Open Brush 1.0

Google’s abandoned VR painting tool Tilt Brush has been picked up and forked, to become the free Open Brush. Open Brush 1.0 has just been released as free open source software. You need a VR headset to sketch and then export sketches as .FBX files. You can also save out a video of the sketch. There are various forms of snapping, curving, mirroring of lines, and so on.

Open Brush is an advance on Tilt Brush, as it lets your strokes be both coloured paint strokes and 3D geometry at the same time. It also has stroke-smoothing. Open Brush 1.1 will soon add a layer system, and let you cut and delete bits of your strokes.

Also, spotted in their ‘experimental’ folder is a version to “use Open Brush without VR, using keyboard and mouse”.

Runs on Windows 7 and up, according to Steam.

AI pictures and copyright

The US Copyright Office has effectively ruled that a named A.I. algorithm cannot ‘own’ copyright on an artwork created by it. A three-person board at the Office considered the claim ‘by’ an artificial intelligence (AI) system to copyright its art.

The Office’s ruling noted that… “the courts have been consistent in finding that non-human expression is ineligible for copyright protection”. Some AI researchers like to aggrandise their work by claiming that AI is “slightly conscious”, but the courts and now the Copyright Office beg to differ.

However the question remains, what happens when the system is not fully non-human? For instance when humans use simple inputs or set constraints for the creation process? As with text-based AIs, where you type “Purple fish flying over an orange ocean at night”, and the AI ‘interprets’ that into a picture? If nothing remains in the output to show that the original input parameters ever existed, is the output then to be classed as a fully non-human picture?

Or might the complexity of the human input somehow modify the status? For instance, if the input is deemed “trivial” might that void the copyright more than if the human inputs need hours of work? But again the ‘traces of the original input’ argument would come into play. It may be that eventually we use the test:

Q: Can this AI art be reverse-engineered or critically examined, so as to reveal all or something of the original human inputs?

A: In ‘no’, then the art was ‘made’ by the AI, and cannot be copyrighted.

There’s a wider questions about the Metaverse, if that isn’t just vapourware intended to keep up the share-price of the failing Facebook. Is it like a giant videogame, where the developer and publisher own everything produced in it, including all screenshots? Or is it like Photoshop or Microsoft Office, where the end-user owns whatever they make with it (in it)?

Amazon Comixology switch-over is a huge disaster

Oh great, I’m hearing all over that Amazon has very badly bjorked the transition from the defunct killed-off Comixology brand. They’ve just moved users over to Amazon Kindle / the main Amazon store.

IGN has the gory details in “Comixology’s New Update Takes All the Fun Out of Reading Digital Comics”…

“This supposed “upgrade” is a huge step backwards for Comixology and the digital comics market as a whole. … the new incarnation is a downgrade in pretty much every respect … It quickly becomes clear that the Kindle reader was designed for prose books first and comics not at all. … All the basic functions of the original Comixology reader, like zooming and Guided View, are gone now. The reader is completely incapable of handling double-page spreads … The [new] Kindle-inspired interface has become more convoluted, not to mention missing some of the organization features.”

And…

“The Comixology website itself is gone now” replaced by Amazon.com. Amazon’s usual ‘confusion marketing’ reigns, and the site is “not particularly helpful in terms of directing users to specific areas of the site.”

Meaning that (as with books) it’s difficult to find what you want and especially what’s new or forthcoming.

IGN further report of the Amazon.com experience that…

“it gives absolutely zero indication of how to access and read the books in your [purchased] library.” and “it’s apparently no longer possible for users outside of the U.S. to subscribe to titles”

It’s said that the new Android reader app is at least not quite so bjorked, yet. Though it’s not great, going by what I’ve seen via YouTube.

Sigh. Amazon we were counting on you to make digital comics a great experience for the masses and to spread readership into the mainstream, at an utterly crucial moment in history. You blew it.

---

Need a second-opinion? Here’s Verve which notes…

“Individual release dates aren’t even listed” … “Any comic that comes with a DRM-free copies won’t” be allowed to offers a subscription option. “Looking for single issues on sale? They’re no longer listed with the collections” … “browsing through titles is now unbelievably complex” not to mention the “the lack of creator credits” on the site.

“Organizing a digital library through Amazon’s portal is a nightmare”. Worse… “Amazon has removed the ability to archive” items. “If you’re someone who’s hidden some of their books to clear space in your library, get ready for all of those books to appear again.”

