First, I can't seem to create a "real" looking flame for the candle on the mantel. Second, the floor shadows do not seem right to me, especially under and around the table. I would think this area would be darker without the far legs of the table being illuminated greater then the forward legs, or the knee of the fat officer, who's knee is under the table and shouldn't be lit at all.  Plus Vickie should have some shadows starting at her foot, I would think. In addition, the stool seems a bit bright.

Could the light of the lantern on the table be passing through the table but its shadows stopping on the table and if so why? Also the pz3 of this scene is over 207 MB, but excluding the "heavy morphs" of the seated Officer plus his pants, coat, and boots there isn't any custom geometry, except all the lights and shadow cam lights which seem to be all made of custom geometry, which I don't think should be in the file at all.

The Blockhouse interior is about 20 foot square with the ceiling at 10 feet. The front and left (right in frame) walls not being used in the scene. The table is in the center of the room with a lantern centered above it and another lantern on the table slightly forward. The left and right lanterns are parallel to each other about 4 feet forward of the center lantern all at the same height. The candle is there to illuminate the picture, which it doesn't seem to be doing well. The left lantern is there to illuminate the deck and flags. And the right lantern is there to add some light to Vickie while still keeping the m3 Soldier in the shadows with all lights.

I would like to get Vickie lit more without drastically effecting the m4 officers except adding a bit more bounce to the standing officer's hat, or adding any more light to the m3 Soldier expect maybe get some highlight going in the bayonet and shako plate and plumb on his hat.

The 5 light sources seen in the image are the only lights illuminating the scene. The fire light is created by a group of "Spot" lights circulating the cinders with a semi transparent cylinder of smoke to dampen their effects on the floor. The lantern lights are each created by 2 Point lights with the inter light set at 100% intensity, 0.0% shadow with a end point of 0.250 feet, and an outer light set at 70% intensity, 0.500% shadow with a start point of 0.300 feet and an end point of 11.250 feet. The each lanterns is turns slightly to make the forward glass perpendicular to the camera to keep the illusion of a flame, if not it just wigs out. The candle light is the same as the lanterns, but with out a "glass" panel in front of the light, I don't get the same effect.

I really like the atmosphere of this scene.  Agree that the lanterns look just right
but the candle doesn't.... you could try adding yellow ambient to the flame prop
if it's not already there.

Shadows from point lights always behave weirdly.  You could try moving the point lights
for the ceiling lanterns down below the lanterns; this would eliminate whatever
effect the body of the lantern might have, and would still look good (I think!)

Pulling the 70% intensity, 0.500 shadow "Point" lights down below the lanterns does warm up the scene, allow the intensity to be adjusted without wigging out the flame in the lantern, and does produce more realistic shadows on the floor.

However, on the down side it eliminates the intense flame in the lanterns I thought added to the scene and kind of removed the dapple lighting effect caused from those lights on the ceiling.

I guess the next attempt should be to add three more points of lights inside of each of the ceiling lanterns and try lowing the "Dist End" values to get the flame back but not cast light much farther past the lantern as possible.

The frustrating thing is that these Point lights don't illuminate anything in the preview mode and it takes several hours to render the scene before I know anything about the effects of any adjustments made to the lights.

just wanted to mention that we seldom see such a well-constructed scene here.

however, when setting the shadow dial on a poser light, as was probly noted above, any setting less than 1.0 allows lite to pass thru solid objects.  setting shadow at 0.5 allows half the lite to pass thru.  setting shadow dial to 0.0 allows all the lite to pass thru. some users suggest that this is a bug, as they feel that lite should not pass thru opaque objects, but it's actually a feature, not a bug.

Miss Nancy,

Well, thank you and thank you for the information. That is very helpful.

I'm a Painter, Cartographer, and Draftsmen and sometimes make 3D widgets. I don't really know Poser that well, except it has been a pain to use for me since version 1. I did not know that about the shadow settings and somehow always assumed it was the intensity of the shadow cast that the dial controlled, not its X-ray powers. Silly me! I always thought transparency of an object controlled its density or opaqueness.

Normally, I would just export the scene to a renderer and figure the lighting out in there just to avoid the frustration, but doing a complete scene in poser is the objective.

You can also try the inverse square shader for the point light from Bagginsbill !

