Daz 3D is part of 
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.Daz 3D is part of
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
No no no, thanks to YOU, Rashad! The file you've offer for free is so rich! There's also a lot of things to learn with your lightning approach.
Well, trying to get back in this great discussion and for those interested, I'm posting my last render of a megascape which is inspired of Rashad's one.
I'm not using EGFPLS lightning here, only T.A with sunlight and a simple bryce blue sky (no IBL).
I've first try to create this scene using EGFPLS but doesn't get good result while I'm always employing transparency for my leaves materials. Render time seams to take ages of age if you do that.
I've got ambiance chanel on for all my materials and that means it's another working way compare to EGFPLS which require no ambiance for its glowing system.
To let you compare both method, I need you to let me know if you'll be interested to download this norvegian cliff. Of course, this offer is for free and will be downloaded out of the forum.
Now some figures about this render : I've use Speedtree for my models and there's about 10 differents. All those trees have been instanced over the landscape which is composed of 5 different plans from the foreground to the background.
The first plan is made of a lattice covered by 8 grasses species. The next plans are only covered by trees. The last plan in the distance is made with spiked trees generated in the terrain editor.
There's just over one billion polygons for these scene and you must run a 64 bits system with L.A.A to be able to load it in Bryce.
I hope you enjoy and like Rashad asked, I will like to see more dialogue here too.
very nice C-ram!
Well, what can I say? It's enough just to express admiration for this work!
Although there are some comments and questions.
1. Mark, why do in your scenes volumetric clouds look overexposed? Maybe it's worth reducing the diffuse value in the material?
2. How did you make the coastal foam and foam near the stones?
3. What is the file size?
4. What considerations were in favor of choosing TA Focused Scattering compared to TA Scattering Correction? I just made an experiment. Render with TA Focused Scattering gives a brighter and more saturated shade of blue, in which all objects on the ground are painted under the influence of the color of the sky. And this, in my opinion, is not right.
Alexey : thank you!
1 the clouds here is looking overexposed because the sun is taking place just over in a low altitude. In some case, clouds are looking the same way in real world.
2 the coastal foam is made of terrains with a foam material from David (screen capture above.)
3 the file size is about 542 mb.
4 I have made some test with scattering correction and focused scattering and I prefer the render with focused scattering. False or not, to me the render is looking better this way.
Wow Marco another breath-taking scene. I would be interested in viewing your files. I still can't make sense of the Rashad's MgScp Files. I'm working with the other files he shared, the Fake lighting ones.
c-ram - this is a very beautiful and realistic looking scene, Marco. I see what Alexey means with the clouds. Personally, I would go with a slightly less bright sky but it is not wrong the way it is - depends on the shutter speed used (if it were a photograph). In the end, it's always the artist's choice. We could also argue about the Bryce 6 legacy setting of focused scattering versus scattering correction. There are always options one should never use but if one does anyway, the result may look better. I have my default settings with all the "no-nos" disabled and as I progress, I may change the settings. In the end, what we are doing is always faking and cheating to get the desired result because our tools are just limited tools, not nature itself. I envy you for your patience and dedication to create such awesome beautiful renders.
Mermaid- Just keep at it. Please report any interesting discoveries. Dialogue would be nice to get going.
C-Ram- I'm speechless. Completely dumbfounded. Every time I think you might have reached a personal limit you go and raise the bar yet again. So much is going right here. The haze is an incredible tool here, establishing scale in a perfect manner. Are all of those trees 3d, or are some of them billoards? Love the coast, love the overall composition in every way. Masterful work, Marco. I'm so proud to see this kind of work produced from Bryce. This image would make for an inspiring promo image for Bryce. I can agree that the clouds might seem a tad overexposed, but it doesn't distract me much at all. There's just so much to appreciate. Did you work to any strict scaling in this piece?
Horo- Yep me too, I keep default scenes with all the no-nos disabled. I almost forget certain options even exist after a time.
