Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
Alexey,
Transparency and translucency at the same time?!! That sounds sweet! I will definitely look into this. The only concern I have so far is that from the examples provided it looks as if a viewer can see distinct images of the sky through the leaf itself. For translucency to operate as intended, light should pass through a leaf causing it to glow on the underside when light is above it; but no distinct images should be visible through the leaf itself. For typical translucency we would employ some degree of refraction, usually water's refraction of 133. When in blend transparent mode I think that refraction becomes disabled, but I will check on that too. Thanks so much for this insight! I always hoped there was a way to have the best of both worlds.
There are two considerations with Light Domes and the Specularity. First, as you observe is the the intensity of the specular output from the light sources, and the second is the Specular Halo setting of the material itself. When colors near white are used for the specular halo color, the discrete highlights blend in together nicely. My advice is somewhat the opposite of yours. You want indirect specular light, this becomes essential for human skin for example, to prevent it from looking too dry when not facing the sun, or from looking too wet when facing the sun.
Most anything with noticebly reflective properties usually benefits from some indirect specular, provided the halo settings are ideal and the quality setting of the domes are adequate. At such low quality of 16 or 25, the highlights will remain highly pronounced. But at higher settings it smooths out.
It is troubling however that there is the unknown rogue highlight from the dome. I suspect it is something going worng with the bias, as I rarely use bias settings as high as 50. Thanks for pointing this out. The Dome features took a lot of development time, more than many people wanted, so it was stopped slightly short of perfection.
Which is another thing to consider. If I'm not mistaken you typically employ a Sphere Dome and you combine that with TA. The TA bounces your ground level light upward, to illuminate the underside of the spheres. My EGDLS workflow basically uses the EarthGlow dome to do the work of the TA, which produces much faster results in most instances. The one thing TA cannot provide under any circumstances is indirect specular light, that can only come from point sources since specular is by definition a reflection of a point light source, instead of a reflection of the environment, as Reflection refers to.
There is no right or wrong way to do things, that makes it more difficult to discuss certain things. We discovered long ago that TA looked much better when it was aided by a few point light sources. It smoothed out noise. But it added greatly to render time. Then the idea came about to optimize point lights to work with TA, treating point light sources as glowing geometry (theory courtesy of our very own Peter Fulford), produced finished images in a fraction of the time than it used to take when the light sources were still points.
To everyone:
I apologize for the long wait to get the thread up and running. Once I started compiling I realized that I needed to provide a good amount of sample content. SCenes, models, textures, light rigs... in the hope of being as concise as possible. Please bear with me. Thanks for your patience.
Yes, Domes can have extremely high quality settings, much higher than IBL strangely.
These are some quick teasers for where the thread is likley headed.
First couple of images are a complex cumulus cloud studies and a celestial backdrop study. These clouds and celestial backdrops will be provided with the tutorial materials. These clouds have purposefully whispy edges designed to appear like clouds in Vue. That's right, Bryce can make cumulus clouds very similar to those of Vue. Yaya!!!!
The next two images are to the point of lighting and the purpose of this thread.
Why bother with Light Domes at all, in a world where we have TA? Won't TA always look better than faked GI? Great questions!
Below is the same scenario rendered in two different ways. The first is lit with my EGDLS fake GI set-up, standard AA. The other is rendered Premium Effect True Ambience with Boost Light, Scattering Correction, and TA Optimization, and an HDRI backdrop. Rpp only 16. Render time 5.5 hours with full volumetrics and translucent leaves on the trees! Not bad! At rpp 36 I suspect it would have taken roughly 12 hours or so. All domes are disabled for the TA render. Also the saturation of the hdri is at only 15, as the scene was overly blue otherwise.
Alexy, you wondered before if perhaps the fake GI version of the Desert Temple was too bright. Compared to the sample renders that Dave Savage provided, yeah, seems very bright. But I will attest that the sample renders are underlit, as far as the indirect illumination is concerned, so our expectations are skewed somewhat. However, I will say that the clouds surely are too bright, the underside should be a bit darker. But as far as the overall light levels on the structure, this is more or less what the indirect light is supposed to look like. I can feel confident of this because if you look at the TA version next to it, you will see that the overall light levels are strikingly similar. What I hope to do in this thread is to demonstrate to everyone just closely a well rigged fake EGDLS can compare to a well implemented full GI. EGDLS also provides the user the opportunity to tweak the levels as they desire, so in this case the EarthGlow dome is a little brighter than the TA version assumed the ground level light to be. But the values are still close enough to be comparable.
