Bumping a texture map

Bumping a texture map

I am trying to bump my texture to get a geo look out of it. I did this before with a grid floor. Although I can not tell too much difference between the bumped and not bumped in my version. Which leads me to question myself.

I am at it again as I have seen others work that looks GOOD compared to mine.

Here is a pic, am I going down the right path here?
If so which do I use Normal or Bump? A combo? What about Geo Map?

I have a png of the colored version (touched in P.S) hi-lit yellow
There is the Normal/Tangent Bake (hi-lit purple) out as a PNG or do I use this one at all?

Kinda foggy on getting a good bump out of this like can be seen in Blender Render when using the GSLR shader and you can have a + or a - value that is very predominant.


This is how I did it with the floor I referred to that I had help on. Is this the only way to get a ‘bump’ and that is it?


Another method, though not as reliable, is to use a math node after the normal map node and plug it into the displacement slot of the material output. The math node then becomes the bump amount. This method come in handy when you do not have a dedicated normal map as you can add a color ramp right after the image map and us it to crush white levels of the area you want to bump more.

you can run the texture into a rbg to bw then into bump then into the normal of the diffuse
and run the normal map into the displace like anthony said

the bump is for the ruff stuff
and the normal map is for the fine stuff but if you do the normal map like anthony said you can control the height of the displace

MidnightCPU Thanks for the clarity on the Bump node Vs the Normal Node SO I assume that the way the floor was done in my post was the way to do it… Other than the enhancements via Anthony’s suggestions Thanks Anthony

I tried Anthony’s description of a node set up. However I could not
follow:

The math node I am unclear on which one that I am to use for the drop down choices… There are a dozen of em’ see pic

Also ‘crush’ the ramp? I do not know what or how you are using that slang…
Once I get that ironed out (with some direction) I can then move onto adding the other node rgb/bw~bump node~Diffuse~then into Displace material out Node

Maybe a screen grab from your end of a node tree showing me??


Follow from the top image input.



Converter>Math>set to multiply : Sorry for not being clearer.
Note that in this example I did not bother with a Normal Map node since I was working with a color image.

By crush I mean that you can tighten up the levels. Making the contrast in the images very stark.
An RGD to BW converter is not needed since the color ramp does it for you.

Here are the nodes in action for reference.

I hope this helps.

Original poster has a normal map and wants to know how to connect it. A normal map does not contain height information so should not be connected to displacement.

The “Color” output socket from the “Image Texture” node connects to the “Color” input socket on the “Normal Map” node. The “Normal” output socket from the “Normal Map” node connects to the Normal input socket on the “Diffuse BSDF” node (and if you are mixing BSDFs, also all other BSDFs which you wish to affect.)

When using a normal map in a texture node the first pulldown (by default “Color”) should be set to Non-Color Data to prevent the possibility of colour management altering the normals.

The OP indicated that they aware of the process for plugging up a normal map and is asking if there is another way of going about it.

Anthony thank you very much for that screen grab. I would have been going crazy trying to read that in print. A pic is worth a 1000!

I will play with this in detail. Any tricks on that Ramp I cannot see?

[Note that in this example I did not bother with a Normal Map node since I was working with a color image.]

So by defacto a B&W would use Normal map Node?

Jandress thanks for direction but, yeah I had that going already. I will pay attention to the ‘drop down’ you mentioned though “non color” pulldown.

Which should I bake out to have height adjustment? The tut I am following sends the bakes out to P.S and then brings them back in. I am either somehow missing a detail in one of the steps or am not baking it right. Or better yet that IS a Tangent/normals bake out you see in the pic post #1
Should this be connected a better way “as asked” to get a bump?

There is no other “better” way to connect a normal map. Converting it to black and white will not give you an equivalent bump or displacement map.

If your normal map setup is not giving you the results you expect, post a screengrab of the problem, or a blend.

If you are having issues following a tutorial, can you post a link to the tutorial? The only reason I can think of off the top of my head for running a normal map through photoshop is - some systems assign x/y/z to r/g/b differently so moving between different software or game engines you may need to swap channels, but baking from blender for use in blender you should be ok.