I looked a bit into transparency and translucency. Rashad has a nice texture that can be downloaded from bryce5.com (Material X-Change, page 2) which I often used on foliage but found it a bit too transparent at times. Slepalex shows in this forum near the end of page 1 a means to adjust the transparency. I spent some time with the transparent control, transparent and volume colours as well as the DTE and complied a 3 page PDF - and also a texture library with linear and squared transparencies at 5, 10, 20, ... 80 and 90% (total 20). It's on my website under Bryce & 3D CG Documents > Mine > Materials > Transparency.
c-ram: Amazing render! Maybe a bit light in the sky indeed, but what a richness in this scene! Truly wonderful.
Yes, I saw those experiments. Like you, I do not find any possible way to have leaf shapes being cut out with trans mapping effects while still retaining the optical effect of translucency. One has to make a choice between one effect or the other. Translucency can be very easy to overdo. I've also included examples of translucency in the Tutorial files of this thread. It's just one of the many interesting things I'm surprised isn't generating any dialogue so far.
David's method for translucency differes from my own slightly, it uses blurry transmission instead of refraction values to fake the appearance of some degree of absorption. It might be smarter to use David's method when rendering with Premium Effect mode.
C-Ram is THE Man!
Thanks to all of you!
Horo : you're totally right. We must cheat with bryce to get the best result we can because we are limited in a lot of terms from lighting to texturing etc..
I'm posting the original canvas from shishkin to let you take a look to his awesome work I have been inspired.
Rashad : the scene is entirely covered by real trees so, no billboard here. For the scale of my objects I'm working simply. All those trees have been scaled to the same proportions, depending on the first imported one in the scene. This is going the same way for the grasses on the foreground. Well, I hope some of my works attract attention from the next bryce development team which maybe will pick some of my pictures to promote the new version.. cross fingers! Working back on nGplant right now, so much interresting things to understand.
I really would like to discuss these things, truly. But first finding time to actually play with it and understand a fraction of everything you are doing. I hope experts like C-ram, Slepalex, Horo and David will help out in the discussion-field.
c-ram: You followed the example very closely! Even more impressing. There is one difference and that is the lack of houses in the painting. To be honest, those houses on the shore in your render were the least impressive part of your work. Maybe delete them and re-render?
Rashad
I spent the last 2 weeks viewing the files you shared with us.
I still don’t understand how you set up these extremely magnificent scenes. They are tooooo complex for my retired brains. Initially I did not see your additional post on “Instructions for viewing the MgScp Files. Then I was very confused because you have all the object attributes as “show as box”.
Nonetheless I had fun viewing the different camera views. I rendered a few views and posted them on my Pinterest account. These renders took between 5 to 8 hours to render. The Sunday Picnic with the Clouds unhidden took almost 10 hours. The Sunday Picnic I also rendered at ½ resolutions. I did not change any settings except hiding and un-hiding the cloud layer. I also did a render, hiding all the light layers and used one of Horo’s Sunless Hdri Skies, just to see the effect. Your scenes work best with the EGDLS or EGFPLS. I jumped from one camera view to another, and played with the trackball and FOV to get different views and let it render for 15-20mins to see what I got. Very impressive work.
A few questions:
Which objects can I save to my libraries to use in my simple scenes? I already saved the Roman building and Stone well.
Will I be able to us the light groups EGFPLS and EGDLS in my scenes or can I just use the individual lights?
What do you mean by this “disable the nodes for the translucency in the material lab?”
I am really amazed at the complexity of these scenes. Files like these are awesome tools for professionals and those into gaming.
Thanks for sharing them with us,
I’ve posted one render here and the rest uploaded here: https://www.pinterest.com/maryole3/bryce-3d-landscapes-and-seascapes/
Mermaid,
Wow! Talk about diving in head first. Your Pinterest page is stunning!! Every image on there looks incredible. You've clearly been absorbing the skills offered by David Brinnen and Horo. You're making an impressive showing for Bryce 3d.
