10 Wednesday Jul 2024
Posted in AI, Lovecraftian arts
HPL would like to read you his latest tale…
Incidentally, August 2026 will be Cthulhu’s 100th birthday.
Generated with Stable Diffusion 1.5, plus a little Photoshop. Pure SD prompt, no LORAs or image-to-image.
08 Monday Jul 2024
Posted in AI, Odd scratchings
An interesting finding, with thanks to the experts at Reddit. Wooden pseudo-mediaeval dialogue, of the type sometimes found in 1870s-1920s historical and fantasy literature, can easily be translated into modern English via large-language model AIs.
By pseudo-mediaeval I mean all those interminable over-uses of thee and thou, plus wherefore goest ye, what meaneth it, we twain, thou wert before me, unto yonder hillock, therewith he laughed, and so on.
William Morris, and many other users of clotted pseudo-mediaeval Victoriana, made readable… with AI assistance. Potentially, it would be possible to build a dedicated translator for entire novels. Which would also aid the production of human-read audiobooks, as these books would no longer be so daunting to either readers or listeners.
27 Thursday Jun 2024
Posted in AI
Newly-spotted AI image generation plugins, of possible Lovecraftian interest. For use with Stable Diffusion 1.5 models running locally in InvokeAI, ComfyUI or similar.
Ivan Aivazovsky style, for atmospheric sea-scape images (just add monsters). Could be used with the new Cinematic Tools – Volumetric Lighting for backdrops needing more eerie depth-haze/fog.
A few creature plugins of note. Auraplanktons for slides of lit deep-sea specimens.
Swamp Mysteries generates ‘somewhat toony but still mysterious’ swamp monsters in swamps.
And today there’s Black Cat. Seemingly meant to be an eldritch mystery moggie so it looks a bit rough, more Lovecraft’s battered ‘Old Man’ cat than kute kitten. Still, all the more suitable for a desperate messenger from Ulthar, etc.
For steampunk and 1920s RPG makers, Style of patent drawing may interest.
19 Sunday May 2024
Posted in AI, Lovecraftian arts
Now crowd-funded and in development, Shakespeare vs. Cthulhu, a solo RPG based based on the well-reviewed 2016 book of stories. This new adaptation will offer an…
“interactive, branching narrative, solo-RPG gamebook”
And the author has created such choose-your-own-adventure / RPG books before, including for Dracula, Beowulf, and Alice. The new Shakespeare vs. Cthulhu sounds fun, and will also presumably work well.
Of course this sort of thing is going to be susceptible to AI in the near future — in terms of real-time generated AI voices / character-depth/back-story, on-the-fly music and ambience FX, and AI control of branching storylines. We’re likely to see these possibilities spawning new low-cost interactive AI audio-adventures, at least for an audience which can bring their own visual imagination to the party and doesn’t need to be spoon-fed with expensive visuals.
Until then, Shakespeare vs. Cthulhu is old-school paper and dice. Due in April 2025.
17 Friday May 2024
Posted in AI, Lovecraftian arts, Picture postals
This week on my regular ‘Picture Postals’, a change from the usual (though, in a way, following on from the recent ‘comic-strip’ themed post). I’ve used Photoshop to isolate Lovecraft’s own sketch drawing of The Great Race (“The Shadow out of Time”) from its surrounding notes and also from the blue lined paper on which it was drawn.
Original
Extracted and cleaned, as a 2196-pixel .PNG file. Still a bit mucky down at the base, but I reckon a mollusc is ‘gonna get gloopy’ down there.
Feel free to try to use this as a Controlnet guide input for an AI image generation. Though, good luck in writing the prompt description.
09 Thursday May 2024
More picks of recent Lovecraft related and (now also) R.E. Howard related LORAs, these being free plugin for models based off the free Stable Diffusion 1.5 AI image generator.
* Stygia, explicitly modelled on Conan’s world and darker than usual ‘mediaeval desert, with pyramids and crypts’ settings.
* Conan trained on old comics plus oil paintings and film-stills. Doesn’t look great, the faces being too anime. But might be worth a try with a different model more oriented to western comics? Beware of commercial use, for a lookalike barbarian named ‘Conan’. Since there are still active Conan trademark-trolls in the USA. Apparently they are underlings of the megacorp Tencent, and can claim a trademark in the name until 2028. So you might call him some adjacent name instead, like Xolan or Kohlan.
* Style of Andreas Achenbach, possibly of interest for sword & sorcery, fantasy-historical.
