Exactly. This 3 to 6-light business has never had anything to do with lighting. It was that you were manually doing gamma correction by adding more lights pointing at the dark parts.

One IBL, and one infinite light is all you need for pretty accurate daylight renders.

plus with 6 lights yourrenders will take forever. of course they will. plus with one IBl and one infinite light you get a better sun effect then with lights. with 6 lights it looks like you are in big bulding with 6 lights. you have 6 different shadows. for sun you only need oen shadow. we all know that the light bounces around you(ground,walls) and so the back is lighter.

so faster and more realistic.

So, does your VSS shader already include GC like you demonstrated in the an the Poser 7 Pro feature Gamma Correction be emulated in P7 or P6 with Python post?

Yes that's what I've been trying to make clear now for 15 pages :) lol

There are four key elements of my skin shader. I think I talked about them before but after 17 pages of thread even I don't remember what I said.

The four key elements are:

Conservation of energy (energy out <= energy in)
Fresnel effect (reflection varies with viewing angle)
Subsurface scattering
Gamma correction

You will find some subset of these in some shaders by others but you won't find all. Plus, how I did them is different than almost everybody. I've done these things 20 different ways in the past, and the current shader network has the widest set of conditions under which it still works right, while still being very computationally efficient.

ok, some of what you're saying just isn't making sense to me.  5% ambient light should, i would think, mean that if it's lit by nothing else, it should be at 5% lumniance, right?.  and all the way up to 100.  nothing should make 5% ambient light translate to say 20 or 50 % luminance, right?

actually, i was mainly going by the atmosphere, but if you think the atmosphere shader needs to be GC corrected, i guess i could see that.

just going by eyedropper, my original gave me colors with a luminance of about 5% when lit by the IBL and infinite in an area only lit by the IBL and the infinite.  i could see where that's too dark.  VSS is giving me colors with a  luminance of  about 23% in the same area.  with just an infinite at 2% to take the shadows away from dead black, i'm seeing colors with a luminance of 15%.

i would think the proper situation would be that if you took a figure and put them in a scene that matched exactly an HDRI or IBL source, and lit them with an HDRI using that source at 100%, you should get exactly the same lighting.  if it has to be at 20%, that seems to me that you're subtracting from the original amount of light.  if not, then why?

that doesn't sound even vaguely correct to me, but i openly admit i'm not an expert like you are.

and i would think that it would be constant, and not respond differently to ambient light than, say, a spot.  as i mentioned, i'm not seeing at all a simple too bright figure.  i'm seeing a too evenly lit figure.  

i've collected a bunch of low light images, i'll add yours to the collection.  and i'll keep testing. but so far i'm still not able to get anywhere near the results i'm trying to achieve or even that i've seen in dozens of photos.

i will say this puzzles me:   shifting the gamma on photos doesn't make them not look like photos anymore.  heck, doing crazy stuff with levels, curves, and blending layers doesn't  generally make it stop looking like a photo.  so i don't see why something without a gamma shift in rendering would make the main difference between photographic or not?  also, i see yellow bloom in lots of photos where the light is too bright or too close to the subject and the skin gets overexposed.  so why shouldn't i see it in renders in the same conditions?

just to say, i really don't think it's the GC i'm running into.  not that i've tried turning off GC yet, but i'm seeing much more predictable results with just IBL.  i just don't see why a 2% infinite should make formerly dead black shadows about 7.5% grey (different place than previous numbers, in an area less lit by the infinite), but i'm going to try a few things to see if i can find out more.

 Even more so, whether you do gamma correction or not is not even a matter of your monitor. It's simply following the standards for the file formats, using sRGB. Take image formats that store RGB values from 0 to 255. 255 = full intensity. 127 = 20% intensity! That's simply how JPG works, how TIFF works, how BMP works. It's also how your digital camera works and how your scanner works.