Thanks again for taking the time to explore these crazy MgScp files. You're having the exact type of fun I was hoping people would experience. Thank you for showing us your studies so far!
5-8 hours isnt too bad of a render time in my view for files featuring this degree of complexity and lighting, especially if you're not running a system that provides 8 or more cores.
Onto your questions.
What you can save and reuse? The stone well was actually made by another Bryce master, I think the name was Muon Quark. Or maybe Bullit. I no longer recall exactly. The original file is free with Bryce 7 in the Premium objects dropdown of the Create menu (Installed). I just updated the materials to better interact with the style of lighting being applied. Everything else in the scene was created by myself and therefore are free to use as needed. Even the textures are assembled from my own cell phone image captures and scans of leaves taken from my own personal scanner, so again, free for you to use. Ideally they are not for sale of any kind, and redistribution wasn't my original intent. However, if someone has a way to get a Bryce scene into a high profile project and it features a few plants from these MgScp example files, I'm totally okay with that.
This means you can use the HoroTrees, BrinnenFerns, ChohLillies, and Fulford Pines in any project you'd like.
"Disable nodes for the Translucency in the material lab" This instruction is related to the MgScp Tutorial files, which are different from the MgScp Artwork files. In the tutorial files (those featuring the Roman Building) I have included some trees which have translucent leaf textures enabled by default. Transparency and translucency are distinct optical effects, as translucency is not obvious to users in the way that transparency is. Both require the use of the Transparency node, but in very different ways. I observe users often employing transparency at times when translucency is the actual desired effect. Translucency boosts the realism of leaves dramatically, but it takes the render times along with it. In simple scenes the translucency is the way to go, but in crazy fully complex scenes like the MgScp artwork files, it no longer is feasible. And transparency to my view doesnt create the correct glow effect to be a subsitution for translucency. Thus, to speed the rendering of the trees in the Tutorial files one can simply enter the material lab for the trees and disable the node in the transparency channel. The way translucency operates is that even though the node for transparency is active, the slider itself remains at 0. This allows light to travel through the leaf in a way that doesnt appear glass-like. The slider must remain at 0 as well, because if you raise it even a smidgeon you end up introducing "internal reflections," which again creates the appearance of glass, all of which are acceptable only when transparency is the desired effect. We can and will discuss transparency vs translucency in more detail as time goes along. Great question by the way.
My suggestion is that you save the light groups themselves to your Library. Create a new library for lighting effects and drop them into there. Indirect lighting is what these studies are all about. EGDLS is a bit more flexible than EGFPLS. EGDLS is less dependent on scale, behaving more like HDRI probes. But EGFPLS, is entirely scale dependent. Thus, it must remain the same physical size in all scenes you import it into. The only edits you can ame to EDFPLS is to the brightness of the 3D Fills. Never alter the size. You also want to save the skies as presets, because you need those to match the indirect lighting rigs.
For the age old question: what comes first, light or textures? The answer is and always has been that light comes first. The main lesson with the "EG-blah-blah-blah" files is that the lights are very strong, and that you need a stong sunlight to match them. And also that we need to consider the light shining upward to be just as important as the light raining downward. Due to the strength of these lights, you must be mindful to lower your Diffuse channel values of your materials to 50%, and then raise them incrementally from there until you reach the desired level of brightness. Under these lights you should rarely if ever need to use 100% diffuse, and you will cetainly never need to boost the brightness with Ambience, which is to be avoided whenever possible. Since most presets and other models feature diffuse values at 100%, they will all appear to be overexposed under these lights. Don't mess with the lights if you can avoid it, always tweak the diffuse values of the textures. This means you must edit the textures for every surface of any model that you import under these lights
Horo's sunless HDRI's should be able to do a decent job of the Alpine Valley scenario. Since there's some "breathing room" between the trees and the ground is not covered ny grass geometry, a "dome" approach should be sufficient. In the Sunday Picnic file however, the dome approach fails because there are too many local obstructions to the light coming from a distant dome. This is the reason for developing the EGFPLS, because the 3D Fill lights float around within the local area providing better coverage. That said, the examples on your Pinterest page featuring the sunless HDRI seem to be missing shadows. Shadows are not enabled by default with HDRI. I suspect that 100% shadows wouldnt be the best either, you'd end up with lots of black pixels due to the geometric complexity hiding many faces from receiving illumination. You'd probably have to lower the shadows for the HDRI down to about 75%. But even then you are correct, it would not hold up as well as the EGFPLS.
