[fuso]

Hi all,

After having searched for a bit I couldn't find anything decent but I remember there was a thread not too long ago. So what's
the latest on those two materials? I need them to render fast and noise free (optimised for version 2.5.1) for an interior project.
They don't have to be complicated with scratch maps, etc. just a plain material would do.

Thanks guys

[Half Life]

Probably easiest to just use the metal wizard for something like that.

Best,
Jason.

[fuso]

Hi Jason,

Yeah that's what I did but I have the feeling the presets are not quite optimized for 2.5.1 as generate a hell of of a noise.
For now I'm just using a single layered BSDF with default colour, roughness 0 and ND 20. Can I go wrong with that?

Cheers,

Jost

[Half Life]

Hmm, I set up a wizard chromium at 0% roughness -- I'm not seeing any noise... are you seeing this in Fire or the Core Render app?

The only issue with that setup would be that you will likely see a slight darkening around the edges that would be less than realistic with metal objects... the Nd and K values of the wizard presets are basically the same as using a complex IOR but without dispersion.

For standard glass mirrors all bets are off since we can't setup a real mirror material directly in MXED anyway.

Best,
Jason.

[fuso]

Hi again,

I did another test render of those two interior shots which I had mentioned in my original post above and even though I used
chrome and glass materials from the wizard, I still get that annoying noise in areas which really don't make any sense to me.
The scene setup is fairly simple. No sky or sun, just a few emitters in two seperate 'sealed' washrooms. No strong bump maps,
no complex materials, no IOR files, lowest possible emitter geometry. The noise appears in the mirror at the opposite end of the
room on the floor as you can see in the images. Floor material is fairly simple as well, two BSDF layers, diffuse and specular, no
bump map, roughness layer1 = 95, layer2= 35, both ND 3.0, no fresnel... I wish I'd know what is not working here!





rendertime on single nodes: 40h, SL 20, benchmark: 135.95, resolution 2000x1500!

Please don't tell me I have to go up to SL 35 or so, as I'd then say that maxwell hasn't become faster or has improved after all.

Cheers,

J

[Half Life]

If you could share the scene/materials via PM I could take a look for you.

Best,
Jason.

[fuso]

Hey Jason,

Thanks for kindly offering your help once more. What's the best way of sending it over? Open mxs in Studio > pack 'n go? I can
zip it all up and send it to you, if you'd provide me with an e-mail address via pm, please? I can't upload from here at the
moment, hope that's no prob?

Cheers,

J

[Half Life]

Yep, pack and go zipped up and I'll PM you.

Best,
Jason.

[kami]

that's really a weird noise ... because everything else is quite clean.

[brodie_geers]

That does seem unusual.

Your file will probably be pretty big. Here's a great site I found the other day. Free site, no downloading required, no size limit, and the file can be downloaded while it's being uploaded (you don't have to wait an hour until it's done uploading it before people can download it). http://ge.tt/

-Brodie

[fuso]

Afternoon all,

I have another question regarding this particular scene shown in the images above. So far the lighting levels are pretty much
done by eye just so that it looks good. But of course we have a lighting designer on board who is now asking for wattages or
lumen. At this stage my setup is fairly simple (working in 3DsMax). For the spotlights I have modeled a simple aluminium casing
with a reflector (using a simple chrome material) and I copied them using instances. The actual lightsource is a tiny sphere with
one emitter mxm attached to all of them but not using instances (I believe that this is still the way to do it?). Those don't cast
or receive shadows but are visible to camera and reflections, of course. Also they are not hidden to global illumination either.

Now, I'm a bit worried to start using IES data for those as I believe the render times and noise levels will increase dramatically.
If I however decide to give it a shot, is it still the case that I have to create a new emitter material (with the same IES) for
EVERY individual spot light?? I have tried it briefly last week using ONE emitter material with IES for all spotlights and they were
hardly visible, so I gave up on the idea... Am I right to assume that there's no way to override the intensity of those using a
common IES emitter? It's just a bit confusing as the emitter preview appears properly displayed in my viewport.

