Here is a test with MrHoo scene.
Area lights…FG (2 bounces)…GI
Some problems with one window… where the light intensity is too much, but its easy to resolve.
The idea of using area lights is to create some contact shadows(not only with occlusion), also to bring some more natural interior light.But we need to be careful with overexposed areas.
Mental Ray skills test scenes
Hi slipknot66:
Great one!! I always had fear of using area lights, because my results were noisy and slow…
But I like your result! In my scene there is one problem with the right window (if you look on your picture), because the window frame has a little space to the wall… My modelling mistake 
Could you upload the scene, that I can have a look into ?
CU
Dominik
Hi slipknot66:
Great one!! I always had fear of using area lights, because my results were noisy and slow…
But I like your result! In my scene there is one problem with the right window (if you look on your picture), because the window frame has a little space to the wall… My modelling mistake 
Could you upload the scene, that I can have a look into ?
CU
Dominik
sorry - double post
Where do you clean up the shadows in the sun/sky thingy. I have some majorly dirty shadows very grainy.PS Thanks MHoo for the advice eairly it helped tons!
Play with your FG ray settings, and throw a shadow transparency on the glass table so the light goes through it and creates the proper shadowing.
Go to Hypershade–Utilities–select the mia_physical sun node, go to samples… there you can start with a value of 16 that im sure you resolve your problem.
Nice interior.You could use some area lights there 
MrHoo, i dont have access to the computer where i created your scene, as soon as i get to that computer i will post your scene back here.
Sorry for the delay answering you 
Thanks SlipKnot66,
I found it pretty much right after i wrote that. Sometimes i get to dependent on these forums.
But now i do have a legit questions that i looked everywhere for but can not find anywhere. I am using IES for the Parabolics in that scene but anywhere i have glass it throws a huge burn out reflection of the point light that is used for the IES. Is there a way to turn lights reflections off?
I will post an example if needed?
Hey guys, I’m glad to see nice test renderings and scenes being shared here. Thank you for your contribution. And sorry for the lack of reply on my side – I’ve been busy lately.
IÂ’m also putting heavy test rendering activities on hold until I buy a faster computer, so that the play time is more fun and efficient. My current computer was due for replacement almost a year ago but I decided to get a new one after Maya has a version written for and tested on Vista with the latest hardware. I was able to afford this since I still do more 2D stuff but IÂ’m gradually transitioning to 3D. Unfortunately Autodesk hasnÂ’t listed yet qualified hardware for Maya 2008 but I expect them to do so soon. Once I get a faster machine IÂ’ll resume my tests. Meanwhile IÂ’ll enjoy your participation and contributions here. Keep the good job up
and post your test rendering and scenes.
With interior scenes with light coming through windows from the physical sun & sky environment, I usually put the sunlight in the middle of the room, make sure that the light is not scaled, (to only display it larger, you can change its Locator Scale attribute in the Object Display section of the AE), let Mental ray calculates automatically the bounding box of the scene, and enable the automatic energy in the sun node. If the light is not facing the windows, I put a helper plane with its primary visibility turned off, outside the window in such position so that the photons shot from the sun direction bounce back into the room from the plane.
To be honest IÂ’ve never been able to figure out how to control the focus of the photons with the bounding box values feature and thatÂ’s why I always use the default values of 0. Here is an example of the problem - I tried to visualize the bounding box effect with creating a dummy poly box without a shader parented to the light.
[img]http://i111.photobucket.com/albums/n129/Emil3d/Photonboundingbox2.jpg[/img]
in the image above, the sun light is emitting photons with automatic energy and the Photon_bbox max values are equal to the size of the dummy box that encompasses one of the rooms. Following the help documentation I would expect that the photons will mostly land in the room within the bounding box, but as you can see the photon map visualizer shows thatÂ’s not the case. I increased the min values of the bbox more than a few numbers but MR did not store any photons. It stored photons when I made min = max values which gave a result identical to the default 0 producing equal distribution of photons on all rooms as shown in the image below.
[img]http://i111.photobucket.com/albums/n129/Emil3d/Photonboundingbox.jpg[/img]
This is all what I can find in the help about it.