You've put a HUGE smile on my face today, Mermaid!! Seeing your images makes me feel the effort I put into the studies was worthwhile. Thank you. Sincerely, thanks!!!
All thanks are due to you, Rashad for sharing these awesome files with us. I am particularly interested as I would love to add more vegetation to my landscapes, and I'm lucky that I spend all my leisure time playing with Bryce.
Thanks for your explanations. I will copy it and have a look later.
Thanks for the nice comments. Did you view the other boards as well? Most of my boards on Pinterest are a window to Horo and David’s products.
Played a bit more with Rashad’s files, using elements from David and Horo’s http://www.daz3d.com/bryce-7-1-pro-environments-inclement-weather
I deleted a few layers especially the trees so that I would attach the snow cube to the camera, and also found that some stones/rocks were floating which I deleted.
Don't know about the others, but I'm have lots of fun with these files. Thanks Rashad
mermaid010: Nice result!
mermaid - looks very nice.
Thanks Hansmar and Horo
Great work again, Mermaid! Very smart way to include the falling snow. That sunless sky fits the weather indications quite well.
OK, I finally did some real work with one of the files. The same one as Mermaid used, but the second camera position. I activated the cloud layer and added a few elements (pillara and a boat). I also changed the position of the sun.
Now, I noted a few things: the sea level is not at the zero y position. Why not? I noticed when trying to land the boat on the water: it stayed quite some distance above it. Now, I could have moved everything (including the camera) up to have the sea at zero, but only thought of that too late.
I also increased the amount of clouds. In my view (as you can probably see in the render) this leads to the render being a little dull (too dark). I have now changed the diffusion level on the terrains and the clouds and made the clouds more white on the light parts, but that still needs to render. So, you'll see that for comparison later. I do think, that the shadow is not too dark (too dark shadows being one of the reasons for Rashad's light system), but in this case, without modification of the textures, the light parts are not light enough for me.
So, what do you think?
Edit.
Seeing that there is no reaction yet, I decided to upload two more render from the same vista. The second render is made by increasing the diffusive level of the various textures (except the columns and boat). And I also made the white part of the clouds more white. The third one kept the diffuse as in the original and the white part ot the clouds as well (if I remember correctly), but now I increased the sun intensity to 250 diffuse and 150 specular. I think both the new renders are better than the first, but I think there is not so much difference between the two new ones to decide which one is better and why.
What are your thoughts?
Edit again.
After scrolling back and forth between no.s 2 and 3 I now see some differences. In no. 2 there is more light on the terrains, but it looks a bit more hazy in total than no. 3. In no. 3, you see more difference of light on the columns, which are lighter to the left than to the right. I think i prefer the third one, based on both (small) differences.
Rashad - thanks, enjoying the files.
Hansmar - nice examples, I like the 3rd render too,
Yup, nice examples, Hansmar.
Hansmar,
Yay! You've had some time to play. Thanks for jumping in and getting some conversation going!! My apologies for not replying much sooner. Multiple days of not being around here is rare for me. Working two basically full time jobs means my personal time for 3d art has dropped to near 0. So I'll just have to MAKE the time somehow.