All of the indirect light you see is coming from the TA, thanks to the fact that the sky is a high dynamic range image, and with TA Optimization of the HDRI enabled, TA can actually see the sky for the illumination it truly provides. In TA Optimization mode, the Hdri no longer outputs its light as discrete points, it instead treats the backdrop as a single surface image, allowing the TA to gather it as it would any other continuous surface, but with the benefit that the light is high dynamic range in nature. All the benefits of an HDRI / IBL without the need for points. Remember you are only getting 1 color per virtual light source. So an IBL quality 16 is only giving 16 of the possible thousands of colors contained within the HDR image.Or better stated, instead of only getting 1 color per virtual point as is the usual case with IBL, with TA optimization you can get every color from every pixel of the Hdri to be included in the render. Hands down TA Optimization with Boost Light is the most accurate way for Bryce to interpret IBL. In fact, this is the way IBL should have been implemented from the start, but since boost light hadnt been implemented yet, TA could not interpet any high dynamic range information, so points were used.
The results of TA optimization are compelling. As you can see there is plenty of light without the need to increase it with extra help of a dome. In my opinion Alexey, the only reason why you currently need to add a dome to your true ambience renders is because your TA isnt gathering enough light from the low dynamic range Bryce native skies so you are supplementing the missing dynamic range light with the Dome of points, increasing render time dramamtically. If there was a way to get the adequate indirect light values you needed while not needing to introduce the Dome on top of the TA; wouldn't that be nice? Or what if you could get the indirect light you needed from domes such that there was no longer a need to enable TA at all? Either solution is better than mixing the two as far as render time is concerned.
That's where EGDLS comes in. I needed a fake GI solution that was ROBUST enough to handle most any scenario. We shouldn't need to start over from scratch with each new scene. I knew that as a community we had long ago come to grips with needing to get light down from the sky, but many of us had not come to realize that light needs to shine back upward as well, which is why I've named it EarthGlow because that really is the concept that sets this rigging apart from typical fake gi rigging.
The EGDLS version took 14 hours mind you, but that was with dome qualtiy of 100 for each. If rendered at only quality 16 the render would have completed in a couple of hours.
In conclusion; It is best to pick an approach and to stick with that consistently. So, if you are going to render with TA; then you want to avoid point light sources altogether, especially cluster forms such as Domes and IBL and Fills. You want to enable TA Optimization for the IBL, include Boost Light, Scattering Correction, and you want to use an hdri backdrop because TA needs that high dynamic range color to illuminate the scene properly. Bryce skies are low dynamic range by default, TA gets alomst nothing to work with. This is why I never use Bryce skies, I always use an HDRI. More on this in time.
If you are going to render in standard AA mode fake GI, then you will want to use something like EGDLS. Because using a single dome for skylight alone is only half to battle, we still need another dome to account for light bouncing and shining upward.
Of course you're right, Rashad! Transparency and translucency is not the same thing.
In the case of procedural texture light is scattered inside the sheet, and then going into the environment. In the case of bitmap texture light is not scattered, but merely penetrates. I do not know why this is happening. Maybe the whole thing in geometry? I think you will be working on this issue.
With regard Specularity and Specular Halo, something about it we will talk later. Otherwise it will mess in my head. First we will look at a global illumination.
Also surprisingly, in this thread has got the theme with foliage transparency. We need to talk about lighting. For foliage and trees, we can create another thread.
I have found that translucency greatly affects the overall lighting of a scene because it allows multiple surfaces to be hit with the same rays of light. Items placed below the tree receive light they otherwise would not, making the scene seem much brighter than if the leaves had been opaque.
It is necessary to complete the theme with transparent foliage. In fact, I forgot about the refractive index. Thanks Rashad!
Here is an example of a translucent foliage for procedural textures. This time I used a white color to Ttransparent and Volume channels.
The same experiment with the refractive index for bitmap texture not give significant results.
Here are some examples of the GI with and without TA. But in each case I use Sphere Dome Light.