If you prefer to go the bump or displacement map route and want to bake the maps from geometry rather than painting by hand you will have to user blender internal to bake.

http://wiki.blender.org/index.php/Doc:2.6/Manual/Render/Bake

Cycles does not currently bake displacement, although you could implement a manual setup for simple geometry with an orthographic camera and a shader which takes a grayscale value from the world z position.

Thank you guys

Anthony, if you could share with me the Image texture assignment to help me better understand what each is doing for it’s Node chain that would be helpful and appreciated.
From the Top Texture assignment node. What I am referring to is Bump map, Normal map, and Color or texture map.
IE:
1.Bump
2.Normal
3.Color or actual

jandress-Thanks for the clarity, I am new to this texturing and while I like the result of B.R for GEO I need some of the Cycles attributes for my scene. Leaving me with trying to texture with a bump in Cycles. (pics above) Yes I have baked from BR already, now I am trying to utilize them in a manual fashion in Cycles to ‘cheat’ some geo as (above pics) posted. As I’ve asked about, and now have been shown different ways to go about it via Anthony’s way…

Painting by hand is a mystery I have not tackled yet.

The video I am following is http://cgcookie.com/blender/lessons/creating-a-sci-fi-panel-part-2/
@ 20min in is where he leaves me hanging and shifts to Unity instead of showing how to use CYCles, he actually goes back into Blender Render and assigns the textures as Maps, Bump,Normal, etc…Which I understand that in BR.

From what I am gathering from this post…It seems that it does not matter if the Bump is a Spec map or Normal Map or Full Render. IT looks like you guys are using a Full Render PNG with info in it to do whatever bumping you are wanting out of it. Is that right?

I am creating Maps and so forth for nothing in Cycles? These Maps are useless?


I should just paint as desired and then hook up like Anthony shared (albeit) I am unsure which is which at the moment? Just use the PNG 3 different ways to achieve the ‘bump/geo’ desired and be done with it…


Your maps are all useful - you can use that spec map as the stencil for your gloss shader into mix node, you can use the normal image into a normal vector input for your shaders, you can use the grey scale image for bump instead, into a vector bump node fed to your diffuse and gloss normal sockets, etc. You are on the right path, but you have to find what works best for your needs. Some things don’t need so much info in cycles, some really does need everything depending on proximity to camera. Whether hand painted or derived from processing, it doesn’t matter - you are the artist, so you have to decide on the look you want and work to get it.

Looks good so far, I’m interested in the final scene render after all this.

As Craig Jones says - any image has it’s use.
Here i have to admit i’m having laziness sometimes. Here’s alternative. Ersatz so to say.
http://www.pasteall.org/blend/32943

Eppo thanks for the blend file,
I have a few questions being at a distance and coming in after you built it…
See pic please 1-5b
1.
2.
3.
4.
5a.
5b.
???


Thanks Craig, I just saw your post. I wish there was a few tutorials (recent) in regards to the manner in which you mention ‘plugging’ them in. Like an artists uses different brushes for different effects Blender has so many brushes and so much must be learned before using it…!

Eppo’s Blend file is appreciated and has shown me a complete new way of doing things… Mighty humbling.

At least with a brush it is bristle side down and usually bristles in the paint. Blender is "you need to plug this in here for this but if you plug this into there, well a complete different result. With not to much for a manual or an example. I go to the Doc Wiki and when asking what time it is, I am told how to make a watch. No examples (or very min) of how to get an effect via Nodes… That is why I come here.

At this rate I am hoping by January to have texturing down ( I spend about 30 hours a week if not more trying to learn it) So I can have my little short done… It is a culmination of all my posts into a 1:30 movie. LOL!I guess I am going to have to roll credits to all you guys who’ve helped me…hahaha

Thanks Jandress I will be reading for a while over there…

Basically in a nutshell, the Fac input on mix nodes of either Shader Mix Node or Color Mix Node are the point where plugging the output of an image node/Attribute node/texture node/etc. can act as a stencil. You can use any black and white image to drive the Fac(tor) of the mix nodes that way, and the bottom input of the nodes are the top layer, top input is the bottom layer.

