief backstory for a D&D 5e character named "Perdin" who is a human druid. Include a past (hermit), choose some sort of conflict or change, which brings Perdin to the city in search of fellow adventurers. Providing the AI with a basic (basic!) structure goes a very long way to getting better results. I won't quote the entire response in full (it's not super long tho) but here's the main bits: quote:Perdin's idyllic existence took a sudden turn when a dark force began to encroach upon his beloved forest [...] Perdin, horrified by the destruction, sought to uncover the source of this corruption and put an end to it. This is a perfectly serviceable backstory IMO. The Eyes Have It fucked around with this message at 20:46 on Jul 14, 2023 |
|
#
?
Jul 14, 2023 20:39
|
|
|
You guys have been very encouraging with your comments, thanks! I've resisted the urge to just start dumping all the random poo poo I generate, but I might drop in a bit more into the thread in the future. This is interesting! What's the prompt? Objective Action posted:So the AI voice stuff has gotten really good over the last two years or so but its finally started to leak out from commercial companies like elevenlabs into the open source world. Yeah, there is some fun AI voice stuff out there by now. I'm unfamiliar with this Neco thing, but these videos are fun! The best creator I've seen doing these is There I Ruined It, there's some good poo poo on that channel. gurragadon posted:AI voice could be really benificial to DM's though. Most people are bad at accents or only have a couple in their pocket. Being able to give NPC's unique voices that you could tag and return to would go a long way to make NPC's feel more real. Yeah, it could be useful, especially for voices you just can't do at all. I needed a classic Biblically Accurate Angel B̶̹̹̤̑E̴͇͌ ̷͚̝̈́͆̕N̸̝͙͊Ó̸̆͜͝T̴͔͋ ̶̮͉̐́ͅȂ̴͈̰͖͋F̷̯̉R̴̤̠̕A̵̹̓Ĭ̵̤̀D̵̪̑̕ voice line once and I tried to record myself and fiddle with the recording, but it didn't turn out great, sigh. I do have a bit of a cautionary tale about NPC voice lines, though. Syrinscape provides custom soundtracks and sound effects for some(?) official 5e adventures and those include actor recordings of all the specific lines by the NPCs in the book. So I joined a campaign of Descent into Avernus and the GM had ALL THE poo poo - the paid Roll20 module, the Syrinscape pack, really advanced script stuff on Roll20, etc, but what he didn't have was, well, GM skills. He was pretty inexperienced, as far as I could tell, and used all of this stuff as a crutch. We'd meet some NPC and he'd dutifully plug in the specific music track Syrinscape fed him and, after a bit of pause as he searched for the file, played the NPC's intro voice line and then we'd ask them something and he's go search for a voice line that he could play that would work. It got old so fast. Doctor Zero posted:And I've seen fellow players commission RPG character illustrations a grand total of ZERO times in 35 years of gaming, although that doesn't discount that some people do. I played in a year-long campaign of Storm King's Thunder and it was a very lovely experience and when we wrapped up, the GM posted the art of the group that he commissioned out of his own pocket for us. It was, I regret to say, pretty poo poo. Whatever he paid, it was not worth it. I appreciated the gesture, of course, so I had to, uh, pump some extra enthusiasm into my voice as I vocally appreciated it. Megazver fucked around with this message at 20:55 on Jul 14, 2023 |
|
#
?
Jul 14, 2023 20:51
|
|
|
Megazver posted:
Ofra Haza with dark blue skin, wearing a hajib, intricate, in the style of a Persian tapestry,detailed The character is an Ashalan Sorcerer from Legend of the Burning Sands. The Israeli singer Ofra Haza is also where I got the source for this Champions character...  Ofra Haza, white hair, wearing a military uniform, flaming wings,intricate,detailed, action pose, in the style of Jim Lee Neither look like her exactly, but the facial structure was what I was aiming for. Humbug Scoolbus fucked around with this message at 00:27 on Jul 15, 2023 |
|
#
?
