Strange color points

Feature requests, bug reports and related discussion
Sinister
Posts: 29
Joined: Mon Apr 14, 2008 3:35 am
Location: Ukraine / Kiev
Contact:

Post by Sinister » Fri Nov 28, 2008 8:11 am

eeem sorry for nooby question (as I am noob) but how to reduce brightness? I use mesh light as emitter, lambert material that emits has (0.5;0.5;0.5) RGB, same with incadence, changing glow intensity do not change anything (as I understood, it is relative value) floor also default lambert..

the one idea I have is to put other emitter whith near the same glow intensiti that directer out of scene to overpower existing..but it should slower indigo

Sinister
Posts: 29
Joined: Mon Apr 14, 2008 3:35 am
Location: Ukraine / Kiev
Contact:

Post by Sinister » Fri Nov 28, 2008 8:36 am

render with nearly black floor..
Image

User avatar
Kram1032
Posts: 6649
Joined: Tue Jan 23, 2007 3:55 am
Location: Austria near Vienna

Post by Kram1032 » Fri Nov 28, 2008 8:38 am

not the emitter's brightness, but the floor-material's!

As long as you're using reinhard tonemapping, the emitterbrightness seemingly wont change at all (indeed, it actually does but reinhard changes, based on the brightest point in the scene, so, you wont see it, usually)

Though, that's anyway not what you want.
The (probably diffuse) plane looks like perfect white. tone that one down to 0.8

(Even if this doesn't work, in future, don't use any value over 0.8 - It'll enhance realism a bit and it will speed things up a lot)


EDIT:
Oh, ok, while I was posting, you where as well xD
It was worth a try :)
Cool glow xD
Well, then, you need to tweak more advanced settings like caughy b, MNCR, RBD... maybe even RND? (How big is the gap between the glass and the floor? are they exactly touching, intersecting or is there a gap? THAT might also have an impact.... Though, it's rather unprobable, I'd say...)
Last edited by Kram1032 on Fri Nov 28, 2008 8:53 am, edited 1 time in total.

User avatar
eulgrand
Posts: 66
Joined: Wed May 09, 2007 1:01 am
Location: Rouen, France

Post by eulgrand » Fri Nov 28, 2008 8:47 am

Psychedelic !

:wink:

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Post by CTZn » Fri Nov 28, 2008 9:11 am

Yes indeed... Presuposedly scene scale is correct, isn't it ?

If you want to have control on lighting and exposure, quit both rgb lights and reinhard tonemapping, as Kram said.

Use blackbody lights, and linear or camera tonemapping. To switch the emissive lambert to blackbody, use a value above 1000 for it's incandescence R channel, or set a precise K° value ie 4300, 5500...

With rgb ranging from 0 to 1 AND reinhard tonemapping, you get very distorded exposure range. Time to switch to serious settings ;)

More seriously, many MtI improvements are on hold until we migrated the whole engine toward new shaders definitions, wich does not come easily as far as I'm concerned :) Amongst them an interface, the Indigo Palette, wich exposes only Indigo relevant attributes to make things much easier.

PS: the glow setting of Maya materials is not linked against any Indigo value.

Sinister
Posts: 29
Joined: Mon Apr 14, 2008 3:35 am
Location: Ukraine / Kiev
Contact:

Post by Sinister » Fri Nov 28, 2008 9:42 am

hmm
advanced settings like caughy b, MNCR, RBD... maybe even RND
where is this settings? and how to swich between rgb and temperature? maybe I should read some manual first ? :)

I have found a flag "use Adv Opt" and it gives a good resuld, but still buggy
maybe I could tweak it? (aaahh it much more slower..)

Scale is normal.
the gap between the glass and the floor...yep, there was no gap..

User avatar
Kram1032
Posts: 6649
Joined: Tue Jan 23, 2007 3:55 am
Location: Austria near Vienna

Post by Kram1032 » Fri Nov 28, 2008 10:33 am

cauchy b is a material setting, which actually should be stated as such in your material's definition, I guess...