Render engines, on the other hand, do their calculations in linear space. For them, half the value = half the intensity. So when reading textures, we need to convert the numbers to the linear scale a render engine wants and when writing final renders to disk, one needs to convert those values back to the non-linear space that the image formats use. The method to convert between those two worlds is called gamma correction.

This is by no means a Poser-specific problem. Many commercial 3d applications ship with gamma correction turned off or require the user to setup shaders to handle that (Go look at Maya or 3ds Max). Even 2d programs like Photoshop get this wrong.

that's interesting you say that. most everything i've ever come across calls that 127, 127, 127 50% grey.  and it kind of matches the 50% grey prismacolors i know of, and is much lighter than the 20% grey markers.  it's also what you get if you bring white and black together and mix them at 50%.  and what you get when you ask for 50% luminance.  i'm sorry i'm being a pest, but could you explain how all the print programs, designers and printed scales i know of are wrong?  i'm sure there's science behind it that makes sense, but i'd like to understand your scale, especially since it seems to differ from that of say, Pantone.

and by the way, i'm definitely not getting the same results with an IBL as an infinite.  GC with a single spot and IBL is giving me precisely what i want.  well, minus all the V2 problems i mentioned.  i still wonder why her eyes and lashes seem to be all messed up.  adding an infinite is just skewing everything out of whack.  so i'm going to try something, see if it works.  if my suspicions are correct, i know what to avoid.  otherwise, i'm kind of stumped.

"ok, some of what you're saying just isn't making sense to me.  5% ambient light should, i would think, mean that if it's lit by nothing else, it should be at 5% lumniance, right?.  and all the way up to 100.  nothing should make 5% ambient light translate to say 20 or 50 % luminance, right?"

Ok, you wrote a lot but its all predicated on this premise. So let's just deal with this one tiny point.

Nobody said 5% ambient light should translate to 5% luminance. It should produce 5% luminance.

Now, on a voltage scale of 0 to 1 volt (1 volt being 100% brightness), how many volts should you send to the monitor, in order to produce a 5% luminance?

Is the answer .05 volts? No it is not.

We've already seen from the GC test image I posted that .68 volts STILL isn't even to produce 50% luminance on your monitor (splat #5).

So how many volts are necessary to produce 5% luminance?

BoP = V ** 2.2

In this case, BoP (brightness of pixel, or luminance as you called it) is supposed to be 5% or .05, right?

.05 = V ** 2.2

.05 ** (1 / 2.2) = V

So the voltage necessary to produce a 5% luminance on your screen is .05 to the power 1/2.2. The resulting voltage is about .256, or roughly 25.6% of the maximum voltage.

So in order to get a 5% luminance, you must send a 25.6% voltage to the screen. What is the RGB value? WIth 8-bit RGB, the maximum is 255, so 25.6% of that gives us RGB(66, 66, 66).

So - when you put RGB(66, 66, 66) onto your screen, it is only visually 5% as bright as RGB(255, 255, 255) on your screen.

Once you grasp the significance of this, you will see why the VSS shader is sending such huge numbers to the screen - that is what gamma correction is.

The other point I'm trying to make is you have trained yourself, through years of dealing with renders, into believing that 5% luminance is RGB(13, 13, 13), which it decidedly is not. That is a 5% voltage, to be sure, sent to your monitor, but it produces a 1.3% luminance. To get a 5% luminance, you must send a 25.6% voltage. As a result, you have also trained yourself to believe that 5% luminance is very very dark, and should barely be perceptible. In reality, it is easily bright enough for you to see things clearly.

You are used to Poser renders, produced in linear space, and viewed on your computer monitor severely darkened. You have accepted it, and now reject the "glowing" ridiculous visibility of a GC'd render.

I have never accepted the dark images, and have worked for two years to find ways to correct what I knew all along did not look like real ambient light. I thought it was Poser's lighting model. It is not Poser's lighting model. It is that we've failed to do what is necessary to encode the final render for the device we intend to see it on.