You can make use of the Vector nodes of Bump and Normal for using those respective image textures, and you plug them into the Shader nodes on Norm input. You can plug the same normal or bump node into both shaders being mixed to get the same value of bump, or you can set up separate nodes to drive the bump in different values.

All this means you can use dedicated textures for getting that really realistic detail at close range, or you can make use of simple images blended together with Box mapping to get basic textures, and plug them into bump nodes that drive the different shaders you might use liek diffuse mixed with gloss, etc.

Tutorials will get you so far, and then hopefully this all starts to make sense. I am still exploring, still learning, and I don’t even attempt to know much at all, I just love paint :smiley:

  1. This was done quick and dirty in BI by normalmapping several “floating geometry” objects. I assume you know this part. It is the only image needed in this example.

  2. Logically separated Inputs. Easy to make adjustments. Borrowed from B.S. way of doing things ;). Noodles while not grouped tend to be quite long and sliding window forth and back can be annoying - 2 windows help here.

  3. Two colors for two “real” paint layers being different. Metal usually is covered in primer, then paint.

  4. This is here created Group of nodes if i correctly understand the question. Tab key to jump inside it, Shift Tab back . Can Link as a Group in other blend file and select from Add menu in node editor easily.

  5. Upper Output is the final material. One below serves as a control to see where different spots are on the render, adjust intensities, colors, mask levels and such.

I use to simplify thinking of all this noodle stuff and tend to bring parallels with any paint program’s layer operations. Difference being Vectors as a Layer movement/distortion effects and Shader mixing, BSDFs… that’s specific to 3d - how material reacts to lights.
Bit of a math that color or bw image is just a matrix of numbers - RGBA values one can sum or multiply if needed and do other math stuff for fun ;). Same goes to vectors.
Built in procedural textures can be thought of as a brush one needs to cut for the paint job before - vectors help to cut and scale them, color mixing allows for variations (black’n white mix mostly).
That’s in a short. Hope helps.

thanks you guys,
I am not going to state “I get it all” I am getting a bit closer to realizing that each node is it’s own ingredient that can either spoil a recipe or enhance it. I am at the point where I now realize there is no “right” way to achieve a look. There are many ways pending on what look I am going for, Like an artist who sometimes uses his mixing tool opposed to a brush to paint. Un orthodox but effective in some situations.

Eppo’s Blend file really opened my eyes. Before that I was thinking there was a set way of achieving a ‘bumped’ look via a certain type of map with info… Ha! Now I see as Eppo pointed out a matrix of numbers and Nodes are the formula used to achieve a look. (So I think at this point)
Eppo, #4 question from above… That is a “created group” of nodes? Nothing that is ‘canned’ needs to be created?

Craig I appreciate your input and trying to simplify the FAC
"You can use any black and white image to drive the Fac(tor) of the mix nodes that way,

“and the bottom input of the nodes are the top layer, top input is the bottom layer.”
You are referring to the BSDF’s OUT into the SHADER node IN (green dot) Which means they are opposite of layout? Top dot is bottom and bottom green dot is Top? Like Layers are in Texturing. Photoshop is Top layer is Dominant and in Blender Bottom layer is Dominant… Right?

I am off to read Jandress’s links… I may come up for air in a day or two… If and when I get a grip on this, I would like to make a tut on just the power of using nodes as bumps and all the lateral given with using, two image in’s, an RBG/BW for my FAC, and how the shaders work opposite of their respective layouts…Then show how Eppo’s Blend file and switching nodes can achieve a skipped step in Photoshop by doing it all local in Blender via painting it there.

One question that may be answered on my link quest via Jandress… To get an emitter map to pop like they are glowing as lights without affecting the the other images… Should that be an image node connected to ?? > Connected to>?? any grab or suggestion on getting a bump out of that or a glow if you will ?

Thanks a lot for the direction and time you all give BTW

Regards
NC

P.S do you guys get your login kicked after taking to long to write out any details?

I mean that a Color Mix Node two color inputs or a Shader Mix Node two mix inputs work in the opposite direction of the Photoshop top-dominant layer like you describe. Much easier feedback to see this is when using the Composite nodes for render layers/images and stacking stuff for output with the Alpha Overlay node.


Kind like here, where you see the purple shader on top due to the stencil mask plugged into the fac of the mix shader…