The PoserSurface preview includes lighting information. From what I can see, I'm guessing you have a green light coming from upper left. 

Unless you're doing a theatrical (stage) scene and you need colored lights because in real life the setup would have had colored lights, you should not use colored lights.

@Michaelab

Are you using the default lights? They tint the scene in a way that (maybe) looks nice on human-centric renders, but certainly not for archviz.

But I suspect that the problem stems from the Tile1 colour.

@Bagginsbill

Your shader is coming out fantastically. I do have some suggestions that could make it stand out even more. My observations stem from my almost 9 years working as chief designer & do-it-all in a marble&stone workshop.

Floor tiles are never 100% coplanar in respect of each other. What I mean is, that a mortar is laid on top of a surface, and the tile (whether ceramic, stone, or whatever) is pressed down on top of it. This makes the mortar... spill on the edges and stand out, and depending on performance-aesthetic choices, is how much of the mortar you then ... scrape away.

Now to the tiles themselves. Laying them on top of the mortar, which is not by any means completely planar nor smooth, and that before hardening behaves like a fluid makes each tile subtly non-planar in respect of each other, or in other words, each has a subtly different x,y, and z rotation. This is minimized a bit with the mason using a long piece of log-metal-plastic plank-thingy, but for practical reasons, it can only be so long, certainly never longer than 2m, which limits how many tiles can be more coplanar to each other. That, and that you cannot lay out all the tiles in one go.

The mortar also usually alters visibly (and sometimes physically) the sides of the tiles. As an example, a polished carrara marble tile (which is quite soft and scratchy) commonly loses a bit of it's polish on the sides because the mortar is a bit abrasive, and when the layer cleans the mortar excess, it scratches the polished surface. BTW, there are lots of translucid stones out there, carrara marble being one of them. Which means that when the tiles are not too thick, the colour of the mortar influences the percieved colour of the tiles. And makes any repair a living hell.

Ok, so maybe all this detail is a bit too much, but I thought you could use some of this information and incorporate it into your shader ;) It certainly is amazing as it is now, so maybe this is overkill.

Be it sunlight, a lamp, a reflection, the color of the surface.
All interact to give the eye the total light and color information on each spot.

In my house, I have rather small windows.
As the sun turns around, the light, and the reflections change.
From the east it is open field, then comes a forest, then some green, to a brown wall in the evening?
As the sunlight passes over each, and gives different reflections, the "light" inside the house changes too.
The floor is a rather dull light grey-brown stone, but it still reflects the incoming light.

Good tips, manoloz and vilters. Some of the things manoloz describes I have not demonstrated yet, but I will.

So nobody told what they learned about softness as displacement? Hmmm. I had hoped to teach a man to fish, not do all the fishing.

I will give a hint - I don't like the shape that is produced. 

But if I just go ahead and show what to do about it, I fear some readers will dismiss the subject because of difficulty with nodes. When we're done with this shader, there are going to be dozens and dozens of nodes. The shape is the first thing to deal with, and it's going to need 5 nodes just to fix it. But first you must see why.

Eventually we will adjust things so much that you may come to recognize that using the tile node AT ALL is undesirable. But for the time being I want to show you how to use the tile node, little by little, so that you can finally appreciate in the end why you would not use it, and why you should embrace matmatic if you really want to do quality shaders.

So far, though, everything I have shown was with the tile node as the core of the shader.

So - what's wrong with the shape?

The leather color of the seat in front is green. (As you can see in le lower left corner, even a tad more green actually.)
But with all the light, and reflections in the house, it turns out to more blue in the middle of the picture.
Green + all the yellowish reflections from the wood and the walls, make it turn to a blue shine.

In this second  picture, you see the green, but every relection on the cussions turn to blue, because of the wallpaper and the shine of the wooden roof.

Ahaaa- a detail.
When a blue relfection hits a white paper, => Simsalabim => blue paper. :-)
Lower right corner, second picture :-)

Ps, just turned around, and "on my screen at least", the colors are true to what they are.