As I've said before, if you treat your scene as always, you will be disappointed by GC. You will screw up your colors and brightness and lose your dark shadows. But this is not the fault of GC. It is because you've just got too much light in your scene, and you never saw it for what it really was. Once the true amount of light is revealed, it is obviously necessary to start using less light.

Quote - also what you get if you bring white and black together and mix them at 50%.  and what you get when you ask for 50% luminance.

No - that's the fallacy - it is NOT. My splat test image above proves that.

The background is black and white pixels, mixed in 50% proportion.

The splats are all above 50% in "luminance", yet half of them are darker than the background.

How do you explain that?

Also, Photoshop is one of the programs that supposedly makes this conversion for you. If you edit a document and associate a proper color space with it, it will do all kinds of conversion behind the scene.

The result is that if you choose to fill it with 50% gray, it will absolutely NOT write RGB(127...) into your image.

Stewer mentioned the sRGB color space. If you actually tell Photoshop that's what you're working in, and that your monitor is in that space, and that you intend to print in linear space, then when you select 50% for a fill, you're actually filling it with 73% white.

Most tell photoshop "don't touch my values - just do what I tell you", so you get what you choose, and it previews exactly the same in other apps.

Quote - i'm sorry i'm being a pest, but could you explain how all the print programs, designers and printed scales i know of are wrong?

Because it doesn't make a difference for them. While non-linear (127, 127, 127) is not physically 50% intensity, it just happens to look exactly 50% gray to the human eye on a calibrated monitor or printer. If all components in the chain agree on what (127, 127, 127) stands for, everything is fine (sometimes two wrongs do make a right).

However, physical calculations like fresnel equations or Lambert's cosine law work only correctly when they get physically correct input. That's why for rendering we need to convert between perceptual values and physical values.

actually, i doubled checked before i posted, and Photoshop literally said B = 50% was L = 53%, was RGB = 127, 127, 127.  L does stand for luminance.  that isn't to say that's right, just reporting what i'm seeing.

but while this is very interesting and i really appreciate both of you taking the time to explain all this, i think i should probably stop asking questions and just research myself some other time.  because, as far as i can tell, GC is not my problem.  i'm trying to turn it off now (i think i found the right node), and i'll see what happens, but i'm not expecting it to help.   turning off the infinite and using only IBL did.  my problem isn't coming from 1) a bright figure, or even 2) ambient light in general.  IBL is working just like i'd like, but the infinite seems way out of whack. i don't see how an infinite light at 1% should make the figure only a little less bright as my spot, eliminating the clear edge of my spot.  if the moon worked like this, you could read by it without straining, even when waning.

it would be nice to just have scattered light.

gamma correction is a big deal. i think a lot of poser users dont even know about your GC tutorial.
plus this thread is for VSS.

i think you should open a new thread here in the poser section so that everyone sees it. and then we can have a debate about this. plus we have a thread about realistic renders. we can share new things.

cobaltdream,

I"m really interested in understanding what you're seeing and I'm sure many others are. There are a lot of people who don't speak up because they're shy and it helps them a lot that a few of my favorite Rendo pals, like yourself, will speak up and challenge me. We get a lot better understanding that way, and the shy ones have a voice.

So don't think about stopping with the questions here. I have no interest in ending the conversation. I want to know what you're seeing and why it isn't working for you.

To that end, you should be posting some pictures, because your words aren't making clear the issue, at least for me.

Going back to your May 22 post you said:

"The original isn't muddy, it's properly dark, as far as i can tell." - I gotta see what you're talking about here because it sounds a lot like what I call the Poser look. Meaning, have you trained yourself to believe that "too dark"= "properly dark"? Show us.

"she's not too bright overall, she's too bright where she should be shadowed" - Again, if I were to describe whats wrong with Poser renders that need GC, I'd say "they're not too dark overall, they're too dark where they're shadowed". See what I mean? It sure sounds like what I call "correction" you call "error". Can you show me some renders?

Did you find the 2.2 part in the shader so you can turn it off?

Another thing we need to do before we wander off into pointless speculation: On my splat test, which one blends into the background ON YOUR MONITOR?

