I know this has been brought up before here.

But cant actually locate it at the moment.

I am using P7 and am going to upgrade to either P8 or PP2010.

Now considering I dont have a 64 bit machine. Not likely to ever network render.

I am asking what are the benefits of getting PP2010 over P8 if their are any

as for what I have read to date the diffs are little. and there has been a multitude

of problems with PP2010 mainly with the libs I believe so any advice would be great.

Trix.

The biggest difference to me has been the renderer gamma-correction. Whilst most people - the point-and-click crowd - find this "feature" more of an annoyance than a benefit, once you understand material/render gamma correction you will see this as a real boon.
To clarify, gamma correction at the software level has nothing to do with monitor settings: if you do a search on this topic, you will find a lot of misinformation on the topic. To my understanding, it's all about what sort of information Poser needs in order to process colour information correctly.
Not going into more detail than that... suffice it to say that those who have explored GC in terms of render quality have found this quite useful.

The fact that it's 64 bit is really what interested me the most. I tend to make very large scenes with lots of high res textures and props and can get up to 5 GB of RAM or more  pretty quickly, so 64 bit makes a huge difference. Of course that doesn't mean anything unless you have a 64 bit machine and OS.

That and the GC like RobynsVeil already mentioned.

Another huge thing though is the ability to load a full body morph in one shot through the figure's OBJ file. That's an amazing feature and a huge time-saver.

There are probably other features I'm overlooking. I can't use the Poser Fusion since they don't have a Softimage version, but that would be a decent pro feature for some people. Well, it would be if it had reasonable translations of Poser's materials into Mental Ray materials. As it is all it does is create a Blinn shader and apply textures, and no "pro" Maya user rendering with Mental Ray is going to be happy with that, but at least it enables one to get the figures and poses and animnation in there.

Which reminds me, there ought to be a Softimage version of Poser Fusion. ;-)

To answer your questions though, I think the library is great. I don't have any problems with it and don't understand why others do. I do understand why some people might not like it. I didn't like it at first but gave it a chance and now I love it. It really is an enormous improvement over the old way.

Since you don't have a 64 bit PC, and unless GC, Poser Fusion, and the other features like FBM loading are important to you, might as well just go with Poser 8.

As a member of the "point and click" crowd, allow me to opine that since you have a 32 bit computer, go with Poser 8. 

GC is just something to turn off unless you know how to go in and modify the shaders of every piece of product you own. If you know how to do it, I'm sure it's wonderful, but I wresteled with it for months before I found that it was not something I choose to use. It's just more time in the work pipeline unless you are already a material room expert.

For me, the over-ridding factor in my choice of Pro 2010 is the 64 bit build. If you are on a 64 bit machine with a 64 bit OS it runs like the wind, and is as bullet-proof as any Poser I've ever used. If not, save your money (and a whole lot of grief) and stick to Poser 8. It has the IDL feature which, along with BB's Envdome will change your life when it comes to ease of lighting and realism!

The OP seems to be implying that the decision between P8 versus Pro2010 is partly motivated by the library. I don't understand that, because the library is the same in both. The only difference is that in the 64-bit version the library is its own application and cannot be docked. But the OP started by saying no 64-bit willl be used. So - the library is not a factor in choosing between the two - P8 vs. Pro2010. It's exactly the same.

In both I am making improvements. Ultimately SM will make a decision about whether to port these improvements into P8 and/or Pro2010. I believe that P8 is done for service releases, but I'm not sure, and certainly I'm not a spokeman so don't take that for gospel. All I'm saying is that Pro 2010 has had only 1 SR to date, while P8 has had 3. Consider the possibility that SRs (improvements) may keep coming for Pro2010 alone.

I really can't stand the results of rendering without GC anymore. While I try not to publicly ridicule artists, when I see the renders of dead-looking skin in the galleries, I am embarassed on their behalf. I understand that many people simply cannot tolerate a lot of material faffing about and would rather just render. Well if you're that type, but you also actually care about your results looking realistic (not artistic, realistic) then get Pro. Pro lets you get more realistic results with GC and IDL, and NO messing with materials. Simple P4-style materials look fantastic in Pro 2010.

As usual I have to expect that somebody will educate me on the fact that not everybody thinks realism is important. Before you do, let me educate you on my point.

I'm not pushing realism on anybody. I'm saying, starting with the premise that an artist's preferred style is realism, that Poser Pro 2010 is the only choice, hands down, for that sort of artist.

If you prefer un-realism, then Poser 5 is more than anybody needs and we would not need to discuss purchase decisions at all.

Poser 2010's library is the best to date and has made working with my extensive Poser content library a breeze. As it is the same library as P8's you need not let it enter into the equation, nor need you worry about it being flawed - it is not. It's highly configurable and you'll be able to set it so that it works with your content in the way you want it to.

Gamma Correction is a valuable facility that you might well want to use eventually, even if you prefer not to initially. Poser 2010 makes using it much easier.

