As a traditional and digital painter, there is one unique feature that is easy to overlook: the colors are selectable as true pigments.
Painters don't categorize colors using standard terms: red, blue, green etc. Rather they categorize according to pigment. Different pigments (i.e. chemical base to the paint) have different properties.
For example, a Prussian blue appears almost black when applied thickly, but is very chromatic when applied thinly. In contrast, a cobalt blue is pretty much the same however it is applied.
ASAIK, this is the only app that supports this feature out the box.
The digital painting toolset has been pretty much in stasis for years, but this app offers node based brushes! I am very intrigued. Downloading it now.
I have had time to browse through the app. Regarding the nodes, it would be good if the node parameters had too tips, as it is they are labeled very mysteriously. The keyboard-powered brush picker is very useful. I wish Photoshop had one. The lighting feature I don't see as very useful. Its basically no more than a temperature adjust. The color picker had a few cool features, but I admit I did not explore it too deeply.
Overall I enjoyed it but nonetheless see it as being in the same family as every other painting app.
I would love to see painting apps stop trying to emulate real media and instead try to do things that are uniquely digital. My dream digi-daub app would feature...
- 16 bit as standard. Do a gradient in an 8 bit Photoshop document and you will see how limited 8 bit info is. (OUR PAINT supports this).
- A brush that can paint both behind and in front of previous strokes. Of course, this would need to be supported by a depth channel.
- Supporting this, I want an adjustment parameter that can adjust based on depth. Depth-based contrast is a uniquely powerful force in image-making.
- Also a brush that increases/decreases neighboring regional contrasts.
- Almost all digital brushes are simply repeated stamps. This is now ancient technology. I would love to see a brush that can paint entire objects or the textured components of those objects. For example, with one stroke I would love to be able to paint a tree, or hair and fur. Of course, such a tool would likely be AI.
> 16 bit as standard. Do a gradient in an 8 bit Photoshop document and you will see how limited 8 bit info is. (OUR PAINT supports this).
For interest, this seems to be an active issue for the HTML <canvas> element in browsers. There's a proposal[1] to extend the canvas data type to include both "unorm8" (the beloved default) and a new "float16" (normalised?) format - which should meet your desire?
Typically, the proposal seems to have shipped already in Chrome/Edge browsers. Documentation around what the new functionality is for and how to make best use of it is (of course!) sparse - MDN barely mentions it. As a canvas library maintainer I find this upsetting (eg: Ignore it and it might Go Away).
(I think for now my unhelpful response is: manipulate your RGB images as much as you like; just do it in the OKLAB color space.)
> (I think for now my unhelpful response is: manipulate your RGB images as much as you like; just do it in the OKLAB color space.)
Good advice. I am often introducing photographers to color editing in Lab. They are always amazed at how much more sensitive their lightness and saturation adjustments are.
> manipulate your RGB images as much as you like; just do it in the OKLAB color space
Be careful, OKLAB also isn't quite handling energy correctly (Or, the gradient of the energy slope in this color space had quite some irregularities). In most cases you can get a more "natural" transition (like blue doesn't visually shift to purple when transitioning to white), so make sure you know what you are doing.
> A brush that can paint both behind and in front of previous strokes. Of course, this would need to be supported by a depth channel.
> Also a brush that increases/decreases neighboring regional contrasts.
I believe you can just use Blender Grease Pencil for that. You can paint with depth and sculpt the opacity/contrast to your desire. (Or honestly any vector drawing program? I believe adobe illustrator does this too)
> Almost all digital brushes are simply repeated stamps. This is now ancient technology. I would love to see a brush that can paint entire objects or the textured components of those objects. For example, with one stroke I would love to be able to paint a tree, or hair and fur. Of course, such a tool would likely be AI.
> An AI powered style randomizer.
