Here’s a quick demo for Tom, who asked about the automatic hatch-shading of Poser imports in MotionArtist…
This is my new “Captain Bromley” M3 head, and you can see that MotionArtist has added hatching in the right places, but it’s not very convincing. Here is a screenshot of the basic controls you have on import…
Basically, you make a 2048px screenshot from this rather clunky little Poser scene mini-viewer. There appears to be zero control over lighting. The screenshot gets inserted into the timeline and.ore the current comics panel, and then you exit the mini-viewer. Fine for making up a quick ‘motion storyboard’ to a deadline, as a guide for filming a scene with human actors, but not for a finished motion-comic. For that you would render in Poser, and import the render, not the raw scene.
Hmmm, I see what you mean. It’s not really hatching (or crosshatching). It’s more like a pencil rub effect, and sadly where there are lines, they don’t really conform to the shape of the mesh in any meaningful way.
I have tried using texture maps of hatch patterns on my Poser V4 figures, but the amount of work to get it to look right was tedious (and still not quite believable). Using algorithmically generated textures e.g. woodcut, didn’t work at all.
There is also the trick of assigning a hatch pattern to a Poser light and then beaming it onto your scene. It has some mixed results, but the lines do capture the contour of the environment. For example, here is an alleyway that has some snow and a broken pole. I used a single Poser light from above to project a hatch line pattern down onto the 3d scene.
https://imgur.com/cbGoXWW
That’s fairly good. There’s a free and easy way to remove the tiny speckles and only keep the lines. See the last comment here: https://sketchbooky.wordpress.com/2019/01/10/how-to-remove-stray-speckles-in-photoshop/
Great write-up on removing speckles, David! I’ve used the Dust and Scratches filter, but didn’t know that Accented Edges could also be useful.
I also sometimes remove any speckles by using a raster-to-vector program to convert the image and that can eliminate those annoying specks (at the expense of also losing some detail too).
Of course, I am hopeful that someday I can just teach a GAN how to draw art like John Byrne and then give it a crude hand-drawn sketch, tell it “Hey AI, turn this into something that looks like John Byrne drew it” and voila! 😀
Yes, the great thing about the free programmatic Paint.NET solution, which is found in the comment at the bottom of the linked post, is that there’s no loss of other details in the artwork.
Tom, I wonder if you might like to be interviewed in VisNews? It’s the sister title to Digital Art Live magazine, and is for comics makers. I’m the editor of both titles. We do interviews by email, sending a list of questions.