Embedded within these scenes are perhaps dozens of unique tutorials that would take to long write out. So thanks again for asking, as these questions combined with the files allow me to truncate entire tutorials into mere responses about these files.
To your questions.
0x0x0?- I'm guessing there's about a 9ft difference in the altitude of the water surface vs the area where the ground terrain flattens out. I had to decide whether to assume 0Y (horizon) was going to be at water level or if it was going to be at "ground level." Assuming these scenes were already going to be confusing enough, I decided to make it easier to import characters from DS into Bryce MgScp scenes on what appears to be ground level. Otherwise imported models would have ended up underground, requiring a lot of repositioning before anything useful would become visible. But there is a secondary reason involving the atmosphere. The Fog function in the Sky Lab bascially falls off above the horizon. By placing the water level well below the horizon, I give myself the option to use fog effects above the water surface. Fog is often an essential tool to helping give water the apperance of depth underneath, but it can do useful things above the water line as well if the water line is below the horizon as it is here.
As you can see, every choice in these projects has at least a single if not dual purposes. So do not hesitate to ask anything at all about these projects, because behind your seemingly simple question is likely an entire tutorial. Just for fun you should turn up the Fog in the Sky Lab and see what happens to your boat! WARNING: The Fog Sliders in the Sky Lab operate in what appear to be the opposite manner of the Haze sliders. Don't ask me why!
Cumulus Layer- It's important to realize that by increasing the clouds density you've essentially changed the weather. The difference between a mostly sunny to a partly cloudy to an overcast sky could be just a few units higher or lower on the Density Slider in the Material Lab. No problem. The whole point of EGFPLS is to provide you will a safe envelope of indirect light even in camera shots where direct sunlight is not available.
But realize as well, that without that direct sunlight influence, the scene will not pop in the manner it did before. This is as expected however, because overcast days in real life also don't give photos that pop.
Looking further, it looks to me like you may have accidentally changed the diffuse output intensity of the sun before you made the first render. In the first render, even with a significant density increase the clouds should still have been fairly bright in the areas exposed directly to the sun. The original diffuse output setting for the sun is 260 with a 520 for specular output. Any lower of a diffuse setting and the clouds will begin to lose the contrast we need. That isn't to say that there isn't any loss of light on the world beneath the clouds, because the clouds might be dense enough now to block the sun from striking this particular area. The clouds cast distinct shadows onto the landscape. Due to the scaling of the environment, and the placement of the cloud slab at the actual relative altitude where it would reside in nature, the clumps which cast shadows onto a given patch of ground might be located far away. The angle of the sun could mean that the shadows from the clouds stretch pretty far, 2 miles or more when the sun is at really low angles. Sometimes one just has to play around with the clouds to get the density you want while still leaving an open space for sunlight to penetrate the area of the ground where your camera happens to find itself. There is a lot of room for customizing these clouds.
The Cumulus Cloud Layer is a master tutorial on its own. One manner of alteration that works well is to randomly dial the Orientation setting in the Material Lab (not the DTE). It's okay to spin the dials any way you'd like in regards to orientation. It will generate new cloud formations each time, offering different positions for clumps and vacancies that can often fit your given needs. Altering the Scale can have interesting effects as well due to its linear functionality. You will notice that the scaling is expressed in the thousandths, so do be very careful and never spin dials randomly on this particular parameter. If you double the repetition of the scale, the cloud clumps will appear to become half their current size in a linear fashion. This is good information to have if you ever decide you want to scale these clouds down to where they are more managebale for conventional Bryce projects.
I experimented a bit with the EGDLS lights. The 1st – my normal setup, using a CGAxis tree. Regular render, Boost Light with reflection correction. Render time 45mins. The 2nd I included the EGDLS lights group, I did not change anything and the render time is almost 7 hrs. The screenshot shows the material for the leaves.
Mermaid - nice tree form. I think using full ambience is not an ideal choice to show how much EGDLS contributes to the light.