Thanks again for any advise regarding this.

[Half Life]

IES can be modified either in the IES file itself or in normal multilight -- I would not worry about an increase of noise as it may very well decrease the noise because you will not need to model the fixture (all that data is built into the IES files lighting profile).

I've looked at your scene and made a number of material changes and am doing a test render now... there are many places where I believe the material attributes can be tweaked for better results and I can clearly see that most of the materials are made in the old style (Maxwell 1.X) and can benefit greatly from proper use of Nd, Force Fresnel, and additive layers (for specular effects).

The material I see as being the most problematic is the "Aluminum Anodized" bits which I replaced with one of my RAL Anodized Aluminum materials.

If the test render performs as expected I'll send the file back to you so you can look over the changes.

Best,
Jason.

[brodie_geers]

The only reason you'd want to use a different IES for each light would be to get different ML sliders for each light (same as regular emitters). But be careful in tweaking your IES lights via the .ies data or by ML. If you're using a 100w light but you've tweaked it in either area, you can no longer tell your lighting guy that it represents a 100w light. Instead you should first make sure that your camera settings are realistic. If you put in real camera settings and real IES settings and your scale is accurate but it's still too dark, it may indicate that you need to have more lights designed into the space.

-Brodie

[fuso]

Hi Jason/Brodie,

Thanks so much for the info so far. I admit that there are a few materials which I sort of kept carrying over from file to file.
But as for most of them I thought I had opened them in mxed and saved them over in order to rebuild or update them. Maybe
I should have used the wizard and tweak/replace textures instead. Did you also check the camera's? I used a shutter speed
of 25, f-stop 5.6 and thought it is realistic for an interior shot using only artificial lighting...? How about the scale of the whole
scene? I am still unsure if the translation via plugin is correct as I work in metric (meter) units in 3DsMax.

As for the use of IES files I'm running a test as we speak, with interesting results. The first one I did, I've created an emitter
with IES using it on only one spotlight, then copied that one as well as the assigned IES emitter mxm and assigned them to all
the copies individually. The result was the same as last week, they hardly showed up but then I could modify them using ML.
In the second one I used that same IES emitter but assigned it to ALL the spotlights - with the same result after tweaking it.
It seems I won't be able to pass on detailed information to the lighting designer after all? That would be pretty bad news.

Jason, you said there's no need to model the casing. Well, I need to see something in the rendering, how is that gonna work??
Oh, and how do I modify the actual IES file?

Thanks a million again guys.

[Half Life]

I am not familiar with the Max plugin or how much direct control it gives you over materials (vs. MXED) -- but in many of your materials I see a specular BSDF blended with a color/texture BSDF... there are several problems with this:

1) the material may not have proper Nd/Fresnel falloff as a result of this approach, which will impact realism and may contribute to a "CG" feel.

2) The material will often become slightly grey and lose the full impact of the texture images you are using.

For a "specular" effect I much prefer to place an additive layer above the "diffuse" -- it is important that most shiny or reflective materials (such as a additive specular layer) have the proper Nd with Force Fresnel enabled for maximum realism. If you are using a "specular" map it is usually better to apply this as a layer opacity mask than in the reflectance 0 slot.

As I say you will see this in the file I send back -- and I do cover these topics in depth in the free tutorials on the THINK! site if you want to know more about my reasons. http://think.maxwellrender.com/tutorial ... thor=73948

I did check the camera and I though the EV was a bit too high for this type of scene. Maybe you might have been able to reduce the EV and then be able to reduce the intensity of your lights... There are some areas that are getting "blown-out" as a result of the light intensity -- But that's a alot of tweaking and I wanted to set the materials straight first.

IES files have a multiplier value in them (it's just a text file) which can be manually adjusted to save from dealing with multilight.

It's been a while since I've tried IES arrays but I seem to remember that each sphere needs to be it's own geometry to work properly... I'll test that later. Also, make sure the sphere geometry is very small for the IES material or you may not see all the light.

Best,
Jason.