46
DF General Discussion / Re: Obsidian: A new DF 3D visualizer toolset
« on: May 23, 2010, 06:53:15 pm »Will you be supporting 40d and earlier versions?
Obsidian will be supporting all versions that DFHack supports
Bay 12 Games Forum
- April 11, 2024, 06:45:49 pm
- Welcome, Guest
News:
March 6, 2024: Dwarf Fortress 50.12 has been released.
News: February 3, 2024: The February '24 Report is up.
News: February 4, 2021: Dwarf Fortress Talk #28 has been posted.
News: November 21, 2018: A new Threetoe story has been posted.
Forum Guidelines
Show Posts
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Obsidian will be supporting all versions that DFHack supports
47
DF General Discussion / Re: Obsidian: A new DF 3D visualizer toolset
« on: May 23, 2010, 02:04:10 pm »
Visual Fortress did a good job of keeping the poly count low, so I don't think it will be that bad.
At least for geometry. But I'll keep having a low poly count in mind. If I find that generating full meshes for every map block starts heaping on the poly's there are always tricks like instancing on the GPU that can help out.
This brings up an important note - the real-time renderer will rely heavily on advanced GPU features, so I'm afraid that supporting non-accelerated hardware (like older PC's and laptops) won't be a priority. For those you'll have to stick to exporting the scene to a 3rd-party software renderer, like a raytracer or something. And having a lot of patience
As for objects like buildings, trees, etc. using a good paged rendering system with imposters should do the trick.
As for the textures, keep Djohaal's suggestion of procedurally tinting base textures in mind. It can cut down on the
VF's approach to the surface rendering is very good and Obsidian will support something very similar, but underground rendering is still something I'll need to think about. Any ideas there would be much appreciated.
At least for geometry. But I'll keep having a low poly count in mind. If I find that generating full meshes for every map block starts heaping on the poly's there are always tricks like instancing on the GPU that can help out.
This brings up an important note - the real-time renderer will rely heavily on advanced GPU features, so I'm afraid that supporting non-accelerated hardware (like older PC's and laptops) won't be a priority. For those you'll have to stick to exporting the scene to a 3rd-party software renderer, like a raytracer or something. And having a lot of patience
As for objects like buildings, trees, etc. using a good paged rendering system with imposters should do the trick.
As for the textures, keep Djohaal's suggestion of procedurally tinting base textures in mind. It can cut down on the
VF's approach to the surface rendering is very good and Obsidian will support something very similar, but underground rendering is still something I'll need to think about. Any ideas there would be much appreciated.
48
DF General Discussion / Re: Obsidian: A new DF 3D visualizer toolset
« on: May 23, 2010, 01:38:59 pm »Why not use the OGRE rendering engine since it now uses the MIT licence and it's pretty good.OGRE is a bit heavy duty for my needs at the moment - but fear not! I am testing OpenSceneGraph at the moment. It came with its own set of self-reflected Lua bindings. So. Yeah. I think I just sidestepped a few weeks of happily rolling my own scene management system...
Something we should implement is the new feature of stonesense where all stone materials can get colored procedurally via a xml list of RGB values.That... is a damn good idea. Hrmmm. That goes straight into my ideas.txt...
There's a thread around somewhere for textures, let me see if I can dig it up.
And here I was concerned that getting enough art content for Obsidian might be a problem...
Might there perchance be a similar project for models?
49
DF General Discussion / Re: Obsidian: Competitions
« on: May 23, 2010, 01:27:03 pm »
Wow...
The first post didn't even have time to settle in, make itself a nice cup of tea, or read the funnies in the newspaper before the first entry sneak peak arrived!
And looking very good too, so far!
The first post didn't even have time to settle in, make itself a nice cup of tea, or read the funnies in the newspaper before the first entry sneak peak arrived!
And looking very good too, so far!
50
DF General Discussion / Re: Obsidian: A new DF 3D visualizer toolset
« on: May 23, 2010, 06:03:56 am »
My idea is to make everything swappable
I'll provide a geometry generator, and a basic renderer (both based on B's work in VF), but the architecture will allow anyone to swap in their own generator, or exporter to 3rd party tools, whatever takes your fancy.
Think of Obsidian more as a tool framework for doing interesting stuff with DF, but one starting out with a set of (hopefully) decent 3D visualization tools.
I'll provide a geometry generator, and a basic renderer (both based on B's work in VF), but the architecture will allow anyone to swap in their own generator, or exporter to 3rd party tools, whatever takes your fancy.
