has anyone noticed this?
gradient bump  or normal mapping doesnt work when you us alternate diffuse.
you need to have the diffuse color turn on.i tryed to plug in a math function substract 1,1 . but it also doenst work.

this wouldnt be a problem if bagginsbill wouldnt make this fantastic gamma correction shader that makes poser renders 100 times better
www.renderosity.com/mod/forumpro/showthread.php

i think gradient bump is very good if you know exactly what kind of bump you want to use and what settings you  want to use. if you want to make changes in the material room with nodes then bump is the way to go. with gradient bump you get very good renders without changing the render settings. you ultra details for skin for example.

You can always do gamma in photoshop or similar. Render gamma correction is a bad idea anyways unless you have no other option.

As for normal maps and alternative diffuse, that just seems to be a limitation of firefly. They also don't work with reflections for example, though specular works fine. It's a pity too as normal maps allow some very nice detailing.

''Render gamma correction is a bad idea anyways unless you have no other option.''
you are aware that in those expensive  3D softwares you have gamma correction set to 2.2 right ?

please explain. 

'expensive 3d software' have gamma controls that look like this:

So it's not quite the same thing as poser's gamma correction. (Note how you can control the result in camera terms which is very important if you want to create photoreal work.) Since we don't have anything even close to this in poser, there's really no argument as to whether poser makes a decent gamma tool. Even tools in photoshop kick its ass in this area.

As for why it's a bad idea, regardless of app, it's mainly cause most renders need at least some postwork before they can be called 'final' and this has to be done before gamma correction. By rendering out with a gamma, you will have to first apply a 1 / 2.2 gamma to the image, then edit it, then apply a 2.2 gamma again. This will seriously degrade your render quality. Also you don't want to put a fixed gamma in your image either cause then you will have to do a rerender everytime you need to change displays. eg a 2.2 gamma works fine for online galleries, but not for macs or other things like printers, film etc. OTOH if you have a raw render without gamma, you can apply whatever gamma number you need to it.

The only reason to use render gamma is if you're just doing experiments and dont care about the work. Anything serious should be rendered to hdr and corrected in post.

i see noone has a problem with this.

ok :) 

Honestly, as I said in other posts, the normal map is something I just don't care to use, since I do my bump maps procedurally. I need to be able to generate the texture with nodes and the Bump channel does that for me perfectly, such as this peeling painted stucco. This is a gamma-corrected shader with procedural Bump.

And I don't agree that every render needs postwork to be called final. The wall here is just fine - I can't imagine how I'd improve it in postwork.

just to say....

i can imagine how i'd improve it in post.  but i like playing.  and i have absolutely no problem doing that with gamma corrected renders.   there is no reason it would need to be done before gamma correction whatsoever.  no matter how you break it down, it's all just pixels.  so you want to change stuff, you put in new pixels.  if you use the gamma corrected image as the source and reference for your new pixels, and match it, well, there you go.  and if you don't match it, gamma correction wouldn't help you.   bagginsbill has already posted pictures that show that gamma correction in PS isn't as good as in Poser.  

you have a nice gallery, and it's old enough that i'm sure you've improved, so i won't make assumptions.  but if you want to start a tech debate with bagginsbill, i warn you that you're very unlikely to encounter issues that he hasn't already researched to death.

edited to add: as i always say, photographers edit their photos all the time.

"there is no reason it would need to be done before gamma correction whatsoever.  no matter how you break it down, it's all just pixels.  so you want to change stuff, you put in new pixels."
  
I'm not sure you understand what 'destructive' means in the context of CG.

"you have a nice gallery, and it's old enough that i'm sure you've improved, so i won't make assumptions.  but if you want to start a tech debate with bagginsbill, i warn you that you're very unlikely to encounter issues that he hasn't already researched to death.*"

My mistake, i didn't realize BB was the Infallible God of 3D that one questions at one's own peril. I'll just heed your warning and avoid 'tech debates' in the future, thanks.

I'd rather we didn't snipe, and actually have the tech debate, because I'm not infallible and I often learn stuff through discussion. Unfortunately, I'm about to go on vacation and may not have access to the net for a week.

Anyway, here are some issues to ponder?

**(after showing exposure controls dialog) "So it's not quite the same thing as poser's gamma correction"

**By "poser's gamma correction" I assume you're talking about Poser Pro, right? Poser (7) doesn't have gamma correction and this thread isn't about Poser Pro, it's about Poser 7. What we were talking about is a gamma-correcting shader technique, which is a very different thing. It allows you to apply gamma to materials individually, so you can get correct results in Poser 5, 6, and 7.

If I showed you a Poser shader+light system that had clearly marked parameters for film speed, shutter time, ISO rating, gamma, contrast, brightness, etc. just like you show there, and you had control over that using a similar dialog, all in Poser 5-7-Pro, what would you say about that? Would you tell me not to use it or what? LOL.

**Since we don't have anything even close to this in poser, there's really no argument as to whether poser makes a decent gamma tool.