When photon_bbox_min and photon_bbox_max are left set to 0,0,0 the photon bounding box will be calculated automatically by the shader. If these are set, they define a bounding box in the coordinate system of the light within which photons are aimed. This can be used to “focus” GI photons on a particular area-of-interest. For example, if one has modelled a huge city as a backdrop, but is only rendering the interior of a room, mental ray will by default shoot photons over the entire city and maybe only a few will find their way into the room. With the photon_bbox_max and photon_bbox_min parameters one can focus the photon emission of the mia_physicalsun\ to only aim at the window in question, greatly speeding up and enhancing the quality of the interior rendering.
I’m not smart enough:scream: to figure out this MR “feature”, if someone understand how this works, please let me know.
Also I have no idea how to visualize or comprehend the difference of the min and max values after reading such explanation. In my imagination in the above example the min values when larger than 0 should cause a lack of photons (indirect light hole) in the middle of the room.
I had a chance to play with the portal light in Maya 2008 but not with its Emit Direct Photons feature enabled. However from the examples in the documentation it seems it works as a much better alternative to the Automatic Photon Energy options of the physical sun which is still available there and IÂ’m not sure how and if it will work with a portal light.
Here your expectation is wrong: Photons would be aimed at the dummy box, but they are not stored there (photons are not stored on their 1st interaction, because direct lighting takes care of that).
So the expected locations of photons is places where the photons would bounce when aimed at that box. Which is precicely what the photon distribution in your image suggests.
/Z
Hey MasterZap, thank you for your help. I really appreciate it.
Actually IÂ’m well aware that the photons are not stored after the first intersection and I expected to see photons inside the entire room encompassed by the bounding box in the same way shown in my second image from my previous post when the values of the bounding box were set to 0.
But hereÂ’s the funny thing, since I didnÂ’t save that experimental test scene, I tried to reproduce the problem again and after several tries I couldnÂ’t reproduce it. Now it is always working as expected. No matter what angles and even different bounding boxes I tried. As long as the bounding box covered the window the result was always as expected. This is the kind of result I expected to see in my first test.
[img]http://i111.photobucket.com/albums/n129/Emil3d/bbox1.jpg[/img]
[img]http://i111.photobucket.com/albums/n129/Emil3d/bbox2.jpg[/img]
You can clearly see the difference between these and the first image from my previous message. I feel so wired because IÂ’ve been experiencing this problem on most of my tries and now all of a sudden luckily I canÂ’t reproduce it:). IÂ’ hope it was some omission on my side but I canÂ’t imagine what I couldÂ’ve done wrongly so many times before and not doing it now. IÂ’ve done stupid mistakes before but this one is kind of surreal. If I experience this again I will save and share the scene or my findings.
Anyway tank you for your help again. I also would love to hear from you a few words about the min values in the bbox and what their meaning is. Can I visualize them like I did for the max values with a dummy box. Leaving them at 0 seems to be OK but IÂ’m curious in what cases one would change them.
Your help is always highly appreciated here:thumbsup:.
Hi Emil3d.
That’s a nice test setup to understand how the photon bounding box works.
The min and max values... Each of those numbers define a plane in a 3d coordinate system, giving us 6 planes in total. Each of these planes contain one face of our bounding box. So those values are not sizes, but coordinates that determine the bbox. The dimensions of the box can be calculated by maxCoordinate - minCoordinate.
So imagine that we have (0, 0, 0) for min values and (1, 1, 1) for max values. We'll have a unit cube that has a vertice sitting on the origin of the coordinate system.
[img]http://img.photobucket.com/albums/v179/KakoCG/physicalSun_BBox_000111.jpg[/img]
One more example: min=(0, 1, 1), max=(2, 2, 3). We'll have a 2x1x2 box.
[img]http://img.photobucket.com/albums/v179/KakoCG/physicalSun_BBox_011223.jpg[/img]
Now, imagine that this coordinate system is actually our light's coordinate system. In this situation, the light will be the origin:
[img]http://img.photobucket.com/albums/v179/KakoCG/physicalSun_BBox_lightsCoordinateS.jpg[/img]
Now we can see where the bounding box would be in relation to the light.
In the image above we're using the cube of our first example, where the min values were all zero. Thus, notice that the bounding box has a vertice sitting on the same position as our light. So it's important to see that if we just leave our min values at (0, 0, 0) (and considering we're not using automatic calculation of the bbox), the light WON'T be in the center of our bbox. If we want this to happen, the min values must be equal to MINUS the max values, for example, min=(-2, -2, -2) and max=(2, 2, 2).