On the app on an iPad Verve says…

“The app is so regressive that the design on the search tab makes me feel like I’ve been transported back to 2010.”

Rife-app

Rife-app, an AI-frame interpolator. The problem it solves:

1) you have an old video-loop of 12 frames per second, perhaps a jerky videogame sprite;

2) you want this at 30 frames per second, and with smoother animation;

3) you can’t just stretch the video loop, and you can’t draw in the missing frames that would smooth the jerky animation;

4) so… you run it through Rife-app, which ‘guesses’ what the missing frames in-between should be, and adds them.

Jerky wind-blown hair on the left, nicely AI-smoothed on the right. Sadly WordPress.com blocks all animated .PNGs, but you get the idea. See the original at the Rife-app link above.

So it’s basically an AI ‘inbetweener’, to use a term from the old hand-drawn animation industry.

AI generated artworks take over ArtStation Marketplace

The ArtStation Marketplace is rapidly being taken over by AI-generated individual artworks…

… it started with just a few, a week ago, and now they’re taking over. Also, many are obviously coming from the service Dream by Wombo, which does not allow commercial use (although, admittedly, the copyright status of AI artworks is doubtful).

It would be useful to have a per-user UserScript or browser add-on for ArtStation, to seamlessly block a user from appearing in results. As you have for DeviantArt, to banish the foot-fetish guys and others.

Or perhaps ArtStation needs a new category, ‘Botworks’, which can be opted out of?

How to remove the “Do you want to open this file?” warning from PNG files

I recently tested some well-known graphics freeware. When uninstalled, it reset all Windows image-viewing defaults. It was kind enough to warn about this, just before uninstall. But there was no option to say ‘no’. Thankfully the problem that this uncouth behaviour presented was very easily rectified, with the usual…

Right-click | Open With… | ‘Choose default…’

…and then by assigning the usual viewer software.

However, the .PNG files proved a more intractable problem. Every time, a Window blocking notice appeared: “Do you want to open this file?”. Un-ticking the “Always ask…” checkbox on this warning message made no difference whatsoever to my future attempts to view an Internet-downloaded .PNG file with IrfanView. Always there was the “Do you want to open this file?” message, again and again. This also affected viewing from the Downloads list in my Web browser.

How to solve this problem on Windows 7:

1. In the Windows Start Menu search-box, type or paste gpedit.msc and this will select and enable you to launch the Windows Group Policy Editor.

2. In the Windows Group Policy Editor navigate to:

User Configuration | Administrative Templates | Windows Components |

… then down to Attachment Manager | Inclusion list for low risk file types

Make sure you have opened the low risk, not the moderate risk panel. They’re easily confused.

3. ‘Enable’ this Inclusion list, then carefully type in the following and press Apply.

.png;.bmp;.gif;.webp;

(I’m here adding a couple more graphics file-types for good measure).

It’s done. You have told Windows that these file types are low-risk, even if Windows has them flagged as being downloaded from the scary Internet. You should now have no trouble double-clicking on a .PNG file that was downloaded from the Internet (DeviantArt etc). You should now be able to instantly view it in Irfanview, or whatever other viewer you use, without any warning notices.

Obviously you should take care if you launch graphics in a Web browser instead of a local image-viewer. It would probably be unwise to automatically open a wild .GIF file with Internet Explorer, for instance.

Sidebar links checked, updated, added to

The sidebar links on this blog have been checked by hand, and updated or deleted as needed. It’s been a few years, and the list was in need of a check-over.

Also some choice new additions to the sidebar links…

Blog: David Revoy (France, Krita and Blender, digital comics and Creative Commons)

Blog: John Swogger (UK, archaeology comics, digital/traditional)

Blog: Tentaclii (Lovecraftian arts)

Blog: The Art Squirrel (UK, Christine Garner, digital for traditional illustration)

Blog: Wild Yorkshire (UK, Richard Bell, digital for traditional comics)

Magazine: VisNews (UK, monthly for digital comics makers)

Org: Imaginative Book Illustration Society (UK)

Org: Professional Cartoonists UK

Software: Poser 11 Pro ($80 one-time purchase, real-time lineart from 3D, and this is the one you want because 12 is only in Early Access and breaks all the old Python scripts)

Software: Style Animator (vectorize and autosketch, $45 one-time purchase, easy to use and nice results)

---

Also new on this blog, and which some may not have noticed yet, my ‘test and list’ software page which is titled “Will it run on Windows 7?”