"Tension maps are such cool thing! Using them you could create worms, characters, art, worm characters, and art of worms"😂 Thank you for great explanations of 3d art principles)
I've seen Chris Jones work and was super interested, but the lack of information made it hard to understand or try to replicate. You made it so much more clearer and accessible... Thank you!
After studying software like blender and game engines for 3 years and watching thousands of tutorial this is the first time I heard tension map (btw I'm no pro me dumdum :c ). That is so freaking cool how come this is not popular there's a lot of uses for it like realistic human animation, probably earthquakes, and ehem.. tentacles
Ok , I ‘m just at 2:57 min and yet I must say than after now 4 years self learning 3D sculpting, texturing,animating ,rendering and a km of great ressources found of the web .THAT is one of the best , easiest, easy to get , cool to know , explanation I get on the dark magic art of texturing .
Very cool video! I'm a 3D artist who makes a living doing 3D art primarily for games, and i've used tension maps for facial expressions quite a lot, as tension maps are a great way to accurately show the way the skin folds on the face without needing to have a bunch of geometry in order to manipulate the vertices into doing those skin folds through vertex animation (i.e. shapekeys in blender). I've also seen tension maps used in games like MGSV, or Red Dead Redemption 2, in order to emulate the skin creasing against the muscles when certain limbs bend. Tension maps seriously go a *long* way in making something feel a lot more alive, even if they're just a small, little thing!
Nice little overview! Well done. One point: Normal Maps do not displace. They actually only "bend" normals and only use the R and G Channel as one can only "bend" in two directions. That's why Tangent Space Normal Maps are always blueish (B is set to 0.5)
Thanks for explaining this new map feature no one's ever talked about in the CG community. Chris Jones' CG demonstrations of skin stretching are insane because of this added feature like 10 years ago.
you need to distinguish between displacement- and height/bump maps. Height or bump maps are monochrome and only carry unidirectional (height) displacement information. While on the other hand a displacement map is similar to a normal map in that it creates a 3d vector for the direction and amount of how to move a given point/vertex
kind of! bump maps dont displace vertices and are black/white, displacement maps do displace vertices but are black and white, and vector displacement maps displace vertices in 3 dimensions so they can do things like create a hook that comes out of a surface.
I had toyed around in my head the idea of weight painted 'crumple zones' a while back for giving more consistent results to cloth wrinkling and skin folds. Especially after looking for similar results and finding some pretty convoluted solutions. I also considered adjacent vertex proximity as a means of driving those results. Glad to see that people smarter and more capable than myself have actually made it happen, though I'd still like to see the addition of weight painting to fine tune the strength of where those folds should occur. Though I'm probably overreaching since this appears to be handled at the shader level and can be accomplished by using masks.
Hi this is an awesome idea. Do you have tutorials on how to do this in Blender 3.4+? Also what would I need to export these simulations/animations into a game engine?
For people that want this in a game engine: If I got the explanation right, the blender addon uses neighborhood information for the mesh which you often do not have in a game engine. However, I think you can implement a feature like this by using the transformation matrix of the given vertex. For models animated with bones, this matrix is different for each vertex. You can then deconstruct the matrix in the shader to extract the part of the matrix that is scaling and use that. This would give you a scale for each axis, which is information you can use to properly align wrinkles or something. I cannot say for sure that this is how it works, but you can give it a try
the only information needed when using this in a game engine is variation in length, typically 3 sets of normal maps are used. relaxed, compressed and stretched, these are then blended using the information gathered from relative edge length...
@@maxmustermann3938 I dont know TBH, lol I dont often work with game engine shaders, however once you DO get this, its the only info you need for a tension map, since the shader doesnt need a vector from each vert since tension maps dont work directionally, they only work to blend between a few different normal maps rather than trying to create a normal map on the fly.
