Winxvideo AI 4.1, now greatly improved and speeded up. Basically, it’s a poor man’s Topaz Video AI for the desktop PC, for a reasonable $50. If you just want speedy upscaling from Wan2.1 480p to 720p with detail enhancement, it’s worth a look, though it can do more. There’s a free trial so you can test it out. You need to let it go online and then the AI upscaling models are fairly speedy to download, they’re not like multi-Gb in size. They have two main models, and there’s no hassle with graphic-card detection — it just works.
Category Archives: Companion software
In Kontext…
Not a bad haul for today, with learning the AI called Flux Kontext. I learned how to…
* Speed up the slow Flux Kontext x 2 (turbo LoRA for 12 steps rather than 24, no noticeable difference in output).
* Combine two images into a new prompted composition. That was a feature I hadn’t yet investigated. Got a cat and dog running along a beach (a stress-test it just about managed, from two random stock photos), but the more mundane use would be two talking heads for comic-book yak-yak dialogue. Poser can do this anyway, but the widescreen Poser render you might ideally need for that might not be suitable for input/output in Kontext.
* Zoom the camera in and out in Kontext via a LoRA + prompt, while keeping the central character fixed and any background more or less similar. Again, you’re duplicating what Poser can do anyway in a render, but it’s good to know how to do it in Kontext.
* Generate convincing ‘rough pencils’ line-art from a Poser render, which can then be combined in Photoshop to firm up outlines on a Kontext watercolour render of a Poser figure. Registration is exact when the layers are blended in Photoshop.
Here the source is a Poser render of Nursoda’s ‘Ronk’ Poser figure and his snail, which I’ve shown here before. The above is the ‘rough pencils’ output at 1024px.
* Earlier in the week I also got a universal Controlnet (mistoLine_rank256.safetensors) working for Kontext.
And I now have saved ComfyUI workflows for the above.
Also, it looks like I moved up to Windows 11 just in time, which over the last few weeks has caused me to go get the best of SDXL, Illustrious and Flux style LoRAs and accessories. CivitAI is to be effectively banned here in the UK, from next week.
Microsoft’s New Ray Tracing AI – now in ComfyUI
Posted on
by
Posted in 3D Utilities, Companion software, Poser, Spotted in the News, The Animation Industry
Leave a comment
Posted in 3D Utilities, Companion software, Poser, Spotted in the News, The Animation Industry
Leave a comment
Life moves fast in AI-land. Last month I blogged here about Microsoft’s New Ray Tracing AI. This month — courtesy of Paul Hansen of Germany — Microsoft’s new tech is now free in ComfyUI. Along with an outstanding install guide and documentation. All free. Currently, 2 seconds of finished raytraced animation takes 22 seconds on a 4060 card. Import of .FBX is coming soon.
Release: Keyshot adds AI
KeyShot Studio 2025.2 is the latest version of the 3D object renderer Keyshot. Mostly used for super-photoreal 3D renders of commercial products in various use settings (e.g. kettle in kitchen), suitable for use in magazine ads and catalogues. The new version has added AI editing. The AI can replace the background with a new one, or give the whole picture a new colour style.
New for Poser and DAZ – June 2025
Time for another survey of what’s new for Poser and DAZ, plus items of interest in AI-land. I’m now running on Windows 11 Superlite, so I now have access to more advanced AI software. Indeed, to the very latest goodies such as Flux Kontext, so my OS change was nicely timed!
As usual, my picks of the new releases.
Science fiction:
Owl Bot, a futuristic robot owl for Poser. Likely an enemy of the Space Coop.
The Owl possibly assists with piloting the new ExoNaut ship.
Moonbase Alpha Uniform for G8.1M. Fan-art, so no commercial use.
The Cube for DAZ. A generic sci-fi mysterious space cube. Possibly similar to the Borg in Star Trek, at a guess, but I haven’t seen that series of Trek.
Retro Future for Genesis 2 Male, currently free at DAZ.
Morphing Cyber-Googles for G8.
Easy Environments: ExoPlanet IX.
