anyone know how to take pictures like this?
by Chris · in Torque 3D Professional · 05/18/2010 (5:47 am) · 11 replies
I want to setup a second camera or in some way take images like the following?
I'm using them for a minimap resource -- to make this image I had to take various screen shots from the mission editor and stitch them together.
If I can just get a over head camera to take a partial picture, without GUI elements or effects(AFX) that would be amazing. I will even pay for a good working patch that does just what I need.
I'm using them for a minimap resource -- to make this image I had to take various screen shots from the mission editor and stitch them together.
If I can just get a over head camera to take a partial picture, without GUI elements or effects(AFX) that would be amazing. I will even pay for a good working patch that does just what I need.
About the author
#2
My main hang up is a piece of code that takes a overhead picture and stores it in a GBitmap or Texture without interfering with the players experience.
05/18/2010 (7:07 am)
Ok I want to do this through automation.My main hang up is a piece of code that takes a overhead picture and stores it in a GBitmap or Texture without interfering with the players experience.
#4
The trick is real time automation. I am willing to pay someone if there is a good working solution that will get us there without manual stitching as our world has dynamic building creation that hurts minimap static images.
05/19/2010 (10:52 am)
Yes I am stitching images right now, we have created a 3000x3000 base map for one of our test areas with great success.The trick is real time automation. I am willing to pay someone if there is a good working solution that will get us there without manual stitching as our world has dynamic building creation that hurts minimap static images.
#5
www.torquepowered.com/community/resources/view/5277
05/20/2010 (5:58 am)
Here's a reasource which may or may not work for you, it seems to be the most suggested reasource for minimaps...www.torquepowered.com/community/resources/view/5277
#6
I may be talkin out of my...well, you get the picture. ;) Anyway, not being a programmer, isn't there a way to "render to bitmap" somewhere in the engine? Can't you use a camera, placed above the characters head, looking straight down, with appropriate FoV, and send that camera vew to a window in the GUI? This is how it would be done in some other engines is all I'm saying.
For 'static' area maps, maybe use a modeling program to create a grid of objects at whatever distance is needed. Import into T3D. Then, script a camera (like above), to move object to object in sequence, taking a screenshot and saving to file. That would give you all the shots in one fell swoop...and you can use it on different maps without having to do much other than perhaps change a few key words in the script?
Again, I'm no programmer, so just going by "what I'd do if I knew how to program". Hope it gives some ideas...
^_^
The Hierarch
05/20/2010 (7:55 am)
Hiya.I may be talkin out of my...well, you get the picture. ;) Anyway, not being a programmer, isn't there a way to "render to bitmap" somewhere in the engine? Can't you use a camera, placed above the characters head, looking straight down, with appropriate FoV, and send that camera vew to a window in the GUI? This is how it would be done in some other engines is all I'm saying.
For 'static' area maps, maybe use a modeling program to create a grid of objects at whatever distance is needed. Import into T3D. Then, script a camera (like above), to move object to object in sequence, taking a screenshot and saving to file. That would give you all the shots in one fell swoop...and you can use it on different maps without having to do much other than perhaps change a few key words in the script?
Again, I'm no programmer, so just going by "what I'd do if I knew how to program". Hope it gives some ideas...
^_^
The Hierarch
#7
I'll have to rewrite all my minimap code to work with the render target stuff but it will work.
05/20/2010 (10:16 am)
I figured out enough to make this happen from the later pages in the commander map.I'll have to rewrite all my minimap code to work with the render target stuff but it will work.
#8
10/27/2012 (8:54 am)
Hay Chris you ever get a mini map to work in T3d? if so love to have a peak at your code. 
#9
10/27/2012 (9:27 am)
Have you tried to place a third person camera way over the players head? (in the model itself). You could use render to texture from there. (or you should be able too)
#10
For our purpose, camera tricks like the above linked resource didn't end up working too well. The approach I'm currently using in TGE for exterior mapping is to directly generate a basemap image by processing the terrain block textures.
First scale the base texture pallet to the desired resolution. Then process the terrain square by square to see what texture blend is used for each and use the scaled base textures to generate a bitmap of the entire terrain. As a final step, go through any water objects and do an overlay with alpha-blend on depth. One downside to the approach (compared to a camera render) is that the map ends up being somewhat "flat" (although, it's possible to do a heightmap overlay that really helps in some situations).
One of my big concerns in moving to T3D (and a big reason why I'm working on completing a demo in TGE rather than porting first) is the significantly different approach to terrain. I haven't had much time to play with the codebase yet and am still trying to suss out differences - so I'm not sure how helpful a discussion of the method's technical details are for those trying to do maps in T3d. From what little I've seen, it looks like the basic approach to terrain is quite similar (blender and such), so I'm somewhat hopeful.
Anyone have thoughts on how easily this methodology will port?
11/01/2012 (12:23 pm)
This is a question as much as an answer. (also, I can't seem to see the stitched-image, so I'm going on descriptions)For our purpose, camera tricks like the above linked resource didn't end up working too well. The approach I'm currently using in TGE for exterior mapping is to directly generate a basemap image by processing the terrain block textures.
First scale the base texture pallet to the desired resolution. Then process the terrain square by square to see what texture blend is used for each and use the scaled base textures to generate a bitmap of the entire terrain. As a final step, go through any water objects and do an overlay with alpha-blend on depth. One downside to the approach (compared to a camera render) is that the map ends up being somewhat "flat" (although, it's possible to do a heightmap overlay that really helps in some situations).
One of my big concerns in moving to T3D (and a big reason why I'm working on completing a demo in TGE rather than porting first) is the significantly different approach to terrain. I haven't had much time to play with the codebase yet and am still trying to suss out differences - so I'm not sure how helpful a discussion of the method's technical details are for those trying to do maps in T3d. From what little I've seen, it looks like the basic approach to terrain is quite similar (blender and such), so I'm somewhat hopeful.
Anyone have thoughts on how easily this methodology will port?
#11
11/01/2012 (1:35 pm)
Wouldn't it make sense to take one shot using a projection that gets rid of fish eye? Isn't there a way to change projection type on the fly? I know the T3D editor can change projection type (ortho?). 
Associate Steve Acaster
[YorkshireRifles.com]
Just to check ... you do know that CTRL+H gives a hudless/noGUI screeny, yes?
Also the more zoom you have on the camera the less "fisheye" you'll get (why many fps eye/camera settings are 65 - current T3D stock is 75).
I can't help with turning AFX stuff off though ... it's not my bag ...