MNCR = Max Num Consec Rejections
RBD = Ray Bounce Depth
RND = Ray Nudge Distance

all of those are changing the way of rendering, directly. MNCR usually doesn't need any change, except for seldom problems with glass and caustics...
RBD might improve renders with complicated light-situations but usually, the default should be fine.
RND should be fine, when all your scale is correct. This is the most critical of the three settings from above. Do NOT touch it, except your scene goes either extremely macro or extremely huge. So, it's not an option anyway, right now...

As those options are highly advanced and the ini-file needed for a cleanup, Ono cleaned all those values out. Though, you can either write them into the ini-file again (in that case, you should remember the settings though...) or you can write them into the igs-files for your scenes, which also works nicely...

___
Damn, didn't change those settings for SO long, anymore that I don't know, wether it's alla
MaxNumConsecRejections
or
Max_Num_Consec_Rejections

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Post by CTZn » Fri Nov 28, 2008 10:42 am

Oh my... Sinister, did you find any sort of documentation in the MtI zip file ? argh... that was lost somewhere I guess :oops:

To switch spectrum between types:

- Click on the color swatch of the incandescence attribute to open the Color Picker
- Switch from HSV to RGB

Now, if:

0 < incandescenceR <= 1
Spectrum is RGB, from 0 to 1 (that is restrictive, ideally any positive value should fit).

250 ~< incandescenceR ~<= 800
Spectrum is peak, with:
incandescenceR: <peak_min>
incandescenceG: <peak_width>
incandescenceB: <base_value>
ambientR: <peak_value>

incandescenceR < 1000
Spectrum is blackbody, with
incandescenceR: <temperature>
incandescenceG: <gain>
ambientG: <power_drawn>
ambientB: <overall_luminous_efficacy>

User avatar
Kram1032
Posts: 6649
Joined: Tue Jan 23, 2007 3:55 am
Location: Austria near Vienna

Post by Kram1032 » Fri Nov 28, 2008 10:51 am

Interesting way of doing it O.o

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Post by CTZn » Fri Nov 28, 2008 10:58 am

Rough way.

Pending is this tool (check the video link), until new shaders defs are up and running (me too small to be fast).

One window to rule them all :twisted:

User avatar
rgigante
Posts: 326
Joined: Wed Jun 28, 2006 6:46 am
Location: Italy

Post by rgigante » Fri Nov 28, 2008 7:22 pm

hmmm... it's quite known to me that effect. I've a similar scene with those colored spots, and they are fireflies. If you try to render the same scene with 1.0.8 (if i remember well) they should disappear.

Here's my 2 cents

Sinister
Posts: 29
Joined: Mon Apr 14, 2008 3:35 am
Location: Ukraine / Kiev
Contact:

Post by Sinister » Sat Nov 29, 2008 2:30 am

There was no any text file at all in MtI 1.0.95_beta3 zip .. even installation documenation :?
I'll download some other version now to read..
thanks you guys

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Post by CTZn » Sat Nov 29, 2008 8:39 am

There is only an embryo of a doc yet, so feel free to ask (in the Maya forum please) for any hint you may need to compensate the lack of documentation, actually.

Note: as specified in the Indigo manual (I think), <gain> in one hand, and <power_drawn> & <overall_luminous_efficacy> in the other are exclusive. Set the one or the other two to zero to disable them in Maya.

User avatar
dougal2
Developer
Posts: 2532
Joined: Wed Nov 15, 2006 8:17 am
Location: South London

Post by dougal2 » Sat Nov 29, 2008 9:04 am

Look also in the wiki for some MtI docs that should still be mostly valid.
EDIT: Ignore that, the wiki appears to no longer exist. Damn I spent a long bloody time writing that document! :(

Big Fan
Posts: 745
Joined: Tue Oct 17, 2006 9:37 am
Location: Nelson NZ

Post by Big Fan » Sat Nov 29, 2008 5:09 pm

yeah well you are not alone in that :evil: hopefully wikis come back soon...ono? :?

Post Reply
61 posts