Fantasy:
Fantasy Helmet Collection for G3F through G9
Gift Guardian stone sculpture.
Ruined Mage Towers 1 for DAZ.
Halloween:
Moreau’s Freaks for DAZ.
Storybook:
1971’s Quiet pier for Poser, also available separately for DAZ.
A handy Old Cobblestone Path as free .OBJ and Vue .VOB file. 8k textures plus displacement.
Toon:
Free Poses for Cat Noodle, the toon cat.
Figures and poses, props:
RA Rory M4 for Poser.
Camper Accessories. See also the Rigged fantasy backpack for G8 and G9.
In Good Hands – Hands poses G9F-G8F-G3F.
Animals:
Nature’s Wonders Butterflies of the World Volume 4, with eight endangered species.
Nature’s Wonders Snakes for Poser and DAZ, plus Nature’s Wonders Snakes of the World Vol. 1 (common snakes). Also Nature’s Wonders Slithering Expressions (paired sperpent / human poses).
If you have the above there’s also Nature’s Wonders Snakes Extras as a freebie pack.
Scenes and places:
Car Scrapyard for Daz Studio, an unsual scene. The car models look as though they were made with a hand-held scanner from real wrecks.
Quick Rocky Vignette 4, a good looking beach scene.
Historical:
Temple of the Nile for Poser, also available separately for DAZ. The free Feathers Conditioner looks like it would match well with this.
Frontier Grace Outfit and Props for Genesis 9. American pioneer outfits with bonnets.
British Rail MK1 TSO Coach by DryJack. His British railway starter freebies are in the SHARECG-backup torrent.
The Workbench for DAZ Studio. (No longer active at DAZ, but now at Rendersosity).
Second World War British Army headgear as the free WWII UK Mk II Airborne Helmet Pack for M4.
Scripts and software:
Remove All Modifiers – DAZ Studio DUF Cleaner.
Pose-Save-Utility for DAZ Studio.
Nomad Sculpt 2.3 for Windows. Formerly popular for digital sculpting on Android, now for Windows.
Stable Audio Open 1.0 WebUI Portable for Windows. A powerful free audio FX generator, distilled from the zillions of public-domain field-recording clips at Freesound. Like Stable Diffusion, but for sound effects, you tell it what you want (e.g. “a distant rolling thunderstorm is heard across a vast plain”) and it generates a .WAV file. Free, tested and working.
Poser matcap script. Blurs the textures so each becomes more of a uniform single colour aligned to the underlying colour. Handy if you want to de-grunge mucky textures, ready for filtering the render into a watercolour look.
Tutorials:
How to make low poly billboards to populate backgrounds in Poser.
Use Collapse to simplify your material templates in Poser.
A vital autosave feature in Poser.
POW! Biff! KAPOW! Comic-book FX upscale and extraction using Gigapixel, Vector Magic and Stable Diffusion 1.5.
Cruising Canals in 3D – Modelling Narrowboats and Water Scenes. A paid in-depth tutorial on 3D modelling for the English canals and narrowboats. Related is UltraScenery 2 – Marinas and Moorings.
Local AI, Poser and Python:
The new Flux Kontext Dev has been released, a new free local AI for image editing and filtering, rather than for image-generation. Perfecting 1:1 watercolour with Poser to Flux Kontext shows how to use it with a Poser render. With a Flux Kontext Dev workflow. Run in ComfyUI Windows Portable, after updating the portable to the very latest version.
My tests show Kontext Dev is no good as a local free auto-coloriser, more’s the pity. But it is excellent at watermark removals from public domain artwork (e.g. old postards on eBay, museum images which have no right being watermarked). It can also do image editing (“remove the rabbit ears, change the dress to green”), style makeovers, place a face in a completely new context, join two characters together in a scene, and probably more that users have yet to discover.
How to speed up CivitAI page loading when browsing and searching the site for free AI models and LoRA add-ons.
SD-Categorizer 2000, a free… “Python script to organize a folder containing all your images into folders and export any Stable Diffusion generation metadata.”