That explains why you got that odd result with max=(23, 11, 21). In fact, the bounding box was something like this:
(Notice that there's a vertice on the same position of the light, since your min values were zero.)
[img]http://img.photobucket.com/albums/v179/KakoCG/physicalSun_BBox_EmilsError.jpg[/img]
A correct constrain of a dummy box to the light would be something like this:
[img]http://img.photobucket.com/albums/v179/KakoCG/physicalSun_BBox_correctConstrain.jpg[/img]
That's also the same unit cube of our first example. The important thing is that if our min values are all zero, then the box is contained within the >=0 semi-axes, as you can see in the picture above.
Some tests now. I'll use a scene like yours, but the hole in the cube will be on the ceiling instead. And let's use only one cube with the light at the center of the hole. I set the photon tracing limit to 1 reflection only to make it easier to diagnose the result.
[img]http://img.photobucket.com/albums/v179/KakoCG/physicalSun_BBox_01.jpg[/img]
[img]http://img.photobucket.com/albums/v179/KakoCG/physicalSun_BBox_02.jpg[/img]
[img]http://img.photobucket.com/albums/v179/KakoCG/physicalSun_BBox_03.jpg[/img]
[img]http://img.photobucket.com/albums/v179/KakoCG/physicalSun_BBox_04.jpg[/img]
In the first three situations, as the bbox covers less of the hole, see how photons get a little bit more concentrated to that corner. The last image shows a uniform distribution of the photons. Notice that the min and max values are opposite in the x and y coordinates.
The z coordinate is not that important in our scene. Anyway…

So… When you draw a dummy cube representing the bbox, that’s actually a way of visualizing both the min and the max values.
Kako.
Awesome explanation Kako:thumbsup:. If a simple illustration like one of yours was in the documentation, for me that would have cleared any questions about what those values mean.
In my opinion the creation of this metaphor and its explanation in the documentation is from and for scientifically thinking mindset. Unfortunately for people like me who are used more to life like experience metaphors this cannot be naturally understood without some aid like your explanations and illustrations.
Thanks god that this is not the case for most of the user interaction in Maya which is based on real life metaphors like for example: creating a cube in Maya is pretty much like in real life building a box in the middle of a room with a ruler that has only positive units of measure:).
Thank you for taking the time to put together all these excellent illustrations and clear explanation. You did a great job with this:thumbsup:
Sorry guys for the crummy hard-to-decipher interface… of course, all this was “supposed” to be hid away in a nice UI (like the “photon radius” aim tool in max) but the real world and priorities shifted that around… besides, the “ideal solution” is to shoot sun photons only through the portal lights… (… maybe next version …)
/Z
Hey Master Zap, I really appreciate your concern.
In my opinion Autodesk should hire people with communication skills like scientific journalists who can quickly understand complex ideas and translate them to the masses in a more popular language pretty much like what Kako demonstrated with his excellent explanation and illustrations.
Geniuses like you should spend most of their time creating beautiful shaders and improving MR than writing manuals which you in most parts have done a very good job too:).
Thanks, Emilian! :thumbsup: It was worth it, then. I'm glad to help. But if you hadn't shared your tests in first place, I wouldn't have thought of writing those explanations. I see you're very committed to contributing to the forum and you too are doing a great job. Congrats for that!
MasterZap, no need to be sorry. We know it's not your fault. In fact, the architectural shader's docs are maybe the best of maya's and mr's docs. And Emil3d pretty much said everything in his last post. We cannot thank you enough.
Kako.
Thanks for all the info guys! helping tons!
One thing you never see in 3d Viz work is interiors with no sunlight comming in. Just a space with (20) 2x4 parabolics and a lighting setup for that. I have tried multiple ways in this as well is there a good way to do it? I have done IES files but you can never get good bounce with those i have tried area lights in every parabolic to represent the lights it self. but then you dont get good lighting on the ceilings and stuff so i am open for a clean and good way to do this type of scene. I have looked at thousands of 3d viz projects and about 99.9% of them have sun light in them. So if you guys could help me with this one that would be great. 
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.