Think of Obsidian more as a tool framework for doing interesting stuff with DF, but one starting out with a set of (hopefully) decent 3D visualization tools.
51
DF General Discussion / Obsidian: Competitions - GFXiNXS for the logo win!
« on: May 23, 2010, 05:44:17 am »
I was thinking the first Obsidian competition (I like those - they stimulate the users into awesomeness) will be:
Design a logo for Obsidian!
Yeah, I need a logo for the splash screen.
Your prize - you can sign the image and go in the credits and one day when thousands of players use this tool (here's to hoping), your work will be the first thing they see.
Your inspiration:
This is an obsidian logo. All craftdwarfship is of the highest quality. This object menaces with spikes of Lua. It is encircled with bands of OpenGL. On the item is an image of a fortress. The fortress is being rendered in 3D.
I don't know how much time the artists will need, so I'm going to wait until we have at least 3 serious entries, before I add a poll to see who is the winner. If more time is needed, just yell.
P.S. You can also use this as an excuse to get any artistic person introduced into the wonderful world of alcoholic suicidal midgets, so spread the word, please?
Design a logo for Obsidian!
Yeah, I need a logo for the splash screen.
Your prize - you can sign the image and go in the credits and one day when thousands of players use this tool (here's to hoping), your work will be the first thing they see.
Your inspiration:
This is an obsidian logo. All craftdwarfship is of the highest quality. This object menaces with spikes of Lua. It is encircled with bands of OpenGL. On the item is an image of a fortress. The fortress is being rendered in 3D.
I don't know how much time the artists will need, so I'm going to wait until we have at least 3 serious entries, before I add a poll to see who is the winner. If more time is needed, just yell.
P.S. You can also use this as an excuse to get any artistic person introduced into the wonderful world of alcoholic suicidal midgets, so spread the word, please?
52
DF General Discussion / Re: Obsidian: A new DF 3D visualizer toolset
« on: May 23, 2010, 04:27:55 am »
Just to prove Obsidian isn't a pipe dream - I present to you a screenshot of Obsidian V0.01 in action. That's me testing the Lua error reporting function by putting a typo in my Hello World test script (you always start with one of those, ya know).
Spoiler (click to show/hide)
53
DF General Discussion / Re: Visual Fortress: another 3D visualisation app
« on: May 23, 2010, 03:04:56 am »54
DF General Discussion / Re: Obsidian: A new DF 3D visualizer toolset
« on: May 23, 2010, 03:03:48 am »
Like I said - I'll just be creating the skeleton to tie everything together via the magic of Lua scripting.
It's most likely going to be up to individual coders/'tweakers to implement the cooler stuff, unless I get to it first
Actually, my render-fu is far weaker than Boobananza's, so I'm hoping either he or a render guru will step up and help with that part
It's most likely going to be up to individual coders/'tweakers to implement the cooler stuff, unless I get to it first
Actually, my render-fu is far weaker than Boobananza's, so I'm hoping either he or a render guru will step up and help with that part
55
DF General Discussion / Re: Obsidian: A new DF 3D visualizer toolset
« on: May 23, 2010, 02:39:27 am »
Some feedback from questions in the Visual Fortress thread:
As to the exporters, I'm only going to start with something like povray. The modular nature will hopefully allow anyone to write any exporter for any format they want to, without needing to touch the core visualizer.
As for the geometry, I'll be implementing the constructions etc. as procedural geometry, like in Visual Fortress, meaning they are generated by applying certain rules to the tilesets. These rules will be done in Lua, so anyone with a text editor will be able to change how statues are oriented, how the crenelations of open topped fortifications look like, etc.
You want murder holes in the roof of your barbicon? Add a procedural Lua script that inserts them, triggered by a marker tied to a burrow, so on and so forth.
Another idea I've got is to do periodic recordings of the whole map's tileset info into a compressed stream, which can be exported as a series of scene files, which you then run overnight through a raytracer and encoder, and voila - you have a time-lapse movie showing the construction of you mega-dam/tower/castle
I'll make the tools - its up to you guys to make the awesome with it.
As to the exporters, I'm only going to start with something like povray. The modular nature will hopefully allow anyone to write any exporter for any format they want to, without needing to touch the core visualizer.
As for the geometry, I'll be implementing the constructions etc. as procedural geometry, like in Visual Fortress, meaning they are generated by applying certain rules to the tilesets. These rules will be done in Lua, so anyone with a text editor will be able to change how statues are oriented, how the crenelations of open topped fortifications look like, etc.
You want murder holes in the roof of your barbicon? Add a procedural Lua script that inserts them, triggered by a marker tied to a burrow, so on and so forth.
