I may just have to start posting variations of this once a month.

You need to use gamma correction. I don't care if it's in your postwork, or you use Poser Pro, or you use gamma correcting shaders, or you use a GC lens. Take your pick.

Your lights are not wrong. You don't need to keep changing your lights. You need to make your image work on a computer monitor.

I shall now demonstrate.

There is already too much light in this scene. The brightest parts are blown out already.

BB - I use the "lens" technique you posted about just a short while back in P6 - and it does work wonders...but I now notice an issue with some things coming out "flat" at times (like hair shine and such). My experiments to add in something for the alt specular channel have thus far not given me very good results (and I'm probably trying the wrong things...but then I'm slightly lost at times in the materials room).

So I'm wondering...what can a person use along with the gamma correct "lens" to bump up shine on things so people don't leave comments that "they're too flat looking"? Or would it just be a setting and futher futzing with the lens material itself?

Ah - I see the comment on your image "Light's Shadow".

It may be that you had too much IBL and not enough directional light. That's what it looks like on the woman.

On the man, to make his skin shinier, you want to increase the specular value on the shader. If you're using a Blinn node, that's Reflectivitiy. If you're using the Specular node or the built-in specular of the PoserSurface, that would be Specular_Value.

Do not be afraid to set it higher than 1. There is no limit. For very shiny things, it is often correct to set it to 3 or even 6.

Here is a render of an object. Can you tell what the object is?

The render is using just an IBL, with no image on the light. The IBL intensity is 30%. Thus the amount of light in every direction is identical, 30%. We end up with a featureless lump of geometry, which is indistinguishable from the ground. It doesn't get any "flatter" than this.

By the way, if you really want to understand lights, you should play with a scene like this. No fancy shaders to distract you - everything is a neutral plastic-like material.

The object's shape is nicely revealed. But the shadows seem too dark.

You might be tempted to add more light or decrease the shadow strength. That would be a mistake.

This is typical.

Here I set my IBL to 15%.

Ok. So I use Poser Pro. Everytime I turn on gamma, EVERYTHING just glows. My transparencies get messed up. I understand I don't know how to use it though, I just haven't bothered to learn. What setting works best?

-Sarah

Ahh...it just might be the fact that I generally don't have AO turned on (I usually forget about it...../blush)

My lights are a simple, color based IBL setup with varied zones for color tone (it generally works for my silly little non-realistic scenes) with an infinite for shadows.

I'll have to play some more now when I get home....

Thanks!

If you use my GenIBL to generate the IBL probe from an environment, then you'll get the proper amount of light.

Here I've used an IBL I made earlier today from an outdoor scene (sky and ground), with 50% intensity. This is obviously better than the flat IBL.

If you're using a lightset that "worked" before gamma correction, it will NOT work with gamma correction. You'll have altogether too much light and everything glows.

So try cutting your main light to 2/3 what it was. Any lights you were using for ambience such as IBL should probably be cut half or maybe even lower. Any fill lights that you added to fill in the darker areas can be drastically reduced or perhaps eliminated altogether.

As for the messed up transparencies, that is a problem, but it can be solved. You have to go into each transmap image and tell Poser Pro not to use GC on that image. Transmaps are data, not visual, so it is not correct to let Poser anti-gamma correct those on the way in.

I've attached a screen shot of where you do this. Find the Image_Map and click on the filename. Then change to Custom Gamma and use 1.0 for transmaps.

When the same transmap is used in multiple materials in a scene, you only have to do this once for each. And in the future, Poser Pro should remember those specific files and always use Gamma = 1.0 for them.

I can do all that, it seems easy enough. What number do I put into the gamma box in rendering options?

I am already experimenting. I have been missing out!

:) Sarah

2.2 for PC, 1.8 if you're on a MAC

Very enlightening to read your thoughts and workflow Mr.Baggins, thanks for sharing. Is there a general rule to follow in regards to the individual light Scene AO options? (settings, does only 1 light get it, etc?)

Quote - Very enlightening to read your thoughts and workflow Mr.Baggins, thanks for sharing. Is there a general rule to follow in regards to the individual light Scene AO options? (settings, does only 1 light get it, etc?)

With the newest Posers (P7 SR3, and Poser Pro), light-based AO is much better than it used to be. Still the default values are not quite up to snuff.