Are you on a MAC? If so, I expect you'll tell me #6 or #7, which means all shaded areas are brighter for you than me. But see - I'm already digressing and speculating. Could you just tell me which splat matches the background, please?

OK ice-boy. I will start some new threads today.

I think we need more than one, right?

One for: what is wrong with this picture? To teach people to see the error of their ways.

One for: basic techniques for GC

One for: advanced techniques for GC (have you considered what to do with reflection and refraction? What if all your shaders are producing GC voltage - what does/should the Reflect node do then - I have the answer - do you?)

One for: how to properly light a scene for linear workflow

Mazak

Hi, I have a vss question, using vss2, I can't seem to control the color of the eyelashes in the millenium figures, ie, they're always black. What's curious, is that the shader seems to ignore this texture zone, adding no special nodes, but the diffuse color for the eyelashes always renders black regardless of what actual color is present. I understand the standard diffuse is shut off, so to speak, for the textures using the special nodes, but with the eyelashes also?

sorry, i took so long to respond, but it's been in the mid-90s or so and our air conditioning is out.  it has been for months, but the heat is just gotten unbearable in the past month or so.  i've been trying to get away from the heat.

first off, it's not so much the monitor that's an issue.  it's literally  the colors i'm getting.  i don't trust my monitor at all, so i wouldn't just depend on that. it's that i'm getting colors with what graphics software are saying are a really high luminance, imho.  for instance, in a picture i just downloaded that seems to be pretty bright compared to what i'm trying for, the darkest shadow is at 9 to 13 out of 240  luminance (according to the color picker). 

but my monitor shows that 5 is the same. 

i did find the 2.2 part of the shader, and tried switching it to 1.  that gave me less what i expected, which means it sort of fixed what i'm talking about, but made everything darker.  so i have my nice defined spot, but it's too dark.  so far the only one i'm thinking is any good is the test with GC, 1 spot, 1 IBL at 10%.  i would even pull the IBL higher, due to how dark some shadows are.  if i had a photo of a nude in a spot, i might be able to judge on the harsh shadow beneath her breasts, because to me they look both way too sharp and way too dark, but that could just be me.

to be honest, i don't see the point in posting pictures now.  because i feel like you're just going to tell me it's right and accurate, and it's just not jibing with what i'm getting from the photos i have, nor the reality i've seen.  i can't read outside at night, but an infinite at 1% gives me an image in which i can see every detail of lit texture.   to get what i'd consider the directional (rather than the tiny bit of ambient ) light i'd say was appropriate for most nights without artificial lights, i'm having to pull my infinite down below that.  so far i've tried .1 and  .4.

actually, personally, i'm happy with it at about those levels.  i still think that's a nuts reaction to the Poser light scale (though maybe it will make more sense if i can find comparisons to sun and moon luminosity), but if my limitation with the shader is just with infinites, frankly, i can handle that.

i think it would be much more helpful to work backwards, because i think arguing perception on something without an actual reference isn't very helpful.  if you could show me how to match at least one photos i've found within the same range of luminance as determined by color pickers, i'll be quite content.

http://gallery.photo.net/photo/5676285-lg.jpg
http://www.sxc.hu/photo/68654
http://farm1.static.flickr.com/26/53737078_da32aa843b.jpg?v=0
http://www.morguefile.com/archive/?display=73109&

not dark, but it has good range and overexposure:
http://www.morguefile.com/archive/?display=206845&

this is a bright one, and probably easy to match due to really good even skin tones so i think matching colors and luminance will show how close you are more easily.  basically i think the people are about as close to 3d looking as you can get real people, so the non-shader aspects of matching reality won't hit the eye as much. http://www.renderosity.com/mod/gallery/index.php?image_id=1680385

edited to add: ok, i'm going back in front of the fan for a while now.  i can't take the heat coming off the computer (though it's temp seems to be fine).  oh, and i double checked that i used the synchronize with the prop from vssPR2ControlProp.zip, but i still got a blank image in my displacement node (hadn't noticed till i was adding back the GC).  is there a way to use the right prop and get the wrong synchronize?  and should V2 be having problems with her eyes and lashes?