I don't know what that is, or were to find that.
I've read a few posts that mention "inverse square" posted by Bagginsbill
but at a great loss to understand what it is.

It is the key to lighting an indoor scene.

Very briefly:

Luminance is like temperature - it has to do with the amount of light energy within a given area. As the light energy moves away from the source, it is always spreading out over a larger and larger area. Since the amount of energy is not changing, but the area is increasing, that light gets spread thinner and thinner, appearing weaker.

The math is simple. The luminance is inversely proportional to the square of the distance from the light. Do not say your head exploded. This is nonsense I hear all too often. If you are in a society where it is shameful to be able to understand a simple math formula with only 3 terms in it, then go hide under a blanket and READ THIS.

Let's say "d" is the distance between some prop and the light source.

d is distance

Let's say "I" is the intensity (strength) you choose for the light in Poser.

I is Intensity

Ready? The illumination from any light should be

I / (d^2)

That's I divided by d squared. Inverse square falloff.

It is neither difficult to understand, nor expensive to compute. It is an enormously obvious physical effect, and without it indoor lighting looks totally fake.

Yet, Poser lights do not implement this formula. We need not explain why such a fundamental fact of the universe was ignored in Poser, despite it being trivial to calculate. We need only find a way to add it to the calculation.

Thus was born my Inverse Square Falloff (ISF) light shader.

I don't know where it is at the moment, but I'm sure somebody can point you to the thread.

ISF lighting will be ruined without gamma-correction. You need to do both for even a modest amount of realism.

Once you do those two things, you'll find that lighting your scene suddenly makes sense. You'll find that the Poser lights do things as you'd expect them to do based on your experience with being an actual human in the actual universe.

Well as a once and future Cartographer, I think I kind of understand simple math. As understanding "Poser's" simple math, that's something else.

I do know know that the suggestion by Miss Nancy to set the shadow setting on all the lights to 1.000 is in fact eliminating the "x-ray" ability of these lights to cast light through walls and tables and such, eliminating the odd lighting under the table. But the down side is it's creating a very harshly lite environment that doesn't seem natural at all. It looses all that subtle inter-play of the lights in the scene.

IsaoShi, if you could post bagginsbill's ISF widget and if he doesn't mind, I would like to try it.

Here is the ISF shader. Save it somewhere in your Materials library folder and remove the .txt file extension.

The shader has four parameters. When you have positioned a light, its position must be copied into the XYZ parameters in the shader. This is so it can calculate the distance from the light source to the point on the surface currently being rendered, and hence the intensity at that point.

The other parameter specifies the distance from the light at which the actual light intensity will be equal to the light's dialled intensity. I usually leave the intensity dial at 100% and use this parameter to adjust intensity - I find this method more intuitive, but you can do it either way.

If you use Parmatic, it will create dials for these four parameters on the light's Parameters palette.

By the way, the 'very harshly lit environment' you refer to is typical of images without gamma correction. Bad lighting is not the cause of this, it's simply because you are displaying a linear colour space render on a non-linear (sRGB) display.

I'll stick my neck out and say that with GC you will only need a single ISF point light at each source to achieve a realistically lit and nicely atmospheric scene. You could add a very low level IBL to simulate general ambient lighting, but my guess is you won't need to.

Unfortunately, applying GC to all those materials is going to be a fair amount of work!

In the second shot using ockham's suggestion to drop the lanterns main "Point" lights below the lanterns themselves and turning off the light in the lantern on the table, the softness and richness of the environment was still there and now adjustable. It gave some lift to V4's rump. However, it did eliminate the illusion of a bright flame in the lanterns, but thought I could add three more "point" lights within these lanterns with a shorter "End Dist" to restore that look. Granted it still had some floor shadow problems and under table lighting problems, but it seamed like progress.

Miss Nancy's suggestion to have all shadow setting set to 1.0 did eliminate that odd "x-ray" lighting below the table, but as seen in this latet test shot created what I think is very harsh lighting compared to the first two. V4's back is black, the face of the standing officer is flat and the seated officer's face is sliced down the middle, the lantern shadow pattern on the ceiling is very dark and the floor is just a maze of almost monotone bright shadows with very little definition under the table.