Use AO only on an IBL - never on Infinite or Spot or Point. For those, you should be using ray-traced direct shadows. Spotlights or point lights that are simulating interior lighting, particularly light bulbs with reflectors, should have some blur. An infinite should be used for sunlight, with a blur of 1 to 2, no more.

As for the AO settings, I have found that 3 samples (the default) is not enough. A good number is 7 or 8. More than that you get only marginal improvements with a lot more render time. I like to decrease the Bias to .1 inch and the max distance is usually 6 to 10 inches. (Pay attention to your Poser Display Units when entering these things. If you're not using Inches, convert, or switch to inches to be like me.)

Sometimes for AO under really big things, like ground under a big truck, you need to use a greater AO distance. But this may ruin other shadows. If raising the distance is necessary but creates other problems, consider material based AO for the ground.

The render setting "Irradiance Caching" will affect the quality (dirtiness, smoothness) of the shadows. I find that for test renders, I like 10%. For draft, 50%. For final, 100%. Be prepared for long renders. After you've done a decent draft at medium resolution, if the shadows look splotchy, increase IC to 100% and let it render overnight.

As for the messed up transparencies, that is a problem, but it can be solved. You have to go into each transmap image and tell Poser Pro not to use GC on that image. Transmaps are data, not visual, so it is not correct to let Poser anti-gamma correct those on the way in.

Don't we have the same problem wth bump,  normal and displacement maps?

Best regards,

Bopper.

Great feedback, thanks.
bm.

BagginsBill any advice for us ancient P5 users? Other than upgrade that is. :lol:

Quote - As for the messed up transparencies, that is a problem, but it can be solved. You have to go into each transmap image and tell Poser Pro not to use GC on that image. Transmaps are data, not visual, so it is not correct to let Poser anti-gamma correct those on the way in.

Don't we have the same problem wth bump,  normal and displacement maps?

Best regards,

Bopper.

Quite right. For any "image" that is actually data, not a picture you look at, you need to make sure it doesn't get modified by the GC system.

Bookmark.  Thanks for the tips.

I use Poser 7 and Paint Shop Pro 9 for postwork, and I am  very interested to improve my knowledge about this subject.

Particularly:

1. How can I say when there is "too much" light in an image, but not gamma-corrected, and when instead I should really add a light?

2. How can I apply gamma correction effectively from inside PSP 9? If I learn effectively how to do it, can I achieve the same results as images rendered with Poser Pro or it would be anyway a cheap workaround?

Giorgio

 HI,
I have a light set which works without GC, and I have been experimenting with the ideas posted here. The problem is that when I crank the value of the lights down to the 2/3 level, I can’t see much in the main posing window. It all gets very dark. Is there a way to get the best of both worlds here? Good lighting in the posing room, and improved GC based rendering in Poser?

Quote -  HI,
I have a light set which works without GC, and I have been experimenting with the ideas posted here. The problem is that when I crank the value of the lights down to the 2/3 level, I can’t see much in the main posing window. It all gets very dark. Is there a way to get the best of both worlds here? Good lighting in the posing room, and improved GC based rendering in Poser?

Simple: turn the light up for working, turn it down just before you hit render.

Lighting is no more WYSIWYG than is, say, bump mapping or a lot of other texture features, so people shouldn't treat it as such.

Good morning people,

I'm off to work so this will be short.

hairydalek:
To help with preview, my technique is to add an infinite light JUST FOR PREVIEWING. I set it to really bright 100 to 150 %. Before rendering, I turn it off with a simple shift-click on the light widget. This way I don't mess with my rendering lights so I don't have to put them back to normal before rendering. Also, I can quickly turn this light in any direction to help with previewing. Where it's pointing during render has no impact because it isn't used for rendering.

giorgio:
I'm sorry but I don't know how to do anything in PSP. And I don't even know how to do GC in Photoshop. Yes PS has light levels, curves, and color management but which of those is the easy way to do precisely the same math as (x to the 1/2.2 power) is unknown to me.

As to when you should add a light, it should be motivated by the real-world scenario you are simulating. Are you doing an indoor scene with windows? There should be an IBL, a point light in each window to simulate the incoming ambient light from the sky, a point light or spotlight for each lamp in the simulated room, plus one infinite for sunlight streaming into the room. Outdoor daylight? One IBL and one infinite. Moonlight? One IBL and one infinite.

