Sun and Detail Enhancement

[dalomar_cg]

Re: Sun and Detail Enhancement

@Renee
In order to adjust the colors it is advised to use the Acolor type in the texture editor. Simply set you color in the Acolor Map as you would normally, then decrease the multiplier (underneath Type, not the Alpha value) and you should be able to get the colors to work. With textures simply decrease the multiplier and everything should be good to go. Sorry to lead you down the road with no helpers, but I wanted to make sure that you watched the tutorial first. Also, are you using the vray sky for your environment/bkgd or an HDRI?

@Micha
I know it isn't a secret that you wouldn't use the method, but I would encourage you to maybe take a new look at the tutorial now that we have the Physical Sun/sky/cam. Lets let users determine whether they want to use it for themselves. If you want to discuss the "side affects" then we can do that via email or something

[Micha_cg]

Re: Sun and Detail Enhancement

Good idea.

PS: Next week two current projects should be finished, than I could do more experiments again.

[merlik585]

Re: Sun and Detail Enhancement

Damian,

I use the vray sky.

It is the reflective background that I can get to reflect.The sky I don't have any problem.

I did test the RGB at .255,and frankly I can't see any differences in the 2 renderings.

Renee

[dalomar_cg]

Re: Sun and Detail Enhancement

I'll have to look back at some of my notes for using that workflow with vfr...You should still get reflections to be fine. Is the background at least looking how it should?

[merlik585]

Re: Sun and Detail Enhancement

I will try to explain a little bit better.

1)Doing a rendering,with vray sun and sky,looks perfect the sky reflect in the glass,no problem.Except all I have in the glass is a gradient with no clouds or buildings etc.

And I want an environment that will reflect,so I enable reflection background and set up a map correctly,and hit the render button,what I get is perfect lighting,but the background is pitch black and the glass of course because they are reflecting the black background.

I tried to disable the sky but keep the sun and the reflective background,same problem.

I must do something wrong,but I can't figure out what.Thanks for your time.
It is a long weekend here so I will try and find a solution.

Renee

[dalomar_cg]

Re: Sun and Detail Enhancement

Okay...I think I've got it...remember in the tutorial where he first explained how to increase the intensity of the background map (I think he was using an Output map, which of course doesn't exist in rhino). He was saying that because of the differences between the intensity of normal hdrs and the Vray sky that the output of the hdrs will have to be increased. In Rhino simply increase the multiplier of the hdr in the texture editor. For the example below I used a multiplier of 15, but it will need to adjusted for your scene/desired results. Let me know if that works for you.

Test with HDR @ regular level


Test with multiplied HDR


I know these aren't the prettiest examples, but it demonstrates what I'm talking about. I used HDRs for these examples, and thats why there is atleast some brightness in the first render. If it was a simple jpg or png, then the reflections/refractions would have been completely black. I still thing the reflection should be better once the Correct RGB bug is fixed

[merlik585]

Re: Sun and Detail Enhancement

Thanks Damien

Got it to work,it really depends on how bright the picture is,I have a normal png that I had to boost up to 75 to get the reflections I wanted.But it is working.

Thank You.

Renee

[dalomar_cg]

Re: Sun and Detail Enhancement

Yes, you'll find that it is extremely dependent on the intensity of the image. Even with HDRs there is alot of variance in the brightness from image to image. Thanks Renee for bringing this to my attention. It was something that I had previously overlooked, but is very important for being able to use this particular workflow effectively. So now that your using this method, is it working any better or worse for you ???

[merlik585]

Re: Sun and Detail Enhancement

well I am not working with the .255
I am not comfortable with that,may be I don't understand it very well.

I work with gamma 2.2 and so far it is working Ok for me.

Renee

[Micha_cg]

Re: Sun and Detail Enhancement

Some of the effects can be reached by other tricks - set independent intesities of background and GI environment, set a lower sun intensity (not working in the last build - bug). Usage of lower sec GI engine multiplier (0.25 is a good value to increase the GI shadow contrast of exterior renderings). For me the 0.255 manipulation is a to big engagement too and the named tricks are better to control.
We should not forget the question: what is the goal of the method and how can it achieve by other, simplier controls?

[merlik585]

Re: Sun and Detail Enhancement

For me the 0.255 manipulation is a to big engagement too and the named tricks are better to control.
We should not forget the question: what is the goal of the method and how can it achieve by other, simplier controls?

set independent intesities of background and GI environment,

This is what is not working for me.I will try your settings.