@Bagginsbill: With Render GC I still use the material room to disable it for transparency maps etc. Am I correct in doing so? If so, then there is still some messing with materials, albeit much less than with P8.

@ Snarlygribbly: Yes, Gc on transparency, bump, displacement maps need to be reset to 1.  However, there is a builtin script that will do that automatically for either a single figure or prop, or the entire scene without entering all material zones in the Material Room.

One other thing to consider, Poser 8 may be on sale now that the newer version is out.  Since you have 32 bit, you may want to consider or factor in price.  I have a 64/32 bit machine and I'm sticking with Poser 8.  I have had no problems what-so-ever with it and I just have just SR2.  As for fbms and Poser 8.  Poser 8 has tons of them about as many as  the DAZ figure series.  Jan

Quote - As for fbms and Poser 8.  Poser 8 has tons of them about as many as  the DAZ figure series.  Jan

Well that's not what I mean by FBM's. I meant that in Pro 2010 you can load a completely morphed entire body and apply an entire full body morph that way, through the figure menu. It's a much different thing entirely than using FBMs that come with a figure. It's loading morphs for every body part at once from an exported OBJ file that has been altered in another program.

I was actually referring to Incantrix original post.  The price difference may need to be considered when you can get Poser 8 at Purplus.net for $180 US (includes shipping and handling).  Jan  

"As a member of the "point and click" crowd..."

So like where were you at the last meeting? I mean you totally were not there. It was just me and the guy who always smells like Pez. Man, when you pledge Phi Alpha Chi, you gotta participate. We had to drink all the brewskies by ourselves and got sooo wasted. Dean Wormer almost busted us, but it was cool, really. Anyway, next week, we’re totally gonna panty raid the Gamma house. I hear the new pledge chicks are hawt!. So don’t be a wuss dude, be there or be square.

Oh, and bring your older brother's ID ‘cause last time we got carded and had to pay this homeless guy to buy booze for us. But dude!, he says he can totally hook us up with Vicky’s digits on account of he knows this dude whose sister used to work with her, or a maybe chick who looks a lot like her. So me and the bro-hams are gonna call her and ask her if she really did all that naked stuff with the sword and all. So be there. And remember the secret password is “PAC Rulez Others Drool!.” Later.

Pardon me for living.

Since it's clear that only "super users" are entitled to an opinion, I'll cease and desist trying to be helpful to anyone from this point forward.

Quote -Pardon me for living.

Since it's clear that only "super users" are entitled to an opinion, I'll cease and desist trying to be helpful to anyone from this point forward.

I wouldn't let one comment turn you off to helping folks here and I personally agree with you on Poser 8 versus Poser 2010.  Jan 

Well, thanks for the kind words. I appreciate it. They tend to be few and far between.

The OP sounded like they really did want various opinions as to how to proceed, and as one who owns both P8 and PP2010, I felt like my experiences might be worth something to them. But, I keep forgetting... since I don't model my own meshes and make shaders from scratch in the material room and do all the magic to make GC work, my opinion is worth nothing. Maybe I'll learn to shut the ____ up one of these days since I obviously know nothing.

And no, Jan... I'm not that kind of person. If I think I can help someone, I'll ALWAYS do it... just because it's the right thing to do.

Again, you are kind to offer comfort! :)

What are you going on about? Nobody said your opinion is worth nothing.

I could just as easily make the same statement - unless I'm a noob who just does point and click, my opinion is taken as criticism instead of valid information.

As I said in my post, if you have no interest in realism (i.e. making sure things behave as in real life) then you have no need of even Poser 7.

My experience with attempts at realism before Poser Pro 2010 (including with Poser 8) constantly required adjusting things, tweaking, making compromises, and reacting to strange unexpected lighting effects.

With Poser Pro 2010, you don't need anything fancy. Plain old color and bump work beautifully. That's not saying your opinion is worthless. If you find that is not your experience, I believe you. But I also would counter that with my actual hundreds of renders in Poser Pro 2010 where no fancy shaders were used, and the realism is very high.

One area that people complain about commonly is that the Daz skin shaders (and the hundreds of similar ones made by other vendors) don't work very well when you turn on render GC. That's because those shaders are implementations of gamma compensation and various SSS and secondary lighting hacks. When you double up on compensation techniques, especially those based on non-linear effects, you get over-compensation.

There is nothing complex about using GC. There is quite a bit complex about the Daz shaders. You do not want to use those with Poser Pro 2010. You want simple 2-node shaders.

Yes you have to remove those. But many vendors supply their textures with mat-poses in two flavors - the fancy shader full of compensations, and another with very simple setups. The simple ones work superbly in Poser Pro 2010.

It is quite fine and valid to note that you get bad results with complex shaders in Poser Pro 2010. It is not fine to say that is the fault of Poser Pro and you should just turn off the new features and stick with P7-style material management. The whole point of PPro is to eliminate all that complexity. So my opinion is the opposite because you're not actually doing things the right way. Doing it wrong and stating your opinion doesn't invalidate your opinion. It just confirms what I said. It isn't that you have to learn new techniques. You have to unlearn old ones.