Another idea I've got is to do periodic recordings of the whole map's tileset info into a compressed stream, which can be exported as a series of scene files, which you then run overnight through a raytracer and encoder, and voila - you have a time-lapse movie showing the construction of you mega-dam/tower/castle
I'll make the tools - its up to you guys to make the awesome with it.
56
DF General Discussion / Obsidian: A new DF 3D visualizer toolset
« on: May 23, 2010, 02:37:54 am »
Hi all!
I am an avid fan of Visual Fortress, and am missing it terribly. With Baboonanza so graciously opening up the source code, I thought I'd try my hand at reviving it in a new guise.
From reading the Visualizer thread, I noticed that the biggest problem was lack of flexibility - with each new iteration of DF the visualizer needs to be updated (or at least a new Memory.xml needs to be figured out), and it isn't easy to add features like exporters for offline rendering (think dumping the geometry into a povray file and running it through a raytracer, etc).
So, the new project will be open-source from the start and based around a Lua framework that ties together modules. Modules would be things like DFHack, Baboonanza's DF geometry engine, and a collection of importer/exporters. Modules will be separately built shared libraries, so updating the visualizer to a new version of DF will be as simple as downloading the latest DFHack shared library binary. And if the API for a module changes all we need do is update the Lua framework files.
The better news is that the project is based completely on open source, cross platform tools and libraries. It uses CMake for configuration, so any build system with a CMake generator can be used (Linux, Mingw32, Cygwin, Borland C++, MSVC, etc.) I'm currently developing it on Mingw32 for Windows, but getting a Linux build up and running should be an afternoon's work.
I'll be making it available via GitHub soon, once I have something workable. So far I have console support, a basic multi-threading framework, a Lua engine and a Rendering engine (using GLUT) up and running. Currently all it does is show a rendered triangle and allows interactive Lua commands in a console window. But it is a start
Next up I'll be working on a basic material and scene manager. Most of it will come from existing code in Visual Fortress, so I'm skipping a lot of trial and error. Yay! Thanks B!
I am an engineer with professional game developer experience (think one of the big studios like EA, Ubisoft, Bioware, etc.) so I would like to think I know what I'm doing. But I'll need a lot of help, especially with artistic content and project management.
Any volunteers?
- Someone to manage the Linux and Mac builds would be awesome (I'm an Archlinux fan myself, but have to use Windoze at home for business reasons).
- A render expert to help with the rendering side of things (I have a bit of rendering experience, but not in OpenGL, and if I have to spend the time to learn it first it will take forever)
- A texture and model artist. Visual Fortress has a fair amount of existing content (and I know how to import it) but the collection will need tweaking and growing.
- A community manager - someone to set up a wiki, bug tracker, keep track of things happening in forum threads, create a gallery of visualizations, organize competitions etc.
The size of the response to this initiative will determine how much time I will commit to it. I'm doing it currently for the technical challenge and educational value (plus I really want to see my megaprojects in raytraced glory), but if enough people pitch in we can make this a big thing.
Oh, I'm calling it Obsidian (water + magma = obsidian, and DF + 3D = Awesome, thus Obsidian = Awesome)
Skeggox
(P.S. Spread the word, please)
I am an avid fan of Visual Fortress, and am missing it terribly. With Baboonanza so graciously opening up the source code, I thought I'd try my hand at reviving it in a new guise.
From reading the Visualizer thread, I noticed that the biggest problem was lack of flexibility - with each new iteration of DF the visualizer needs to be updated (or at least a new Memory.xml needs to be figured out), and it isn't easy to add features like exporters for offline rendering (think dumping the geometry into a povray file and running it through a raytracer, etc).
So, the new project will be open-source from the start and based around a Lua framework that ties together modules. Modules would be things like DFHack, Baboonanza's DF geometry engine, and a collection of importer/exporters. Modules will be separately built shared libraries, so updating the visualizer to a new version of DF will be as simple as downloading the latest DFHack shared library binary. And if the API for a module changes all we need do is update the Lua framework files.
The better news is that the project is based completely on open source, cross platform tools and libraries. It uses CMake for configuration, so any build system with a CMake generator can be used (Linux, Mingw32, Cygwin, Borland C++, MSVC, etc.) I'm currently developing it on Mingw32 for Windows, but getting a Linux build up and running should be an afternoon's work.