I feel the same way you do about this,I sure does it involve a lot of manipulations,

Renee

[merlik585]

Re: Sun and Detail Enhancement

set a lower sun intensity (not working in the last build - bug).

it work perfectly here (beta 2),go with the object property and you can lower the sun intensity.
just make sure that in the sun environment,you have the right sun chosen,and that the settings there don't override the setting in the object property box.

Renee

[dalomar_cg]

Re: Sun and Detail Enhancement

We should not forget the question: what is the goal of the method and how can it achieve by other, simplier controls?

The goal of the method is to make Vray behave in more of a physical manner, and this leads to using the physical sun and sky model (at the intensity that it is ment to be used at) and the physical camera. From Lele's findings it becomes necesarry to question the actual intensities of the materials themselves. Once you go through the tutorial and really think about what is being proposed, I think that you will find that Lele's conclusions of having darker materials is the "missing link" in the goal of achieving more physically based renders. I know that this tutorial is allot to digest and also a bit unsettling, and to tell you the truth, when I first watched it, it got me scratching my head and thinking that this is proposterous. But after watching the tutorial 15-20 times it does make sense, and things fall into place. It is hard after doing CG for x-number of years (insert your experience here) to change the mode of thinking from RGB (0-255) to thinking in terms of light intensities, because that what this tutorial is really challenging. If you think about it that way, then I don't think the .255 thing is really that crazy of an idea anymore.

As far as the answer to how to make this simpler to implement, I will propose this. First step(after adding the sun, groundplane and all that stuff) would be to expose your sky. So don't worry about how bright your white ground plane is(make sure you do have a white groundplane though) and just expose your scene till you are comfortable with the appearance of the sky. Do make sure that you have gamma correction enabled now as this will affect the appearance of the sky if you do it later on. Second is to decrease the multiplier on your white groundplane. All you have to do is read the float values (make sure your colors aren't clamped) and use that to determine how much to lower the intensity of your materials. Once you determine the factor needed to correct the white material you can use this for all you other materials. Now for every color use the Acolor map with the multiplier factor (not the alpha), and for every texture simply use the factor for the multiplier. You can still pick your colors and set your maps as you normally would, as long as you use the above method(Acolor and factor). If you want to use an HDR image now add the image to the environment or background and increase the multipler until the HDR is properly exposed (there is a way to calculate this, but it requires HDRshop). This is obviously a brief overview, but I think its a good punch list. Lele has a script to preform the material conversions in Max, and once vfr is scriptable then we will produce a script for this as well. The rhino version of the tutorial is in the works, and once it is fully together, I think it will clarify a number of issues as well as outline this in an even simpler format.

I remind you that Lele took alot of time to research and prepare the tutorial. He ran it by Vlado and it has been approved by ChaosGroup, so I would say that is method is not a shot in the dark. I am open for discussion about this method and by all means I am not an expert on this, but I feel that it is worth a deep consideration and if you want to contribute than do your best to understand and use the method as it is proposed.

[Micha_cg]

Re: Sun and Detail Enhancement

Damien, could you write down a conclusion of the effects of the 0.255 method? So we can talk about it.

My conclusion is:
(1) the relation between material color and background is changed, the background will look much brighter and all emitter materials are four times brighter, but the emitted light is the same like befor. The problem, that a visible emitter plane is very bright and show jagged seams will increase.
(2) diffuse bounced GI light get a higher attenuation
(3) If reflection colors are corrected too, than 100% reflective chrom materials dosn't work anymore.

I don't like effect (1) and Effect (2) can be achieved by GI multiplier. How handle Lele issue (3)?

@Renee: the sun multiplier change the intensity of sun and(!) sky. I missing to control the sun only.
But I found a hidden way: the texture multiplier of the environment dialog control the sky intensity only. So, if you need less sun intensity, than set texture multiplier at 4. Now, the sun stay the same, but the sky is four times brighter, a nice smooth sun effect can be achieved (like at soft autumn days). Only the general color mapping multiplier must be decreased to get the right image brightness.
Joe, please don't fix this "bug" or try to implement a true "only sun intensity" control please.

[dalomar_cg]

Re: Sun and Detail Enhancement

I'm not sure if this is a "conclusion" or not, but I will put it out there and hopefully this is what your looking for.