On the hillside
Bryce 7 Pro. Render Premium 36 rpp, time 13:38:11.
Lighting: 500% sun (shadows of clouds included) and Sphere Dome Light.
Mountain trail
Bryce 7 Pro. Render Premium TA 16 rpp.
Lighting: sun and Sphere Dome Light.
File size 34,7 MB.
On the river bank
Bryce 7 Pro. Render Premium TA 16 rpp.
Lighting: sun and Sphere Dome Light.
5 Bryce of trees, a grass and flowers from a private library and small flowers from Wings 3D are used.
Instansing Lab and manual cloning is applied. The size of the file is 30,3 MB.
I first used at the same time render TA, volumetric clouds, Sphere Dome Light, transparency map for the "User Leaf". Render time several tens of hours, especially at night. If I had used more soft shadows, the DOF and transparent foliage, you can not wait until the end of render.
Rashad Carter - How can I change lights colour, i.e. the head lights of the truck to be a more yellow shine and lights inside the interiors to be not too white (seen on the left building in foreground of attached image)? I thought if I have the light (red wire frame of the light) selected and change the colour it will change but it just stays the same! I have little experience with lighting.
launok,
1. Make Alt - click on this button, select the palette and select a color.
OR:
2. Open the Light Lab, open any color picture over the Bryce window, click LMB on this button, without releasing the mouse button, drag it to any color on your image, release the mouse button and you will get the color of your light source.
Thank you for your great help. I have used the second option as to use the Light Lab.
I could not underdtand option 1. What do you mean by "Make Alt"? Are you referring to the ALT button? And the palette, is it the box next to the light wireframe (colour family)? I am not familiar to lighting. In fact, today is the first time ever I am using Light Lab.
I have only add a very slight yellow to the head lights as it affects the ladies and wall also in the scene. Add a bit blue to light in the house in foreground.
Launok, Welcome to the Light Lab! Now the fun begins!
SlepAlex,
I am always in awe of your work. Your Bryce trees are very good looking. I love your sense of depth, I am pulled into each of these renders. I love the colors. So much to appreciate. I'll go into more detail at some point soon. Thank you for these examples. Your methods are definitely effective.
Sorry for my English. This Google-translator's fault. :-)
1. Move the mouse cursor over the button to select color of the light source in the Light Lab as on my picture. You see this cursor?
a) Press LMB (left mouse button). You will see the usual palette of WINDOWS. Without releasing the mouse button grab the pipette desired color swatch from any area of the palette. Release the mouse button. But better
b) press the "Alt" key on your keyboard. Without releasing the "Alt" key, click LMB on the color selector button. Palette window appears with four tabs, RGB, HSV, HLS, CMY. Select the desired palette. I prefer the HLS (Hue Lightness Saturation). Drag the sliders of the palette to the desired position or specify numerical values from the keyboard. Confirm your selection.
In the same way you can change the color of the sky, haze, fog, etc. On the main interface or in the Sky Lab, as well as in Mat Lab and in general wherever it is necessary to change / assign color.
I continue my experiment.
We drop our "Sphere Dome Light" below the ground plane. Nothing changed! The lighting remains the same. This means that regardless of the position "Sphere Dome Light" virtual light sources are uniformly distributed around the top of the sky hemisphere (see. The left part of Fig. 1).
Let's uncheck "cast shadows" in the material properties of the plane. Settings "Dome Light" is still as follows:
Diffuse = 10
Specular = 10
Quality = 16
Bias = 50
That is, regardless of the position "Dome Light" (above or below ground) and the presence of checkboxes "Cast Shadows" in the material of plane, is lighting everywhere the same.
This does not apply to negative values "Bias". But we'll talk about later.
*****
Now, do the following. "Cast shadows" is unchecked in the material of plane. Let us return "Dome Light" to its original position and set Quality = 16, Bias = 75 and 10. Compare the two results. Left "Bias" = 75, right side 10 (see. Fig. 2). In the case of Bias = 10 light sources are lowered to the horizon, so the lower part of the object is illuminated more. At the same time on the horizontal sections misses less light.
Rashad, although I enjoy thread wander, this has really gone all over and into the long grass. Maybe you should fully compile your presentation and then issue it all at once in a new thread (where reserved posts could serve a purpose other than you practising your one chat-up line ;-) ). Kind of like a lecture that can be thrown open to discussion afterwards.
