cinera_handmade.network/cmuratori/hero/code/code103.hmml

46 lines
3.0 KiB
Plaintext
Raw Normal View History

[video member=cmuratori stream_platform=twitch stream_username=handmade_hero project=code title="Card-like Normal Map Reflections" vod_platform=youtube id=-LWk-TVfGTk annotator=dspecht annotator=Miblo]
[1:23][Recap the bug]
[3:09][Blackboard: Sprite cards and casting directions]
[10:04][Take a look at the offending notion in handmade_render_group.cpp]
[12:01][Base DistanceFromMapInZ on HeightOfCard]
[13:49][Blackboard: Figuring out the height offset for the cast]
[16:36][Calculate ZDiff]
[18:41][Turn DistanceFromMapInZ into a more sane equation]
[20:36][Initialise values]
[22:48][Position the EnvMaps]
[23:46][Take a look in-game]
[24:49][Disable the sampling debug visualisations for now and check out the reflections]
[26:06][Setup EnvMaps values for testing]
[27:39][Set the Origin to the middle for FixedCastY]
[28:33][Set ScreenSpaceUV to 0.5f, 0.5f and investigate the warping reflections]
[35:00][Alpha out the area around the sphere]
[40:05][View our sphere and verify that our BounceDirection is not changing]
[40:41][Consider what else may plausibly be changing]
[42:10][Blackboard: Coordinate system considerations]
[42:59][Set OriginY in order to pass it to ZDiff]
[44:06][Turn off the sampling overdraw and darken the sphere]
[44:59][Check out our stable sphere and allow it to move around again]
[46:34][Q&A]
[47:08][@d7samurai][The reflection is "inverted" - the edge should be the center and vice versa]
2017-12-24 00:16:27 +00:00
[47:13][@garlandobloom][Q: Reflection I think is backwards, at least vertically]
[47:53][@teknorath][How long do you anticipate it will take before we get into more game-designey type topics rather than building up the engine?]
[48:39][@stelar7][There are 12 mins left of normal stream time!]
[49:10][Investigate the sphere's apparently inverted reflection]
[53:13][Blackboard: Accentuating tFarMap's falloff by squaring it]
[55:27][@d7samurai][My guess is that the normals at the edges are now pointing "straight up" while the normals at the center are pointing "sideways"]
2017-12-24 00:16:27 +00:00
[56:32][@garlandobloom][Q: Move it and then it looks wrong]
[57:49][@lincomberg][Where are the walls relative to the sphere? If it's above it as in out of the screen, then it's wrong]
2017-12-24 00:16:27 +00:00
[1:01:33][@garlandobloom][Q: Reflect Casey face]
[1:01:49][@stelar7][It feels like the red and blue should swap places]
[1:02:05][@naikrovek][Can you make the checkerboard a bit tighter? I think the problem, if any, will show up better if there are more color transitions]
[1:03:17][@andsz_][Do you only want to support standing cards and lying cards, or also any other angle?]
[1:04:19][@grumpygiant256][I think part of the problem is that it might not be 'sitting' on the floor properly]
[1:09:14][Casey is happy, considering the software renderer]
[1:10:52][@miblo][Isn't this with optimisations on?]
[1:11:57][@lauripoika][This is not threaded yet?]
[1:12:28][Kill it for today and glimpse into the future]
[1:13:07][The green dude is tough][quote 68]
[1:15:27][@ttbjm][What about Fresnel!!!]
[1:16:45][Wrap it up for today]
[/video]