As to "Gamma Correction" I'm not sure exactly what that entails or where that is adjusted. I run a mac and back in os9, Adobe had a system extension to change the gamma setting of one's monitor for use with PhotoShop, but osx 10.4.9 doesn't seem to have that option in its system preferences anymore.

There's a whole bunch of information on Gamma Correction here in the forums.

But put very simply, an unadjusted render coming out of Poser 5, 6, or 7 (Poser Pro has built-in gamma correction) does not have a correct Levels curve applied to it for display on any computer monitor -- because all computer monitors are calibrated to receive and display images with a specific Levels correction curve. This manifests itself in darker areas of the image being much too dark - exactly what we can see in your above image.

There really is sufficient light in those dark areas, but unless you gamma correct the image you won't be able to see it properly on your monitor. This is not to do with the gamma correction of your monitor, but of the image itself, so that it displays correctly on your monitor, and on everybody else's monitor.

(Having said that, I'm also on Mac, and I have adjusted my monitor gamma correction to 2.2 in the display calibration utility. With the release of Snow Leopard, Apple will set 2.2 as the default gamma correction in OS X. Up until now they have always used 1.8. This change will have the effect of making your images look even darker, but the point is that they will be closer to how most other people see them).

So what do you do? Well, the simplest thing is to apply a gamma-correction curve in Photoshop, using a Levels Adjustment layer. That will get closer to the correct lighting levels, but your material textures may begin to look a bit washed out. The reason for that is that the original colours of those textures were already gamma corrected in the image maps you are using, and now you are doing it again on the final render -- so they get too bright, and lose colour depth.

The correct way to apply gamma correction within the Poser material shaders is to remove gamma correction from the source image maps for the textures, and then re-apply gamma correction to the pixel colour values in the final render.

This is the method that is covered in depth by bagginsbill and others within the Poser forum threads on this subject. Most recently, bantha has posted a Python script which applies the required gamma correction nodes to a material shader.

now if we could get that shader to read IES Light files...

lakota, in tiger, use colorsync utility to see what yer image looks like on their PC monitors.  it allows one to change the gamma from 1.8 to 2.2, which (in the simplest sense) will make yer image look darker.  aside from bantha's gamma script, another way to spruce up the shadows and lighting is to enable the indirect diffuse calculations in poser's currently available FFRender, but in a 208 MB scene, such a render would take at least a week, hence the quicker way is to correct the textures as mentioned above.

You should be using IBL as well. In this sort of environment, there would be a lot of secondary lighting bouncing off surfaces. Some IBL will decrease the harshness of the light/shadow comparison.

Personally, I would generate an IBL that matches the direct lighting you've produced, using my GenIBL tool. This would allow you to include some approximation to the indirect lighting of the scene that you would get if you had a render with global illumination.

Well, I do something with the gamma and setting it to 2.2. I don't think I used the colorsync utility, 'cause that is asking me to agree to its terms before it even opens, so I must not have evened it before. I opened "help" and typed in gamma, hit a link, then another link and walked through something. I did set that up before at some point when I installed a new monitor a few years back.

Fondly remember the "classic" old mac os when you knew where everything was and what everything did.

Now the images are a bit darker, but not by much. The scene is now 207 MB down from a high of 213.5 MB. It takes my old g4 Quicksilver ppc about 3 hours to render a 803 x 570 image of it set low in raytracing in firefly rendering. At finish rendering settings at 1170 x 708 with hair on the seated figure, that's a rendering best left to when you're sleeping and be surprised in the morning as to what it looks like over a good cup of coffee.

I need to experiment with bagginsbill's IFS shader. Maybe just rebuild the scene again, start afresh for the fouth time. Throw a lantern in an empty Blockhouse and see its effect then start adding the stuff again. It's all chunked together in my libraries, all parented together in some way or another, not hard to reload.

Excluding the figures and such, all the props and clothing are my stuff, so redoing the texture maps won't be that hard. Excluding the clothing, they are all in the same folder, well the textures for the Bockhouse are in one folder and the prop textures are in another. V4 Vickie's outfit is in another, the m4 officer's uniform is in another, and the m3 Soldier's textures in another. And all those folders are in one 1812 folder.

The point now is to get a handle on this effect, make adjustments to refine it, and then fill the room up with stuff again.

