Tiny method but done from shaders so it only affects the rendering and not real scale, you also get to control FOV separately if you do it via materials
Just use 2x Render Objects (custom passes) and set to "After Rendering Opaques", set depth test in both, one on "always" and another one on "less equal", now your SSAO will mess up the model, just tick another option on it that i dont remember right now and fixed.
The issues with all these is that when it comes to some players (me) going up to NPC's face, we'd get hands and guns fitting entirely within the head. It looks really goofy Example: play Borderlands 2
2:30 how do you make it so that the material doesnt override the textures, or do you have to make a 2 separate materials for each weapon, one for when the weapon is in your hands and another when its dropped down on the ground
2:36 you can make different fov for weapon and environment by just stretching the weapon model like they've done it in Half-Life 1 Colt Python. Other thing is they stretched it too much :)
Why does the first method affect performance so much? Why does a second render target (is that the right term?) bomb so much, even if it just has a gun model and two arms?
For Example your scene has a forest with 30million triangles. Now u have 2 cameras. So now the game is rendering that 30mill twice which is 60mill triangles so basically it lowers your fps as you are rendering the world twice.
@@lalukaka493 I understand that much, but in the case where you have a separate weapon/arm viewport, it shouldn't be as bad, right? You have the main camera drawing 30 million triangles as it normally would, for example, and a second camera that's not drawing any of that. The second camera is only drawing a gun and two arms. I know there'd be some overhead in a second camera, and you're wasting the pixels that the arm viewport covers (overdraw), but other than that why did it tank performance? Is the overhead just that high to draw three objects in another render target and overlay it?
Skip to 3:00 for the only correct solution to this ENTIRELY self-imposed problem. Three minutes of wasted breath to describe 1.) a problem that only exists if you MAKE it exist by cutting corners and 2.) a bunch of silly non-solutions that look like shit and make gunplay feel flat and boring as fuck.
This has only two real solutions, don't be dumb 1. two cameras 2. separate animation for hands and weapons when close to a wall Everything else is just stupid solutions
Or you could have a anim stat where the character lowers their gun when too close to walls. 🤔
03:03
that's the only real way it doesn't ruin the sense of scale.
@@olehkulys Missed that. Thanks. 😅
@@atDigso i said lower, he said raise.
@@OdysseyHome-Gaming thats literally the same thing
Tiny method but done from shaders so it only affects the rendering and not real scale, you also get to control FOV separately if you do it via materials
Just use 2x Render Objects (custom passes) and set to "After Rendering Opaques", set depth test in both, one on "always" and another one on "less equal", now your SSAO will mess up the model, just tick another option on it that i dont remember right now and fixed.
you can also just use a shader to make it always render ontop or use a matrix to make it render in a really weird way that makes it always ontop
The issues with all these is that when it comes to some players (me) going up to NPC's face, we'd get hands and guns fitting entirely within the head. It looks really goofy
Example: play Borderlands 2
2:30 how do you make it so that the material doesnt override the textures, or do you have to make a 2 separate materials for each weapon, one for when the weapon is in your hands and another when its dropped down on the ground
2:36 you can make different fov for weapon and environment by just stretching the weapon model like they've done it in Half-Life 1 Colt Python. Other thing is they stretched it too much :)
what about just changing the sorting layer?
This video is a hidden gem, thank you
Why does the first method affect performance so much? Why does a second render target (is that the right term?) bomb so much, even if it just has a gun model and two arms?
For Example your scene has a forest with 30million triangles.
Now u have 2 cameras.
So now the game is rendering that 30mill twice which is 60mill triangles so basically it lowers your fps as you are rendering the world twice.
@@lalukaka493 I understand that much, but in the case where you have a separate weapon/arm viewport, it shouldn't be as bad, right? You have the main camera drawing 30 million triangles as it normally would, for example, and a second camera that's not drawing any of that. The second camera is only drawing a gun and two arms. I know there'd be some overhead in a second camera, and you're wasting the pixels that the arm viewport covers (overdraw), but other than that why did it tank performance? Is the overhead just that high to draw three objects in another render target and overlay it?
Skip to 3:00 for the only correct solution to this ENTIRELY self-imposed problem.
Three minutes of wasted breath to describe 1.) a problem that only exists if you MAKE it exist by cutting corners and 2.) a bunch of silly non-solutions that look like shit and make gunplay feel flat and boring as fuck.
ı use the tiny method actaully. another camera is a good way to performance lost
vay karşim sende burdasın bea
@@bertadev what are you saying brother i dont understand your language
@@bertadev he burdayim bea
Hey, is there any way to contact you? Wed love to sponsor you! I own 3D AI Studio and i think your channel would be a good fit!!! Viele Grüße ; )
This has only two real solutions, don't be dumb
1. two cameras
2. separate animation for hands and weapons when close to a wall
Everything else is just stupid solutions