Once you do, the point of Poser Pro 2010 becomes crystal clear.

I'd also point out that if you're not even going to enable shadows in your renders, then your opinions about how hard one technique or another is is biased towards click and render, which is hardly the point of Poser PRO.

My issue was not with you, BB, but with lmckenzie.

However, Who said anything about no shadows? I ALWAYS use raytraced shadows.

That aside, I stand corrected.

I think lmckenzie was just doing some light-hearted joking around. He and I have a similar sense of humor, I think.

One of the things I often talk about is that in order to be funny, you have to take risks. Sometimes the joke isn't funny at all or offends some people. I do that all the time, both in this forum and in real life. I'm reminded of the time I said something like "All you Mac f***ers are lucky to have a Poser at all" back when I was struggling to make P8 drag and drop work on the Mac. That pissed some people off really bad, although I thought it was a hilarious overreaction on my part that would be understood to be humor. It wasn't.

In fact, that very thing is busted again and I'm currently trying to fix it for SM, so this is a good time to say "All you Mac friends are lucky to have a Poser at all." SM is very committed to supporting the Mac, much to my chagrin. (I get paid a fixed amount to make it work, regardless of how many days it takes me. Grrr.)

Meanwhile, sorry about the shadow assumption. I glanced quickly at your one-and-only gallery image and saw the ear that is facing the ground is lit and assumed there were no shadows. Looking at it again, I see some shadows under the figure, but strangely not correct ones. Maybe your shadow min bias was set too high and the shadow got skipped in some places. Particularly troubling is under his waist, where clearly some of his body is throwing a shadow, but not all of it. Weird.

Thanks for the reply and the pointers. I'll take it back to the drawing board and see what develops.

On-Topic:
For me there are essentially 3 questions, that remain for deciding if PoserPro 2010 is more worthy than Poser 8, and that could not be answered by the otherwise omniscient internet:

• Can PP2010 output more than 8 effective bits per channel?
  On the smith-micro site they write that PP2010 can output hdri images. That does not necessarily mean that the images produced contain more levels per R/G/B than an 8-bit png file, but it probably does, at least for the main image. But what about the multi-pass rendering with PSD- output? For example, an output of a higher resolution depth-channel would really be useful.

• Can PP2010 read textures with more than 8 bits per channel?
  The best case would be, if PP2010 could read hdri images, just as it can write them. However the product page only says that it can write (explicitly), not read; but can it, for example, read 16-bit tiff?

• Is the gamma-correction facility an all-or-nothing switch in PP2010?
  Assuming that PP2010 handles images with more than 8 bits per channel, the one (and only) point where GC still would be beneficial is, when displaying something on the screen (render window and material room in particular) even when gamma is otherwise disabled, in the other cases a converter (like e.g. imagemagick) would probably suffice. From what i have read it sounds more like that there is a big GC-on/off switch where one could enable or disable GC, but not partly.

As a side note i am also wondering what the color-dialog from the material room might look like when GC is enabled. In Poser8 i get a color dialog that lets me choose RGB values in the range 0-255, which have at least a small resemblence to what i get without gamma correction, like: "Choose the color that your totally diffuse material, lit by a white lamp, would encode to in the resulting png-file.", so i can e.g. enter a 127 for a material that reflects 50% of the light. But with GC, do i have to enter 186 to get the same effect (since gamma(186/255) = 127/255), or can i enter 0.5 (meaning that i can enter decimal values), or is it something different (would not surprise me, if 127 still works, because of compatibility)?

Great questions.

Quote - Can PP2010 output more than 8 effective bits per channel?

Yes - in render settings you check the "HDRI-Optimized" output option.

Quote - That does not necessarily mean that the images produced contain more levels per R/G/B than an 8-bit png file, but it probably does, at least for the main image.

Yes you get full floating-point output with no loss of precision. In 16-bit TIFF you get some loss of precision and can't go over 1.0, but you 65536 shades. In HDR or EXR format, you get the actual number with the full precision allowed in those formats. Numbers greater than 1 are not clipped.

Quote - But what about the multi-pass rendering with PSD- output? For example, an output of a higher resolution depth-channel would really be useful.

Yes - full resolution for all passes. The post-processing options are maximized.

Quote - Can PP2010 read textures with more than 8 bits per channel?

Yes, but this is true of Poser 7 and 8 as well. No need for Pro just to get this feature.

Quote - Is the gamma-correction facility an all-or-nothing switch in PP2010?

Yes all or nothing on output, but there's more to it than just output gamma. Incoming material is linearized - anti-gamma correcting the incoming colors from images. On each image, you control the individual anti-gamma value. This is a point of confusion for those unclear on what a "linear rendering equation" is for and why you care, but it's very important. One of the things perceived as "extra work" is the necessity to think about and specify the anti-gamma value for each image. It assumes you want the render gamma to apply as anti-gamma on incoming material. This is wrong for bump, displacement, and transparency, but there is a 3-click Python script to fix that for the whole scene automatically.