Pictures, resized to 25% and sharpened once.
Saved as 10% reduction in jpg, using Picture Publisher 10.

Hmmm.  I have a project I'm working on at the moment ... but I will definately ponder the "softness as desplacement" when I get the chance.  That really intrigues me.

Boni 

"What's wrong with the shape?

Looks like some accidental positive displacement making a raised ring at the tile taper to joint. It's ugly! I'd use your talents on marble or one color ceramic. ;) Btw, you've not mentioned sss in tile but I'm sure there is tons of it.

Also judging from the large size of the taper at the tile edge, it makes the tile small. Say, 2 inches like a shower tile rather than a floor tile of 12 to 30 inches. trivia  but important

Right - there is an S-shaped modulation of what would otherwise be a linear ramp.

Now what happens if you run that through a math:Gain node with gain of .32 or thereabouts?

This is assuming that you have the soft tile mortar color = black, tile color = white. This is important, because we want the tile edge ramp to go from 0 to 1.

Guessing with no Poser... a neutral gray with no displacement?

Look at this render. I set up two small sections of tile, showing only a slice of the tiles. This reveals the shape as a cross section.

The top is the natural shape you get by using softness = 1 alone.

The bottom is what you get after running it through a Gain node. The number was .36, whereas I earlier suggested .32. Earlier I had just eyeballed it vaguely but I was more careful this time and concluded that Gain of .36 is close to perfect to cancel out the funny shape.

Now you may think this isn't such a great shape either, but I can work with this because it is a linear ramp, f(x) = x, going from 0 to 1. I like linear inputs and I like 0 to 1. I have a very large collection of math functions that manipulate ramps of this form. Any other sort of ramp makes it harder to figure out what to do. But f(x) = x, x = [0, 1] is awesome.

Starting from this I can show you many interesting things.

Up top is my linear ramp, f(x) = Gain(theTile, .36)

Given this rounding Corner function: (4 nodes)

def Corner(x, k) = (1 - (1 - x) ** k) ** (1/k)

I then used:

Corner(f(x), 4)

to produce the bottom shape.

I'm showing you k=4, but lots of different values of k are possible and interesting results abound.

This is very handy for making various edge shapes on tiles, by manipulating just one number.

You might have already tackled this and it's something you're going to post later on, but I've found that using non-square (i.e., tile width and tile height are not equal) tiles makes the mortar width all screwed up because that's controlled by one value whereas the tile size has the two values. Result being that if, say, the tile width is twice the height, then the horizontal mortar line is half the size of the vertical mortar lines. Which, for most tiling situations, is not ideal.

EDIT: Actually... disregard that! I've just tried it out in PP2010 and it works okay, with consistent horizontal and vertical lines. It never used to work way back in... whenever it was I last tried it and gave up because I couldn't work out how to even out the mortar width. Poser 6, perhaps? Anyway... just shows I should check what the latest behaviour is rather than relying on outdated data!

Quote - You might have already tackled this and it's something you're going to post later on, but I've found that using non-square (i.e., tile width and tile height are not equal) tiles makes the mortar width all screwed up because that's controlled by one value whereas the tile size has the two values. Result being that if, say, the tile width is twice the height, then the horizontal mortar line is half the size of the vertical mortar lines. Which, for most tiling situations, is not ideal.

Yes!!!! Oh how joyful it is to have readers who appreciate the problems.

We can only use the Tile node for homogeneous coordinates - quite right.

For other systems, we will have to abandon the tile node altogether.

Never fear, a correct and far more useful version of the fundamental tile shape is one equation away. But if I just show that equation, people will barf, as it is not easy to understand unless you approach it slowly, with head down and gaze averted. Otherwise it will skitter away.

Well I see your edit, but you're not wrong - there are cases where it will not work right.

Interesting. K=2 looks like a nice chanel leather pleat. An earlier poster who worked as a mason made a good point about uniformity or the lack thereof in a real world.

Can you produce an algorythm to pitch these multiple faces a degree or so from each other, orwould it be better to beat a one sided square up in a modeling prog? I mean, won't you get little cloned specular spots lined up like soldiers?