I should say, I tried lower settings of 100% Intensity, but that wasn't popping through the lantern's glass at all, unless something is dramatically wrong with the texture maps which have not been change or gamma altered in any way.

So here is the scene with only one lantern over the table in the room. The lighting is now an inter point light set at 200%Intensity, 1.000 Shadow, with a Dist End at 0.005.

And an outer Point light using the ISF  shader set at 300% intensity, 1.000 Shadow, Dist Start at 0.500, Dist End at 0.000, and the Intensity Distance in the ISF shader at 500.000 for its value_1.

There is also a sky dome in the scene, but it's outside the room and shouldn't have that much effect.

I don't know if this is a correct setting with in a 20 x 20 x 10 foot room, but it's a starting point.

You've missed the point completely. The Dist Start/End business is Poser's way of doing falloff and should be disabled. At Dist End, the intensity goes to 0. No light does that in real life - never goes to zero. That's why we can see stars from billions of light years away.

You must turn off all built-in falloff, then set up the ISF light.  You will also need to use a gamma correcting shader to get these low levels if illumination to appear accurately on a computer monitor.

I know that doesn't seem to make sense - you've never need to before, right? But you've finally hit the wall where you can't ignore these things or hack around them with more light. Using more light spoils your scene.

Maybe others can chime in here to help with details. You've chosen to need help on 4th of July weekend - LOL. Those in the UK are around.

I"m about to head out for a day of sailing and drinking. They go together.

The ISF shader I gave you, which was my own 'tidied up' version of bagginsbill's original shader, had some mistakes in it. The Ratio and PM:Intensity_Distance nodes were incorrectly set up. It does not give anything like the intensity of light that it should. I attach the corrected version.

Sorry, sorry, sorry a thousand times, to you and to bb, for distributing something of his with errors made by me. I think others have downloaded this incorrect version of the shader, and I'll raise another thread to try and put this right. I feel really bad about this... sorry.

I also forgot to mention to you that the ISF shader must have all its parameters entered in inches. I'll do another post in a minute showing what it should look like with the settings you are using.

(edit) "sailing and drinking go together...".  As long as you are not at the helm, bb!

bagginsbill, respectfully, I don't think I'm missing the point completely.

I'm running tests on a shader node that I've never used and don't know anything about. I don't know much about Poser's material nodes to begin with. First tests created a more or less solid darken room with no illumination at all.
So I set up some parameters that lit the room and will work back from that.

Enjoy the day, lake is a bit calm here for sailing

This setup would give an intensity of 4,800% at around 10.5 feet - far too much light for anything.

There's no point going into what's happening with the surfaces in my image, particularly the reflective tiles that are coming out black. The light is just so intense that my shaders are blowing fuses left right and centre. (In fact, this is a good demonstration of inherent inaccuracies in my shaders, that they should react to very intense light in this way).

Well, that might explain something.

But no need to fall on your sword over it. I often load my own wrong props from the wrong folder thinking that they are the right or newest version forgetting that I never throw the old one away.

I d/l this shader and will continue testing, but inches? I'm in feet or does just the node settings to be in inches

If you want to enter these four parameters in feet instead of inches, you need to change the three values in the P node from 0.0969 to 1.1628. According to bagginsbill, the P node returns its position values in one-tenths of a Poser native unit, equal to 10.32 inches. This needs to be converted to whatever units you want to use in the parameter nodes.

Note to bagginsbill: while checking and double-checking the corrections to my version of the ISF shader, I have come across various versions (all posted by you) with different values in the distance Pow node and in the final Divide node. Can you confirm which values are correct, please? I don't want to post a thread with a corrected version if it's still wrong. Thanks.

(edit) the version I used for the above-posted correction is shown in the image above. The only difference is I did not round up the P-node conversion factors to 0.1

Here is a "test" shot of the interior of my Blockhouse again, now using the correct version or one of the corrected versions of bagginbill's ISF shader.

It has only one light source simply to test the effects of the shader and to just see what's what.

The setting for this ISF light are as follows:
Dist Start 0.000, Dist End 0.000, Shadow 1.000, Intensity 100%, Shadow Blur Radius 0.5, Shadow Min Bias 0.8, Atmosphere Strength 0.01, and in the ISF shader the Intensity_Distance Value_1 is set at 60 inches. I did not change any other settings.