Quote - As a side note i am also wondering what the color-dialog from the material room might look like when GC is enabled.

Excellent question. All the color chips on nodes are assumed to be that what you see is what you want, but in linear value. So PPro anti-gamma corrects these. They look the same, but the numerical value is the linear one, not the one necessary to see what you see.

Generally when a color chip is used to say what color you want to see, the anti-gamma corrected value is what you really want and that's what you get. When you really want .5, you should connect a math node and type .5. Or use the User_Defined node, in which you enter a color as floating point numbers in linear scale.

Also, there is a new node (I've never talked about it, nor have I seen anybody else do so) called the "Gamma" node. This node let's you specify a color and a gamma value to control how that color is interpreted. So you can use this node to enter gray 127 directly from the color picker and you can be sure that if you tell it that is the Gamma=1.0 value, it will not be anti-gamma corrected, even if you enable render GC.

You can also connect things to the Gamma node and choose which gamma to apply to its input, and which direction (correct or anti-correct).

Note: Failure to understand that color chips are anti-gamma corrected leads to confusion amongst those new to linear render equations. The Daz shaders put a light blue value in the Diffuse_Color. When anti-gamma corrected, this is no longer light blue, but dark blue!!! Thus, they look horrible. These shaders need to be edited because they are making an assumption that they are NEVER going to be used with the linear rendering equation, which is the wrong assumption.

I usually suggest that people just discard those shaders and use mine. Applied via VSS with one click, you get a linear rendering equation compatible shader that is compatible with any version of Poser. As well, mine looks more real, often producing a credible SSS effect.

Currently the VSS shader needs to be told to set gamma = 1.0 when render GC is used, but there is a thread somewhere showing how to add two nodes to make it automatic. And - I will be publishing an updated set of VSS template shaders that are yet another improvement in realism, have more options, and automatically detect render GC and switch off the shader GC automatically.

Note also that a color chip that is set to pure white is the same RGB 255, 255, 255 regardless of gamma. Those are safe to use in both the linear and non-linear rendering equations.

Wow, many thanks for this detailed answer, Bagginsbill.

I think i want to have it :-)

Sorry it took me a while to get back. Basicwiz, I indeed did not mean to offend you with my bit of silliness and I am sorry that I did so. I’ve seen your work and it definitely does not fall into the ‘point and click’ category. You obviously put a lot of time and care into it. My own dabblings OTOH are much closer to that, so I had a flash about a fraternity in which I was the only member. The rest, as they say is history – that went about as well as Ferdinand’s visit to Sarajevo in 1914. But comedy as BB (thanks BB -:) said is risky business and sometimes you bomb – mea culpa.

Wow Thankyou all for your replies many more than I expected.

And basicwiz I was also a member of the point and click society for many years as well. Until I got bagginsbill's vss and started playing around with that. which got my interest going for much improved renders. But due to working way out beyond the black stump (any aussie can tell you where that is) with very limited net service havn't had a chance to post anything. But the thought of the GC sounds very interesting. And there is always a chance in the future of getting a 64 bit machine (dreaming lol) .

Thanks BB for all the extra imput I do like to get some realism into what I and you have given me more thought to process. looks like

I think I will be going with PP2010 from what I have read so far.  Also the texture problem to Mentalray was that the render issue and only through Maya or does 3DMax suffer as well. As I was thinking of that being my next purchase.  (tax refund time is soooo good lol)

Once again thanks to everyone for their input it is very much appreaciated by myself.

Trix.

Lmckenzie,

No hard feelings. See my sitemail.

To all who were hurt by my thoughtless point-and-click remark: I apologise.
I had no intention of pointing fingers or making anyone feel bad about their approach to this tool. I tend to complicate my life artificially, so doing things the busy way is natural to me, but it certainly does not make it qualitatively superior by any means.

Poser is what you make it. Enjoy it for what it offers you.

Quote - @Bagginsbill: With Render GC I still use the material room to disable it for transparency maps etc. Am I correct in doing so? If so, then there is still some messing with materials, albeit much less than with P8.

You can use the "material mods" python script. (BB told me about this). You can select by material, by figure, by all, and set the gamma. It will then ask you which inputs it applies to, so you can choose it to apply to what goes in the transparency and displacement etc.

Sorry to bring back this month old thread, but where is this material mods python script. (yes, I have PP2010 )

Menu -> Scripts -> MaterialMods -> changeGamma

I can go to Menu -> Scripts, but there's no MaterialMods there. Searching my mac I found a MaterialMods folder. There's no changeGamma script in it. I opened the materialModButtons.py in text editor, and it references changeGamma.py but it's nowhere on my mac. materialModButtons.py does refer to a writeMats.py that I do have which goes with button "Inventory", and a set_mats.py. This folder is not in the Scripts folder so it doesn't show up there. Maybe I have a bad install. PP2010 seems to work at any rate. Though I wonder what else I'm missing.

BTW, I did the SR1 for PP2010 yesterday. Anyway to tell it worked. It doesn't mention SR1 in the about window.