Not crits but looking to push the effects

Quote - Interesting. K=2 looks like a nice chanel leather pleat. An earlier poster who worked as a mason made a good point about uniformity or the lack thereof in a real world.

Can you produce an algorythm to pitch these multiple faces a degree or so from each other, orwould it be better to beat a one sided square up in a modeling prog? I mean, won't you get little cloned specular spots lined up like soldiers?

Not crits but looking to push the effects

As long as you're willing to put up with the math, I can show the following randomized imperfections:

Tiles not identical size.

Tiles have sides not perfectly straight.

Tiles rotated or off center slightly.

Tiles with too much or too little glue under them.

Tiles with uneven amounts of glue under them.

Edges chipped.

Surfaces that are not flat occasionally. (bowed)

Surfaces that are bumpy.

Surfaces with divots taken out.

Large chunks missing.

Grease and grime on the edges and packed into the mortar.

Streaks and scratches.

Whoops - once again I'm called to go out to dinner. I am flying home tomorrow - I may not be back for a while. I don't know how late I'll be out tonight.

Yes, I've just found one.

The image attached to this message shows tile width 0.1 and height 0.5 (to show some contrast between horizontal and vertical) and the mortar width is 0.5.

Looks okay for having consistently spaced tiles and it appears the 0.5 mortar width is using the horizontal spacing to determine what percentage of the tile to occupy.

That's fine; we just have to keep in mind that the horizontal size determines mortar width and orient things appropriately.

However...

And the mortar width has not changed at all!

If it were really using the horizontal size to determine the width of the mortar lines, you would expect some change as the horizontal and vertical tile sizes are switched around. But there's no change, so it's obvious that something non-obvious is determining mortar width, and not just something as simple as the horizontal size.

I will be posting more. Sorry about being away. I suddenly had six days of emergency-need-help-right-now work from a client. I finished that today. But now I'm a little behind on the other project I was previously scheduled to do this week and have to do that tomorrow.

Sorry you've got an avalanch of work! I have now bookmarked this discussion. I should have in the first place.

Experiment with it. Ask me questions.

A couple notes about the shader.

The Math:Divide at the bottom right is setting up how many tiles you want per unit of U/V space in the second value. It is dividing the space into 1/8th size tiles. For most floors or walls, 1/8 is not the right value. I was just doing that so the preview was legible. You probably want somewhere in the range of 1/20 to 1/100.

If you have a known size for the U/V space this can be precise. For example, if I have a floor that is 10 feet across in U/V, and I want 3 tiles per foot (4 inch tiles), then I need 1/30.

I'm not using the Tile node for color or displacement directly. I'm just using it to create a gradient around each tile. Remember from an earlier post that I need to adjust that to get it to be a linear gradient, using Gain. That's Math_Functions_2 which I forgot to open in this image. Use the Gain value .36 I showed earlier.

The Math_Functions_3 Subtract .1 is shifting that to choose what part is grout and what part is tile. Where the gradient value is below .1, that will be grout. Try adjusting that value. Also adjust the Mortar_Thickness in the Tile node itself.

The Math:Clamp (Math_Functions_4) is making a sharp but anti-aliased edge gradient between grout and tile. That's my master mask that tells the rest of the shader it is doing grout or it is doing tile - two different materials.

On the right/middle is my Fresnel approximation - an improvement over Schlick's approximation, but I designed it before knowing about Schlick's so it was never inspired by a desire to improve Schlick's. I found it myself and it's very useful. LuxRender should be using this approximation and not Schlick's, because it is much more accurate.

The Blender_2 node is connecting my Diffuse (Clay) reflection with my mirror reflection. By combining them in this way, conservation of energy is guaranteed.

The .25 blending factor is an easy-to-use control for the maximum shine from reflections - in this case 25%. Few floor tiles shine brighter than that and most shine less than that. Bathroom wall tile can shine more.