For what it's worth, I think this test is doing what's suppose to do, figuring out things.
If I keep the Intensity_Distance Value_1 low enough, I eliminate the need for a second point of light within the lantern to emulate a bright flame. I did run a test with a Intensity_Distance Value_1 set at 120 inches, but the light seemed to bright within the lantern.

And that is where things are at the moment.

The center light is still centered and centered over the table set at 7.3 feet above the floor. The other two ceiling lights are at the same height 6.55 feet forward and 3.64 feet to the left and right of center. The light on the table is set at 3.6 feet above the floor, 2.5 feet to the left of center, and 1.1 feet back.

All the ceiling lights are set at 200% Intensity with the ISF's Intensity Distance set at 30 inches. The table light is set at 50% Intensity with the ISF's Intensity Distance set at 30 inches as well.

The scene's lighting is improving, but still needs work.

It has lost the illusion of a bright flame in the lanterns which was there when the two outer lanterns were set back from their present location 3 feet with all lights were set at 100% Intensity with the ISF's Intensity Distance set at 60 inches. However, this setting burned to white the top of the table, but lit the walls and ceiling nicely, casting real sweet shadows although the room in general was a bit bright for the desired mood.  

At the moment, I can best describe the lighting as something you might see on a stage in a theater  in some scene from a play.  It is dramatic, which is the wanted effect, but here, I don't know why the light would be so white under the table.

I would guess some of the problems are caused by the panes of glass in the lanterns themselves. They are not one sided, they have depth.

In addition, I have yet to change any of the texture maps to "correct" the Gamma point as has been suggested to improve the realism of the scene.

The lighting isn't right yet, but the unexplainable bright shot on the floor is almost gone. I thought I had that solved, but Poser crashed and I hadn't saved any of the numerous experimental changes to the ISF Intensity Distances settings made to try to eliminate it.

Yet, I still have no clue as to why Vickie's foot under the table is lit so strongly. To me it makes no sense, if one of the spot lights was the cause, one would think that the forward legs of the table would be effected as well.

Did you read off the coordinates of your lights in inches, and type those into the ISF shader as inches? What are your Poser display units set to? What values did you put into ISF for the light coordinates?

Suppose your light is 6 feet off the ground, and you used feet as units, and typed in 6 for the ISF Y position. That would be 6 INCHES, and so you'd get maximum light under the table instead of above it.

My units in Poser's preferences are set to FEET.
With in your shader they are set to INCHES (I believe).

The main center point light as well as the two spots are set at 7.326 above ground, the floor is 0.02 above ground. It is set at 70 inches (5.83 ft) for Intensity Distance.

The spot lights are both 3.64 feet left and right of center, 10 feet forward and angled down 12 degrees.

But the shader is set up for inches. That is what I was talking about.

You entered your light position into the shader using units of feet. You typed 7.326. That is in feet. But the shader uses inches. That's why you're getting the brightest light 7 inches from the floor. You told the shader the light is 7 inches from the floor.

You need to do the conversion to inches, i.e. 7.326 feet * 12 inches/foot = 87.91 inches. Or, as I always suggest, set your Poser display units to inches and then just read the posiiton off the dials.

Thanks bagginsbill,

I was just looking at that screen shot and pondering that possibility.

I wonder if there is some misunderstanding about the units.

Poser display units only change the meaning of numbers that are in Poser parameters that are used exclusively for dimensions or distances.

The math nodes in a shader are unitless - they do not change they're units just because you're using them to represent a distance. Poser has no way of knowing these things.

So regardless of what you've chosen for Poser display unit, the numbers you enter in the shader position nodes MUST be in inches.

Oh, and Izi, you said to support the shader in feet, change the .0969 values in the P node to 1.1628 (.0969 * 12). You got it backwards. That would make it give data in units that are 1/12 of the inch. You need to DIVIDE by 12, not multply by twelve. If you're starting with a measurement in inches, you DIVIDE by 12 to get the same measurement in feet.

The correct multiplier for the P node in feet is .0969 / 12 = .008075.

But then we'd have to express the reference distance in feet as well, which would require more changes. This is confusing enough as it is. Just everybody use inches for everything and the shader will work right.

