Cycles: Physical Metal

I search for some physical background of steel material.
I know that some light is reflected and some light is “absorbed”.
but how many of them for typical machined steel?

Here is a rendering image i found:
This steel look realistic - but i could not reproduce that.
http://pressgate.files.wordpress.com/2011/02/02-dmu85monoblock_grundaufbau.jpg

and here is a real image for reference:
Notice that here are also some strong light reflections on metal looking nearly white
http://pressgate.files.wordpress.com/2011/05/02-duoblock-fd-technologie.jpg

My pure glossy steel look more like a rough mirror.
I set glossy color to 0,456 for RGB (0,7^2,2) and it reflect too many light back (so i see reflecting spots).

It look not so natural… i think i need some mixing with diffuse to fix that.
My material look also a bit too dark

It is not metal surface on picture, it is metal coated with some complex paint (maybe as car paint with dielectric main component with complex metal particles inside).

No storm… most surfaces are painted - but here is also a pure metal without paint.
You can see it in second image better

On other side - in second image metals look like glossy with roughnes
On first image - like mixed material…

As a good start for coated metallic surfaces you might want to use a mix shader with fresnel input as factor. Something like this very simple node set up I’m using as a quick and dirty car paint shader:


Decreasing IOR makes the surface less reflective (becomes essentially matte at 1.000), higher, more reflective/metallic-looking. Adjusting glossy shader roughness makes the reflections more or less pure depending on the setting.

I made this quicly just to see what I could get


A pretty good result compared to the images you posted, but was just a quickly test. But if you spend just a few minutes you can achieve that good looking metal :smiley:

What node setup have you used for that?

wooooooow~! you are soo good in materials can you post a node plisss <3

Everything set to 128 (bounce max, min, diffuse, etc), 200 passes.

And its a very very simple material, but maybe TS1234 ask for it, then here it is:
#EDIT: Whops, before I post, Kramon already asked for it xD


Lighting was just gray background and two emitter planes.

While i probably could not set better nodes, source images are photoshoped 100%.
Gimp, 5 min work


Original here

A pretty good result compared to the images you posted, but was just a quickly test. But if you spend just a few minutes you can achieve that good looking metal http://1.1.1.1/bmi/blenderartists.org/forum/images/smilies/sago/biggrin.gif
that looks really good - but this is more try&error
what is the physical base?
how many reflect or absorb steel?

the same for glas? how many absorb glas?
i readed in a article it absorb 2% of light
and reflect 4% an each surface point - so total 10% of light is “lost” and not transmitted
i mean such physical informations

glas article in german (just translate per google)

i found some infos here for aluminium
http://people.csail.mit.edu/jaffer/FreeSnell/metallic.html#Thick%20Reflective%20Metal

and some for iron in soil
http://web.pdx.edu/~emch/rs/vh4.html

or here:


http://ars.sciencedirect.com/content/image/1-s2.0-S0022072805004407-gr7.jpg

Fig. 7. UV–reflectance curves for iron under different conditions: a – polished; b – iron treated with 0.5 M H2SO4; c – iron treated with 0.5 M H2SO4 in the presence of poly(diphenylamine).

it looks that iron reflect only 50% (but this is not steel)