Quote - Menu -> Scripts -> MaterialMods -> changeGamma

Use this script with caution. Some materials have the same image map plugged into both the Diffuse AND the Bump channels, and this script will turn affect both channels equally as the settings are applied to the image map.

Of course, materials shouldn't be made this way, but unfortunately some are.

Wow thanks again to everybody. A lot of information here and very much appreciated. Plus it has lead me onto other avenues of things to look at which i had not done previously.

I read this

http://www.runtimedna.com/forum/showthread.php?43536-Outdoor-lighting-with-poser-8-IDL&p=470003

I take it the idl in p8 is the same as in pp2010. And looks like I have a lot more to learn in relation to it my first attempt at using it gave an excellent scene but with severe glow under the armpits. But reading BB's earlier points have a fair bit of tweaking to do as the whole scene was just directly imported from p7. Did manage to do a couple of renders on the weekends with reasonable results which I posted once again old scenes that i had been working on in p7 that needed a lot of adjustments to get right and still need a lot looking at them. Am away at work again so will have to wait for the weekend before I can play around some more.

While reading this thread, I still wonder about the GC thing. Can anyone educate me a bit more on this?

I do understand the gamma distortion by monitor (printer, etc), and I do understand that it's compensated for by Color Management. A basic sRGB setup gives a rough start, Apple (gamma 1.8) and PC (gamma 2.2) are different, and actually I run color spider (datacolor.com) every 3 months or so to get profiles that match my monitors accurately so the same image gives the same result on each box and each tube. Each monitor gets is own profile indeed, even on my multi-monitor platform.

This way I try to avoid the trap that I compensate my image itself for any flaws in any hardware. And I comfortably exchange images and video with other people, running calibrated systems as well.

Now, why should I use (or: definitely need) Poser 2010 GC in order to get any photoreal results out? Why can't I do without? Do I need it for fully handmade (photoshop painted) textures too, or is it for photograph-based textures only? Is it for IDL/IBL lighting conditions or for normal spot lighting as well? (I never do IBL for indoor studio portrait shots, I've build multi-spot softboxes instead - I've found that the best way to get photoreal results is to mimic photographers working conditions).

I guess Bagginsbill is right (as always) and I'm missing something (as usual), but what?

Render GC is actually not very different from color management in things like monitors or printers. It's the same idea of getting accurate colors. And using GC alone will not give you photorealism any more then a well calibrated mmonitor will make a shitty photo look nice. The accuracy from using GC just stops you from making mistakes, which leads to better quality renders and (if you want) better photorealism.

Which, incidently, is why it's needed for pretty much all textures that are from real or pseudo photographic sources. Like say you do a 100% hand painted skin texture and paint it to look exactly like human skin would look like in a photo, you need GC for that as well. And of course any textures that are not mimicing photos (like bump maps) can be safely ignored.

As for lighting, generally without GC you'll get a render that's too dark for the amount of lighting in it. Many people that work without GC will then add more lights/brightness to compensate, which works, but is IMO an inferior method then just using GC in the first place. Why complicate your scene more then it needs to be ?

hi ghonma

thanks for the effort, but I still don't get it. I'm just too slow.

Gamma Correction is all about getting more detail out of the dark areas at the cost of loosing detail in midtones and highlights. Just take Curves in Photoshop, leaving the end points alone and move the midpoint up (or down). This is some exponential curve, and gamma is the name of the shape driving parameter. Gamma =1 produces a straight line (lineair), Gamma between 0 and 1 produces a hollow curve with less detail in the dark and more in the whites (great for clouds), and Gamma >1 does the opposite (great for church interiors). That's how I understand it.

My calibrated monitor does not make beauty out of shit, but prevents me from correcting any image itself for box and tube specific hardware flaws. As a result, the shitty shot looks the same on all calibrated whatever, even when viewed over the net.

When I use photographic material to build a texture, of course darks and highlights should have been avoided and all texture areas should be in the same color range. My hand painted parts should match the color tones, of course.

Then I map the textures onto the object, apply lights, and render. And as a render is as good as a camera shot or scanner capture, it needs to pass the usual levels (histogram) corrections, and sometimes a gamma adjustment as well. And especially Vue shots can be taken with different camera settings, and can be combined using HDRI techniques afterwards. Great fun, all like photography.

But I still fail to see why I should want additional gamma correction on any of my (color/diffuse) textures after I brought them into Poser, and before taking up the rendered result. And therefore I miss the "added value" of this feature. Apparently it solves a problem which I'm not aware of, yet. But which?

Quote - My calibrated monitor does not make beauty out of shit, but prevents me from correcting any image itself for box and tube specific hardware flaws. As a result, the shitty shot looks the same on all calibrated whatever, even when viewed over the net.

Sure, but why does it look the same ? It's cause your image has been color corrected to a standard (sRGB most likely) and all displays that follow the same standard, know how to display the image properly, yes ? But note that a renderer is not a physical device, it has no faulty tubes or cheap LCDs that need to be compensated for. So when you take a photo that has this hardware correction in it, and plug it into something that doesn't need correcting, what should happen ?

