How can a mesh be referred to as having "pixel density"?
Roman_K2
Posts: 1,206
Woah... hold it, stop it right there! Thanks.
Ok, I located a mention of adjusting the bump values for rendering in Iray: http://docs.daz3d... snip ...shader/shader_general_concepts/start#normal_map
This has been a long-term thing for me (bump and Iray) because I like to play with surfaces if I can.
From the above documentation Wiki: "The bump is fairly sensitive to both the pixel density of the mesh and the resolution of the image map. A typical example is the Genesis 2 Figure. The face is more dense than the head in number of pixels..."
This must have to do with different material zones or something, or compressing more graphic information (like the face) into a small area? I guess I'm at a loss as to the ins and outs of filling a simple mesh with "more", or "less"... pixels. Huh?!!! To me, a mesh is composed of vector-like "lines" and "vertices" and there's nothing "in between" any two vertices except a single line. There's no pixels that I know of !
At any rate I hope someone could suggest a range of bump values to try (or not?) with Iray - like skin, porous concrete or pumice, leather or brushed metal, jute weave etc. I gather that some of these settings are more important or hold more power in certain situations or instances, and a lot of things may safely be ignored? I'm fascinated by the idea of the Oren-Nayar values in Blender, (wiki.blender.org link) and what this sort of thing is good for.
Thanks, I'm done now. 
Daz 3D is part of 
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
It could mean Texel density which is simply the scale of unity texture space. I haven't heard the term "pixel density" used before, but texel density would make sense in this context. Then again, it's only a phrase i hear in the maya community, not so much elsewhere.
My best guess would be that they are talking about texture map resolution and mesh density in the same breath and inadvertently mixing up their terminology. Presumably they meant to say polygon density of the mesh, which would make more sense.
I thought it might be something like that... too bad that as a newbie I'm kind of hitting this brick wall, with Iray texturing.
There seems to be a lot of typos in the category - I moved on and downloaded something from the "iray render" dot com web page and it said to stay away from absolute white and white values, suggesting that "RGB 1,1,1" should be used only for the brightest possible SNOW scenes, and even then "almost never".
Huh??? RGB 1,1,1 is the code for black!!!
I think for my next experiment I'm going to try rendering some legacy M4 and V4 surfaces with 3DL, and then I'll try and repeat the exact same scene with Iray after applying the Uber base shader.
RGB 0,0,0 is black.
There are two ways of showing RGB values...0-255 (the way the 3DL shaders use, in Studio) or 0-1 (the way the Iray shaders use). They are talking about the 0-1 way.
As mjc2016 said, there are two ways of representing RGB values. The range of 0 to 255 are integers and represent the exact same range as 0.0 to 1.0. Thus 0 is the same as 0.0 and 255 is the same as 1.0. A floating point/decimal value of 0.5 is the same as an integer value of 128.
PBR renderers like iray tend to use the floating point/decimal range for various reasons. Older renderers generally use the integer (0 to 255) range to represent the same thing.
As far as the range goes, I've normally seen recommendations that black shouldn't be less that 20 or 30 (in integers) and the brightest values should rarely be above 216 or 0.7 (70%). The latter is more important as it has a big effect on how quickly the simulated photons die.
A completely black color, i.e. RGB values of 0/0/0 or 0.0/0.0/0.0 means the surface reflects absolutely no light at all. A surface with that type of material is unlikely to cause big problems in your scene, but isn't physically plausible.
A very bright color, i.e. a color having one or more RGB values above 216 or 0.7 means the surface reflects a lot of light. Very few surfaces in the real world reflect more than 70% of the light that hits them. A surface with that high of reflectivity is going to keep the simulated photons alive longer, giving them more bounces, which can make your scene converge slower. Also, a highly reflective surface like this is unlikely to be physically plausible.
When computing the amount of reflectivity a surface has, one needs to include both the texture map, if the channel has one, and the RGB color assigned to the channel.
NOTE: The above is dealing with the Base Color (diffuse) parameter. Other parameters, like Glossy Color, have their own rules, but generally should be pure white for non-metal surfaces.
The map is multiplied by the color...so to represent the colors on the map unchanged, values of 1 could be used, for the color. But if the map contains very 'white' whites, it may be better to drop the color value some.
Oh-oh... having trouble with this concept, big time. RGB 123,123,123 is the format I usually use when entering values into an image editor or vector drawing program. If I'm not mistaken this allows several million variations which makes it analogous to mixing paint the old way, back in the old days, eg. brush and palette, oil paint, acrylics, watercolor, gouache etc... that sort of thing.
Duh... at the risk of sounding stupid, it seems to me that 256 choices is not the same as having millions of choices, when you're rendering something and you're trying to have some subtle shading and so on. [Edit: Ok, ok... just read zaz777's explanation above, AFTER writing most of THIS comment. So around midnight it was starting to be clearer... except the term "PBR" is something I haven't encountered before.]
About "0,0,0 being black" - I know that, but there is so little variation between the first eight choices in that spectrum, that I usually lump them all together -- for me then, "8,8,8" is usually my first choice for when I want something real dark, but not necessarily the "blackest black".
Attached, a screen shot of NVIDIA's documentation for this stuff, and I've circled the part where they say that RGB 1,1,1 is white which to me is clearly wrong, although I have enough experience with paint to understand where they are going with "colors in nature" and so on. And the part where they mention a sort of decimal notation -- "0.7,0.7,0.7" -- is news to me; I have never seen that before. What is that, a way of expressing a fraction of a number between 0 and 256? Can't I do it the old way if I'm more comfortable with that??? Again the programs that I use, including DS 4.8 (eg. the backdrop value in the "Environment" pane) all seem to use the format 123,123,123
I think part of the problem with explaining/understand this stuff is that it is sooooo procedural: since I have really have no clue as to what I'm doing, I have to accept anything you folks say as gospel truth - I must type in exactly what you say, exactly the way you say it, and when you say to do it that way -- or risk disaster. Thanks for your patience.
Also attached, a screen shot of me messing around with one of Jen's Iray nature shaders. Heh heh. Here I've applied it to one of Maclean's primitives... the shader seems to have some sort of NULL SPACE or something built in to it, or at any rate this is what I got when I flattened out the resulting TIFF file. Very cool!