Set up Microsoft Visual Studio Code for Poser python coding. Speaking of which, I’m disappointed to learn from several tests that local sub-14B AI’s won’t cut it as Python script-coding assistants on a 3060 12Gb card. It seems one really needs one of the big beasts (30b and above), that runs in the cloud. However, note the Translate ComfyUI workflows into executable Python code free node for ComfyUI. This is local and could be used, I think, to have Poser call and run Comfy from a script. Watch this space.
Oh, and I reckon that Msty is the best ‘local AI library and model-runner’ for the desktop. I tried several. [Update: here I meant Msty 1.9.x, not the flaky new Studio version].
That’s it for now. More later in the summer.
Perfecting 1:1 watercolour in Flux Kontext – getting better watercolour with a LoRA
Posted on
by
Posted in Companion software, Natural media emulation, Poser, Poser to Stable Diffusion
1 Comment
Posted in Companion software, Natural media emulation, Poser, Poser to Stable Diffusion
1 Comment
Further to my Flux Kontext Dev experiments of the last few days, for filtering Poser renders… here I show how to get better watercolour by using a LoRA. Add the Aurelle v2 LoRA at 0.8. Specifically, this is designed to give an imperfect ‘human-made watercolour’ look for Flux. No garish day-glo colour (as with the default Flux Kontext), and no colour instructions or later Photoshop adjustments needed. And it’s much more subtle. It does want to swish down the hat-brim, and the 1:1 registration is lost there. But that’s easy to fix and otherwise it’s great. We also get rid of the dark 3D shadow under the hat.
The 1024px is slightly fuzzy, fuzzier than the default watercolour output. Possibly that’s because I’m using a regular Flux LoRA, not a Flux Kontext LoRA. But even so, a Firefly line-art render could be layered and blended in Photoshop, to bring back a little harder definition of the shapes and on the edges. That’s not been done here, on this quick demo.
I was however using the real-time character render from Poser that was .PNG and masked with transparency, which seems better than one with a white background. Another test showed 2048px gives clearer and larger results with more detail (e.g. fingernails), but takes far longer and is not so watercoloury. Try working in 1024px to test ideas, with 2x upscale. Then move to pure 2048px for a second pass, which is later blended back into the first in Photoshop?
The ComfyUI workflow shown is the official GGUF demo, adapted for a LoRA and with elements moved around. Note that FluxGuidance (CFG in Stable Diffusion -speak) is at 1.0 rather than 2.5, and apparently this favours a traditional artwork style.
MediBang Paint no longer free
MediBang Paint, a free rival to Clip Studio, is no longer free in its latest version. It’s now $50 on the Windows Store. There is however still a download link to the “old version” for now, which is v29.1. I seem to recall that the free version was ad-supported, though.
Perfecting 1:1 watercolour in Flux Kontext
Could Flux Kontext Dev handle a backdrop as well as a character, thus bypassing the need to composite later? To find out I threw together a basic garden around Nursoda’s Ronk figure and his snail. Obviously, one would spend a lot more time constructing a garden that was destined to appear in many scenes in a storybook or comic. But this is just for a workflow demo.
Pretty ugly from Poser (Comic Book mode lineart and a bright light preset helps it along, but like all 3D it’s desperate to go ‘dark and grungy’). Yet Kontext handles it nicely. Note the new word at the start of the prompt, ‘Filter …’
The problem is then the garish day-glo nature of the colouring on the new image. But because we have 1:1 registration with the Poser source-image, we can easily lay the colours back in by using it as a colour blending layer in Photoshop. Here that’s been done. Then just a little of the Kontext colour has been brought back in. The layer was then flattened and auto-contrast applied, then desaturated slightly to take account of the colour-boost caused by the auto-contrast. The final result…
And since it’s come from Poser, we can have easy-select masks galore via a clown pass / toonID render, should any further postwork be needed. And if a holding-line around the character, or a blurring or fading of the background, is needed… then Poser can also supply the masks needed.