* UFO Alchemy, which looks like it could be de-UFO’d and made into more of a Lovecraftian ‘cosmic map’.
* Style of Norman Ackroyd, moody British 1970s aquatint, possibly useful for Innsmouth-type scenes to which you’d add text to make a ‘widescreen’ storybook. Apparently Norman Ackroyd was one of the core artists in the SD 1.5 initial training. His ‘look’ was…
* 1900 style photographs, likely to be useful for RPG pictures and the like.
* There was also a new LORA that attempted to emulate Weird Tales covers, but the samples looked so bad that I’m not linking it.
17 Wednesday Apr 2024
Posted in AI
A few more recent Lovecraft-adjacent new free LORA add-ons, for use with a local install of Stable Diffusion 1.5.
Meteor Shower v1.0. Artistic meteor showers. It can also do just one or two. The same maker also has the earlier similar Fireflies from last Halloween.
Meteor Shower
Fireflies
Dark Core Style. It “creates comic-book-like inked illustrations, dark, gloomy and simplistic with lots of sharp angles”.
And Death and Decay, looking very suitable for generating the basis of a heavy metal album cover. The same maker also has a new Celtic / Gaul Warrior today.
15 Monday Apr 2024
Posted in AI
Testing out an AI that claims to do reasonably good text generation. But… kinda. Heavy Photoshopping was needed.
09 Tuesday Apr 2024
Posted in AI, Kittee Tuesday, Lovecraftian arts
New on YouTube, “In Lovecraft’s Dreamlands” by The Terrible Old Man and his Band of Cats.
Song lyrics…
In the shadows of ancient time
Where dreams and nightmares intertwine
The cosmic dance in the heart of night
The stars bear witness to the unfathom’ble night
The celestial lands, in eerie blend
Where horror and wonder never seem to end
Nyarlathoptep lurking in the abyss
A haunting presence you cannot dismiss.
Moonlight over Ry’lh, where darkness reigns supreme
In Lovecraft’s dreamlands we are free to dream
Embrace the eld and let your spirit roam
In the depths of madness forevermore…
In the depths of madness forevermore…
Moonlight over Ry’lh, where darkness reigns supreme
In Lovecraft’s dreamlands we are free to dream
Embrace the eld and let your spirit roam,
In the depths of madness forevermore…
Others are way ahead of me. Check out Plan 9 From Outer Space – the AI Album.
04 Thursday Apr 2024
Posted in AI, Lovecraftian arts
More new(ish) LORAs which may appeal to Tentaclii readers. LORAs being free plugins that ‘guide’ the AI images made with your local install of Stable Diffusion 1.5.
* New this week is the somewhat Solomon Kane-ish Witch Hunters, which by the look of it was also trained with some over-the-top corporate card-art. Still, you can stack LORAs, thus you could blend it with another such as James Purefoy (Solomon Kane) which was trained on movie stills.
* One I missed, from a year ago, CthulhuTECH. Every image, no matter what the subject, becomes somewhat Cthuloid. This one is a LYCORIS, but is used in the same way as a LORA. CthulhuTECH is very well documented, so make sure to click “Show more” and then save a .MHTML page alongside your download.
Also has a cthulhuTECH books add-on as a LORA.
* Looking at the CthulhuTECH maker’s other stuff, I see I also missed his Pulp Ladies – In Spaaaaace. Again, it’s a LYCORIS but could still be combined with a LORA to give the image output a more ‘aged pulp cover’ feel.
26 Tuesday Mar 2024
Posted in AI
You wait ages for new Lovecraft-ish LORAs to come along, then two come along on the same day. As usual these are for local generation of AI images with SD 1.5, and are free.
* Classic Film Noir (Concept) – v1.0. “This LORA is designed to reproduce the feeling of classic noir aesthetics. The clothing, the people, the lighting, the camera work and the time period.” And it’s not censored for the ubiquitous smoking of the era, either.
* Marine Biology v1.0 LORA.
* Also a most useful helper today, Vertical View Angle Slider LECO – v1.0. A LECO just goes in the LORA folder and acts the same. Use with From_Above or From_Below or (from below) placed early in the prompt, and then tweak the slider a bit for precise and reliable camera-angle control.
19 Tuesday Mar 2024
Posted in AI, Lovecraftian arts
New on YouTube, “H.P. Lovecraft Poetry, Converted into Goth Alternative Rock by AI”.