The top Blender is choosing tile color versus grout color. We will be plugging lots more things into that in the future. For now, I just used a solid color for grout, and a little bit of variation from a gray-scale Clouds node into the tile color which is currently dark blue. Experiment with that blue - change it to lots of different colors - get a feel for how they end up looking when shiny.

For this shader I did not do the cool rounded edge I showed earlier. I wanted to focus on the techniques for shine and color management. Adding the other nodes I posted earlier to make the round edge is simple, but distracting.

You know what the material room needs? Comment sections for each node so you can explain to yourself what it does for when you go back to look at it and can't remember what it's doing. (Yeah, my memory isn't what it used to be!)

For instance Math_Functions needs a comment section to say it's affecting how many tiles there are. I know I won't remember if I come back to use it again a year from now. 

Quote - Study this shader. It's not every trick I know, but it's a basic high-quality tile shader from which many additional variations are easily made.

Experiment with it. Ask me questions.

Is there something not shown in the material here that I don't know about. As setting it up just like the example doesn't quite look like the example and just renders a flat non-tiled floor? World units set to inches. Raytraced light and raytracing enable in render setting. 2 Raytraces. Using Poser Pro 2010. Tried to add a screenshot but it failed. I am not going to try again with the screenshot don't have enough time in the day to figure out why it failed. When it was smaller than recommeded.

Or is it not going to work with Poser Pro 2010 as I have had other things just not work in it as I would work with 7 or earlier.

@ BB > Unreproducable

Math_Functions 2 is colapsed.
No one can see what is in it.

Well, BB did accidentally leave one of his math functions closed and you had to read his text to find out what numbers to put into that function. But I had some problems getting this set up to work as well. Part of it turned out to be my own mistake because I had a wrong number in one of the math functions and it took me about 15 minutes of eyeballing my set up and the screen capture to figure out where I went wrong. The other problem has got to be a rendering problem because I'm not getting displacement to render well at all. I have Use Displacement maps checked. I haven't messed with this in a couple of days to figure out where I'm going wrong.

It's working fine for me, check message on math function.  Actually I got it to work somewhat earlier by guessing that it was an Add 1.0, but gain works much better.

"That's Math_Functions_2 which I forgot to open in this image. Use the Gain value .36 I showed earlier."

I keep coming back to this thread as I love these floors.  I do have a question since I am collecting "sets" as scenes in PP2014. My background image is NOT being reflected on the floor through the windows as they are on page one in BB's renders.  I've copied his mat room settings and it just isn't happening.  Is this a "billboard" background rather that an "import background" image?

thanks. 

It's an environment sphere. My environment sphere.

https://sites.google.com/site/bagginsbill/free-stuff/environment-sphere

There are 5 sub pages full of more info and quick-start scene is among them that is ready to use. You only have to pick an equirectangular image off the web, download it, and load it into the EnvSphere.

Tried the "billboard" .... disaster.  Any ideas?

Awesome floors.   I can't wait to use them.

Ah, I missed those sub pages.  Going right over there.  Thanks, BB. :)

Boni

Quote - Tried the "billboard" .... disaster.  Any ideas?

post screenshot of billboard shader/obj.  use 2 - 4 raytrace bounces, with IDL/raytrace enabled.  in BB's case, it was envsphere with hdri applied, as mentioned above.

Perfect!  Thank you so much!!! I'm jazzed to get going on this!

Wonder if you could do a procedural floor for wood?

Quote - Wonder if you could do a procedural floor for wood?

Yes but I have not revealed the shader yet. 

http://www.renderosity.com/mod/forumpro/showthread.php?message_id=3827875&ebot_calc_page#message_3827875

http://www.renderosity.com/mod/gallery/index.php?image_id=2365050&user_id=374541&np&np

http://www.renderosity.com/mod/gallery/index.php?image_id=2364530&user_id=374541&np&np

http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2853901&page=4#message_3990719

Hmmm.  I've just taken a "free" wood floor and am rendering it with high reflection now.  I'll share tomorrow ... it's time for the world to come in on me. ;)