1:1 watercolour in Flux
A quick Poser experiment with the new Flux Kontext Dev. Nursoda’s Ronk and his snail, in Poser. Render to real-time Preview at 2048px, with high texture quality and a little Comic-Book applied. Lay this Poser render on white in Photoshop, reduce to 1024px and use this as the seed image.
The prompt gives a pencil and watercolour effect, but does not cause the layer-registration to shift. It remains an exact 1:1 match, despite the style change. In other words, Kontext can act exactly like a Photoshop filter would. Takes about 70 seconds on a 3060 12Gb graphics-card, at 1024px. This speed is comparable with intensive Photoshop filter plugins such as Reactor or G’Mic. There is a ‘turbo’ version from a third party, said to give a 2x speed up, but it appears to require intense Python wrangling and lots of tracking down dependencies to get it to work.
A 1:1 match means we can restore the Poser colour, by using the original render as a colour-blending layer in Photoshop. Which means we can have consistent colour from panel to panel and page to page, when storytelling in a comic or storybook.
We get a little drop-out of definition. For instance, the spiral of the snail’s shell is lost. If we had a lineart only Firefly render from Poser, we could bring it back by layering in Photoshop.
Update: It appears that if you go back to it then next day, and experiment with style descriptions, then try to go back to the original prompt, the earlier styled generations somewhow adversely affect the later output (more hard and cartoony than it should be). Possibly old latents are being partly re-used? Anyway… start from a fresh launch of Comfy, then go to the workflow and don’t tinker or change anything before starting your output.
Update: It seems a Poser .PNG render with transparency is the best to drop in as the seed image. Rather than needing to first place it onto a white background. Also, “filter” rather than “convert” seems a better choice of words for the prompt.
Exact 1:1 registration in Flux Kontext Dev
Kontext is wayward in terms of wanting to resize things. But it’s a matter of getting the prompt right, for exact 1:1 registration.
Add a layer of simple black and white lineart, while showing the photo beneath and keeping identical subject placement, camera angle, framing and perspective.
Layer the result in Photoshop, and blend via Multiply. Brush a soft eraser over teeth etc. Then filter the base photo to lighten it up. I imagine it would blend nicely with a Poser real-time comic-book lineart render, for added line variation. You’re welcome.
Source image from the official test workflow. Using the official default GGUF workflow, but made compact by moving things around and with upscaling nodes removed.
Flux Kontext
Posted on
by
Posted in Comics, Companion software, Poser to Stable Diffusion, Python scripts, Spotted in the News
Leave a comment
Posted in Comics, Companion software, Poser to Stable Diffusion, Python scripts, Spotted in the News
Leave a comment
I’m downloading the free Flux Context and its various dependencies now, and it looks like it will void all previous local Stable Diffusion attempts at ‘style makeover while keeping the content fixed’. In other words, ‘work like a Photoshop filter’.
But, a super-powered one that also knows how to make precise image edits (e.g ‘work like a Photoshop filter, but add a hat and change nothing else’). It’ll thus be very interesting to see what this new form of local AI can do with some basic real-time Poser renders.
Oh, and I also have a ComfyUI node which outputs a workflow as a Python script, which would seem to offer potential to have Poser grab a just-made render and feed it straight into Flux Kontext. The drawback on that is the render-time (1min+ for Kontext, per 1024px image), which is not so turbo-charged. So, near-instant basic Stable Diffusion style-change methods will likely still have their place.
Watch this space.
Windows 11 on a HP Z600
Given my new aim of a stripped-down no-bloat no-nonsense Windows 11 superlite install, the next question was… can the latest version of the OS be installed on a HP Z600? The old workstation has the SSE 4.2 on the CPU, required by the latest version. And the superlite ISO can bypass the hardware checks by default. But a Z600 lacks UEFI in the BIOS, which is a roadblock to getting the bootable USB recognised. Turns out the OS can be installed, and it’s a myth that Windows 11 requires EUFI and GPT. You just need to configure the bootable USB maker (the freeware Rufus) to use legacy BIOS and MBR. Then use an installer ISO that turns off the hardware requirements.