Horo - thanks for your comment and reminder about the ambience. I think Rashad did mention something about. I wonder if reducing the Ambience will improve the render time.???
mermaid010 Horo - thanks for your comment and reminder about the ambience. I think Rashad did mention something about. I wonder if reducing the Ambience will improve the render time.???
It is necessary to reduce the ambience to 0. This does not affect the rendering time, but it will be realistic from the viewpoint of lighting by Rashad.
Mermaid,
Several things to comment upon here. Firstly I'd like to thank you for taking the EGDLS rig and applying it to your own works, and asking questions along the way. The image examples and the questions are essential because you are not the only person who will have the exact questions you are posing. Once again, your seemingly simple questions allow an opportunity to increase understanding of the way the render engine works as a tool, why it works this way, and what we must do as artists and content creators to get the job done with the greatest efficiency
Alexey and Horo are both more than familiar enough with my thoughts to fill in for me at any point along the way. They know everything I know. No need to wait for me, fyi. I consider them both to be correct in their observations. There are two things I'd like to discuss in this instance; Blend Transparency and Ambience. Let's start with my personal favorite.
Ambience...
"Ambience is evil!" was my forum mantra on lighting for many years. I suspect it will be inscribed on my Bryce tombstone. It was the search for alternatives to this very effect that has led to the obsession with lighting I find myself still carrying to this day. Had lots of debates about it over the years and might have even lost a friend or two along the way. But eventually, many people started to see for themselves the problems ambience presents. I'm much less dramatic about discussing ambience nowadays. While it may seem silly to have once anthropomorphized and infused a Material Lab effect slider with human behavioral characteristics; at the time it seemed necessary to do in order to shake the community members of the day away from their unquestioned reliance on Ambience. The ease with which total reliance upon Ambience takes root in new Bryce users is what led me to call the effect a bit of evil. Most users give up the quest for better indirect lighting methods once they get comfortable with the speed of ambience and go blind to the obvious drawbacks. The going blind to the obvious drawbacks part is most fascinating to me. More on that some day long from now. I'll explain.
Due to technological limits of the time, Ambience was once considered essential for all materials for the simple reason that it was fast enough to keep pace with the slow processing power of the time. Without it, all the shadows in the scene would render as fully black and totally featureless unless one added dozens of point lights leading to crazy render times. More on that in a bit. While better, more accurate methods of indirect light transport were already known, they were not yet feasible especially at the hobbyist level. But there was also another, more deeply rooted reason. That reason being that Ambience looks a lot like the results we've observed our entire lives from cartoons and canvas paintings and other man-made 2d image representations. It wasnt until one sat down to try to recreate an actual photograph that they'd find there was very little if any tolerance for Ambience allowed if true realism was the goal. Why does Ambience disagree with realism? It's a simple issue of thermodynamics and blackbody curves.
Ambience: Its Intended Purpose
1. Indirect Lighting; In particular, from the sky.
2. Thermal Excitation; in particular that of hot materials like fire/ lava / light bulbs / laser beams / Magical effects (blackbody curves) where the warmer a surface the more brightly it radiates light as it attempts to cool down
Consider this; On the one hand we expect Ambience to represent the incoming light received on a surface from the surrounding environment. Yet on the other hand we expect Ambience to represent light being emitted by a thermodynamically excited surface into the surrounding environment. These two ideals are the complete opposite of one another. Are we asking too much of a single tool to perform both of these highly divergent tasks at the same time? I'd say yes, most definitely. The problem with Ambience is that by itself without help from other lights it cannot do both of its two jobs at the same time, it is usually better when Ambience is used for one thing or the other. Either as indirect light from the sky or as a state of thermal excitation for some super heated object. Trying to do both at once usually fails in my view.
But there is another HUGE problem, in that Ambience is only capable of performing half the job of a true skylight, and only half the job of a true glowing object. So since it fails at producing either of its intended effects reliably, why should I invest time in ambient materials when I'm obviously going to need other lights to help me sooner or later?