If you're working with dumb tools, like poser was a few versions back, they would just read the 'fixed' colors of this photo (which the renderer doesn't actually need) and happily render out inaccurate color values. Smart tools like recent poser versions though will let you remove this hardware correction before rendering and thus render accurate colors. After which you can, if you want, apply the 'fix' back so that you again end up with a nice calibrated photo.

Of course poser is a low end tool so it doesn't do full sRGB calibration and you have to settle for basic GC instead, which works ok most of the time.

At the risk of starting another debate, I'll just say this: you don't need anything.

This is what I understand. Images such as those in jpg format are gamma-corrected, or you really couldn't see them all that well. Poser 5-6-7-8 and Poser Pro and Pro 2010 use linear colour information when they process. (Shoot me down, but that's what I understand). If you give the colour processor non-linear (i.e., gamma-corrected) information, the colours won't process correctly.

Gamma-correction has nothing to do with photo-real or anything like that. It has to do with how Poser processes colour. In non-pro versions, you linearise corrected colour in the material room using a set of nodes and then gamma-correct after processing has been done. In the Pro versions, the programme does it for you.

Again, nothing to do with photo-real. Nothing.

ETA: there are formulas that have been written that are more accurate in terms of colour processing along the luminance spectrum than simple GC. BB came up with that formula too. It's around here, somewhere.

hi ghonma,

thanks again. We do disagree on the calibration part, but the debate helped me to figure out the Poser-GC thing. Hence, I write it out to inform anyone else interested. All comments are welcome, of course.

On the PoserPro2010-GC:

When one takes various shots of the same object, with different zoom and other settings, and Level/Gamma correct each individual image, one gets slight differences that has to be adjusted for when combining all those shots into one final object texture. This way, one can get photoreal textures for the girl, the bikini, the car and the background, for example. Or photoreal textures for skin, hear, eye, leather and lace.

But all those textures have received different corrections, and in order to use them in one scene one has to balance them against each other, for even more photorealism. One can do that upfront, or even in post using object masking, but now PP2010 can do it internally as well. Good feature.

On the calibration:

All physical devices have characteristics that put them into trouble producing detail in either dark or light areas. Paper / ink interaction, voltage to light emission, all non-linear outside the midtone range. One doesn't need any faults or cheap stuff for it. 

For that reason one never sends an image directly to a device, but to the device driver instead. This thing tries to make an inverse translation, in order to get the image out of the device as it went into the driver. Basic gamma adjustment is the rudest way, applying a standard (like sRGB) profile comes next, and using an up to date monitor specific profile is best. Such a profile exactly inverts the devices behaviour.

As a result, the image will look exactly the same on each monitor that uses an up to date device specific profile. There is no hardware related information embedded in the image any more. I can see it right in front of me, by moving a single image around over a single large scale multi-PC multi-monitor desktop. Or by displaying it in large on this monitor-wall in a 2x2 display setup. No color changes.

All this becomes different when only native of generic (say sRGB) color management is applied. As is usually the case when people view each others gallery images. When the result is overly dark, the creator very likely is on a Mac while you're on a PC. When the result is overly light, it's likely to be the reverse.

The renderer indeed is not a physical device, and has no limitations to be corrected for. So, the result usually will need a Levels (Dynamics) adjustment, but not a gamma correction: it's linear between the images white- and blackpoints. But one might want to add in a gamma correction, to make the image look like a real camera shot, eventually compensated for outdoor-clouds or indoor-church conditions. That's personal, as there is nothing to fix. Or one might want to adjust the image to make it not too much off for people with a non-calibrated tube, on Mac as well as PC.

Happy Posing.

You make valid points that must be thought about, but the essence of render GC has nothing to do with monitor calibration or compensation for device variance.

It's much simpler than that.

8-bit per channel digital images encode colors using very few bits - in fact just 8. This is a serious limitation. (One that does not apply to HDR or EXR file formats.)

With only 256 levels possible, a linear mapping of the numerical values to luminance would result in unnecessary detail at one end and severe loss of detail at the other.

Therefore the industry settled on the sRGB standard. This is, as you know, a non-linear mapping.

For the most part, all pictures, whether from camera, or renderer, or scanner, or hand-drawn in Photoshop, are encoded this way.

Poser render GC (or LuxRender GC or many other apps) exists for the reason that it is naive and wrong to treat digital images as linear data.

Shine a line at a texture at various angles that has been interpreted as linear diffuse color and the result is wrong.

But convert the incoming texture (anti-gamma correcting it) to linear luminance, and the linear diffuse color equation comes out right.

Then it gets linearly combined with specular effects. (The entire rendering equation is linear, in fact.)

Then the final result, in order to be looked at with fidelity on a monitor, must be converted to sRGB color space. That means gamma correction at the end.