This is not a complaint to you or anyone else in the thread, by the way. Just a suggestion for your ideas being able to come over more clearly.
Or continue as is, which is also fun - if a bit schizoid!
Either way, it's good to see lots of busyness on the forum.
You're right. It's taking longer than I expected. I have a new motherboard I'm installing as well as some other computer issues so it's lagging behind. The content compiled so far is well over 1 gb, and could easily rise to 2gb. Yes, I'll need to start a new thread. Might as well make use of this one for the time being.
When in the Distant mode, the position nor the rotation of the dome light matters at all. For this reason it is sometimes best to simply use a super large dome with the distant settings disabled so that it can still be rotated and moved around.
And yes again, that is why I suggest two domes. One to supply light from above the horizon, and another to provide light from below the horizon. But there also needs to ba a large amount of area around the horizon where the two dome light influences overlap, so as not to have a sharp contrast of missing light from certain angles.
Rashad! It is about two light domes will be in the following messages!
10% of the time I spend on the creation of images, 90% on a double translation (to and fro) and drafting the text.
It will be named accessible and understandable, without unnecessary theorizing:
Sky light (Dome Light 1)
Earth light (Dome Light 2)
Slepalex – I did not know we could save bitmap textures in a library. Thanks for the screenshots and your simple explanation about the Dome light.
Your renders are awesome, did you use 1 or many sphere dome lights for these landscapes?
I have used before one dome, recently two. Wait my posts, I will tell everything.
In the previous message, we found that, if the "Bias" is large, the individual light sources are concentrated mainly in the zenith. When the lateral surfaces of objects poorly lit, but the upper surface and the horizontal areas in particular illuminated stronger. If the value of "Bias" is small, then side surfaces lit more, and horizontal areas are darker. When set to "Bias = 50" individual sources are evenly distributed throughout the sphere as above ground and underground. And if uncheck "Cast Shadows" with infinite ground plane (terrains, water surface etc.), the light will penetrate through them, and to lit all other objects irrespective of the position of the "Dome Light" on the scene. Sometimes, for small values of "Bias" is necessary to increase the value of the "Diffuse", and for large to reduce. It all depends on the orientation of the camera. For example, if the camera is low and looking up (on the trees), it makes sense to reduce the amount of "Bias" and increase the "Diffuse".
Look at the picture with three options. The material of plane is unchecked "Cast Shadows". I prefer the third option (right) of illumination. Although it depends on the particular scene, the camera position and personal preference.
***
In the scene with the trees I added sunlight. As we can see, the smooth surface of spheres well lit from the standpoint of the GI. However, very little light penetrates the tree crowns despite the Quality = 256. This is because adjacent leaves prevent the penetration of light from all sources, and the leaves themselves do not reflect or scatter light.
Mermaid- While one dome can surely do the job to some degree as IBL proves, a two dome system works out better, because it provides the user with more control and more control is always better.
Alexey- Thanks for these diagrams. As you've stated, the trick is that the ground plane must have shadow casting disabled in the material. But it also means that we might want to exclude any ground covering terrain or water plane from the Influence of the Earthglow dome, since we do not want to disable shadow casting completely for the terrain or the slab. All we want is for the light from the EarthGlow dome to make it into the scene successfully. In general most water mats dont need to be lit by the domes, so they can be excluded.
I should warn you that there are some bugs. Yep. There are situations where objects that have been excluded from receiving light will still cast shadows onto nearby objects. It adds yet another issue for consideration. To test, you can cover the ground plane with a grass model and notice that even if excluded it will still cast shadows onto the sphere. And in the case of Instances, we have an even bigger problem. And that is the fact that for some odd reason each instance has it's own session reserved in the Light Lab, instead of sharing the one session with the original source object as it should be. Excluding the source from being lit doesnt affect the instances as it should. If you observe you'll find that if instances are visible in the wireframe they will each be listed in the Influence dropdown of the light lab. This is BAD. because no one will take the time to assign lighting to thousands of instances individually. It also makes the dropdown list impractical, since "copy of" occupies the entire list. There are workarounds, but still, frightening. In related news; I've also seen material connections between sources and Instances to break meaning each instance now has its own reserved session in the Material lab, increasing the memory footprint of the scene dramamtically. As always, ahead of myself. But its hard to hold back.
