How about that baloon? It's supposed to carry cameras. I'd prefer to not retexture stuff. It may be fun to see how a prism would look as a terran unit, but these days I don't have much time on that.
Also, pick the screen model from the standard assets.
Yea, I experimented with this earlier, but portrait and dialog model cameras don't show rtt on models they see. This is why we need it to be the game world model.
Glad you have map ideas for this feature, will be interesting to see it used in game.
There are 7 RTT channels, so we can have 7 pictures simultaneously.
To not add cameras to all models, there's a way: one model with only a camera will be used for all units as a main model. And then the real unit models will be attached to this model.
We've been trying to implement picture-in-picture system for a long time, since RTT for game world was discovered. The idea is to have one or more additional pictures from differnet cameras above the standard camera view that we see by default. Understand now?
It looked easy, it looked like there are lots of ways to do it, but there have alwasy been appearing awkward limitations, that never let us reach this possibility. Even Blizzard employees who used to visit our irc, were confused and didn't find a way to do it. So, I gave up a long time ago. And now it's just funny, that its first, possible, second, easy, and third, still tricky and implemented via unexpected mechanism.
Yes, there are 2 know flaws:
- the screen doesn't cover the whole map, so if we scroll far enough to reach its edge, picture will disappear
- screen doesn't have render priority over other models, so terrain mesh or high height models simply intersect the screen plane and seem to be above the rtt picture
You brother noticed one of these things, right? First one can be fixed by scaling the model up, and the second I don't know how to fix, but maybe it's possible.
so, he's pretending to not getting the topic, and you're pretending to not getting his obvioulsy trolling intention? probably I should pretend that I don't get that you only pretending, and explain that he's just trolling
@willuwontu: Go
I've got a better idea.
Made camera + attachment point model. Use it as the main unit model. And attach the desired model to it.
Also, made 3 screens for 1,2 and 3 channels. Attach them to anything you want.
@willuwontu: Go
How about that baloon? It's supposed to carry cameras. I'd prefer to not retexture stuff. It may be fun to see how a prism would look as a terran unit, but these days I don't have much time on that.
Also, pick the screen model from the standard assets.
Yea, I experimented with this earlier, but portrait and dialog model cameras don't show rtt on models they see. This is why we need it to be the game world model.
@willuwontu: Go
Portraits don't support rtt, unfortunately.
@willuwontu: Go@Delphinium1987: Go
Looks like simultaneously 6 channels won't work either. The 4, 5 and 6th seem to be dependant on 3rd. So, only first 3 are ok.
Also, it's laggy when all 6 are exposed. You can try it with this example map.
@willuwontu: Go
Glad you have map ideas for this feature, will be interesting to see it used in game.
There are 7 RTT channels, so we can have 7 pictures simultaneously.
To not add cameras to all models, there's a way: one model with only a camera will be used for all units as a main model. And then the real unit models will be attached to this model.
@DrSuperEvil: Go
Maybe it's because far cut parameter for the camera is 50 units, and "doodad far cut" - only 30, so they disappear on shorter distance.
@Oelfrachter: Go
We've been trying to implement picture-in-picture system for a long time, since RTT for game world was discovered. The idea is to have one or more additional pictures from differnet cameras above the standard camera view that we see by default. Understand now?
It looked easy, it looked like there are lots of ways to do it, but there have alwasy been appearing awkward limitations, that never let us reach this possibility. Even Blizzard employees who used to visit our irc, were confused and didn't find a way to do it. So, I gave up a long time ago. And now it's just funny, that its first, possible, second, easy, and third, still tricky and implemented via unexpected mechanism.
@DrSuperEvil: Go
Yes, there are 2 know flaws:
- the screen doesn't cover the whole map, so if we scroll far enough to reach its edge, picture will disappear
- screen doesn't have render priority over other models, so terrain mesh or high height models simply intersect the screen plane and seem to be above the rtt picture
You brother noticed one of these things, right? First one can be fixed by scaling the model up, and the second I don't know how to fix, but maybe it's possible.
@DrSuperEvil: Go
so, he's pretending to not getting the topic, and you're pretending to not getting his obvioulsy trolling intention? probably I should pretend that I don't get that you only pretending, and explain that he's just trolling
Finally, found a way to implement nicely working picture in picture
How is it done?
A huge plane covers the whole map. It has RTT and Screen UV coordinates. Units have cameras. On "select" event unit's camera is used for RTT channel.