Ambience:- the GOOD
1. The fastest form of lighting available in Bryce, or any other software for that matter. It provides a clean, pure, noiseless and fully uniform illumination. When used in this manner the Sky Lab controls are essential.
2. Heat. In Bryce at least, Ambience has control options at both the global (Sky Lab) and local (Material Lab) settings arenas. When creating hot surfaces, it is in the Material Lab settings where Ambience is tweaked.
Ambience:- the BAD
1. Ambience fails at representing Skylight in a realistic manner. Ambience can only provide perfectly uniform light from all directions with no occlusion whatsoever; it cannot provide any geometric shading nor can it accept shadows from itself (self shading) nor from other objects. This is why Ambience is so fast, because it skips all shadow related calculations...very very good for speed. But without shading effects (especially self shading), Ambience appears as a uniform lighting that has a "flattening" effect on the 3d render, making it seem more 2d, more like a cartoon or painting, which is a waste of 3d potential. Shading is the master key that unlocks most everything else in a realistically inclined render. Realize that in real life all lighting effects, even soft light from the huge sky still casts some form of shadows. Even if those shadows are extremely soft, they are still present and essential if true realism is the goal.
2.Ambience fails at representing states of Thermal Excitation in a realistic manner. Ambience may add brightness to its own surface, but it cannot lend any light to the surrounding surfaces in the way a real world emitter would. Hot lava not only glows itself but it also lends light to any nearby surfaces, that's how we know it's hot! Just as Ambience doesnt accept shadows from other surfaces, so then does it fail to lend light rays to other surfaces. Multiple surfaces in the same scene with Ambience will "glow" in complete isolation from one another; in completely different universes. As anyone who studies realism will tell you, Interaction is the key to realism. So any effect that lowers the degree to which surfaces interact with one another via the four main interactions (shadows/reflections/transparency/radiance) will cause the render to appear less realistic.
Now look closely at the two examples you submitted. Even with the offending Ambience at 100, the EGDLS rigged version demonstrates a powerful point. Notice how the light dome approach spreads highlights in a manner almost as uniform as the Ambience, but it has the added benefit of casting shadow rays as well. The benefit of those shadow rays is that the EGDLS render, while fried with Ambience, still has a more 3 dimensional feel to it than the one lit with only Ambience. The leaves remain more distinct, instead of meshing into an "Ambient Soup" as they do in the first render.
Conclusion
TA, IBL, and EGDLS are all replacements for Ambience. They are much slower rendering than Ambience, because they do not skip all shadow calculations while they work. And if you ever look at the numbers of a Render Report from Bryce, you'll see thatt the shadow ray firings greatly outnumber the primary ray firings (light rays) by a 10 to 1 margin often times. There's no escaping it, shadows COST! When using these more advanced lighting methods, Ambience serves no real purpose anymore. The only thing you'll need it for is to brighten surfaces which are to be perceived as hot (thermal excitation), but you'll still need a point light or two to create the outgoing light to affect the surrounding environment.
Remember: Cool to the Touch
For realism, you should be able to imagine reaching into the screen and handling the objects of the 3d world. Feel them make contact with your skin. Does this object appear as if it would be cool to the touch if I reached for it? If not, then there is too much Ambience. What I see when I look at these two trees are leaf surfaces that are so thermally excited (hot) that they literally glow...becoming brighter than they would be if only reflecting the surrounding incoming light. I think I'd get a burn. Ambience (to my eye) raises the apparent "temperature" of any surface you apply it to, so use Ambience sparingly if at all. Use it as a final tweak perhaps while in conjuction with other methods doing mosty of the work (TA / IBL / EGDLS / EGFPLS), but never rely on Ambience to do the heavy lifting alone if realism is the goal. If fantasy is the goal, then more ambience is ideal.