Jul 15, 2023 00:22
|
|
|
Experimenting with first-stage worldbuilding just by trying to cultivate an aesthetic - the sort of thing I'd do by trawling pinterest first. All of these are "(short desc), Byzantine and Incan aesthetic, colorful matte fantasy (portrait/illustration) by Delacroix and Osman Hamdi Bey" - with occasional negative prompts to make women not too sexy and men not always Delacroix himself - with SDXL 0.9.            
|
|
#
?
Jul 15, 2023 12:17
|
|
|
Oligopsony posted:Experimenting with first-stage worldbuilding just by trying to cultivate an aesthetic - the sort of thing I'd do by trawling pinterest first. All of these are "(short desc), Byzantine and Incan aesthetic, colorful matte fantasy (portrait/illustration) by Delacroix and Osman Hamdi Bey" - with occasional negative prompts to make women not too sexy and men not always Delacroix himself - with SDXL 0.9. Yeah, these are really nice. I like messing around with aesthetics as well. Sometimes I get something interesting trying to generate something completely different and go "oh poo poo, I need to explore this further". For example, just trying to generate random D&D art but also prompting for 'Wes Anderson'. "symmetrical wes anderson portrait, a male greenskinned 30-year-old fantasy D&D orc as a D&D paladin wearing bright plate armor, "  I thought "huh, can I do more with this aesthetic?" and generated a fancy-pants warlock aasimar for a one-shot      This would be something like "symmetrical wes anderson portrait, a slick male colorfully skinned 30-year-old fantasy D&D aasimar, wearing colorful fancy nobleman outfit, wearing small round sunglasses, nimbus behind his head, pastel". (It didn't want to do the colorful skin lol.)
|
|
#
?
Jul 15, 2023 13:00
|
|
|
Another player in the group who doesn't have MJ asked me to make a portrait for her Daughter of Cacophony. Basic request - "blonde vampire Alyssa Milano with a guitar". Okay. "Blonde 25-year-old Alyssa Milano as a vampire, holding a guitar, clipart, pop art, flat shaded, clean and sharp inking, high contrast chiaroscuro portrait"  Tried to add 'snarling' 'ominous' and 'black and red palette'.  Cut out "clip art, pop art, ominous"         She was already pretty happy with these. Then I decided to gently caress around a little. "screaming blonde 25-year-old Alyssa Milano as vampire 1960s flower child hippie, holding a guitar, clipart, pop art, flat shaded, clean and sharp inking, high contrast chiaroscuro portrait"   Well actually, there should be some reflection of the 'cacophony' in the portrait, right? "singing blonde 25-year-old Alyssa Milano as vampire, holding a guitar, distorted, glitchpunk, vhs, art by john van fleet" (John Van Fleet is the guy who did the covers for the Revised clanbooks. I'm not sure MJ recognizes him.)  "singing blonde 25-year-old Alyssa Milano as vampire, holding a guitar, distorted, glitchpunk, glitch art, vhs, circular moire pattern lighting, Light leaks, Double exposure, noise and grain, color degradation"  "singing blonde 25-year-old Alyssa Milano as vampire, holding a guitar, distorted, glitchpunk, glitch art, vhs, circular moire pattern lighting, Light leaks, Double exposure, noise and grain, color degradation, background: glitchy broken misaligned lettering AAAAAA ' (no lettering in the background, bah.)   "three of screaming blonde 25-year-old Alyssa Milano as vampire holding a guitar, triple exposure, superimposed, overlapping, distorted, glitchpunk, glitch art, vhs, color degradation, "    "three of screaming blonde 25-year-old Alyssa Milano as vampire holding a guitar, triple exposure, prismatic art, superimposed, overlapping, extreme distortion, glitchpunk, glitch art, vhs, color degradation, art by mike mignola"  Also tried "snarling blonde 25-year-old Alyssa Milano as a vampire, ominous, evil, a spiral Moiré pattern shadow overlapping, greyscale, mostly black, high contrast lighting, stark contrast of light and shadow, photograph by anton corbijn, expressive light and shadow, optical illusion" but it's not that interesting.  
|
|
#
?