@@Great.Milenko yea, my comment was about how one could potentially get that info in a game engine, but I am not sure that what I described would work. You'd usually use rigged and skinned meshes and thus you have a transformation matrix for each vertex. Rasterization pipeline focuses on vertices, so you really don't have neighborhood info like edges if you don't manually make sure of that. Which is why so thought just using the vertex-local transformation and grabbing the scaling portion of that might work. This way you essentially generate your "vertex colors" on the fly in a vertex Shader without the need for any of that stuff. Could work but I really don't know lol. You do get more than one scale, as in one scale per axis, which I guess you could just combine into a single scale or somehow make use of for whatever you want if you can come up with something
For me the most difficult part about 3D is rendering ... You have to learn what are matrices, and it's really not that easy. (Maybe i should use unity instead of OpenGl ...)
You really have to understand transforms (in matrix form, or in vector/quaternion form) and vector spaces if you want to do any 3D, be it in a game engine like Unity, Unreal, or a raw graphics API like OpenGL. You just can't escape them. Good news it that once you understand, a world of possibilities opens up :)
@@VirtualMethod i dont think you need to understand matrix to use unity, ive made 5 games and ive only used matrix once in one line of code out of tens of thousands.
@@XxxTheGoldenApplexxX Not talking about Matrix4x4, but about Transform, which you have to use in pretty much any game (unless it's extremely simple). Once you understand things like world/local space, converting data between spaces, transform composition, rotation order, etc matrices are just one way of representing them. Imho what's difficult is not understanding what a matrix is, but how transforms work. Once you can juggle transforms, putting them in matrix form is easy: get the identity matrix, slap the orthonormal axis in the first 3 columns (or rows, if using row major) and the position in the fourth, that's it. Projection matrices are a different beast, but dealing with them directly is much more uncommon.
@@VirtualMethod i dont think you need to understand transform as a matrix in unity. Its just 3 variables, one for each axis. Also what kind of game mechanic requires you to convert transform from local space to world space? Also dont unity already has specific variables for both local and world space? Idk its been a while since ive touched unity.
@@XxxTheGoldenApplexxX " i dont think you need to understand transform as a matrix in unity. Its just 3 variables, one for each axis." Nope. A transform is a position, a rotation, and a scale, so 10 variables if the rotation is in quaternion form (3+4+3). 9 if the rotation is euler angles. 16 in 4x4 matrix form. With 3 variables (one per axis), you only get a position. "Also what kind of game mechanic requires you to convert transform from local space to world space?" Something as basic as shooting in a FPS: the direction you shoot is "forward" in the character's local space, but you have to convert that to world space to shoot (as the bullet is typically not parented to the character). Same for strafing, you move "left" or "right" in local space, but you need to convert these to world space. If you know about matrices, and how they're built, you can get the local axis expressed in world space from the character's matrix first 3 columns/rows, though. Another typical example is a character standing in a moving platform. Usually you'd want to convert data from character local to platform local, so that's character->world->platform. Can be done much more efficiently by multiplying matrices together to get a "character to platform" matrix. Any non-trivial UI: you drag an icon from one panel to another (think inventories) and you need to convert the click/release mouse positions from screen space to panel local. Many camera setups (arcball, third person, etc) also require you to deal with these. My point is that in all the years I've been making games there's not a single game where I hadn't had to convert stuff between vector spaces. So if you're serious about making anything 3D matrices are a extremely valuable and powerful tool, investing time in them is totally worth it. If you write your own shaders or modify the render pipeline (not everyone needs or wants to, but many intermediate-advanced stuff requires it), matrices/spaces are everywhere as you need to understand object, tangent, world, camera, and clip space. "Also dont unity already has specific variables for both local and world space?" For convenience, Unity gives you local and world space rotation/position for each transform. But that does not help with converting to a different local space (as in the platform example), and in many other cases. Once scale is involved, things get much hairier as composing rotation and scale gives rise to skew (which is generally undesired) and doesn't help there either.
Oh well, the 3D animation essentials were better explained than in most actual tutorial videos. Good work there! I don't suppose tension maps are available for use in Blender 2.79 internal?