You said earlier that gamma correction is all about getting more detail in the darks. That's incorrect. The sRGB standard is what is all about getting more detail in the darks.

Render GC is all about recognizing the sRGB standard. Failure to do so results in luminance and color distortion. The distortion is often attributed to "the Poser look" but it's there in other apps as well, whenever you do not pay attention to the sRGB encoding of digital images.

The parochial view of the Poser community is at fault in constantly explaining render GC as some sort of Smith Micro peculiarism, or that it has to do with monitor variation, and also claiming it is just an artistic choice. It is a choice, but only in the same sense that deciding not to poke a chopstick in your ear is a choice.

to RobinsVeil (posts crossing when writing the previous one)

I agree, we even don't need Poser in the first place ;-), but I just like to understand what's going on and like to grasp why Bagginsbill took his very firm position on PP2010, the GC-feature, photorealism and not-needing P7 or up when you were not into this.

I understand it now, and I can see his point.

To my knowledge there is no gamma distortion in JPG by definition, but Photoshop (and cameras?) can add profile info (say sRGB) into the file to tell which was used to view the result. Note that hand-painting and scanning produce linear color info right from the start.

When you use something different (say Adobe RGB, or a device specific one as in my case) some software (eg Photoshop) tries to compensate for the difference to make me see the same thing as you were when making the image. It's some attempt to have the experience and workflow of a calibrated environment, but without the need of calibration itself.

And perhaps PP2010 uses this info for linearizing the colors before rendering, because that's what's required indeed. But I would be surprized, as lots of processes loose this extra info on the way.

JPEG is sRGB *** **BY DEFINITION *.

You said:

Quote - To my knowledge there is no gamma distortion in JPG by definition,

This is where your wheels fall off - your knowledge is incorrect.

Quoting wikipedia:

Quote - Many JPEG files embed an ICC color profile (color space). Commonly used color profiles include sRGB and Adobe RGB. Because these color spaces use a non-linear transformation, the dynamic range of an 8-bit JPEG file is about 11 stops; see gamma curve. However, many applications are not able to deal with JPEG color profiles and simply ignore them.

Almost all apps ignore the color space and assume it's the standard sRGB. In any case, whether sRGB or Adobe RGB, neither are linear.

I could write a shader to decode ARGB in Poser but I never bothered.

hi bagginsbill,

thanks for comment. As I stated in my first post in this thread, you're always right (well, about) and I miss something (well, not always).

JPG indeed can have a profile like sRGB or AdobeRGB embedded, which then is used for the colormapping for exactly the reason you explained: otherwise we would have not enough detail in the low and high ends. But... actually most JPGs do not have a profile embedded at all. In that case, a gamma 2.2 is used (and before Snow Leopard 2009, Mac used 1.8 instead). The difference between sRGB and gamma 2.2 is limited, but nevertheless. In any case, there is no linear colordata in JPG. Actually, there is no real RGB in JPG either, it's more Photoshop Lab alike.

Other file formats, like BMP (and TIFF? PNG? AVI) do not have any colormapping included and hence contain linear RGB-like colordata, while MPG (as on DVD's) has gamma=2.2 embedded.

(It's all on the Wikipedia pages you were quoting from, on JPEG, Gamma and more).

So, PP2010 uses the sRGB profile instead of gamma 2.2, when sRGB is included. What if another profile is included? What if no profile is included? What if the texture is in non-JPG format? Can GC (a gamma value) be set on a per-texture basis or is it an overall setting? Just asking.

Gamma Correction is a generic action on images which can be performed in Photoshop, it exactly does as I said. When using Levels, the middle number denotes the gamma correction, and a value of 2 brightens a 50% gray to 71%. Using curves, raising the midpoint (127,127) to (127,180) does exactly the same. As can be seen, a small range in the dark is mapped onto a large range, enabling to view more detail. 

sRGB includes (an close approach to) gamma, but essentially it's a color definition which tells us which shade is meant by extreme red, as in (255,0,0) or any equivalent. Extreme red in sRGB is different from extreme red in AdobeRGB, or RAL, or LAB, or Pantone, or whatever colorspace you're referring to. Such is relevant when you've got to produce the same color on various media, like plastic, wood-paint, textile, paper and so on.

So, whether one needs gamma or sRGB to get more detail out of the dark, is more a case of semantics to me. One can do gamma correction without sRGB, but when one maps onto sRGB, the gamma is included.

wrapping up, at this moment the GC-think boiles down for me as follows:

When you use JPG for diffuse/color texture, and it contains a colorprofile (different from the JPG default one, which I presume is gamma=2.2 but might be sRGB) and you run PP2010, then it will use the profile instead of the default while P8 (and earlier) will use the default instead, as will most renderers around. Correct?

When this JPG default is gamma=2.2 indeed and sRGB is applied for adjustment, then PP2010 will show a lighter texture in the dark areas of it, and a mildly darker texture in the light areas of it. hence you'll experience a slightly reduced contrast and more detail in the dark areas, especially in places where the scene is somewhat dark too. Thats good.