Jul 15, 2023 13:21
|
|
|
For the Stable Diffusion crowd, has anyone else run into a problem with After Detailer where it will try to turn people's hands into faces?  I had great luck with ADetailer on my first couple of serious "learn how to SD" projects, but now it's just going berserk and I can't figure out why.
|
|
#
?
Jul 16, 2023 19:35
|
|
|
Kestral posted:For the Stable Diffusion crowd, has anyone else run into a problem with After Detailer where it will try to turn people's hands into faces? No but can I see this please lol
|
|
#
?
Jul 16, 2023 19:50
|
|
|
Someone was testing on the poster for “the mob doctor.”
|
|
#
?
Jul 16, 2023 21:21
|
|
|
feverish and oversexed posted:No but can I see this please lol I deleted the original batch because they were unsalvageable, but here's another set showing the effect:   BE NOT AFRAID (of the facehands) What's really funny (and infuriating) is that you can see it happening in the step preview. It'll get through the initial generation, then After Details will take over and zoom in on each hand, and you get to watch it transform them into this:   It's actually much less nightmarish this time around for whatever reason: last night I was getting bulging fractal clusters of faces where the hands should be, instead of... Whatever the hell this is. Don't mind the very haunted faces, I haven't even started working on fixing that yet. This image is actually worth talking about to highlight some of the difficulties - and opportunities - in Stable Diffusion that I'm encountering as a new user. Hopefully it'll help others getting into this. My "Learn To Stable Diffusion" project is creating the PCs and major NPCs for a long-running campaign from some years back, since I have clear mental images and descriptions of them, and enough emotional attachment to them to want to get it right instead of just settling for "good enough." I did what I figured would be the easy ones first and learned a lot in the process, so now it's time to try a character who needs a specific pose, one that is iconic to the character and which no portrait of him would be complete without: glaring at the viewer while calling lightning from the sky like an Old Testament prophet. I'd used Controlnet's pose function a bit on the earlier ones, just to see what it was like, but Ihsan here was going to be my first attempt at getting exactly the pose I'm looking for. Now, for whatever reason, SD really did not want to cooperate for poor Ihsan. To pull off what I had in mind, I needed four elements: 1) A young man, ~18-19, with medium-tan skin, any expression between "grim and determined" and "outright hostile," and the build of an artisan: no crazy muscles, which a lot of SD models desperately want to put on men. 2) A savannah background, with lightning-riven stormclouds overhead, and bolts of lightning coming down from on high. 3) White robes (or white with a bit of gold) in a reasonably practical style for traveling long distances by foot. 4) A pose with one hand outstretched. Exactly how the hand is positioned doesn't matter much: a clenched fist, a palm upturned and either flat or with fingers curled, anything like that. And folks, let me tell you, aside from the savannah and the lightning, SD wanted none of it. Fought me every step of the way. It was especially cantanerkous about skin tone (it super, super wanted him to be either pasty caucasian or black), and on clothes that didn't show rows of chiseled abs and were shades of white. Fun fact, in several models I tried, telling it to make him wear "all white clothes" or "pure white clothes" turned Ihsan into a white girl I tried five different models, using the X/Y/Z Plot script to compare them side-by-side, in a matrix with different sampling methods. Often I was able to lock in one major element, but at the expense of another. The more reading I do on this, the more it confirms what I'd noticed in my own experimentations: a lot of these models are overwhelmingly trained on photos/paintings/drawings of white and asian women, and if you want something other than that, you may need to either find a model that specifically gets around it, or add LORAs. ControlNet consistently gave me a workable pose based on this:  ... but I found that the OpenPose preprocessor would often pick up the trousers, giving the character modern clothes instead of robes. After much fiddling, I managed to produce this:  ... which would have been nearly goddamn perfect, except I'd screwed up a ControlNet setting and accidentally turned the Resize Mode to "Crop and Resize" instead of Just Resize, and it cut off the top of his head!  