[djflod]

great and very informative video jason ! thank you a lot for that workflow

teaching simply runs through your veins, huh ?

[Half Life]

One thing I forgot to mention in the video -- the wood veneers/floor sets are different in that their specularity should come from a clear-coat finish of some type... so make sure you lower the Nd of the specular layer to between 1.491 (acrylic) and 1.768 (aluminum oxide).

Also, for the record, I've already got the MXM automation pretty well covered for all the veneers sets -- so that basically just leaves the "stonework" and "tiles" sets where I've only got partial automation... which isn't so bad. Those veneers MXM sets should be done and posted fairly quickly.

You know, the funny thing is I never sought out being a instructor -- I just kind of fell into it, but I have found that doing these tutorial videos really suits my temperament. One of my local friends teaches at 2 schools (medical anatomy at both) and, while we do trade notes, I would never want his job in a million years... way too structured.

Best,
Jason.

[kami]

hi jason
sorry for not keeping up, i've had too much to do during the last days. we got mxms for concrete, edition1, stonework, tiles, wood floors, wood veneers 1
just say, if you need any of them. I think they are a bit differently built up than you showed in your tutorial (especially the additive layer differs a bit - I did not use any texture on the diffuse there)
best
christoph

[jespi]

Thanks a lot for the video Jason!. Now I fully understand why I was having those issues regarding that strange noise. You're an invaluable support for maxwell's community.

Have a nice day,

josé

[Nova66]

OK here's the video I promised
...
http://youtu.be/mvOZ1BOA3kw

This stuff is fantastic Jason!!!
I can't tell you how much you are influencing my workflow


Andrew.

[gotoxy av-media]

thanks a lot jason for this very helpful tutorial video!

munch

[Mihai]

Hey Jason, it's great you're doing this library to quickly use the Arroway textures. I watched the video and thought of a few things:

- when they say gloss, don't they mean the glossy parameter in biased renders meaning how sharp/specular the surface is? So a gloss of 10 would mean a pretty diffuse surface so higher roughness for us.

- I would lower the default ND of 3 when using force fresnel on the specular layer as this would make the surface equally reflective from all viewing angles and perhaps with a wooden floor it's not what I would want generally speaking.

- regarding the point above, in some cases I would want stronger speculars and not have them look 'faded' like they can look by lowering the layer weight (which just lowers the opacity of everything, including the strongest speculars which I don't want to look faded). Instead I would use a coating and control the reflections with both a map and the nd (I would still get strong reflections at the 90 degree viewing angle). With coatings I can also play with coloration effects for old laquered look.

[Half Life]

The reason I settled on a Nd of 3 for the wood in that example is because the most specular parts are actually metal nails in it -- there is no perfectly realistic way to have settings over a whole line like this so I was looking for a way of getting the most consistent overall look for the texture line as a whole rather than any individual texture.

Of course individual textures can be made into whatever the end-user desires -- the point is to get them into a good "base-state" so that users who are not familiar with MXM parameters have a highly useful material.

Coatings need some serious modification (as I have suggested many times) before I would use them for anything like this since they cannot have mapped roughness or mapped opacity, they are pretty well useless in all but a few cases.

Try my settings over an Arroway line and then you can try to come up with a better formula that can be applied uniformly line-wide, if you can succeed then I will adopt it as my new standard. I never said these were infallible, just the best combination I have found after weeks of trying every possible variation.

Best,
Jason.

[Mihai]

I would really not discard coatings as useless. I prefer doing plastics with them instead of additives. Try it, it makes the plastic look much more alive, than fading the speculars with layer weights. You can still make the layer the coating is on with opacity as usual, but better use the ND (with FF on) to control the reflectivity amount first of all, then layer weights as a second option. You can also apply a bump just to the bsdf that has the coating component and it will affect the coating also (and also global bump affects the coating). So it's still very flexible. It would be more practicle for coatings to have a roughness parameter but there are easy workarounds. One is to create a generic noise map in PS and use that as 'roughness' (put it in the bump on the bsdf that has the coating).