Why would Blender ever want/need to remove such a super cool/awesome and advanced feature...? I don't see any reason for that, they should bring the tension map back to Blender to make it an even better 3D free software.
Never heard about this, thought this is going to demonstrate vector displacement. But I think one should stick to elaborated terms and not make up new ones that confuses people. Tension describes exactly what it does, wrinkle doesn't. It's like calling displacement maps terrain maps, because most people use it for that, but it's not limited to.
Lol yeah, this was made before geo nodes, I’m so glad blender is progressing as much as it is!! Hopefully one day it gets to the level houdini is on but it’d need to be ALL nodes for that
Blender didin't had it, we had a addon that was very old and quite clunky, Steve did us a solid and ported it to the new API of blender and it works great
nope, blender 2.79 and earlier had tension maps, even if it was a little used features rarely mentioned with basically zero documentation. Edit, it was called stress and was a kind of texture space
fun fact from a 3D modeler: source 1 models from half life 2 have very bad topology. they are made up completely of triangles, which is usually referred to as bad because they are hard to work with.
Blender removing the wrinkle map odly says to my mind that the blender team did not want to take part on hyper realistically animated porn. I'm honestly kinda mentally ill so dont take my word for it but I've seen some fucked up animated nightmare stuff in the internet. So it's out there, maybe the blender team noticed and decided to not take part in it.
Sorry but millions of vertices is totally something even a laptop can render today. I know, it sounds insane! But it's not uncommon for a game to show millions of vertices per frame. And at 60fps that's around 16ms! Insane. Also, I thought this video was aimed at 3D folks when I clicked! (It starts explaining some stuff that's a bit... patronizing if it was a video aimed at people who use 3D in any form)
millions of verts is easily rendered during gameplay in a modern game engine, however if every object were using displacements instead of normal maps, the number of verts on screen would be nsane even for the highest end graphics cards. good luck throwing 15 billion verts around at a high framerate, this is why hardware tesselation was developed, you can get adisplacement from the GPU without needing to internally render every vert. but hardware tesselation and displacement maps are very different even if they look similar.
This already exists in houdini 😅 simple point cloud vex lookup and apply it as material blend... "tension" map is definitely not remotely similar to normal or bump maps... if anything it's more similar to curvature of an object.
@@mitch9244 well yeah, you could use a texture mask or anything to blend between normal maps or any other part of a materials makeup. the reason for tension maps is thats its dynamic and done on the fly using a deformation of some kind, instead of needing to change it manually, it changes it automatically, making animating with them super easy. for example a characters face can have 3 normal maps, stretched, compressed and relaxed, these can be blended proportionally using a tension map to instantly change normals add wrinkles, ect. you wouldnt need to hand animate the material.
All I can think about is “Sqworm”
Wrinklers....
@@36jbf-as031lb wrinklets
Sqworm
Yes, we all know who we’re talking about.
chris jones gang
Chris Jones Huh?
"Tension maps are such cool thing! Using them you could create worms, characters, art, worm characters, and art of worms"😂
Thank you for great explanations of 3d art principles)
Well, flying pen*s just apeard in my reccomendations :/
😂😂
ITS A SQWORM
We all know it’s Sqworm but seriously this video is actually interesting.
I've seen Chris Jones work and was super interested, but the lack of information made it hard to understand or try to replicate. You made it so much more clearer and accessible... Thank you!
After studying software like blender and game engines for 3 years and watching thousands of tutorial this is the first time I heard tension map (btw I'm no pro me dumdum :c ). That is so freaking cool how come this is not popular there's a lot of uses for it like realistic human animation, probably earthquakes, and ehem.. tentacles
its great for skin deformation in faces, and clothing.
I've been using Blender for YEARS and I didn't know about this. Great vid!
Besides learning about "wrinkle maps" I also really liked your explanation of normal and displacement maps. Thanks!