You won't want that in this context. If that being an asset production tool or a diagram tool then maybe yes, but otherwise nope. This tool is intended to create images, and human (supposedly) perceive an image with spatial arrangement of shapes and gradients, and the way artists interpret and represent shapes and edges is mainly what natural painting process is all about. So the basic structure of an artwork in this sense is just a bunch of abstract shapes arranged in a certain order, not a statistic probability of pixel values.
If you haven't already, I encourage you to subscribe this guys channel - Asianometry. How he finds the time to reserach in such detail such a wide range of tek-related subjects is beyond me.
That’s a good point, he did a video in my space recently and while it sounded impressive it was clear he only had a surface level understanding of the domain. Still impressive, but I find it hard to trust folks that represent themselves as experts on topics they’re not really experts in.
It depends on what your best hopes for yourself are.
You don’t need a lot of creative skill to model something geometric such as furniture or mechanisms. Having modeled it you would need to texture, light and render it. That can required good visual judgement, but god knows there are lots of examples to learn from.
Making organic forms is another matter and would require some sculpting skills, which are not easy to acquire.
Finally, there is the procedural environment of geometry nodes. With a good mind for maths, you could make some super cool abstract animations.
I have worked in the capacity of an artists technical assistance. I am an artist myself but am also good with fabrication. Working for someone who knows what they want and can clearly express it can be rewarding. Less rewarding is working for someone who doesn't. One young artist i worked for asked me to cut a sheet of board to a set of dimensions she was to supply. She got these dimensions wrong at least four times. Worst still was that she seemed to think her ineptitude was charming... laughing at my increasing desperation. I could fill a book with such stories.
It's [not really] surprising how well this dovetails with software development:
I have worked in the capacity of a [Production Software Developer]. I am an [architect] myself but am also good with [writing shipping software]. Working for someone who knows what they want and can clearly express it can be rewarding. Less rewarding is working for someone who doesn't. One young [founder] i worked for asked me to [develop an app] to a set of [requirements] she was to supply. She got these [requirements] wrong at least four times. Worst still was that she seemed to think her ineptitude was charming... laughing at my increasing desperation. I could fill a book with such stories.
I'm not entirely sure what op means by good taste in this context.
If we can seperate good taste from best practice, it's that good taste is commonly associated with restraint and economy. Hence, no one can say that Liberace or Elvis had good taste. On the other hand best practice is primarily governed by efficiency and is driven by commonly practiced principles.
The most significant reason that Blender is in its current position is because of the significant refactoring it undertook starting in version 2.5 I believe. 7 years brewing in the pot! It cant have been easy... but the outcome (Blender 2.8) is when we sersiolsy starting thinking about using it in our uni.
I have always found it odd how in VFX we spend a lot of time degrading our perfect 3D renders: motion blur, film grain, sensor noise and lens blur, which I would call a defocus. I am interested in the application of this research, and imagine a library of typical cameras and their associated blurs. We have similar libraries of film grain and sensor noises.
The use of Prussian blue is rampant in most painting programs. The reason is that you are essentially getting two paints for the price of one. When applied thinly, it is light and saturated. When applied thickly it is almost black. Contrast that with cobalt blue, which looks the same pretty much however you apply it.
Alizarin crimson behaves similarly.
Painting with either of these pigments it is relatively easy to get superficially impressive effects.
As a painter and also a digital artist, I am always amazed by such physical dimensions of oil paint. Another example is the huge difference between zinc white (low in coverage, good for transparency, slightly cold), and titanium white (high in coverage, good with mixing with other pigments, more neutral).
Painters don't categorize colors using standard terms: red, blue, green etc. Rather they categorize according to pigment. Different pigments (i.e. chemical base to the paint) have different properties.
For example, a Prussian blue appears almost black when applied thickly, but is very chromatic when applied thinly. In contrast, a cobalt blue is pretty much the same however it is applied.
ASAIK, this is the only app that supports this feature out the box.
The digital painting toolset has been pretty much in stasis for years, but this app offers node based brushes! I am very intrigued. Downloading it now.