Saturday 26 February 2022

Tabulae Mortuae V (Archives XV): Digital paintings die too...

 Every now and again I show an image from the Creature Vaults, those hidden domains where old sketches, failed paintings and discarded designs find their final resting place. 'Final', unless they are dragged out to be presented to the world, usually for the first time.

Click to enlarge; copyright Gert van Dijk

This image is one such, and it is the first to come from a vault without physical form. Other vaults consist of large cardboard folders, or of stacks of oil paintings carelessly stacked against the back wall of a closet. This vault is digital.

I started the conversion of the Furaha project from oil paintings to digital art some 11 years ago. The project, now nearly done, changes as time passes. The Great Hexapod Revolution had as a result that legs, heads and jaws or earlier hexapods no longer followed my self-imposed rules. The changes were too large to be solved with moderate cosmetic changes (I tried), so many paintings are now seeing a 'Mark II". In fact, some started as oil paintings (MkI), were later redone as digital paintings (MkII), and are now revisited to become MkIII. Mind you, most paintings these days are entirely new.

Click to enlarge; copyright Gert van Dijk

Here is some more detail of the head of this now defunct animal. It is a pity that I had to discard it, as I rather like the painting. But I kept the overall design and colour scheme for the MkIII version, which is nearly finished, and looks just as well or better, I think.

The animal is a 'thresher', with the Latin name 'Ira tarda'. That means 'slow anger', a name that was inspired by memories of an old teacher of mine. Threshers are solitary, grumpy and are best left to their own devices. They do have to meet from time to time, in view of the perpetuation of the species, but their behaviour at such times gives little indication of a mood upswing. Best not talk about it, really.



9 comments:

Petr said...

Intriguing!
Always happy to get a glimpse of your older works, just to see how the project has evolved over time.
The close up on the face made me think, what kind of eyes did your hexapods have at this time? Were these beasts meant to have compound eyes?

Keenir said...

tis a cool Tarda!

(if i remember correctly, for at least pre-Revolution hexapods, they were all supposed to have compound eyes that were at least the equal of Earth dragonflies)

-Anthony Docimo

Sigmund Nastrazzurro said...

Petr and Anthony: the eyes are still of the compound type. At one point they were all placed on stalks, like crabs' eyes, but that is no longer the case.
The idea is that a lineage is stuck with the type of eyes it first evolved, and for Hexapods that means compound eyes. The resolution of compound eyes is fundamentally inferior to that to camera eyes, but there might be a workaround. Hexapod eyes are generally much larger than arthropod eyes, leading to more ommatidia (not larger ones), and therefore leading to an overlap of the visual fields of ommatidia. I have started looking at papers on artificial vision to see whether clever filtering algorithms might improve resolution in such cases.

Petr said...

That is so interesting! I don't even think super high resolution is necessary. Was there something about compound eyes being more sensitive to movement or am I imagining things?
That feels like a much more important thing for survival than the "pixel" resolution of each type of vision.
But like you said, large hexapods feel like they would have millions of "pixels" in each eye just by virtue of their size, which means they would be able to discern fine detail anyway.

I wonder whether there is a precedent on earth for a compound eye where each omatidium has a fine control of its lens.

That feels like an incredibly powerful tool for focusing on any point of interest within the entire field of vision, no matter where it's located.

There was a time a few years ago where I was absolutely obsessed with trilobites and the things they have done with their eyes.
Truly fascinating animals and if compound eye is the way you want to go, they are a treasure trove of inspiration even just in the sense of what a compound eye apparatus can look like.

Sigmund Nastrazzurro said...

Petr: I thought that it would be easy to find relevant material that that proved more difficult than I thought; hence the delay. Anyway, you are right: compiund eyes are often stated to have a large visual field
, little chromatic aberration and good motion sensitivity. The drawback is moderate spatial resolution,and as you said, that can partially be compensated for by making the eyes as a whole larger. But the ommatidia must have a minimum size to let in enough light. I found mention of the ability to discern between 'pixels' 0.25 degrees apart, or 15 arc minutes. Not bad. It is mentioned in this paper:
https://journals.biologists.com/jeb/article/224/4/jeb215541/237187/Stark-trade-offs-and-elegant-solutions-in

Here is a source about biologically-inspired artificial eyes:
https://www.nature.com/articles/s41377-020-0261-8
It also mentions a fascinating visual system in Strepsiptera, insects with a weird life cycle.

I'll keep looking...

Petr said...

Ah, wonderful! Thank you for linking these resources for further reading! 💙

Abbydon said...

It's interesting reading about the Strepsiptera eye as that is EXACTLY what I have envisaged for my world. I had imagined a chiton like organism with simple image-forming eyes on its back. Initially they would be evenly distributed but over time the majority (but not all) would cluster at the front to produce a single array of eyes like the Strepsiptera eye. This would actually produce a light field camera that has a lot of interesting properties with sufficient processing power.

This seems somewhat plausible and yet suitable alien. Conveniently, it also matches well with my professional background in optics and novel sensors.

Sigmund Nastrazzurro said...

Petr and Abbydon: it turns out there is more to say about Strepsiptera eyes. later studies toned down the optimistic tone of the earliest reports considerably. Anyway, the basis idea seems to be that an ommatidium turns into a small eye with some spatial resolution, and that the spatial information is themn combined to enhance the 'image'. The big question, I think, is to which extent this can improve the spatial resolution of a compound eye. I like that approach, as it seems like something biological evolution would do.
I was thinking about a different approach to reach the same goal: if you start with compound eyes, how can you improve the spatial resolution? II think I will send you an email to talk about this (but that may take a few days)

Abbydon said...

An array of small cameras can actually be used for more than just spatial resolution improvement. There are quite a few papers on this subject and while it is outdated now the Stanford Multi-Camera Array is perhaps a good place to start.

However, some of the advantages are achieved due to how video cameras work, i.e. they are taking static photos using a shutter at a high rate. Organic eyes simply don't work like that and they actually work in a similar way to neuromorphic event cameras. Each pixel reports a sufficiently large change in brightness as soon as it happens completely asynchronously from the other pixels. Just as soon as I can persuade work to buy several neuromorphic cameras I'll try investigating what you can achieve with an array of them!