Just mentioning this to make people understand that you have a base option for starting to work with these textures (or just use the plastic wizard and hook in the textures) but settings Arroway suggests there, I don't think we need to be so strict especially as it's a base starting point - it makes you develop a boxed in approach and calling it finished, when there are many other approaches you forget about eventually, or don't learn.

[Mihai]

The reason I'm against finding these systems and values, is you think these become quasi universal, that if you don't do this as a base, you have a wrong starting point. When instead you should be learning simply what the parameters do - then at least you have a reason for doing something and you know it. Take for example this "rule" about adding a brighter version of the texture in the refl90. That has a purpose, but not for this case. You have become so formatted to set them up this way that you forgot an important thing, and I'm pretty sure you are aware of it too. With the settings you used in the video it will have absolutely no influence. Because at roughness 95, unless you turn on the r2 parameter, you can change the refl90 to your hearts content - it will have zero influence. Or at least, such an infinitesimal influence that you can just ignore it. If you DO also turn in the r2 parameter, then it will have an influence and it pays to read what it's ment to do, why it's there and ONLY then have a reason for adding it.

To make this visual, here is a screenshot of the refl90 set to bright red, and also the relf0 texture activated in refl90 and brightness increased. Absolutely no difference.




Just another simple example with a coating (although I would have made a second BSDF instead and opacity mapped it to have better control of where the coating appears). This is just an example though, and I'm starting to prefer it to additives because it keeps me thinking more in real world and I have much finer and accurate control of reflections with the ND of the coating - meaning even in that very faint reflective example I still get a good increase in reflectivity controlled by fresnel alone which is what I guess happens in the real world. Not an additive mix-up of layer weights, map contrast etc...Mainly I like to keep things as simple as possible

[Half Life]

Good points, and I do use coatings I used them on my RAL Shiny Plastic and RAL Car Paint lines instead of using additive layers for the very reasons you outline.

Actually my very first post to the forums was a question regarding why coatings are not the solution we want them to, here: http://www.maxwellrender.com/forum/view ... 97&t=34365 so this is a topic I've been thinking about and talking about since the very beginning.

Using a high-frequency bump for roughness introduces poly-faceting in many cases... so it is not the optimal solution. If you care to know, I can give you the full rundown on what I have suggested to others at Next Limit for increasing the utility of Coatings and decreasing the need for "Additive" layer mode (or eliminating it altogether)-- I'm actually probably the most pro-coatings user on these forums, but I'm also realistic about what the current implementation can do.

I'm very far from someone who has "boxed in" thinking on the subject of materials -- for example this recent thread here: http://www.maxwellrender.com/forum/view ... 74#p353774 I made 4 different looks with the same Arroway textures (2 of which you can download), just to show it can be done with ease. I think you are making assumptions about what I do and why I do it that are unfounded in any facts -- I make great effort to be as thorough as possible when designing materials for others to use.

The wizard materials by comparison of very often a disaster, and most of my RAL lines were created to be a direct alternative/solution to the wizard equivalents.

Best,
Jason.

[Half Life]

The brighter version of a texture in Reflectance 90 is useful once you drop the roughness even a little bit -- and I like to set up materials so I can make changes to settings (like roughness) without having to readjust my reflectance mapping -- it hurts absolutely nothing to put it in there and it only takes a few seconds to idiot-proof the material for future modifications.

I didn't realize I have to deeply explain every one of my actions in a video I clearly said was not meant as a tutorial -- if you want to know the full reasons for everything watch my tutorials here: http://support.nextlimit.com/display/tu ... e+from+VTC

I was asked to explain how I convert the Arroway materials that's all -- nothing more, if you read anything more into the video that is your own fault.

[Half Life]

Oh and by the way, reducing the Nd of your coatings is a bullshit solution and you know it... in the real-world a metal nail has an Nd, the wood has an Nd and any type of coating has an Nd... these are real things and as you deviate away from those real world values you become less and less realistic.

Your examples are much more misleading than anything I could say.

[Bubbaloo]

If you have a material that is supposed to simulate multiple surface types (wood and metal), you should create proper masks so that you can control the nd separately per layer. That's a benefit of a layered material system.

[Half Life]

That would be the province of Arroway (who makes the textures), not me -- I have several thousand of these I am doing and I have to optimize the process or I would never get done.