I do want to explain why the theorizing is important. EGDLS is a way of thinking, it isn't just an arrangement of light domes. There is a logical reasoning behind this approach, which is why it works in most every single outdoor scene. People need to know the exact functions these two domes provide. People need to know how to improvise. My goal isn't to instruct people on the one way to do something, but to explain to them how to think critically about the situation, to know where they need more light and to know where that light should ( or should not ) be coming from. All outdoor scenes share the same concerns, sunlight, skylight and ground level light whether lit with TA or domes or whatever. How we attend to those three concerns is the key. But it helps to know ahead of time that those are the three major contributors to the lighting of any landscape, and to know that if one of those three is left unaddressed the scene will not appear as realistic as it could.
Also you mention about adjusting the lights based on where the camera is located. I feel that a global illumination should be very robust and flexible, and it should not require tweaking based on a particular camera angle. The way I build scenes is to imagine that it might someday be animated, and that I could find myself placing the camera almost anywhere within the world. I need lighting that can handle that. If we set up the domes and other lights properly, we should be able to render any camera perspecitive at any point in time with full confidence that the lighting will be convincing.
As a general rule, we underlight most of our Bryce scenes. We don't tend to notice when shadows are too dark, or when contrasts are too distinct. Many users do not yet know the tricks required to represent the different imes of the dsay, the color the sky should be, the color the haze should be, the intensity and color of the lights that should be applied. Mastering time of day must occur for a user to truly feel ready to go off on their own.
The reasoning behind naming the domes as Skylight and EarthGlow, is because when we set out to create the look of different times of the day, we need to know how much intensity and what colors to assign to each of these domes. Skylight domes should be roughly the color of the sky...whatever that happens to be. EarthGlow domes should be more or less the color of the ground plane or terrain. And then there is a third term, that is non shadow casting that finishes off the illumination in areas blocked by the two other domes. In a geometrically simplified scene with only a sphere we dont run into the same problems that we run into with more complex geometry.
Rashad!
I read your long message and understand what you mean. I keep on disk your old file 2011.
I do not agree with all the conclusions. We're not talking about the animation and copies of grass and difficulties with materials and a long drop-down list. We are talking about the most simple method of producing fake GI in a static scene in terms of acceptable quality and render time.
If the sky is blue, it does not mean that it stains the ground in shades of blue. In the air, there is a diffuse scattering of sunlight, which is white in color. The air looks blue, but falls to the ground white sunlight. This is evidenced by the experience. If in a shadow it is somehow manifests, then not any human eye can distinguish subtle shades. Yes, at sunset the sun objects can be painted in shades of yellow or orange. So you just have to change the color of the sun at Bryce.
Earth and the objects are on it disperses in the sum the reflected shade of gray color. Especially at distance.
About two domes. Give me two or three more days and I'll get to the issue.
The third dome isn't necessary at all! Put Radial Light in camera coordinates (Alt + C on the selected camera, Alt + V on the selected Radial), bind Radial with camera as the parent link, disable the shadow and animate the scene to health. This is equivalent to the use Ambience in all materials of scene. I have used this reception. It is also effective when rendering with TA. Let's just follow from simple to complex in this matter, and to theorize and discuss the nuances will be then. After all, this thread simple inexperienced users too are watching.
****
Well, I spent an extra hour to write and a double translation of this message...
Alexey,
Don't spend too much time on translations, the images matter most.
Not just the most simplistic method, but the most robust method. The method that gives anyone, especially a newbie, the necessary starting logic that can be applied to any outdoor lighting simulation. A lighting method that can scale up or down with the needed complexity of the scene or the demands of the final output, such as animation, if that is ultimately desired. What I'm saying is; after tons of hours of testing including under the stress of animation and still images alike, EGDLS covers all the bases. Not that it is the only way. but if one follows these assumptions they will always end up with a plausible result.
Think of EGDLS as if it were a product being sold in the store. It is a light set for outdoor scenes to make it easier. But people still need to know how to use it.