Here’s proof that it can install and run.
A London executive is selling their dual-Xeon Z600 this week on eBay… “My IT team installed Windows 11 by CMD command as an Administrator, skipping the software compatibility tests.” Which means they bypassed the hardware requirements. In her seller images she shows proof of having the latest 24H2 version booted and working…
And from back in 2022, a forum post stating a successful install of an earlier version of 11 on a Z600…
So it looks promising. There’s nothing to be lost by taking out the existing primary SSD, slotting in a new blank one and seeing if the Z600 can install/boot into Windows 11. Of course, a nice brand-new £600 PC would be nice, but I can’t afford that right now.
So… maybe not Linux
After looking closely at Linux Mint for two weeks, I’m coming to the conclusion that Linux has just one too many drawbacks when it comes to drivers, Windows software etc. Not to mention all the learning of its arcane ways which would be needed, and how apparently easy it is to break the OS entirely just by trying out the wrong bit of software.
So…. I’ve recently discovered how far the ‘debloat’ of Windows 11 has come in the last few years. I’m impressed. What used to be a little script that turned off some settings, has become a range of robustly de-bloated ISO installers. Installers that rip out all the crap, not just politely flick at a few switches.
These ISOs go by names such as Nano11, Tiny 11 and Ghost Spectre Superlight. The latter appears to junk the most junk, along with the need for a Microsoft account, telemetry, privacy invasion, ads, hardware snootiness and more. Not just ‘turning it off’, but (judging by what I read) actually ripping it out. Also taking down all the hardware barriers, so it can basically be installed on any x64 PC (even old ones, though note the CPU does need to support SSE 4.2). And yet still run very well. Automatic forced Windows updates can even be turned off until 2077!
Hmmm… all sounds very nice, and videos and benchmarks show it leads to a very lean fast OS (these ISOs are aimed mostly at gamers who want that ‘Win 7 feel’ to their PC). But… such ISOs don’t sound all that trustworthy.
Perhaps better to run the freeware debloater utility NTLite on the official Microsoft ISO, which creates your own custom debloat installer ISO. Doesn’t look too hard.
So that looks like the way to go. The freeware Rufus is thus the initial starting point, just as it would be for Linux Mint, being a free utility for putting the installer ISO onto a properly bootable USB. Then you’d disconnect the existing SSD, slot in a fresh new SSD, boot and install the OS.
And, handily released this week… BetaNews reports “Windows 7 Reloaded solves Windows 11’s biggest problem”…
“Windows 7 Reloaded Edition is a custom theme built for Windows 11 that transforms the look and feel of the OS, restoring the classic left-aligned layout and other design elements. It’s completely free.”
Requires Windhawk, a handler for Windows tweak-mods. Not perfect by the look of it on the video, but Reloaded appears to fix several likely Windows 11 annoyances. The Start menu needs instead to be the Windows 7-like StartAllBack (included with Ghost Spectre Superlight SE version). Then OldNewExplorer would be needed to modify Windows 11 File Explorer. But all the above is now looking like the most viable route — a fast ripped-down Windows 11 with a light Windows 7 makeover.
KAPOW!
A practical tutorial in comic-book FX upscaling and extraction using AI, using local Windows tools…
Tyler GND for Poser – works with an eyetarget
Just confirming that Tyler GND for Poser appears to work fine with an eye-target script (free here in ‘More Poser 12 helpers’). This one-click Python script gives the figure what’s otherwise lacking, an easy way to move both eyes together and have them sync. Just grab and move the non-rendering cube, and the eyes move to follow it.
Tyler GND was a highly advanced male figure, with masses of morphs, built on top of Ryan 2 which was the flagship figure which shipped (only) with Poser 9 / Poser Pro 2012. Reminds me a little of Roger Moore (who played James Bond, at one time).
As you can see he looks very good with his default head, in real-time Preview in Poser. He never caught on (awkwardly caught between M4 and Genesis 1) and thus had hardly any clothing made for him, but a Crossdresser 4 converter module is still available for $3.99 and will convert most M4 etc clothing.