Quote - Note to bagginsbill: while checking and double-checking the corrections to my version of the ISF shader, I have come across various versions (all posted by you) with different values in the distance Pow node and in the final Divide node. Can you confirm which values are correct, please? I don't want to post a thread with a corrected version if it's still wrong. Thanks.

The value in the reference distance pow node is different because it's whatever I want it to be. For example, if I know that I have an object that is 42 inches away, and I want the luminance to be 80% at that distance, then I put 42 in there and 80% intensity. If I want 60% intensity at 90 inches, then I put 90 in there and 60% intensity.

I'm not aware of differences in the final Divide node, unless those were cases where I was not using a Pow node for the distance. Sometimes I used to just put the square of the reference distance in the divide, but that was harder to understand. So the divide is always 1/3 and the 1 is connected to the reference distance node which squares it. [ Pow(d, 2) ]

Sorry, bb, I wasn't clear. I was referring to the Value_2 dials in both the distance pow node and the final divide node, not the Value_1 dials.

Your second paragraph explains some of this. But I also saw a version with (I think) 1.3 in Value_2 of the final Divide node. I vaguely remember something to do with simulating light coming in through a large window... the Pad, maybe. Don't worry about it, if I ever find it again there will probably be an explanation to go with it.

Quote - The correct multiplier for the P node in feet is .0969 / 12 = .008075.

Oops.. I struggled with that conversion, first one way then the other, and eventually came down on the wrong side. A blind spot!
I think you only need to change the three P node multipliers, and then enter all four parameters in feet.
But I agree.. we should all use inches and be done with these conversion issues.

Quote - Sorry, bb, I wasn't clear. I was referring to the Value_2 dials in both the distance pow node and the final divide node, not the Value_1 dials.

Your second paragraph explains some of this. But I also saw a version with (I think) 1.3 in Value_2 of the final Divide node. I vaguely remember something to do with simulating light coming in through a large window... the Pad, maybe. Don't worry about it, if I ever find it again there will probably be an explanation to go with it.

Oh, I remember that. Yes you may want to change that.

From my physics classes, I recall that the inverse square [ pow(distance, 2) ] relationship is only for an idealized point source. For an infinite plane, the relationship is inverse linear [ pow(distance, 1) ]. For any real light source that is neither an infinite plane, nor an infinitesimal point, the best approximation is a power law somewhere between 1 and 2. That's why I wrote a 1.3 in some cases.

aha, that explains it - thanks!

bb, I spent the whole journey home this evening puzzling over what we are doing with that P node in the ISF shader.

I can't believe you got it wrong, but neither do I understand why it is right. Another blind spot, perhaps. Can I break down my thinking, so you can put me right.... (sorry to ask, I know how little time you have, but it should be quick).

Thought 1: The P node, I believe you said, returns its values in strange units equal to one-tenth of a Poser unit... or 10.32 inches. So if our scene has a point on a surface which happens to be positioned at XYZ = (10.32, 10.32, 10.32) inches, the P node when rendering this point would return a value of (1.0, 1.0, 1.0).

Thought 2: In order to compare the above P-node position with the point light position that we enter in inches, we would have to multiply the P node values by 10.32. So, for example, if our ISF light position was also at (10.32, 10.32, 10.32) inches, the difference would come out at zero, and our calculated intensity would theoretically be infinite.

Thought 3: Instead of which, the ISF shader divides the P node values by 10.32 (multiplies by 0.0969) before comparing with the inch values.

One of these thoughts must be wrong, but I can't see it. Is it simply that the P node returns its values in 1 / 1000th of a Poser Unit, not 1 / 10th? Please help me sleep tonight!

Approximately tenth of an inch, not a tenth of a PNU. Or, more accurately, 1/1000 PNU.

if lakota is still reading this thread, should be apprised that they augmented the indirect diffuse algorithms in poser 8 to the point where it'll now be practical to render OP's scene with much more pleasing results in re: lighting.

Thank you for the information, Nancy

I'm still trying to get the scene lit. Now I'm having the problem of the over all scene lit well, but now the lantern's pure white. It I get the lantern looking good, the rest of the scene is dark and unaffected by the enclosed light. I can't get the light to pop out from behind the glass unless it explodes the lantern glass and top to bright white

I still think it is easier to just export the geometry of a scene to a better renderer, 'cause I've been working on lighting this scene in Poser for three weeks now. I don't recall it being this difficult in other programs.