The blue sky above us does indeed cast a very blue tinted light onto the world beneath. We do not notice it on a conscious level because our unconscious minds tend to filter it out. This allows us to perceive other colors more naturally. But it is a trick of our brains. In real life the blue tint is there. It becomes noticable again in photographs, such as any image of a snowy environment where the skylight isn't being mixed with other ground level colors because the entire ground level environment is white. A CG render is closer to the perceptions of a photograph than those of the human eye, and that is why the blue should be visible from a realism standpoint. And that is also why there need to be a ground level light that corrects for that blueness from the sky. The result is that our unconscious again fail to notice the blue. Adding the blue isnt a problem, so long as it is balanced out with another ground level light color. Thus the emphasis on the theory of the EarthGlow dome.
https://www.google.com/search?q=daytime+blue+sky+blue+shadows&source=lnms&tbm=isch&sa=X&ved=0ahUKEwjcheK9nfDRAhUr_IMKHcUDCCgQ_AUICigD&biw=1920&bih=945#imgrc=ptouex70VZfCwM:
https://www.google.com/search?q=blue+sky+casts+blue+shadows&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwj8nviWnvDRAhVj1oMKHQ46BlAQsAQIGw&biw=1920&bih=945#imgrc=Fy29SLmtj5-dVM:
https://www.google.com/search?q=blue+sky+casts+blue+shadows&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwj8nviWnvDRAhVj1oMKHQ46BlAQsAQIGw&biw=1920&bih=945#imgrc=aPKXon0rO4649M:
https://www.google.com/search?q=blue+sky+casts+blue+shadows&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwj8nviWnvDRAhVj1oMKHQ46BlAQsAQIGw&biw=1920&bih=945#imgdii=i9WCwQCDJtYuQM:&imgrc=kopsdik5hZZTUM:
https://www.google.com/search?q=daytime+blue+sky+blue+shadows&source=lnms&tbm=isch&sa=X&ved=0ahUKEwjcheK9nfDRAhUr_IMKHcUDCCgQ_AUICigD&biw=1920&bih=945#imgrc=3NWEqU-ad0xhqM:
Also, I use a proxy method for cameras. I usually save around 100 different camera placements per scene. Using a radial placed at the camera position is not feasible for me, but I think it works well for most. I'd consdier that something akin to using a camera flash, which is very good. But most of my shots wouldn't require that.
Crikey, Rashad; you have moved on. On the old forums I argued this point to you until I was blue in the face, and finally gave up because you just wouldn't accept it (and were quite obnoxious about it). And here you are proclaiming it!
Progress.
Round and around we go, as knowledge dances about!! Yes, it's hilarious!!! Where is Len when you need him? Good to have someone to enjoy a good recollection with. What is the point if not to make progress, old friend? You are one individual to whom I owe a great deal of gratitude, as does a lot of the Bryce community whether or not they realize it. Most everything I ever learned about TA I learned directly from you, as well as so many other lighting lessons. You are the brainchild behind TA Optimization, an idea I was able to pass to the development team because you had once shared it with me in one of those contentious debates. I felt honored to pass those lessons along to developers then and feel proud to share them again today. I consider you a true friend, and obviously one capable of keeping it all in perspective. Admittedly I'm pretty obnoxious about most things, whether right or wrong, and you handle me justly at every turn.
I think this issue of skylight blueness you managed to prove me on fairly quickly. I'm completely certain any blue on your face came from other things I was much less willing to budge upon, such as the accuracy of True Ambience itself as it was presented and implemented in Bryce 6 (with TA intensity goverened by the actual ambience channel among other issues) at the time. With blaring problems like that at the foundation I just never could find any reason to trust it as a solution moving toward realism. I kept trying to prove to you that TA was too inaccurate to rely upon, and you were trying to convince me that regardless of its problems it was still the best option availble at the time. I think in some parts we were both correct. Luckily Bryce development was soon to address many of those issues due in no small part to input and discoveries made by you.
Realism can be limiting from an artistic perspective, and I can understand why you didnt want me to get too bogged down in seeking realistic results at any cost. Don't retire just yet!
I have watched this thread with interest but kept quiet. It's Rashad's Thoughts about Lighting, not mine. Like Peter above I noticed that the thread goes off its main direction (as I understand it). The contributions are all valid and interesting to follow, even to comment, but they lead away from the main topic.