Quite a few more rolls of the dice and tweaking the prompt later and I had this:  And that I can work with! The flaming horizon is actually what sold me on it, because this character's defining moment from the first arc of the campaign takes place while the plains are burning in the background behind him. That showing up was a total accident and I knew I wasn't going to get that effect again without a lot more work, and while the face and outstretched hand (and to a lesser extent the feet) are messed up, they're within workable range. I threw it over to IMG2IMG and started employing the various techniques I've seen people discuss for After Detailing hands and faces. People regard AD as nigh on miraculous, and you can watch it perform some incredible feats: https://www.youtube.com/watch?v=ZNcz4k5JCCo (Skip to 7 minutes) I'm fond of that one because the method he uses makes it impossible to stack the cards in his favor to make it look easy for youtube: he randomly generates his prompt and throws it straight into SD, then does the whole correction process with After Detailer in one take. The results are amazing, and in my limited experience with it I was seriously impressed. Unfortunately, now it's going horribly wrong and I have no idea why! None of the techniques I've used up 'til now have been able to salvage that outstretched hand, which either becomes a gnarled lump of flesh and knuckles, or turns into faces. It's very haunted and I'm nearly at my wit's end. Ah well, time for more research. Kestral fucked around with this message at 06:06 on Jul 17, 2023 |
|
#
?
Jul 17, 2023 06:04
|
|
|
Kestral posted:
this is about what I was imagining, perfect. Hopefully someone can help you with some hands-on experience, I'm not familiar with stable diffusion myself.
|
|
#
?
Jul 17, 2023 07:07
|
|
|
Kestral posted:Unfortunately, now it's going horribly wrong and I have no idea why! None of the techniques I've used up 'til now have been able to salvage that outstretched hand, which either becomes a gnarled lump of flesh and knuckles, or turns into faces. It's very haunted and I'm nearly at my wit's end. Ah well, time for more research. Thanks for the write-up, that'll be handy. It's late, I'll watch the video tomorrow and try to wrap my head around what could be going wrong, offhand I can't quite put my finger on it. Face it, it's hard enough to get ahead of these problems. I want to make sure I can help and not just give out second-hand information. Small tip though, lock in a seed when you find something you like. Much less randomization that way, helps to narrow down exactly what changes are being made by what setting.
|
|
#
?
Jul 17, 2023 08:51
|
|
|
feverish and oversexed posted:this is about what I was imagining, perfect. Hopefully someone can help you with some heads-on experience, I'm not familiar with stable diffusion myself. Fixed that for you. 
|
|
#
?
Jul 17, 2023 14:07
|
|
|
This is really cool. I haven't paid attention to SD for a few months and turns out there's an entirely new wondertool available.
|
|
#
?
Jul 17, 2023 14:45
|
|
|
KwegiboHB posted:Thanks for the write-up, that'll be handy. It's late, I'll watch the video tomorrow and try to wrap my head around what could be going wrong, offhand I can't quite put my finger on it. Face it, it's hard enough to get ahead of these problems. I want to make sure I can help and not just give out second-hand information. Any insight would be appreciated, but no worries if it ends up being a project! I don't want to assign anyone here homework  I'm not actually salty about the experience, honestly, it's just part of learning a new set of tools, and I felt like documenting it here so that other folks would know how that learning process can go - it can be frustrating to try to get into something complex or difficult and only see other people posting their Ws, IMO. Megazver posted:This is really cool. I haven't paid attention to SD for a few months and turns out there's an entirely new wondertool available. It's amazing, isn't it? Literally every time I do more research I stumble across at least one new thing that meaningfully adds to its capabilities. Like, Segment Anything! Holy poo poo, you can just... choose areas for inpainting, already neatly divided up for you like it was a paint-by-numbers. This was enormously helpful in one of my earlier projects, where I'd managed to get a fairly complex composition down (a man riding an unusually large Clydesdale-esque horse, turned toward and looking down at the camera), but it kept generating the character as a pinup calendar version of himself, shirtless and abs-out. I could have just inpainted it, but it had also generated this neat necklace that hung down to his chest, and my hands aren't steady enough to preserve that detail while inpainting. But Segment Anything immediately knew what was chest, necklace, horse, etc., blocked it out in colored regions, and let me select just the exposed skin of the chest to iterate on that until he finally put some drat clothes and armor on.
|
|
#
?