I'll be making it available via GitHub soon, once I have something workable. So far I have console support, a basic multi-threading framework, a Lua engine and a Rendering engine (using GLUT) up and running. Currently all it does is show a rendered triangle and allows interactive Lua commands in a console window. But it is a start
Next up I'll be working on a basic material and scene manager. Most of it will come from existing code in Visual Fortress, so I'm skipping a lot of trial and error. Yay! Thanks B!
I am an engineer with professional game developer experience (think one of the big studios like EA, Ubisoft, Bioware, etc.) so I would like to think I know what I'm doing. But I'll need a lot of help, especially with artistic content and project management.
Any volunteers?
- Someone to manage the Linux and Mac builds would be awesome (I'm an Archlinux fan myself, but have to use Windoze at home for business reasons).
- A render expert to help with the rendering side of things (I have a bit of rendering experience, but not in OpenGL, and if I have to spend the time to learn it first it will take forever)
- A texture and model artist. Visual Fortress has a fair amount of existing content (and I know how to import it) but the collection will need tweaking and growing.
- A community manager - someone to set up a wiki, bug tracker, keep track of things happening in forum threads, create a gallery of visualizations, organize competitions etc.
The size of the response to this initiative will determine how much time I will commit to it. I'm doing it currently for the technical challenge and educational value (plus I really want to see my megaprojects in raytraced glory), but if enough people pitch in we can make this a big thing.
Oh, I'm calling it Obsidian (water + magma = obsidian, and DF + 3D = Awesome, thus Obsidian = Awesome)
Skeggox
(P.S. Spread the word, please)
57
DF General Discussion / Re: Visual Fortress: another 3D visualisation app
« on: May 23, 2010, 02:29:49 am »
Kewl ... I like the interest
I'll repost into a new thread, and maybe throw in a very first look at what's cookin' so far.
As to the exporters, I'm only going to start with something like povray. The modular nature will hopefully allow anyone to write any exporter for any format they want to, without needing to touch the core visualizer.
As for the geometry, I'll be implementing the constructions etc. as procedural geometry, like in Visual Fortress, meaning they are generated by applying certain rules to the tilesets. These rules will be done in Lua, so anyone with a text editor will be able to change how statues are oriented, how the crenelations of open topped fortifications look like, etc.
You want murder holes in the roof of your barbicon? Add a procedural Lua script that inserts them, triggered by a marker tied to a burrow, so on and so forth.
Another idea I've got is to do periodic recordings of the whole map's tileset info into a compressed stream, which can be exported as a series of scene files, which you then run overnight through a raytracer and encoder, and voila - you have a time-lapse movie showing the construction of you mega-dam/tower/castle
I'll make the tools - its up to you guys to make the awesome with it.
I'll repost into a new thread, and maybe throw in a very first look at what's cookin' so far.
As to the exporters, I'm only going to start with something like povray. The modular nature will hopefully allow anyone to write any exporter for any format they want to, without needing to touch the core visualizer.
As for the geometry, I'll be implementing the constructions etc. as procedural geometry, like in Visual Fortress, meaning they are generated by applying certain rules to the tilesets. These rules will be done in Lua, so anyone with a text editor will be able to change how statues are oriented, how the crenelations of open topped fortifications look like, etc.
You want murder holes in the roof of your barbicon? Add a procedural Lua script that inserts them, triggered by a marker tied to a burrow, so on and so forth.
Another idea I've got is to do periodic recordings of the whole map's tileset info into a compressed stream, which can be exported as a series of scene files, which you then run overnight through a raytracer and encoder, and voila - you have a time-lapse movie showing the construction of you mega-dam/tower/castle
I'll make the tools - its up to you guys to make the awesome with it.
58
DF General Discussion / Re: Visual Fortress: another 3D visualisation app
« on: May 22, 2010, 05:31:23 pm »
Hi all!
I am an avid fan of Visual Fortress, and am missing it terribly. With Babanoonza so graciously opening up the source code, I thought I'd try my hand at reviving it in a new guise.
From reading this thread, I noticed that the biggest problem is lack of flexibility - with each new iteration of DF the visualizer needs to be updated (or at least a new Memory.xml needs to be figured out), and it isn't easy to add features like exporters for offline rendering (think dumping the geometry into a povray file and running it through a raytracer, etc).
So, the new project will be open-source from the start and based around a Lua framework that ties together modules. Modules would be things like DFHack, Babanoonza's DF geometry engine, and a collection of importer/exporters. Modules will be separately built shared libraries, so updating the visualizer to a new version of DF is as simple as downloading the latest DFHack shared library binary. And if the API for a module changes all we need do is update the Lua framework files.
