I used to post here and am glad to see this amazing program is still not only keeping on but has grown quite a bit.
I was looking through my collection of books a day ago and had an idea for someone who might be up for a challenge. I have several things like this is my photoshop actions list I just did myself but sadly this is the exent of my "programming" ability. Its also not quite the thing but close. I like doing art, my logical abilitie don't even extend to knowing where I keep my car keys between trips.
I was looking at engravings by Dore' and thought "I wonder if its possible to make a "crosshatch renderer" that unlike what is in my photoshop actions, takes the 3rd dimension into account. A good ingraver or etcher can hint at a third dimentsion by subly (or less subtly) curving lines and layering his hatch lines at different angles.
Not like a post-processor of the flat art after the render but actually takes the artwork in 3d and makes judgements about how to hatch it like an artist might. Not a great one, but a pretty good one ;-)
I use the outline feature of AOI's scanline alot like I used to used adobe dimensions/illustrator a long time ago and this would make a nice add-on to AOI. If you could call patches of hatch you produce in photoshop or some vector program to be angled, curved or layered in AOI it could be a feature that goes well beyond just hatching and would make a very style for animation also.
I mention this because I do see some problem solving thinking going on around here and suspect that some of you might see this as a worthy sport.
Bryce
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
There may be those with better ideas/skill-sets, but I will take a shot that a "cross-hatchy" procedural texture that includes the angle-incidence module (?) modulating the hatches' quanities and/or thicknesses might create a similar effect (for now?).
Richard
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Yes, I think the trouble would be getting it to have a "decison set" that would do an artistic skill like crosshatching well and make it believeable.
This is not just lines but small dots, circles curves etc that are deformed somewhat and layered to hint at light and a 3rd dimension. Dore's bible illustrations are an amazing example of this. Thats what gave me the idea since I think this is in the abilities of software.
Maybe you could have a set of hatches like you say named arbitrarily for shadows at 30% then 40% is an overlay at some angle decided by the shape of the object or the lighting etc.
Its an interesting concept problem I think.
Bryce
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Another approach would be to write a camera filter. Since filters have access to the depth map, it would be able to adjust the output based on the surface curvature.
A huge amount has been written about "non-photorealistic rendering", and simulating etchings is one of the popular subfields. So if you do a web search, you can probably find lots of information about different algorithms.
And thanks for posting the link to the Dante images. Those are beautiful!
Peter
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
>You should see on paper. Those pics are so low-rez you can't really see.
I can believe it. I went to an exhibit of Rembrandt etchings last year, and they were selling magnifying glasses at the door. And you really needed them to see all the detail.
> Is this difficult? Is it like object scripts?
Well, you would write it as a plugin rather than a script. You create a class that extends ImageFilter, and implement the filterImage() method to modify the image in whatever way you want. That's pretty easy to do, but the algorithm you implement might be quite complicated.
Peter
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I used to post here and am glad to see this amazing program is still not only keeping on but has grown quite a bit.
I was looking through my collection of books a day ago and had an idea for someone who might be up for a challenge. I have several things like this is my photoshop actions list I just did myself but sadly this is the exent of my "programming" ability. Its also not quite the thing but close. I like doing art, my logical abilitie don't even extend to knowing where I keep my car keys between trips.
I was looking at engravings by Dore' and thought "I wonder if its possible to make a "crosshatch renderer" that unlike what is in my photoshop actions, takes the 3rd dimension into account. A good ingraver or etcher can hint at a third dimentsion by subly (or less subtly) curving lines and layering his hatch lines at different angles.
Not like a post-processor of the flat art after the render but actually takes the artwork in 3d and makes judgements about how to hatch it like an artist might. Not a great one, but a pretty good one ;-)
I use the outline feature of AOI's scanline alot like I used to used adobe dimensions/illustrator a long time ago and this would make a nice add-on to AOI. If you could call patches of hatch you produce in photoshop or some vector program to be angled, curved or layered in AOI it could be a feature that goes well beyond just hatching and would make a very style for animation also.
I mention this because I do see some problem solving thinking going on around here and suspect that some of you might see this as a worthy sport.
Bryce
Welcome back, Bryce.
There may be those with better ideas/skill-sets, but I will take a shot that a "cross-hatchy" procedural texture that includes the angle-incidence module (?) modulating the hatches' quanities and/or thicknesses might create a similar effect (for now?).
Richard
Yes, I think the trouble would be getting it to have a "decison set" that would do an artistic skill like crosshatching well and make it believeable.
This is not just lines but small dots, circles curves etc that are deformed somewhat and layered to hint at light and a 3rd dimension. Dore's bible illustrations are an amazing example of this. Thats what gave me the idea since I think this is in the abilities of software.
Maybe you could have a set of hatches like you say named arbitrarily for shadows at 30% then 40% is an overlay at some angle decided by the shape of the object or the lighting etc.
Its an interesting concept problem I think.
Bryce
http://jade.ccccd.edu/Andrade/WorldLitI2332/Dante/gallery_dore.html
A few examples. Seeing it on paper is better but...
Bryce
Another approach would be to write a camera filter. Since filters have access to the depth map, it would be able to adjust the output based on the surface curvature.
A huge amount has been written about "non-photorealistic rendering", and simulating etchings is one of the popular subfields. So if you do a web search, you can probably find lots of information about different algorithms.
And thanks for posting the link to the Dante images. Those are beautiful!
Peter
You should see on paper. Those pics are so low-rez you can't really see.
Is this difficult? Is it like object scripts?
Bryce
>You should see on paper. Those pics are so low-rez you can't really see.
I can believe it. I went to an exhibit of Rembrandt etchings last year, and they were selling magnifying glasses at the door. And you really needed them to see all the detail.
> Is this difficult? Is it like object scripts?
Well, you would write it as a plugin rather than a script. You create a class that extends ImageFilter, and implement the filterImage() method to modify the image in whatever way you want. That's pretty easy to do, but the algorithm you implement might be quite complicated.
Peter