I've been playing around with the VSS and some textures I made, except lashes (they're Milan's anyway).  And using Poser 7 and 4BlueEyes rebelmommy's IBL light rig, I think I'm coming up with something -- really good for a 6 year-old computer.

Well here it is.   Simple render, no GC, no postwork.

One thing to ask BB:

Is the VSS prop itself have the templates (e.g. skin, eye nodes, etc) while going in the Materials Room, and we can play around the variations of it?  I'm just curious to know.

Thanks for sharing this, it will be worthwhile I believe.

Quote - OK ice-boy. I will start some new threads today.

I think we need more than one, right?

One for: what is wrong with this picture? To teach people to see the error of their ways.

One for: basic techniques for GC

One for: advanced techniques for GC (have you considered what to do with reflection and refraction? What if all your shaders are producing GC voltage - what does/should the Reflect node do then - I have the answer - do you?)

One for: how to properly light a scene for linear workflow

sarcasm.......i like it he he :) 

Quote -

this is a bright one, and probably easy to match due to really good even skin tones so i think matching colors and luminance will show how close you are more easily.  basically i think the people are about as close to 3d looking as you can get real people, so the non-shader aspects of matching reality won't hit the eye as much. http://www.renderosity.com/mod/gallery/index.php?image_id=1680385

to bad. you are not using teh VSS shader to the maximum. this could have been done with anything.
its a nice pic. but VSS can make poser look like a PRO software.

Indeed. But my intention was just to see whether a "quick 'n dirty" application of the shader would look like, and I liked the result.

Quote - > Quote -

this is a bright one, and probably easy to match due to really good even skin tones so i think matching colors and luminance will show how close you are more easily.  basically i think the people are about as close to 3d looking as you can get real people, so the non-shader aspects of matching reality won't hit the eye as much. http://www.renderosity.com/mod/gallery/index.php?image_id=1680385

i think you misunderstood.  that's just a random picture.  i mean duplicating the image.  so you get shadows in approximately the right places.

Quote - Hello cobaltdream Most of the image examples you have given are using non-standard exposures, lighting techniques and perhaps film stock inorder to achieve their effect. I'm not certain how to effectly adjust exposure settings within poser, but I can safely say that somethings you are going to require a certain amount of post work. Here is a quick atempt to mimic one of your examples, simply by taking the original render and then adjusting over-all brightness from within an image editing program. good luck jdc

non-standard exposures, lighting techniques, and yeah, even various film stock, are what make a photo more than just a snapshot.  if what you're basically saying is we can't duplicate any more than a very narrow and fairly boring type of photos without postwork, then personally, none of this matters and what i need to master is painting skills.  because i've been there and done that, and i'm looking for more right now.  and i don't mean more realistic.   the one and only reason i'm interested in realism is to achieve the (imho) interesting effects light and shadow i've seen in reality and in photos.

i don't buy that it's that impossible to get closer than we have before, and get a better reaction  to light than we have in the past.  personally, i have faith both in bagginsbill's abilities to create incredible shaders and to say no when it's too much trouble.   so thanks for the advice, and i'll keep building on my 2d skills (which are not bad).  but i'm still interested in seeing if bagginsbill has any 3d advice or techniques to offer.

Well said cobaltdream. My attention is very much on this thread right now.