Ok , I ‘m just at 2:57 min and yet I must say than after now 4 years self learning 3D sculpting, texturing,animating ,rendering and a km of great ressources found of the web .THAT is one of the best , easiest, easy to get , cool to know , explanation I get on the dark magic art of texturing .
Very cool video! I'm a 3D artist who makes a living doing 3D art primarily for games, and i've used tension maps for facial expressions quite a lot, as tension maps are a great way to accurately show the way the skin folds on the face without needing to have a bunch of geometry in order to manipulate the vertices into doing those skin folds through vertex animation (i.e. shapekeys in blender). I've also seen tension maps used in games like MGSV, or Red Dead Redemption 2, in order to emulate the skin creasing against the muscles when certain limbs bend. Tension maps seriously go a *long* way in making something feel a lot more alive, even if they're just a small, little thing!
That’s so cool! It makes a lot of sense that they’re being used out there, I just wish they were more accessible to freelance artists.
Question. Are tension maps used real time in games, or are they best used to bake better animation blends?
I really appreciate the way you demonstrated everything regarding 3D Animation.
You've gained a new sub.
Nice little overview! Well done. One point: Normal Maps do not displace. They actually only "bend" normals and only use the R and G Channel as one can only "bend" in two directions. That's why Tangent Space Normal Maps are always blueish (B is set to 0.5)
*Me* : "It's an Earthworm. It's an Earthworm. It's an Earthworm. It's a-
*My brain* : "hehe flying genital"
child
@@fireball75677 uhm.. Ok higher being
@@antonkoekemoer4136 lol
Why was it removed from Blender?
When I saw it on the thumbnail I sweared it was somethings meater
You know we're in a lockdown when the thumbnail got me like 🤤😍😜😳😳
😭😭😭 dude
Girl you thirsty...
🥵 ok chill ma’am
thats a really good video, wish i had seen that before starting with blender :D
First time understanding all these properly. thanks.
Thanks for explaining this new map feature no one's ever talked about in the CG community. Chris Jones' CG demonstrations of skin stretching are insane because of this added feature like 10 years ago.
Best explanation I have seen on UA-cam. Great content on this channel.
Great explanation. I'll try using tension maps soon.
am i the only one who clicked cause of chris jones`s worm on the preview?
Actually that’s my worm! I recreated his worm with the tension map addon.
@@kextnext wow, nice blender knowledge!
Tension maps? That's going right my tool belt. Thanks for the informative video.
you need to distinguish between displacement- and height/bump maps.
Height or bump maps are monochrome and only carry unidirectional (height) displacement information.
While on the other hand a displacement map is similar to a normal map in that it creates a 3d vector for the direction and amount of how to move a given point/vertex
kind of! bump maps dont displace vertices and are black/white, displacement maps do displace vertices but are black and white, and vector displacement maps displace vertices in 3 dimensions so they can do things like create a hook that comes out of a surface.
I saw the thumbnail and immediately thought Chris jone or something like that
Really interesting. Didn’t know about this. Nice channel. Subed.
Came here from Chris Jones
Ah yes, Chris Jones.
yea boi
Sqworm
1:18 Isn't there a wrap modifier for this kind of animation?
Not quite sure what you mean, but that was just a cloth simulation
It was a low-poly cloth sim just to demonstrate what I was talking about
I had toyed around in my head the idea of weight painted 'crumple zones' a while back for giving more consistent results to cloth wrinkling and skin folds. Especially after looking for similar results and finding some pretty convoluted solutions. I also considered adjacent vertex proximity as a means of driving those results. Glad to see that people smarter and more capable than myself have actually made it happen, though I'd still like to see the addition of weight painting to fine tune the strength of where those folds should occur. Though I'm probably overreaching since this appears to be handled at the shader level and can be accomplished by using masks.
yes, the addon can output vertex groups, these groups can be combined with manually painted vertex groups to get the adjustable effect you want.
Nice man. This will definitely help course. Keep up the good work and stay safe.