The better news is that the project is based completely on open source, cross platform tools and libraries. It uses CMake for configuration, so any build system with a CMake generator can be used (Linux, Mingw32, Cygwin, Borland C++, MSVC, etc.) I'm currently developing it on Mingw32 for Windows, but getting a Linux build up and running should be an afternoon's work.
I'll be making it available via GitHub soon, once I have something workable. So far I have console support, a basic multi-threading framework, a Lua engine and a Rendering engine (using GLUT) up and running. Currently all it does is show a rendered triangle and allows interactive Lua commands in a console window. But it is a start
Next up I'll be working on a basic material and scene manager. Most of it will come from existing code in Visual Fortress, so I'm skipping a lot of trial and error. Yay! Thanks B!
I am an engineer with professional game developer experience (think one of the big studios like EA, Ubisoft, Bioware, etc.) so I would like to think I know what I'm doing. But I'll need a lot of help, especially with artistic content and project management.
Any volunteers?
- Someone to manage the Linux and Mac builds would be awesome (I'm an Archlinux fan myself, but have to use Windoze at home for business reasons).
- A render expert to help with the rendering side of things (I have a bit of rendering experience, but not in OpenGL, and if I have to spend the time to learn it first it will take forever)
- A texture and model artist. Visual Fortress has a fair amount of existing content (and I know how to import it) but the collection will need tweaking and growing.
- A community manager - someone to set up a wiki, bug tracker, keep track of things happening in forum threads, create a gallery of visualizations, organize competitions etc.
The size of the response to this initiative will determine how much time I will commit to it. I'm doing it currently for the technical challenge and educational value (plus I really want to see my megaprojects in raytraced glory), but if enough people pitch in we can make this a big thing.
Oh, I'm calling it Obsidian (water + magma = obsidian, and DF + 3D = Awesome, thus Obsidian = Awesome)
Skeggox
(P.S. Spread the word, please)
I am an avid fan of Visual Fortress, and am missing it terribly. With Babanoonza so graciously opening up the source code, I thought I'd try my hand at reviving it in a new guise.
From reading this thread, I noticed that the biggest problem is lack of flexibility - with each new iteration of DF the visualizer needs to be updated (or at least a new Memory.xml needs to be figured out), and it isn't easy to add features like exporters for offline rendering (think dumping the geometry into a povray file and running it through a raytracer, etc).
So, the new project will be open-source from the start and based around a Lua framework that ties together modules. Modules would be things like DFHack, Babanoonza's DF geometry engine, and a collection of importer/exporters. Modules will be separately built shared libraries, so updating the visualizer to a new version of DF is as simple as downloading the latest DFHack shared library binary. And if the API for a module changes all we need do is update the Lua framework files.
The better news is that the project is based completely on open source, cross platform tools and libraries. It uses CMake for configuration, so any build system with a CMake generator can be used (Linux, Mingw32, Cygwin, Borland C++, MSVC, etc.) I'm currently developing it on Mingw32 for Windows, but getting a Linux build up and running should be an afternoon's work.
I'll be making it available via GitHub soon, once I have something workable. So far I have console support, a basic multi-threading framework, a Lua engine and a Rendering engine (using GLUT) up and running. Currently all it does is show a rendered triangle and allows interactive Lua commands in a console window. But it is a start
Next up I'll be working on a basic material and scene manager. Most of it will come from existing code in Visual Fortress, so I'm skipping a lot of trial and error. Yay! Thanks B!
I am an engineer with professional game developer experience (think one of the big studios like EA, Ubisoft, Bioware, etc.) so I would like to think I know what I'm doing. But I'll need a lot of help, especially with artistic content and project management.
Any volunteers?
- Someone to manage the Linux and Mac builds would be awesome (I'm an Archlinux fan myself, but have to use Windoze at home for business reasons).
- A render expert to help with the rendering side of things (I have a bit of rendering experience, but not in OpenGL, and if I have to spend the time to learn it first it will take forever)
- A texture and model artist. Visual Fortress has a fair amount of existing content (and I know how to import it) but the collection will need tweaking and growing.
- A community manager - someone to set up a wiki, bug tracker, keep track of things happening in forum threads, create a gallery of visualizations, organize competitions etc.
The size of the response to this initiative will determine how much time I will commit to it. I'm doing it currently for the technical challenge and educational value (plus I really want to see my megaprojects in raytraced glory), but if enough people pitch in we can make this a big thing.
Oh, I'm calling it Obsidian (water + magma = obsidian, and DF + 3D = Awesome, thus Obsidian = Awesome)
Skeggox
(P.S. Spread the word, please)