OK, let me expand on my thought.... There are a number of dynamics at work here. You have the light, the subject and the instrument used to take the picture(s). All these dynamics MUST be taken into consideration. In the example I tried to mimic you have the Sun light, the room, the subject and the camera. The subject leans over into shadow and the photographer has to adjust the camera's exposure settings to properly capture the image. If the subject stands up straight, her face will be in direct sunlight and thus force the photographer to re-adjust the camera's exposure settings accordingly. What has changed between leaning over and standing up straight? The amount of sunlight entering the room remains constant. The light bouncing around that room remains constant. The subject (for the most part) remains constant. What has changed is the amount of light relected into the camera. The photographer MUST adjust the exposure settings (ie: f-stop, shutter speed) inorder to deal with changes in the amount of light entering the lens. The dynamic of "camera settings and exposure levels" is VERY important with regard to the examples you have given. The VSS process does not deal with the dynamics of light intensity, it does not deal with the dynamics camera exposure - it deals, mostly, with how light bounces/reflects off the subject matter. (...continued next post)

you weren't unclear.  i understand what you mean.   i know what exposure is, but that's not the point.  of course exposure settings, if you have them, affect the final render.  but that's just an adjustable setting.  in the real world, that photographer would have constant light, so he'd probably adjust exposure.  but in our much more adjustable 3d world, we can change the amount of light and not the exposure.  in the example you posted, i'm fairly sure you could have created those permutations with different light settings, too. 

and if it doesn't work, what i said before still stands.  either we can do interesting pictures straight out of Poser, or shaders don't really matter and it's all about postwork, and i personally won't worry about any of this any more.  because it's all about the end result, and for me there's not much of a purpose in duplicating  snapshots with inaccurate models. 

Sorry I took so long - I had to take a break from VSS - I was tired of it. :)

I forgot to do anything with the Template Lash shader. I left it shut off. Doh.

You can fix it easily. Go into your VSS prop and edit the Template Lash material. Turn off Specular_Value, turn on Diffuse_Value=1, and put what color you want into Diffuse_Color. Then Synchronize and Render.

@cobaltdream,

Sorry to hear about the heat. I become non-functional under those conditions. Which is why I have air conditioning everywhere I go, or I don't go. :)

I'm still not dialed into all the recent discussions. I was partying all weekend and now I'm high on allergy relief medicine - pollen counts here in New England are unbearable this time of year.

The blank displacement map has not been fixed yet - that's the scripts fault, not the control prop, and I haven't released a new script yet. Will do so soon.

Regarding V2 zone assignments - I didn't do anything about that either. You should be able to edit the Shader Rules to suit your figure, though. You just need to add (or maybe change) a rule here and there. I have a lot of rule changes I need to perform, what with so many strange zone names out there. Vince even has French ones - I'm going to change VSS to use internal names - hopefully that will mean I don't need to pay attention to the French zone names.

Quote - I think I have here a nice example how useful GC in PoserPro can be. In my example I use the DAZ cyclorama as backdrop scene. With GC off the scene is total overshot.
VSS2; light is Outdoor Beach.

Mazak

Looks great, Mazak. Post some more.

Quote - Hey there bagginsbill et al.

I've been playing around with the VSS and some textures I made, except lashes (they're Milan's anyway).  And using Poser 7 and 4BlueEyes rebelmommy's IBL light rig, I think I'm coming up with something -- really good for a 6 year-old computer.

Well here it is.   Simple render, no GC, no postwork.

One thing to ask BB:

Is the VSS prop itself have the templates (e.g. skin, eye nodes, etc) while going in the Materials Room, and we can play around the variations of it?  I'm just curious to know.

Thanks for sharing this, it will be worthwhile I believe.

jartz,

Looks good. Yes the VSS prop shader templates are there for you to change as you like. Those are just a starting point. You can edit them, save them as new materials, load them later, or you can save the whole material collection and load it later, or you can save your entire custom prop and load it later. Do whatever is most fun for you :)

Quote - sarcasm.......i like it he he :) 

ice-boy:

LOL I wasn't being sarcastic. I was not saying the opposite of what I mean. I figure that, like me, most artists don't even know they have a problem, to begin with. They're OK with burning hot lights. We need to show them why that is not a good idea first.

Otherwise, if I go straight into "look, you need to edit 50 materials and put 20 nodes in each like this, in every scene you have" they're going to tell me to go to hell. :) LOL

The exposure thing raises an interesting question.