Am I the only one who thought of Chris Jones' sqworm animation?
no)
Dang bro this video is Epic. Just subbed
This is the first time I've ever heard about them. Thanks. : )
Ive been trying to make a tentacle for a few days now this helped so much thx!
Tension maps are still on blender's new versions. They can be found in the animation nodes :)
Intresting.... We had this in rigging classes with Maya.
very nice explanation of basic 3d essentials!
Hi this is an awesome idea. Do you have tutorials on how to do this in Blender 3.4+? Also what would I need to export these simulations/animations into a game engine?
Using tension maps. "We all know where this is going". >.>
TL:DW
they create vertex color maps in tension areas and stretch areas
blender used to have them but now you need a plug-in.
well then how to do it for unity and unreal engine?
"if you were playing a game with millions of vertices, you'd break your laptop!"
*looks at my pc/vr setup* yeah... laptop...
For people that want this in a game engine:
If I got the explanation right, the blender addon uses neighborhood information for the mesh which you often do not have in a game engine. However, I think you can implement a feature like this by using the transformation matrix of the given vertex. For models animated with bones, this matrix is different for each vertex. You can then deconstruct the matrix in the shader to extract the part of the matrix that is scaling and use that. This would give you a scale for each axis, which is information you can use to properly align wrinkles or something. I cannot say for sure that this is how it works, but you can give it a try
the only information needed when using this in a game engine is variation in length, typically 3 sets of normal maps are used. relaxed, compressed and stretched, these are then blended using the information gathered from relative edge length...
@@Great.Milenko where do you get relative edge length from in a standard rendering pipeline?
@@maxmustermann3938 I dont know TBH, lol I dont often work with game engine shaders, however once you DO get this, its the only info you need for a tension map, since the shader doesnt need a vector from each vert since tension maps dont work directionally, they only work to blend between a few different normal maps rather than trying to create a normal map on the fly.
@@Great.Milenko yea, my comment was about how one could potentially get that info in a game engine, but I am not sure that what I described would work. You'd usually use rigged and skinned meshes and thus you have a transformation matrix for each vertex. Rasterization pipeline focuses on vertices, so you really don't have neighborhood info like edges if you don't manually make sure of that. Which is why so thought just using the vertex-local transformation and grabbing the scaling portion of that might work. This way you essentially generate your "vertex colors" on the fly in a vertex Shader without the need for any of that stuff. Could work but I really don't know lol. You do get more than one scale, as in one scale per axis, which I guess you could just combine into a single scale or somehow make use of for whatever you want if you can come up with something
Underrated
Thanks youtube algorithm
I'll check this one out, thanks for the vid
Why was it removed?
"Floppy hotdog" uh huh
Chris Jones is a wizard
"meat"
-vinesauce
0:00
Doesnt Chris jones sell a wrinkle map thing?
Love this. Great video!
little do you know Wrinkle maps are something the SFM and Source Modding communities have been using and making for over 10 years
I use these a lot, theyre also found in many modern game engines.
Skip to 3:32 for the subject of the video
Normal maps don't display displacement. Normals tell where a suface is pointing to.
2:09 Normalmaps have use in industry too. This method gives detail for free.
its what we mostly use it for, free detail
1:25 Normal maps are not displacement bud
lol
Best expaination ever. Thank you!
1:45 legacy sphere
great video, i can't believe you only have 3.5k subs
Valve just calls this wrinkle maps. They've been using it since TF2.
For me the most difficult part about 3D is rendering ... You have to learn what are matrices, and it's really not that easy. (Maybe i should use unity instead of OpenGl ...)
You really have to understand transforms (in matrix form, or in vector/quaternion form) and vector spaces if you want to do any 3D, be it in a game engine like Unity, Unreal, or a raw graphics API like OpenGL. You just can't escape them. Good news it that once you understand, a world of possibilities opens up :)
@@VirtualMethod i dont think you need to understand matrix to use unity, ive made 5 games and ive only used matrix once in one line of code out of tens of thousands.