My main question is: which reality? The blue light from a cloudless sky is a reality, a physical fact, we usually do not perceive as such. The eye is only a contrast detector; the brain does the main work for "seeing". The sensitivity of the eye is different for each individual; differences may be very small but also profound. The brain corrects this from experience and creates the image. The brain also does the bulk of the work for adapting brightly lit and low lit environments; the pupil diameter has only a minor impact on this. There is no need to keep an image with blue shadows on the snow, it can be filtered out. We see what we expect, what we expect is influenced by experience but also by our mood the moment we look at nature; fond memories of a sunset with a loved person makes the current sunset look more beautiful than if the beholder remembers when crashing into a tree with the car because of the low sun blinding him/her. Snow in the shadow is neutral but when consciously comparing a patch of snow lit and the patch in the shadow next to it, the blue tint can be perceived. Walking on in the shadow, the snow becomes neutral coloured again. I did that experiment several times because I thought my photographs are too blue in the shadow.
A photograph records the light within the range the film or electronic sensor is capable of, and the colour depends on the white balance setting. Digital cameras process the sensor data already before it is stored. The blue shadow on the snow on a cloudless day is evident and also physically correct provided the white balance was correctly set. This would be photo-realism. How much a photograph we see has been processed, we usually do not know. Processing started with chemical film when developing and again when creating the print. The dynamic range of a film is much higher than what can be brought on paper. Equally so, the dynamic range of a photograph made with a moderately good digital camera exceeds the capability of the computer display. Photo-realism is therefore also subject to the intention of the photographer; what should be shown, which impression it should leave for the beholder, etc.
A traditional paint of a landscape is also far from any realism. It is just not possible to create the whole gamut with colours blotches on a white canvas. The artist strives to bring in the objects, colours and light in such a way that is pleasing and real looking for most beholders.
Lighting is key to any visual artwork. Discussing different lighting methods is important, helpful and can lead to new insights; there are several ways to accomplish a very similar result. The right method to use is the one that makes the result look as the artist expected and is pleased with. Reality is finally a subjective experience
Well, guys! I see that the discussion inevitably turns toward color. But I would like to understand at first with light. I want that with the minimum losses of the time of a render to achieve illusion of GI or indirect illumination of objects which are in a shadow from all light sources. For example, it refers to a dense forest or dense urban areas.
And yes! On the color theme in 3D and photo for a long time and much has been said. For example:
http://www.itchy-animation.co.uk/tutorials/light01.htm
I corrected some errors in the message from 31 January.
In the previous message, we found that, if the "Bias" is large, the individual light sources are concentrated mainly in the zenith. When the lateral surfaces of objects poorly lit, but the upper surface and the horizontal areas in particular illuminated stronger. If the value of "Bias" is small, then side surfaces lit more, and horizontal areas are darker. When set to "Bias = 50" individual sources are evenly distributed throughout the sphere as above ground and underground. And if uncheck "Cast Shadows" with infinite ground plane (terrains, water surface etc.), the light will penetrate through them, and to lit all other objects irrespective of the position of the "Dome Light" on the scene. Sometimes, for small values of "Bias" is necessary to increase the value of the "Diffuse", and for large to reduce. It all depends on the orientation of the camera. For example, if the camera is low and looking up (on the trees), it makes sense to reduce the amount of "Bias" and increase the "Diffuse".
Look at the picture with three options. The material of plane is unchecked "Cast Shadows". I prefer the third option (right) of illumination. Although it depends on the particular scene, the camera position and personal preference.
***
In the scene with the trees I added sunlight. As we can see, the smooth surface of spheres well lit from the standpoint of the GI. However, very little light penetrates the tree crowns despite the Quality = 256. This is because adjacent leaves prevent the penetration of light from all sources, and the leaves themselves do not reflect or scatter light.
Slepalex – Thanks for the simple explanations and images, I’m beginning to understand how to use the Dome Light, as I’m doing the examples you mention.
Rashad – Thanks, I never had any luck with using the Dome Lights or 3D fills thus far except for doing one of David's tutorials where he uses the Fill light, I wondered how you guys use it.
Both Slepalex and Rashad are masters in their own field, and personally for me Slepalex’s way is good start to get the basics while waiting for Rashad’s more advanced theories. I'm reading and learning Thanks