However, I'm still chugging along

In review:

Using standard point lighting with no nodes and using the intensity and shadow dials to control the lighting, I was able to achieve a somewhat realistic look to the scene until the light reached a point below the table and then it just didn't work. The shadow dial settings allowed Poser's odd "x-ray" lighting to light areas that should not be lit and the low shadows setting were ghosting the shadows on the floor to almost nothing.

After being informed of the true effects of Poser's shadow dial, numerous attempts were made to correct this by keeping all shadow setting to 100%, but this created a very harsh environment with no subtleties to the shadows on the floor.

After this, bagginsbill's ISF shader was used. Although by mistake I was inputting the wrong location of the lights in his shader node, after numerous trail and error tests  I was able to achieve most of the desired effects until the light hit the floor with bright spots of light and the shadows were off due to the wrong information as to the lights locations.

After this, a forth series of trials were started using the correct location information entered into the ISF shader (I was entering the info in feet when it should be in inches). In these tests, I was able to get the room lit within an acceptable margin to just be able now to tweak the lighting settings with all shadows behaving as they should, except now I have the problem of the lantern washing out to bright white.

In this example the point light in the lantern is set to 50% intensity with an ISF intensity distance set to 60 inches. It illuminates the figures and table well and creates a nice lighting effect on the ceiling. The only problem is that the lantern is bright white. A test was made with the lantern not visible and the lighting was still good.

My best guess now is to add a shot light pointing up within the lantern to create the ceiling lighting effect and another spot light pointing down just under the lantern to light the figures and table.

Obtaining this flame glow in some form or another within the lantern's glass structure and lighting the room is the goal without doing any post rendering in PhotoShop. I know cutting a lantern out of one pic and adding it over another is not difficult, but that is not the point.

Did you ever start using gamma-corrected materials? If not, you're just spinning your wheels. As we said on page 1, this sort of subtle lighting is impossible to get right without gamma correction. Bantha posted a python script that will assist with converting each material to a gamma-corrected material.

Without gamma correction, you are choosing to increase the light intensity in order to compensate for how dark other things are appearing at a distance. This forces the local luminance inside the lamp to very high levels. As a consequence, the lamp is becoming over lit. Once your materials in general are responding to lighting correctly, then you will be able to light the room quite easily with a much lower intensity.

As for the ISF falloff, you mentioned trying to set the intensity very high at 2 inches. This is just forcing you to guess what setting to use to accomplish what luminance you want at other distances. But various combinations of distance and luminance produce equivalent values that you'd get the same results with at other distance and luminance settings.

For example, the following are identical:

2 Inches 1000%
60 inches 1.1111 %

See what I mean? Whatever you choose for 2 inch luminance will result in 900 times less than that at 5 feet. Similarly, whatever you choose at 5 feet will result in 900 times more luminance at 2 inches. ISF is an approximation that is only accurate at close distances if the light source is an infinitesimal point.

In real life, a flame is not a tiny dimensionless point. It has a definite non-zero volume. In other words, the total energy from the flame is not concentrated in a single point. If the tip of the flame is two inches from the top of the lamp, then all the other parts are more than that. The real-life result is that the local luminance inside the lamp is never 900 times the luminance at 5 feet. But that's what the math here produces.

As a workaround, I suggest that the shader on the lamp be adjusted to produce less diffuse reflection. Instead of a Diffuse_Value at or close to 1, I'd suggest a much smaller fraction. What I'm saying is, get your gamma correction going, get the luminance at 3 feet or greater looking right. Then, decrease the reflectivity of the lamp until you've compensated it enough to look like it is lit by a local 3 dimensional flame, instead of a dimensionless point with infinite light density at the origin.

You will also want to employ some IBL because in a closed environment, the primary light source gets bounced around so that all the objects in the room become secondary light sources for all the other objects in the room. With Poser 8, this is automatic, but you don't have that, so you need to use a very low level IBL to approximate the overall secondary lighting.

Note: Don't even waste your time incorporating IBl until you get the GC set up. Low levels of lighting are off by a factor of 10x without GC. Without GC, you'll try using an IBL that is much too bright, and the whole thing will fall apart again.

Thanks, bagginsbill, and thank you for your patience.