I have two point lights, one for the window you see, and one for the skylight above. I have an infinite for the sun (look behind the couch on the floor). And I have an IBL for general ambience, custom made for this environment.

Out of curiosity, for your daytime window point lights are you running shadows off them?  I would think not because it would look wrong.

Attached Link: How is gamma handled in video, computer graphics, and desktop computing?

In PSP7, it's Color > Adjust > Gamma Correction, and I assume there's something similar in 9. As I understand it, gamma correcting a final render is not going to be accurate. You need to "un-gamma-correct" your source material, in particular colour maps which are usually already gamma corrected.

Despite a lifetime of designing electronics for TV and computer video, this subject still makes my head spin - so don't take my word for it. Wait until BB gets home from work, and see if he tells me off or not. ;)

Out of curiousity (and because I haven't seen anyone ask it), what are the best ways to gamma correct in poser 6 or 7 if you don't have pro?

Attached Link: http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2754029&page=1

Thus far seems to work ok...and now that I have lighting straightened out a bit more, it looks even better on things.

Thanks for the link, I'll give that a try.

Thanks, bagginsbilol, for your suggestion. Why didn’t I think of this.... doh! 

Quote - 2.2 for PC, 1.8 if you're on a MAC

1.8 is the default gamma setting for Macs when you take them out of the box, which is a figure set for print work. However it seems to be fairly common practice these days to set Mac gamma to 2.2, especially if you are working primarily on web sites and you need to be sure that your images will work properly on Windows machines. Working to a gamma of 1.8 can make your colours seem very dark on a Windows machine.

Should the gamma settings in Poerr Pro’s render settings match the gamma setting for your monitor profile rather than the "out of the box" figure? I’m giving this a go, BTW, and so far I think it’s making a positive difference.

Quote - Should the gamma settings in Poerr Pro’s render settings match the gamma setting for your monitor profile rather than the "out of the box" figure?

Yes, they should.

In fact, they should match the gamma settings of the monitors that your audience is going to use to view the pictures, as well. Can anyone detect a can of worms here?

Attached Link: http://www.appleinsider.com/articles/08/10/25/new_snow_leopard_seed_leak_confirms_cocoa_finder_more.html

Quote - > Quote - Should the gamma settings in Poerr Pro’s render settings match the gamma setting for your monitor profile rather than the "out of the box" figure?

Yes, they should.

In fact, they should match the gamma settings of the monitors that your audience is going to use to view the pictures, as well. Can anyone detect a can of worms here?

And no one has mentioned profiling monitors and ICC files yet ... :-)

Quote - For Poser6, I use the Artistic Lens idea BB posted a short while back....I use the second gamma correct one with the hsv node.

Thus far seems to work ok...and now that I have lighting straightened out a bit more, it looks even better on things.

it is a good way but i think the rendertime is longer because we havea lens so we need raytracing . rendering glass is is taking very long to render.
i think having GC in the materials works faster. at least to me.

Thanks for a wealth of info.  :O)

 Just one, maybe related, question. Mec4D gave the advise to effectively half every ambient color when rendering with her (double bright) light sets. The results were quite good when I tried this. Would that have an effect on the gamma value too? Or is that totally unrelated?

@ bantha: To my understanding, the answer is no.  The two are totally unrelated.

I wasn't sure about this. If you half the ambient value, you effectively use only the lower half of all possible intensities. So you effectively move the whole color range down, shouldn't that half the effect of the missing gamma correction?

Would you like the answer or an explanation? I know some people get tired of wading through paragraphs of math from me.

---- Answer
The short answer is: no it will not effect the gamma correction that is still necessary to see what you have done. Light levels and shader response levels will change what you have done (changes the outcome), but it will not change your inability to SEE what you have done.

Gamma correction is necessary to correct for the fact that your computer monitor is displaying a lie. It does not show you what you rendered. All cameras are aware of this, too. You must convert your image to the sRGB color space. I don't care if your image is at night and every pixel is below 25% or it is the inside of a blast furnace with every pixel over 75%. If you have any pixels above 0% or below 100% you need to gamma correct.