@@XxxTheGoldenApplexxX Not talking about Matrix4x4, but about Transform, which you have to use in pretty much any game (unless it's extremely simple). Once you understand things like world/local space, converting data between spaces, transform composition, rotation order, etc matrices are just one way of representing them.
Imho what's difficult is not understanding what a matrix is, but how transforms work. Once you can juggle transforms, putting them in matrix form is easy: get the identity matrix, slap the orthonormal axis in the first 3 columns (or rows, if using row major) and the position in the fourth, that's it.
Projection matrices are a different beast, but dealing with them directly is much more uncommon.
@@VirtualMethod i dont think you need to understand transform as a matrix in unity. Its just 3 variables, one for each axis.
Also what kind of game mechanic requires you to convert transform from local space to world space? Also dont unity already has specific variables for both local and world space?
Idk its been a while since ive touched unity.
@@XxxTheGoldenApplexxX " i dont think you need to understand transform as a matrix in unity. Its just 3 variables, one for each axis." Nope. A transform is a position, a rotation, and a scale, so 10 variables if the rotation is in quaternion form (3+4+3). 9 if the rotation is euler angles. 16 in 4x4 matrix form. With 3 variables (one per axis), you only get a position.
"Also what kind of game mechanic requires you to convert transform from local space to world space?" Something as basic as shooting in a FPS: the direction you shoot is "forward" in the character's local space, but you have to convert that to world space to shoot (as the bullet is typically not parented to the character). Same for strafing, you move "left" or "right" in local space, but you need to convert these to world space. If you know about matrices, and how they're built, you can get the local axis expressed in world space from the character's matrix first 3 columns/rows, though.
Another typical example is a character standing in a moving platform. Usually you'd want to convert data from character local to platform local, so that's character->world->platform. Can be done much more efficiently by multiplying matrices together to get a "character to platform" matrix.
Any non-trivial UI: you drag an icon from one panel to another (think inventories) and you need to convert the click/release mouse positions from screen space to panel local.
Many camera setups (arcball, third person, etc) also require you to deal with these. My point is that in all the years I've been making games there's not a single game where I hadn't had to convert stuff between vector spaces. So if you're serious about making anything 3D matrices are a extremely valuable and powerful tool, investing time in them is totally worth it. If you write your own shaders or modify the render pipeline (not everyone needs or wants to, but many intermediate-advanced stuff requires it), matrices/spaces are everywhere as you need to understand object, tangent, world, camera, and clip space.
"Also dont unity already has specific variables for both local and world space?" For convenience, Unity gives you local and world space rotation/position for each transform. But that does not help with converting to a different local space (as in the platform example), and in many other cases. Once scale is involved, things get much hairier as composing rotation and scale gives rise to skew (which is generally undesired) and doesn't help there either.
Oh well, the 3D animation essentials were better explained than in most actual tutorial videos. Good work there!
I don't suppose tension maps are available for use in Blender 2.79 internal?
you could've chosen anything as the thumbnail and you went with the thing from "slither"
This was really helpful!
Why would Blender ever want/need to remove such a super cool/awesome and advanced feature...? I don't see any reason for that, they should bring the tension map back to Blender to make it an even better 3D free software.
Amazing video! An interesting decision to make to video understandable for none-cg artists, why did you chose that?
Not sure. Just a change in pace from my usual content, I guess.
Good job on this!
my computer crashes just trying to render small models
Never heard about this, thought this is going to demonstrate vector displacement. But I think one should stick to elaborated terms and not make up new ones that confuses people. Tension describes exactly what it does, wrinkle doesn't. It's like calling displacement maps terrain maps, because most people use it for that, but it's not limited to.
Good video.
Bro could use another material, but chose exactly this one
Wrinkle maps are back in blender though geometry nodes. Here's a video by Bad Normals: ua-cam.com/video/bNGGwaHOzHY/v-deo.html
Lol yeah, this was made before geo nodes, I’m so glad blender is progressing as much as it is!! Hopefully one day it gets to the level houdini is on but it’d need to be ALL nodes for that
Excellent
Blender didin't had it, we had a addon that was very old and quite clunky, Steve did us a solid and ported it to the new API of blender and it works great
nope, blender 2.79 and earlier had tension maps, even if it was a little used features rarely mentioned with basically zero documentation.