Jul 17, 2023 21:23
|
|
|
https://www.reddit.com/r/rpg/comments/1528d1z/i_use_ai_art_in_my_own_draws_more_easy_is_i_a/quote:AII use AI Art in my own draws. More easy is I a wanna prepared the scenarios, characters and artist style for my writes and my images. (self.rpg) solemn nod Eloquently put, Reddit user /r/AjLukan___, eloquently put.
|
|
#
?
Jul 17, 2023 23:22
|
|
|
Kestral posted:People regard AD as nigh on miraculous, and you can watch it perform some incredible feats: It's funny, the same thing is happening in the video you linked. I went and gave this a shot since I need to learn this extension anyways, and I must say, this is quite a powerful tool here! It looks like everything could be done manually, but having the mask selection and Inpainting settings done for you is a huge time saver. Here's what I came up with and some  about what I found out. young tanned male artisan, white robes, grim determined expression, hand outstretched, burning savannah background scenery flaming horizon, lightning bolts storm clouds, 8k intricate detailed Negative prompt: nude, bare chest, open shirt, fleshpile, zombie, captcha, blurry, monochrome, greyscale, lineart, draft, minimalist Steps: 32, Sampler: DPM++ 2M Karras, CFG scale: 11, Seed: 2, Size: 512x512, Model hash: 4199bcdd14, Model: revAnimated_v122, Denoising strength: 0.48, Version: v1.2.1, ADetailer model: mediapipe_face_full, ADetailer prompt: "(young:.6), (grim:.6), (determined:.6)", ADetailer negative prompt: "old, grizzled, happy", ADetailer confidence: 0.3, ADetailer mask min ratio: 0.001, ADetailer dilate/erode: 4, ADetailer mask blur: 8, ADetailer denoising strength: 0.32, ADetailer inpaint only masked: True, ADetailer inpaint padding: 32, ADetailer use separate steps: True, ADetailer steps: 96, ADetailer model 2nd: hand_yolov8n.pt, ADetailer prompt 2nd: open palm extended, ADetailer negative prompt 2nd: closed fist, ADetailer confidence 2nd: 0.3, ADetailer mask min ratio 2nd: 0.001, ADetailer dilate/erode 2nd: 4, ADetailer mask blur 2nd: 8, ADetailer denoising strength 2nd: 0.45, ADetailer inpaint only masked 2nd: True, ADetailer inpaint padding 2nd: 32, ADetailer use separate steps 2nd: True, ADetailer steps 2nd: 96, ADetailer version: 23.7.5, Hires upscale: 1.5, Hires steps: 32, Hires upscaler: 4x-UltraSharp - In the Detection settings, there's a dropdown menu with 'Detection model confidence threshold', that should be the setting to automatically determine which part get masked and then Inpainted. This can be adjusted per model, maybe that will help fix things? My guess is that 1st and 2nd model are getting crossed. - In Mask Preprocessing settings, there's an option for Merge, that too might be getting mixed up. Or why it's getting mixed up. Mask min ratio of around 0.01 should exclude the background according to the github page. - In Inpainting settings there's "Inpaint denoising strength" this determines how far from the original image you go when you regenerate it. It also is the fraction of steps used: i.e. .5 strength at 64 steps will actually take 32, .25 strength would take 16. Feel free to up the steps used at low denoising. A lot of experimentation is required to find sweet spots for 'same but different enough'. Inpaint mask blur will prevent the hard boundaries between masked area and what you regenerate, changing this can help blend everything seemlessly. - Oddly enough I couldn't get the hand detection to work, at all. This pictures hand is the result of the Hires fix. - Negative prompt is an incredibly useful tool. When you find a base image to start working with, add more terms to the negative prompt to get rid of the parts you don't want. Things like "bare chest" can help put that robe back on. Likewise for "woman" or "asian" if the model you're using is overtrained on those. Hopefully this helps some and thanks for bringing this extension up in the first place. I will definitely be using this in the future.
|
|
#
?