In my opinion the whole tutorial is challenging our typical way of thinking in terms of RGB values and shifting that towards thinking in illumination values. If you think about cg in this manner then materials are no longer dark or bright, but they are either properly exposed or over/under exposed. What the tutorial is pointing out is that when the Sun/sky system (which is based on research) is kept at its proper level and exposed via the physical camera, then the materials are no longer exposed. And if we expose for the materials, then our sky is dark as night. From this the idea of "darkening" the materials resulted. Keep in mind that when materials are prepared it is supposed to be as if they were prepared without any consideration of light at all. If that is the case then we cannot trust our eyes (I sound like Descartes) because our eyes obviously need light, and have the ability to expose for a very wide range of light.

I am just going to paraphrase some of Lele's conclusions as I agree with almost all of them.
1. With keeping the Sun intensity@1 and keeping the material color @255 white causes the intense light energy of the sun to be preserved and bounced in a manner that causes issues. Mainly this intense light energy is producing caustic like effects in places which contain allot of bounced light. Not only does the rendered result look horrible, but rendertimes are now significantly longer because Vray has to calculate all of that bounced light. I would say your trick of lowering the secondary bounces is a direct measure to counteract that, and this would also be working against the proposed set up in vray.

2. Now if you were to use the sun @1 and leave your materials @255 and expose for that, then you still have the problem of having a the sky being very dark. Logic would say that you should just increase the intensity of the sky and all will be good. Well, assuming your GI and Background are linked (which is the most physically correct way), then increasing the skys intensity will throw off the proposed ratio between the intensity of the Sun and the Sky. This of course is not very physical, and causes the influce of the sky's illumination (which is blue) to be out of balance with the illumination of the sun (which is yellowish). Now we have the issue of the colors of our materials not doing what we want them to do because of the of balance colors of our lighting.

I think thats all I have for Lele's conclusions at the moment, so now I will move on to reponding to the ones you had.

1. I think wether you consider Background the vray sky or an HDR. If you consider it an HDR there are two issues. First is that the Hdrs are prepared in much the same way as materials are, only with the addition of some dynamic range. I say some because I doubt some one would spend their time taking the 60-75 exposures per position to go from the true intensity of the sun (385 float) all the way through the visible spectrum, just to get a visibly pure white image that only is properly exposed via a physical camera in a rendering software. What I'm trying to say is that hdrs are still interms of how our eyes see them, so we have to treat Hdrs as something that is relative and not an absolute. The second issue is that the intensities in an HDR are not similar to the intensities of the vray sky, therefore the intensities of the HDR must be increased. I'm not sure why you feel that the background will look brighter considering how the method is showing how to bring the environment back into proper exposure from being under exposed.

As far as your emmitter materials being four times brighter, I would actually recommend to reconsider how you make emmitters. If you think about it, the colors/textures of the emitters should be set in the exact way that they are set for the materials (either with Acolor and the .255 multiplier or texture and the .255 multiplier) because if the aren't, then they are already significantly brighter than nonemitting materials even with a intensity of 1 due to the difference in color. Also if the colors and textures are not set the same way as the regular materials, then the emitter will respond to intensity multiplier in a kind of exponential manner. Using the same method for setting the color/texture is going to keep the result of the emmiter inline with the responses from other materials as well as keep the intensity values as they were ment to be.

2. I would say the having materials be at their normal values would cause a GI to have a higher attenuation. The .255 method is actually a way to manage this strong light intensity so that the strong intensity does not have an adverse affect on the bouncing light.

3. To be completely honest, I don't know what exactly is the treatment for reflections. I did a few quick tests and noticed that the reflections were acting seamingly to strong, so I went through an corrected the colors in the reflection slots and I can't say for sure what the proper treatment should be. I will go back and try to do more testing and look over the CG thread a few more times. I can't post on the CG thread, so maybe I will ask Corey/Joe to add something for me, or just have them tell someone @CG to let me post.

I don't like effect (1) and Effect (2) can be achieved by GI multiplier

If your refering to your trick of decreasing the secondary bounces, I don't think that that is a very good way to go about things. To me the lowering the secondary bounces is a much more unsettling idea than the .255 thing. I can believe that my eyes percieve things as brighter than they really are, but it seams like a bit of a stretch that after the initial bouce of light the intensity suddenly decreases by a 1/4.

I hope this answers afew concerns you had about this. I'm sure you'll have a few more thoughts about this, and I look foward to hearing them. I will say that this discussion is having me think about this method much more in depth manner, and for me at least its giving me a better understanding of it.