**But nobody said Poser is a gamma tool. We are saying that a Poser shader, a material, can be developed that is much more sophisticated with regard to tone mapping and so on, and that such sophistication requires the use of alternate diffuse, etc., and that such configuration does not function correctly when using normal maps in the newest versions of Poser. The topic of this thread is

• sophisticated shaders cannot be used with normal maps *

The topic of this thread is not "I want to use gamma correction in shaders, tell me why that is inappropriate".

Having said that, you raise an interesting point, so let's go with it.

I think you would agree that tone-mapping an individual material is, in fact, MORE sophisticated than your average high-end 3D "photographic terms" tone-mapping dialog, right? So, just on the basis of capability, I have more chest-thumping righteousness than a Maya/Max user who can only tone-map his entire render. Poor Maya/Max user, can't tone-map his sky without also tone-mapping his figure. LOL I'm being theatrical here, but I hope my point is understood. For whatever reason you want, if you are into controlling things, then more control is better than less control. Tone-mapping shaders give you more control. If you think someone shouldn't use a particular tone-mapping feature, such as gamma, that's fine, but a general dismissal of tone-mapping shader techniques is actually pretty naive. Have you ever tried any? I've only done, oh, maybe a thousand such shaders. I can tell you they are always better than the ones that have no such capability.

If you were to look at the effects shader for my environment sphere, you would see tone-mapping on steroids. Brightness, contrast, hue, saturation, gamma (in and out) as well as umpteen zillion different color transformations all in one shader, just for the sky. Do you really want to try to do all that in postwork? No you don't, trust me. Even if you were to try to "mask" select the sky in post, you'd never be able to get the reflections of that sky to be correct for the parts of the scene that are reflecting the sky after you post-work tone-mapped the sky. Nor could you fix the anti-aliased edges that have tiny bits of sky color mixed into them.

In general, this notion of postworking sections of a render always shows flaws to me. Any effect having to do with mixing or multi-sampling, which happens wherever a pixel spans the visible boundary between two materials in the render, will be corrupted in post if you try to make any non-homogeneous manipulation.

As for why it's a bad idea, regardless of app, it's mainly cause most renders need at least some postwork before they can be called 'final' and this has to be done before gamma correction.

As I said, I disagree. If you correctly generated the reaction to light in the first place (which I always strive to do) there will be no reason to tone-map the finished product in any way. Now if you want to move pixels around, say to correct a bad armpit, then it makes no difference if you move linear pixels or gamma corrected pixels, does it? I'm seriously asking.

If you did not correctly generate the reaction to light in the first place, then I agree you'd want to do some tone-mapping. However, I would say (and have demonstrated elsewhere) that tone-mapping a 24-bit color render is inferior to getting it right in the first place. There is not enough precision in the 8-bits per color to let you make such transformations accurately. As well, if you did not tone map the incoming material in the first place, then you will get hue and saturation shifts that were not planned.

The beauty of the tone-mapping tool you showed above is that it is not operating on an 8-bit render. It is operating on the internal results in floating point. That is a good thing.

If you were to export to EXR format, then you could succeed in post-work tone-mapping the same as in-renderer. However, for Poser 7 and less users, that is simply not an option. In other words, if you are using Poser 7 or less, and compare material tone-mapped in shaders versus tone-mapped in Photoshop (using the 8-bit linear render) the results are vastly different, and the Photoshop result is not the better one.

**By rendering out with a gamma, you will have to first apply a 1 / 2.2 gamma to the image, then edit it, then apply a 2.2 gamma again. This will seriously degrade your render quality.
**
Are we talking about 8-bit values or floating point values? If floating point, this is nonsense. For any linear pixel value, x, and some gamma correct version of that y = x ** (1/2.2),  we can recover x from y by using y ** 2.2.

Now consider some tone-mapping function of that value, F. You argument is that F(x) is substantially different from F(y ** 2.2). For EXR format this is not true.

For 8-bit integers the differences are larger, but the question is how much larger? Let me ask you, since you spoke of this difference so confidently, can you say for which integer values, x (0 to 255), is the difference F(x) - F(y**2.2) largest, and what that difference is?

Of course it will depend on F, so maybe it is only valid to ask what is the difference between x and y ** 2.2 (for integers)?

Then you have to ask yourself, since even linear RGB integer values are off by as much as .5/255, how much of your F(x) is wrong to begin with?

What I'm getting at (and you should research for yourself) is that encoding your render (linear or gamma corrected, either way) as an 8-bit image, results in round-off or quantization errors. Depending on whether it is linear or gamma corrected, the distribution of these quantization errors are changed, but there are still errors. Now in post-work, you apply functions to these erroneous data points. You are doing garbage-in-garbage-out. If you have no option to export your render as EXR, then the only honestly quality work you can perform is to do no tone-mapping postwork. Any postwork tone-mapping you do will compound the errors.