Edit, it was called stress and was a kind of texture space
@@Great.Milenko Rly? i only knew about it after the addon
@@kendarr yeah it was super confusing to use though and while it could get the same effect as tension maps it was done entirely differently
epic video thx for the info
Wheres the creds to Sqworm?
It’s created by Chris Jones, who I’ve referenced in the video. As for the thumbnail, I recreated it myself using the addon, so it’s my image.
fun fact from a 3D modeler: source 1 models from half life 2 have very bad topology. they are made up completely of triangles, which is usually referred to as bad because they are hard to work with.
But it's very common to triangulate your meshes before exporting
Harder to model, easier to develop a game engine for.
4:19 nice
Wish I watched something like this when I started with 3d
sacrifice for the skin worm
So cool. Thank
The thumbnail looks umm. It doesn't look sfw
It’s a worm ;)
but do the pores stretch?
they can, thats what a tension map is useful for. create normal map. create stretched normal map, use tension map to blend between the two.
@@Great.Milenko I know, I was just referencing a dumb joke from the Corridor Crew channel who are big 3D/VFX guys hahaha
@@sqoooge not seen that video so i didnt get the refernce but i love those guys, their content is awesome.
Thank you
Brilliant!
You had to make it flesh colored, huh
someone needs to recreate the Alaskan bullworm from spongebob
4:20 "potential" hmm yes
This is epic!
wrinkle maps makes me think of the ability to have the pores on skin squash or stretch
now i can make Thomas the tank engine
Blender removing the wrinkle map odly says to my mind that the blender team did not want to take part on hyper realistically animated porn.
I'm honestly kinda mentally ill so dont take my word for it but I've seen some fucked up animated nightmare stuff in the internet.
So it's out there, maybe the blender team noticed and decided to not take part in it.
Yeah no that's a stretch
Enjoy your nightmare CGI hentai though
Good old times we had "3d dancing baby" for representing 3d
better writing skills than 80% of youtubers
you speled super wrong
Sorry but millions of vertices is totally something even a laptop can render today. I know, it sounds insane! But it's not uncommon for a game to show millions of vertices per frame. And at 60fps that's around 16ms! Insane. Also, I thought this video was aimed at 3D folks when I clicked! (It starts explaining some stuff that's a bit... patronizing if it was a video aimed at people who use 3D in any form)
millions of verts is easily rendered during gameplay in a modern game engine, however if every object were using displacements instead of normal maps, the number of verts on screen would be nsane even for the highest end graphics cards. good luck throwing 15 billion verts around at a high framerate, this is why hardware tesselation was developed, you can get adisplacement from the GPU without needing to internally render every vert. but hardware tesselation and displacement maps are very different even if they look similar.
Sqworm personally offends me.
My virgin eyes
now i can make the squom
NICE! This will be used. And no, im sorry I cant put on instagram what im going to use this for. muhawhahahaha
This already exists in houdini 😅 simple point cloud vex lookup and apply it as material blend... "tension" map is definitely not remotely similar to normal or bump maps... if anything it's more similar to curvature of an object.
no but a tension map can be used to blend between different normal maps, this allows easy wrinkling based on pre baked normal maps.
@@Great.Milenko you could... blend between normal maps with anything right?... I just don't understand the necessity for them.
@@mitch9244 well yeah, you could use a texture mask or anything to blend between normal maps or any other part of a materials makeup. the reason for tension maps is thats its dynamic and done on the fly using a deformation of some kind, instead of needing to change it manually, it changes it automatically, making animating with them super easy. for example a characters face can have 3 normal maps, stretched, compressed and relaxed, these can be blended proportionally using a tension map to instantly change normals add wrinkles, ect. you wouldnt need to hand animate the material.