----- Explanation (skip if you don't care)

(Leaving aside gamma correction for the moment.)

In a typical real-life material we generally model and combine two distinct types of interaction with incoming light - specular and diffuse reflection. In each of these there several terms multiplied together. One term is the intensity of the incoming light. Another term is the reflectivity of the material. Then there are some other terms, which we can ignore for the current question.

Please bear with me, this is really simple. Let's call the light level L, the material reflectivity level M, the other factors K, and the resulting rendered pixel R. So the equation is this:

R = KLM

Notice the alphabetic sequence KLM? I did that on purpose so you could remember it.

The material reflectivity for Diffuse light is the product of the Diffuse_Color and the Diffuse_Value in Poser. I typically set Diffuse_Value = 80% and leave the Diffuse_Color = WHITE, which is 1. That means my M is 80%.

OK. Now suppose you have a light at some level, perhaps 70%. And a material which reflects at some typical amount, 80%. For this situation, the rendered pixel would be:

R = K * 70% * 80% = K * 56%

Remember K has to do with angles and such but what this is saying is whatever value K produces, we then use 56% of that because of the light level and the material level.

Now some other Poserite comes along and tells you you're doing it wrong. You need to double your light intensity, and then change your Diffuse_Color to 50% of what it was and you will get better results. Let's see what that does. We were using 70% light, but we're told to double it. Ok we'll use 140% light. We're also told to put a 50% gray into our Diffuse_Color where there used to be a 100% gray (also known as white). OK that changes the reflectivity of my material in half, from 80% to 40%.

R = K * 140% * 40% = K * 56%

Uhhh - wait a minute. I haven't changed anything. Is this true in all cases? In other words, is this equation true?

K * L * M = K * 2L * M/2

Uh - actually - yes it is true.

<sarcasm_alert>

However, if you enjoyed that little detour, I have another one for you.

Triple your light intensity, but cut your shader reflectivity to a third what it was. It will look even better.

Or, if you have the time, increase your light intensity to 10 times what it was, and change your shaders to 1/10th the reflectivity. That will be 10 times better.

Wait, wait, increase your light intensity to 1 MILLION times the current value, and then in your shader, set Diffuse_Value = .000001 (1 / 1 million) and render. You'll be amazed.

</sarcasm_alert>

Attached Link: http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2712423&page=7#message_3086316

She never talked about ambient, BTW. It was Diffuse_Color.

Any data on the ambient channel has nothing to do with lights. It is just sent directly to the render. This is useful for making things glow without a light.

http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2712423&page=3#message_3074280

Another one where she required you to go around and change the diffuse color of every one of your materials, instead of simply reducing the light intensity by half.

also here:
http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2712423&page=4#message_3075810

and here:
http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2712423&page=2#message_3074141

and here:
http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2712423&page=7#message_3086183

She kept talking about using the RGB 127 127 127 to bring things into balance. But "balance" involves two quantities. You can bring items into balance by decreasing the heavier, or increasing the lighter, or both.

While I'm all for solutions and tricks involving materials, this one is a fallacy and requires you to do extra work. If you want to cut the brightness of a render in half, do it at the light, not all the materials. Or use the lens, which is even easier, if you have a lot of lights.

Using the lens you can change the overall brightness of the entire render and gamma correct in in one step.

Well, the work was a short script, which I mainly did to see how to manipulate materials with Phyton. I did not use her lightsets much afterwards, mostly because I found Oliviers masks whith the color nodes on an IBL to be more versatile. I just remebered the whole thing and wanted to ask if there is an effect at all. I did not make a detailed comparison, I just asked out of curiosity. But at least I find it good to get rid of my misconceptions, even if that means I have to ask a dumb question.

I'm working with your lens in the scene I'm doing at the moment, basically from the moment you mentioned it here in this thread. I haven't posted the picture because I still have to get the hair of one figure right,. Some of the colors from the clothes became a bit too prominent, but this was easy to fix, the overall result is remarkable good. Great stuff, very versatile, incredible usefull.

Not a dumb question, believe me. After all the hype, who wouldn't think it was true? In general, as far as I can see, only face_off and I question these things and actually test them to see if they are useful principles or just a coincidence. So far it is all coincidence.