In other "good" renderers, not only do we have emulation of shutter-speed and f-stop, we also have the ability to define light sources in terms of watts per square meter, and similar physical settings.

It begs the question, what is a Poser infinite light at, for example, 80%? What is that? Is that the brightness of a candle at 1/15th second exposure with f 2? Or is that the sun at 1/120th second with F 4.5? I have no idea.

Leaving that aside, I have no idea what light intensities and shader "Diffuse_Value" actually mean, in terms of the math going on. 

Now I'll have to go experiment. :) I'll see you guys in a month. LOL just kidding.

It sure would be nice if Poser had those final exposure dials like Kerkythea has. Right now, if you get it wrong, for example, too bright, you have to go edit every single light you have and render again. You can't just fix it in postwork, because some information was already lost. 

Shouldn't it be possible to correct that in Poser Pro, when rendering to an HDRI image? All needed informations should be there - but you would need another software to extract the LDRI image, probably. Am I wrong here?

bantha, that is correct. If you render to an HDRI, then you have floating point numbers for each pixel. Tiny numbers as well as numbers well over 1 can be represented with excellent fidelity, and can then be adjusted to an integer in the range 0 to 255 via manipulation of "exposure" controls.

We only have this problem if we're saving as an LDRI - in which case the smallest delta we can represent is 1/255 and the highest value is 1.0.

I just did some little experiments and the diffuse lighting is as we'd expect in Poser, based on the simple Lambertian diffuse lighting model for directional lights.

Given
I: the light intensity, expressed as a unit number (0 to 1)
V: the "diffuse value" (also 0 to 1)
C: being the color value of the surface
A: the angle formed by the surface normal and the direction to the light (with 0 being the light is pointing straight at the surface)

then I find that the rendered pixels (as a unit number 0 to 1) is precisely:

I * V * C * cos(A)

This calculation is performed for each color channel, R, G, and B.

So, for example, assuming I is .8, V is .6, C is .902 and the angle is 45 degrees, I expect my pixel value to be:

255 * .8 * .6 * .902 * cos(45)

which comes to 78.06. When I set this situation up in Poser, I read out 78 for the LDRI value. Yay.

Hmmm, but when I set Poser Pro GC to 2.2, that should render as 149. It renders as 141. What is up with that?

Never mind - I just remembered that when I set the Diffuse_Color to 230,230,230, that got gamma uncorrected first.

pardon, but can you show how the GC affects that equation?  that is to say, GC of 2.2 should give you 149?

Converting our 8-bit 73 out of 255, that is a unit value of about .3059.

The gamma corrected value is .3059 ** (1 / 2.2) which is .5837.

Converting back to 8-bit, you get .5837*255 = 148.8, rounding to 149.

um, wait.  what do you mean by "**" ? to the power of? and you mean 78, right?

78 is the integer value, as expressed as a fraction of the maximum 8-bit RGB value, where 255 represents maximum intensity. So 78, in unit form, is 78/255. What I mean by unit form is using numbers from 0 to 1, where 1 means full intensity.

Gamma correction of 2.2 is expressed as the intensity, in unit form, raised to the power 1 / 2.2, or .4545454545.

And yes I said earlier that ** means "to the power of".

To do the math with the 8-bit form as well, you convert to unit form (divide by 255), do the power, then multiply with 255 to get back in 8-bit form.

((V / 255) ** (1 / 2.2)) * 255

Attached Link: http://www.poserpro.net/blog/poser_pro/2007/11/poser-pro-and-hdri.html

If you look at the Poser Pro blog, it wrote about doing exactly that with HDR exports from Poser:
http://www.poserpro.net/blog/poser_pro/2007/11/poser-pro-and-hdri.html

Quote - Converting our 8-bit 73 out of 255, that is a unit value of about .3059.

i understood about converting the raw numbers.  what i was making sure of was that you meant 78, and not the 73 you subsequently posted.  and i'm sorry, but i missed where you previously said that "**" means to the power of.