Jul 18, 2023 01:54
|
|
|
I hesitate to put this in the Solo RPG thread, but I was playing with Claude and it's much better at admitting when it doesn't know something, although not perfect. When it does make up something, it's pretty true to the source material. If you've been disappointed with Chat-GPT for gaming, maybe try it.Claude posted:Unfortunately I don't have specific information about a group called the "Crusaders" in the Villains and Vigilantes RPG. But I asked it about Ironsworn and it understood most of the basics, but it clearly didn't know the book text of the world Truths. HOWEVER what it did generate was loving rad and would fit perfectly in the setting. Claude posted:Here is a sample world truth I generated for The Old World setting in Ironsworn: Claude posted:Here is an example Firstborn Truth I generated for the Ironsworn RPG setting: Claude posted:Here is an example Beasts Truth I generated for the Ironsworn RPG setting: Claude posted:Here is an example Religion Truth I generated for the Ironsworn RPG setting: Claude posted:Here is an example of an Iron Truth I generated for the Ironsworn RPG setting: Claude posted:Here is an example Horrors Truth I generated for the Ironsworn RPG setting:
|
|
#
?
Jul 20, 2023 14:24
|
|
|
Speaking of language models, LLaMA 2 just released. It probably has already leaked but the signup was practically instant for me. Meta released a version trained for chat right from the start with it, but it's very censored. The "normal" version isn't censored though so I'm sure people are already finetuning it with uncensored datasets. Only problem is there's no 33B model yet, that got delayed. Apparently that model for some reason was very "unsafe." Regardless, we're looking at another big jump in open source language models most likely. The largest version is 70B, which is probably, I think, runable on 48gb vram or 2 3090s. And when the 33B or whatever it will end up being model releases, that'll target 1 3090, with the 13B and 7B models the ones where all the attention goes because they can run on anything. This one, unlike LLaMA 1, has a free for commercial purposes license too. Not that that will matter much for traditional games, but hell, it could. For image gen stuff I've been playing with the leaked SDXL 0.9, 1.0 is supposed to release in about a week. It's very powerful. SD 2.* And 1.* were pretty useless for battlemaps, it just gave you weird unusable things. SDXL though is as good as midjourney at it. I'll post some examples when I get to my computer. As far as characters go, I tested it with some prompts identical for stuff I've used for our DnD game. The output is about as good, but where it really improves is consistency. With this prompt in SD 1.5 I got a bunch of different things that weren't all too similar, with SDXL they were all consistent and matched the prompt very closely. All variations of the same character basically.
|
|
#
?
Jul 20, 2023 15:22
|
|
|
|
| # ? May 2, 2024 07:52 |
|
|
For the sake of comparison this is my attempt at creating a battlemap with SD 1.5. As you can see this is useless. I was using a dreambooth model specifically trained on battlemaps here And here are four examples of battlemaps I created with SDXL 0.9, they are just simple prompts that included things like "battlemap" and "top down perspective" with just slight variations between them     And here are some examples I tried to create of one of our NPCs. It's a completely normal AI generated portrait, but the thing is, they are all incredibly consistent. Here are two examples of this prompt in SD 1.5. I actually like the 2nd one the most out of even the SDXL ones, but still you can see the two pictures are very different from each other. Both pictures, certain parts of the prompt actually dominated the rest of the picture and it lost some details like her hair color and that shes elven.   And here's that same prompt in SDXL    None of these were made with much effort besides the SD 1.5 ones. The SDXL ones were really just me screwing around copying over the same prompts and quickly grabbing the first results so I could compare.
|
|
#
?
Jul 20, 2023 15:38
|
|