[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k] [s4s] [vip] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Board
Settings Home
/3/ - 3DCG



Thread archived.
You cannot reply anymore.



File: look_at_that_cunt.png (464 KB, 449x401)
464 KB
464 KB PNG
> he uses crazy bump to make normal maps
>>
File: archkek.png (270 KB, 354x357)
270 KB
270 KB PNG
>>538042
>keekkk
>>
>>538042

Cant believe I bought that.
>>
Wha-what's wrong with Crazy Bump?
>>
I'm fairly new to this whole 3D business and I'm just enrolled in a course where the teacher strongly suggested to use crazy bump and another online app thingy. What's the best way to make normals if it isn't with software like these? Please bear in mind I literally just entered the 3d design world a couple of weeks ago.
>>
the fuck you need crazy bump for doesn't every 3d modelling program bake normals from high to low poly
>>
But how would you make normals from an image otherwise?
>>
>>538065
drop out of that class and save yourself a lot of time and money. quixel suite has some of the best normal and height map generators you can use for photos. you have a lot to learn though, i don't think learning 3d in general is a very good thing to do in a class setting but that's just my opinion. you can learn it all on your own in the same amount of time it'd take you to do it in college but you'd have more to show for it if you self taught
>>
>>538070
You dont
>>
>>538065
Generally you want to create high poly versions of your models and bake that information onto the low poly. Generating good normals from images requires a very specific set of circumstances: no highlights or shadows, subject parallel to the camera, very high resolution, among other things. It's not impossible, but creating the high poly version yourself gives you complete control over the bake.

I've worked on projects that were previously outsourced and the people knew fuck-all about creating textures. They took pictures of a tile wall, then generated specular and normal maps from those. The flash from the camera left big bumps in the normal map and rough spots on the spec. It was ridiculous.
>>
File: normal 2.png (1.11 MB, 1024x1024)
1.11 MB
1.11 MB PNG
>>538065

Baking from high poly to low poly is the preferred option, but normal map generators have their uses.

I use Mindtex2, its cheap on steam, does the same job as crazybump - after I use it to create the normalmap and specular I then blend it with a baked map via Gimp - giving me a normal map that has the best of both worlds.

pic related - its a normal map made for a low poly rock that was retopo'ed from a photoscan from both the photoscan and the texture.
>>
If you are baking from high to low poly, just fucking use your modeling program. If you need to make a normal from an image, who the fuck doesn't have the Nvidia normal map Photoshop plugin installed?
>>
File: 1475954742394.jpg (44 KB, 720x540)
44 KB
44 KB JPG
>>538100
LITERALLY THIS
The Nvidia plugin is all you need
>>
>>538065
Whilst learning other basic concepts using CB is fine, it allows you to focus on other things.

It's like when teachers tell children "i before e except after c" when teaching them spelling: it isn't always correct but it's a good stepping stone when you're just getting started.
>>
>>538065
Your teacher must be completely incompetent...
>>
>>538100
>>538228
>not using xnormal for baking from hp
>not using ndo2 instead of a crappy plugin
>>
>>538322
not for baking moron, just to make maps from scratch (black and white information)
>>
>>538242
You must be completely retarded. For a newb at 3D crazybump is adequate.
>>
I sometimes use CB to blend low frequency details on a highpoly baked normalmap. On rocks textures for instance. To me it add some naturalness to your normals. I sometimes use the Nvidia plugin to do the same thing, but I feel like it does not do the same thing, it often looks too flat for what I'm trying to achieve

Also a good habit is to process your photo before send it to CB, to like remove the highlights and edit stuff that would look incoherent. For instance when some dark parts should actually protrude instead of making holes I make them bright and vice versa
>>
File: crab.jpg (873 KB, 1446x1446)
873 KB
873 KB JPG
>>538371
Example

Notice how the lighting was fucking up the shape recognition and how the antennaes were bumping in instead of bumping out. Same for the spots on the claws and legs.

While I do recognize the value of high poly baking instead of using that kind of program, I think it can definitely be a time saver if you dont juste use it cluelessly thinking it will magically be good and put some effort fixing the input image first.
>>
>>538371
>blend low frequency details on a highpoly baked normalmap
is what you're describing just blending in a blurred low opacity copy of the texture? i tried insane bump or something, a free clone of crazy bump, and it had sliders or fine detail and large shapes and that's what it seemed to do
>>
>>538337
You sound pretty stupid. Learning 3D is primarily learning the programs you're using, and learning the right ones is incredibly important early on. You don't learn an "easy" program and advance to more complicated ones. That's not how it works - I've never seen someone not get complacent, or be very enthusiastic over learning an entirely different program later on because they learned a shitty, less capable one first because it was "easier".
>>
File: testcb.jpg (1.67 MB, 1490x1646)
1.67 MB
1.67 MB JPG
>>538391
I meant high frequency details, sorry
Here is an exemple I quickly put together. Sometimes I make a sculpt in zbrush and bake it, and I'm happy with the broad shapes, but If I want to add finer surface details I would use a normal generated from a tiling rock photo and overlay it with my baked normal (I use Ndo scripts for photoshop for this, you can also just Ndo, there's also ways to do it manually on your photoshop layer).
You should also use the rock photo you used in crazybump in the making of the albedo for more coherency.
>>
>>538431
There are better, more easily controllable ways to do that, like using Substance Designer.
>>
>>538432
For sure but with Substance designer you've got to spend time figuring out how you can mimic the natural quality of a surface, and sometimes it's really hard to do it right, while with this trick you can quickly try different maps and get the results in seconds. Once again, it's really to save time.

Maybe a good idea would be to use this technique to quickly prototype, and then go to Substance designer to reproduce the same look but with more control, If you need that
>>
>>538431
looks neat
>>
>>538328
>not using ndo2 instead of a crappy plugin
>>
>>538100
Should it work for CC2014? Or without a graphics card?

I've used CB in the past and I've noticed very uneven results with it- Sometimes it works okay, others it's sub-par and yet others completely unusable.
>>
File: Aesthetics.jpg (947 KB, 1024x1024)
947 KB
947 KB JPG
Aesthetics.
>>
>>538437
This is true but if you're going for a job it's a lot better to know how to mimic those details procedurally in a powerful program than using crazybump for everything.
>>
>>538042
GIMP or gtfo
>>
File: depth.png (87 KB, 540x810)
87 KB
87 KB PNG
>Not producing a depth map from a scribble to get a sensible normal map instead of a glorified edge filter

http://3dstereophoto.blogspot.com/2015/12/depth-map-automatic-generator-9-dmag9.html
>>
File: bad but neat.png (355 KB, 512x512)
355 KB
355 KB PNG
>>538744
think i wasted 90 minutes on this
>>
File: 1472102722783.jpg (36 KB, 800x600)
36 KB
36 KB JPG
>He doesn't create his displacement/normal maps from electron microscope imagery



Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.