[fv]

I was reading the tip of the day 12 on tod12, by Jason (half life),
I got a great result out of it, nice leafs, better than I had before. TX. But I am beginning to feel uncomfortable with Maxwellrender and its strategy towards materials.

I was trying to explain this to my collegue at work and I realized how complicated texturing with Maxwell is getting to be. Why really can't sss not be used in a single layer for instance. Do we really know what we are doing or are we exchanging the outcome of endless trial and error. Which is good in itself but is there not a better way, more logical way to set up materials.

Francois

[Half Life]

That argument holds alot of weight -- and I would agree, but people have been happily using AGS without question for years... to me that means that most Maxwell users are results oriented and not really concerned about how much things make sense (in relation to how things are in reality).

For the record, I dislike both AGS and ThinSSS because they seem non-realistic to me -- I think real dielectrics and SSS with proper volume are how things should operate, because they makes sense in a "real world" way. However going back to the results-oriented concept -- you cannot deny their utility given all the limitations we currently face in both modeling power and processing power.

I'm a big advocate for more powerful material systems because I think they free the artist from the tyranny of the polygon (which is unrealistic in itself) -- but they should be made as logical and "real-life" as possible. Particles are the way to go in my opinion...

An example of why polygons are a bad thing is the concept of "normal smoothing" -- no such animal exists in nature and it leads to unrealistic reflective surfaces more often than not... at some point in time polygons will have to be left behind to create true realism.

Voxels make much more sense to me, but they are still underdeveloped as a modeling tool.

Best,
Jason.

[Half Life]

I re-read my previous post and realized it is a bit fragmented -- maybe not so easy to understand what I'm getting at.

ThinSSS and AGS both work well for single thickness surfaces -- on an infinitely thin polygon... which is great for ease of use. However no infinitely thin plane exists in reality to my knowledge so the very foundation that we are developing on is flawed.

Then you add into the equation compromises for poly count like smoothing angle which approximates a smooth surface where none exists and we are working with a foundation for our materials that bears hardly any relationship to how things actually are in reality.

So you could say the materials are unrealistically made, but the reality is they are conforming the pre-existing rules of the substrate they are applied to... therefor they can by definition never really be realistically designed.

I find it odd that we work in a 3 dimensional medium with ultra 2D parts (polygons) -- which is why I think Voxels make a bit more sense... they have 3 dimensions themselves and therefor are already that much more 3D.

I think if you look at particle simulation (like Realflow) you will see the future of the future in 3D -- after all everything is just made of particles in the real world.

Unfortunately I see the reign of the polygon as being very strong and so my hope for the immediate future is that more powerful material systems will lead the way into our particle based future.

One example I can envision is a much more powerful displacement system that would work like the thickness map in ThinSSS but by generating actual volume at render time.

A dream now, but technology does not sit still... so maybe soon.

Best,
Jason.

[Half Life]

I finished my deadline yesterday for the Google Sketchup to Maxwell Render Workflow video series and I'm taking today off to recuperate.

Starting tommorow I'll start updating the MXM tip of the day thread again -- I'll have plenty of time as I am effectively unemployed... anybody else who wants to participate is more than welcome.

Best,
Jason.

[fv]

Hi jason,
I did not respond to your comments on my reply since I agree with what you write.
I hope the tip of the day thread does not suffer from these insights since by itself its a great help to resources and info.
tx, Francois

[kami]

hi
I got a problem with the custom preview scenes. most of them render only black or pure white.
what could be the reason for this? (i'm using the rhino plugin)
cheers, kami

[brodie_geers]

Hard telling. Has anyone else out there had this problem?

Can you tell me what your process was? What I would do is copy your current Previews folder in case you ever need it again, then delete everything from the original and extract the contents of the .zip folder into that Previews folder. If the previews aren't working at that point, you might open up some of the .mxs files and try running a render from within there and see if you get any errors that pop up.

-Brodie

[kami]

Hmm ... It's weird.
I tried opening a scene (Floor Distant) and rendering it, then I got an error message concerning the write path. So I changed it (was "C:\Program Files\Next Limit\Maxwell 2default.mxi" to a relative path like it is in the scenes that are working (eg. defaultpreview = the simball). I tried to render the scene and it turned out good. But when I loaded the scene in Rhino, it was rendering all white. (some are all black and sometimes the camera seems just wrong ...)
I tried to load the scenes as preview in studio and the are all working fine, so I guess this problem belongs to the rhino section.

[brodie_geers]

Only thing I can think of is that native MW apps seem to be better at locating textures. I've noticed that in plugins, if a texture path isn't just right, it tends to give up and cause issues.

-brodie

[lord.saddler]

Great tips here, big thanks to all the posters!


I have a question regarding Jason´s Tip #3, or to be more specific the Gold Nd preset he is using:

It has a Nd of 0,290. Most lists online show a value below zero (look here for example). Wouldn´t this imply gold is a meta-material with an Nd lower than vacuum? Or does the K value "correct" this?

[Half Life]

Hello, thanks -- I've had some unforeseen issue that have kept me from updating the tip of the day thread but I'm glad you found it useful.

The Gold is just the Metal Wizard Gold and it is based on complex IOR measurements as you can see here:

http://refractiveindex.info/?group=METALS&material=Gold - Input .589 as the wavelength.

The Extinction coefficient adjusts the Nd sufficently upwards to compensate for low Nd... we need more measurements of material like these for better realism

The advantage of the wizard Gold over complex IOR Gold is you have no dispersion calculations and you can easily adjust to suit if you want something else based on golds reflectivity. ( for instance I used this as the base for my RAL Anodized Aluminum line)

Best,
Jason.