Exponential Color Mapping

[YoyoBoy]

When/why does setting the Color Mapping to either of the Exponential choices make my renders look like I reduced the contrast in PhotoShop? There have been times when this doesn't happen but I don't remember what my other settings were at the time.

Thanks.

[natty]

yeah this happens to me aswell .....
looks slightly greyed out ????

[rhino]

Exponential sucks IMHO, it's a quick fix for inadequate light distribution. Vray suffers from a short light throw, it's like working in 256 colours...well maybe not that bad. Photon Mapping distribution is far better than Irradiance Mapping but has it's share of problems/limitations that make it suitable only for limited scenes. Secondary falloff need to be far more linear, or support higher multipliers in the time being. The fact I need to use skylight portals or secondary lighting to get adequate lighting leads me to think the falloff formula is too high or just too blunt.

[losbellos]

I think also that would be much better if secondary multiplier could be higher than 1. Its a limitation.

[sunny]

I may be completely incorrect in my understanding of Exponential/HSV but here's how I perceive it to work:

Exponential essentially increases your "dynamic range" so to speak of your image. The purpose of this si that you can increase the lighting a lot further before your white coloured objects start blowing out.

At least that's how I use it. Seems to work for me much better than Linear.

Sunny.

[losbellos]

I'll try this, but what about the colors? Will see

[sunny]

what do you mean about the colours?

Use HSV instead of exponential.

SunnyC

[flipside]

The fact I need to use skylight portals or secondary lighting to get adequate lighting leads me to think the falloff formula is too high or just too blunt.

I think that too, even with big windows light doesn't travel far enough. It would be nice if we could set another formula ourselves for the falloff of bouced light. I have the same problem with vray lights, they also have such a strong falloff.

[vlado]

I think that too, even with big windows light doesn't travel far enough. It would be nice if we could set another formula ourselves for the falloff of bouced light. I have the same problem with vray lights, they also have such a strong falloff.

VRay lights have inverse square falloff by default - exactly what you get from real-world lights. You can override this by turning on the "No decay" option, in which case they will behave like standard Max lights (which are not physically correct by default).

The Exponential color mapping tends to wash out colors; basically this is the way it should work - it conveys the feeling of stronger light by pushing colors towards white (thus saturating them).

The HSV color mapping overcomes this by keeping the saturation of colors, and only limiting their "value".

Best regards,
Vlado

[metroberlin]

yes, but sometimes the the HSV looks very strange. don´t know exactly why, just a feeling. i normaly use linear for indoor when there´s no dirct light. with direct light i use the exponential maqpping with a higher multiplier for die direct light.

btw. there´s a german football club with the name HSV (Hamburger Sport Verein) )

[flipside]

I think that too, even with big windows light doesn't travel far enough. It would be nice if we could set another formula ourselves for the falloff of bouced light. I have the same problem with vray lights, they also have such a strong falloff.

VRay lights have inverse square falloff by default - exactly what you get from real-world lights. You can override this by turning on the "No decay" option, in which case they will behave like standard Max lights (which are not physically correct by default).

The Exponential color mapping tends to wash out colors; basically this is the way it should work - it conveys the feeling of stronger light by pushing colors towards white (thus saturating them).

The HSV color mapping overcomes this by keeping the saturation of colors, and only limiting their "value".

Best regards,
Vlado

But still, physically correct or not, something in between inverse square and std no decay would be handy sometimes.

[taboo]

are we going to be able to play with all of these color mappings, interactively??? (After the render?)

Maybe in 1.1?

[tom schuelke]

Hi

for me it seems, the hsv exponential for my purposes also is best. in Realworld for example, sunlinght is much much stronger than interior lights. using the linear , the white areas always seems to be blown out. but could the problem of the light which seems not to travel far enough be a material problem. ?

i think it would be easier, not to use the gi multiplier of objekts and tell multiply it to 2 ore 3, ( what ever this means.) in realworld, lightplaners use a percent number of how many of the incoming lihgt always gets reflectet. for example white diffuse collors reflect usualy about 80 % back to the scene, wood about 60 %, dark stones about 30 to 40 %.

i think it would be much more comfortable, to make it such way.
So i think the problem of light not traveling far enough could be a material ore object settings problem. (what would be also nice if you could control the amount of color bleeding, a material produces). and also better to control this in the material and not in the object settings.

(one of the few things the standart max5 radiosity seems to make eysier than Vray i think )

[Yadla]

I experimented this exponential mapping today, and if my first impression was to ask myself what is that mapping useful to. I had a good surprise when i tweaked lights and GI emmitting intensities !
I could achieve a more realistic rendering with exponential.
(sorry i dont have picts, i didnt save my test project)

[YoyoBoy]

>> even with big windows light doesn't travel far enough <<

I'm starting to get some good results by jacking up the Global Photon Map's Multiplier.

It's increasing the render time alot though