I did kind of figured your system was akin to a camera's f-stop and shutter speed relationship. Or water pressure and pipe diameter, not scientifically maybe, but a good visual reference.

I am still toying with it.

The extreme test settings were just that, tests, tests of one variable while keeping the other constant. I get something working in one method that doesn't apply to the next method used.

Basically in all these tests I'm been just moving problems from one corner of the room to another. Each test rendering takes about three hours to render on my Mac, so progress is slow. It is all been kind of like sweeping the floor but never using a dust pan to pick anything up. The best hope has been to make the pile as small as possible and then sweep it under the rug somewhere.

As to materials on the lantern, there are no texture maps. There is a transparency map to create the holes in the top to produce the effects on the ceiling. I think the lantern frame is black, highlight gray, ambient black, reflection white. In my book, given that the scene renders relatively well without the lantern there, it must be in its material settings or its construction. Or I am trying to accomplish two distinct effects not necessarily related to each other a node setting.

I don't know, as I wrote, generally I just move a scene out of Poser and work on the environmentals and lighting in another program. I'm an old school illustrator, I'm visual.

This is all reverse from my thinking on the functions of math. Given an aerial of tire tracks in the sand, I was taught to figure out the probability of what type of vehicle produced them, how much cargo their might be carrying, and what that cargo might be. I was not given a set of  numbers, equations and variables and asked what the image in the sand would look like if it were a specific vehicle with a know cargo.

As to a glow, it is the reflective glow or halo of a light bouncing off the glass that I was referring to. That is all based on the reflective qualities of the object behind it, that object's angle to the light in relationship to the camera angle.

If it were a painting, I would get out an airbrush and do it. Throw in the artistic concept of studying the physical work, the paint on canvas and winging it with soul.

For example, it's quite "illuminating" (forgive the pun) to visualize the curve of the inverse square luminance as a function of distance, as pictured above. I have this image in my head when I think about lighting.

The X axis is distance in inches. The Y axis is Poser light intensity. I chose to graph a light that is 50% intensity at 5 feet (60 inches).

Have a look at 10 feet (120 inches). The intensity drops to 12.5% - still a respectable amount of light.

But the surprising thing is the extreme slope at the low intensities. At 30 inches, we're twice as close, therefore four times brighter at 200%.

At 10 inches, this same lamp is producing energy at 1800% intensity!!!! Imagine what the value is at 2 inches.

So why does lantern glass look relatively dim? Because it only catches a tiny amount of the light passing through it. The "glass" needs to be not 98% transparent, but something like 99.99% transparent, otherwise it's going to look hot as hell.

I downloaded Bantha's Gamma assistant as suggested and used it to convert most of the major components  in the scene. I didn't do all of them, 'cause these are tests, but I did correct the officer's uniforms, the majority of the m4 textures, floor and walls, flags and picture. I did correct the table, but not what is on it. I used the GC button.

I then deleted all lights except the main center light. I deleted the lantern as not to interfere with the test rendering. I did try setting the transparany of the glass to 99.98%, but it was still burning to white.

I set the intensity to 100% and the ISF intensity distance to 30 inches.

However, something is again is creating light under the table and not producing floor shadows. Now my ISF intensity distance might be off to produce shadows now that the scene is brighter, but prior to using the Bantha's python script and changing the gamma setting on most items, that issue was fixed and although a dark  scene without this correction at least it was functioning as I thought it should.

I have to think you've got the coordinates wrong. Try disconnecting the node from the point light intensity and just render with it normal, no falloff. Do you get shadow under the table?

For speed testing purposes, I'd hide the figures and just deal with the walls, floor, ceiling, and table.

I'd also change Pixel Samples=1, Min Shading Rate = 3, and render only about 200 pixels wide. That's all good enough to see how much light falls everywhere. Should be 50 times faster.

Looking closely, I don't see shadows ANYWHERE. Are you sure you have shadows enabled and raytracing enabled? Point light shadows only work with raytracing.

Bagginsbill, That's it shadows were off. Poser is running so slow, it doesn't click on very well and may click off by clicking on twice and not knowing it. While writing this, its clicked off three times which might be from clicks I made minutes ago.

Never mind

I'm going to start from scratch and just reload the scene again.
maybe have a good stiff drink first