For example, if I make a fine gradient, with small steps x, x+1, x+2, etc. Suppose you postwork tone-map that. The new gradient will have gaps in it at some points, i.e. x+5, x + 8, x + 9, particularly if the tone mapping you do is gamma correction. If you don't believe me, when I come back from vacation, I will show you this from real results. Quantized input to gamma correction is very very bad, which means trying to gamma-correct 8-bit RGB is very bad.

**Also you don't want to put a fixed gamma in your image either cause then you will have to do a rerender everytime you need to change displays. eg a 2.2 gamma works fine for online galleries, but not for macs or other things like printers, film etc.

**Hmm. And when a Mac user looks at my web gallery what does he say - this looks like crap? 

I think the difference between gamma = 1.8 and gamma = 2.2 is small enough to fall within the notion of "subjective quality" - i.e. you could easily see the artist wanted it either way, and you can't really tell which is the one he didn't really want. Or you could adjust for gamma = 2.0, the middle ground, and have it slightly dark on PC and slightly bright on Mac.

Or, you could just render twice! If you're really that anal, then you render at both gammas, and publish both. I know this takes longer, but compared to post-worked gamma correction of a 24-bit color image, the results are often noticeably superior.

OTOH if you have a raw render without gamma, you can apply whatever gamma number you need to it.

Only if you have EXR, i.e. floating point numbers. If you don't, then this workflow produces demonstrably inferior results.

Pseudo off topic.
Bagginsbill, what do you think of the node materials in Lightwave? Just an innocent question, no animostiy or software war intended...

wow, you sure took that personally. 

first of all, i'm not sure you understand painting and color, so i guess we're even.

second, i'm not saying bagginsbill is infallible.  i've watched him correct and advance a lot of his theories over a couple of years.  but i've watched long enough to know a little bit about how he works.  he tests.  excessively and more methodically than the professors i know.  i've never seen anyone else here or anywhere else go through permutations of settings and situations in such a methodical fashion.  and this is on top of his mathematical abilities and knowledge of physics.  please, go ahead and debate.  i'm sure we'lll all learn a great deal.  but since i've only ever seen postings (both here and at CG Society) by technicians who were much less accurate, i thought it would be fair to warn you that you're about to debate someone who is incredibly thorough about testing an immense amount of factors dealing with lighting, rendering and materials.  it is certainly possible that you are at the theoretical level of working with and deriving your own equations and the technical level of testing your theories with the precision of a scientist.   but even one of those qualities is rare in  these forums, both are almost unheard of, and i can't guess at how much you know about a single member of this community.  since it doesn't seem like you have even seen his comparison of gamma correction with PS and rendered gamma correction (because it would be an obvious talking point), i was guessing your knowledge just might be less than complete.

Quote - I'm guessing that the number of Poser users who have even have any normal maps to use is less than 100. Of those, maybe 5 use shaders of any complexity whatsoever. Of those, none are regular Rendo forum users.

Honestly, as I said in other posts, the normal map is something I just don't care to use, since I do my bump maps procedurally. I need to be able to generate the texture with nodes and the Bump channel does that for me perfectly, such as this peeling painted stucco. This is a gamma-corrected shader with procedural Bump.

And I don't agree that every render needs postwork to be called final. The wall here is just fine - I can't imagine how I'd improve it in postwork.

i use normal maps in photoshop. there is a free plug in for photoshop. you put in a bump map and photoshop makes normal map.
it renders high quality faster. at least this is what i found. and since i need short render time because of animation this would help me very much.

you have poser pro. is the same in poser pro? doest it work with alternate diffuse?

same in PPro - doesn't work

to bad.
but what can you do. we still have bump.

According to the Poser 7 manual, when I look up Gradient Bump, I'm told it's for using P4 bump maps when rendering in the P4 render engine. Gradient Bump: Previous versions of Poser used a proprietary file format for bump maps (.BUM). If you want to use these older file types in Poser 7, connect an image map node to the Gradient_Bump attribute and load your desired *.BUM file into that node (see “Image Map” on page 335 for information on loading images). Then, select Poser 4 Renderer in the Render Settings window and check the Ignore Shader Trees checkbox. Please refer to Chapter 37: “Using the Poser 4 Render Engine” on page 364 for more information on the Poser 4 renderer. You cannot use .BUM files with the FireFly renderer. Is the manual wrong again?

Bah. Tried to copy and paste from the Poser Pro manual, but it didn't paste correctly.

There are additional modes used for the Gradient_Bump channel, defined by the new parameter, Gradient_Mode. Using this, you can choose to use Normal Maps in the Gradient_Bump input.

its working now  with specular in poser pro. but if you use the alternate diffuse it doesnt work.

Dark, look when he last posted in this thread will ya? last year!

Quote - Still banging away on the normal maps are no good, huh BB?

I'm guessing that the number of Poser users who have even have any normal maps to use is less than 100. Of those, maybe 5 use shaders of any complexity whatsoever. Of those, none are regular Rendo forum users.

Normal map render no postwork.

No specular either. Pay attention. I said the normal map DOES NOT SHOW UP IN SPECULAR.

I didn't say it doesn't show up at all.