When you apply a different image format, or you apply it to a different channel, or it does not include an explicit color profile (*) or the JPG default already is sRGB, then PP2010 produces results similar to P8 - with respect to this subject at hand. Correct?

(*) just open the JPG in a text editor, it will show some readable ICC profile info in the first part of the file. Or use a hex editor instead. The profile takes about 3kb file space.

You're saying things for which I have zero evidence, and zero contradictory evidence.

I'm not aware of Poser (any version) paying any attention whatsoever to color profiles embedded in images. That doesn't mean I know it doesn't pay attention. I mean I'm not aware of it - I've never used a JPEG with knowledge of it containing an explicit color profile.

In Poser Pro it uses the render gamma setting or it uses an explicit gamma on a per-image basis. I'm not aware of it ever correctly implementing sRGB.

Not Pro versions of Poser don't pay any attention to any of this. They simply take the numerical values from the images at face value - which is wrong.

Wow one simple question. Started all this.

Been a pleasure reading it.  As always BB a wealth of information. Understanding of it all may not be up to speed. Me being one that just plays with everything till I get something that looks half decent, which more often than not doesn't happen? But all part of the learning process.  Linear information and sRGB go way beyond me and unsure as to how much I really need to know about them anyway.  But if it helps to improve what I do then well worth learning about.

Bagginsbill what difference does it make to the render when using idl in conjunction with your env sphere. Which I have but have not as yet ever used. BUt have a couple of outdoor scenes that I am thinking of doing and have read various tutes etc on the benifits of using it. but wondering with the idl and what effects etc it would have on it.

Quote - You make valid points that must be thought about, but the essence of render GC has nothing to do with monitor calibration or compensation for device variance.

It's much simpler than that.

8-bit per channel digital images encode colors using very few bits - in fact just 8. This is a serious limitation. (One that does not apply to HDR or EXR file formats.)

With only 256 levels possible, a linear mapping of the numerical values to luminance would result in unnecessary detail at one end and severe loss of detail at the other.

Therefore the industry settled on the sRGB standard. This is, as you know, a non-linear mapping.

For the most part, all pictures, whether from camera, or renderer, or scanner, or hand-drawn in Photoshop, are encoded this way.

So, in a way, sRGB is very much like Dolby. (In audio recording the basic idea of Dolby is to compress the dynamic range on recording and expand it back on playback. For example, a symphony orchestra has about 70dB, but standard Chromium Oxide cassette tapes (which dates me) had about a 55dB range between tape hiss on the quiet end, and distortion on the louder end. Dolby C compresses the dynamic range by about 15dB on recording, and expands it back on playback. Most of the compression occurs at higher frequencies (since most of the noise on tape is moderately high frequency), and low amplitudes.

hi all,

before this thread takes all directions (after I sort of hijacked it), and to save Bagginsbill the trouble of repeating himself for the tenth time in this post and others (sorry BB, as I said, I'm slow - but determined) I'll wrap up the CG issue, present the basic approach as in PP2010, and discuss some issues. In a next post, and some following.

wrap up and basics.

at the start, the question was about the main differences between P8 and PP2010. It turned out that PP2010 is a major step forward for 64-bit environments. On top of that, BB (baggingsbill) pointed out that the GC (Gamma Correction) feature alone was worth the buy for a 32-bit environment, provided one was into photoreal work. This is when I got interested, as I'm running Vista-32 and P8 at the moment and I'm looking into improving on photorealism.

For various reasons, it took me a while to really grasp the issue at hand. Here it is, in my own words and in an attempt to leave out the techy details.

1. output devices are seriously handicapped by their inability to handle fine nuances in the dark areas. This is due to their non-linear behaviour. Linear means: when you feed it twice the input you'll get twice the output, and the other way around as well. For output devices (monitor, TV, printer, camera or phone display, ...) it's different. When you half the brightness of the image, the display will not generate half the light intensity, but less (at high brightness levels) or more (at low brightness levels). Hence these devices are non-linear. Their behaviour can be measured, and put into a graph which is known as the gamma-curve. Input devices (camera, scanner) do have similar characteristics in the light areas but far less noticeable, and as the chain is as strong as the weakest link, I'll focus on the output side.

2. unless your machine has Color Management implemented (more on this later), or your painting program is not Photoshop (which had onboard CM, more on this later too) this will effect your behaviour when hand-painting a texture. As your monitor is not displaying the dark color that well, you'll have the tendency to pick brighter colors than necessary, because it looks good. This way, you are baking the compensation for the monitor characteristics into your texture. The same thing occurs when you're improving on a photo: you will pump up the darks more than necessary in the ideal linear world.

3. the Poser renderer is assumed to behave linear as well. This means: half the light on a full red color produces the same result as full light on a half-red color. This is not the case when the inputted texture is non-linear itself, so you'll get color mismatches and artifacts. Shadow gradients will be too harsh, lighting differences produce different results where they shouldn't, and lit areas won't match with unlit areas like background images. To say the least.

4. so PP2010 provides the answer: the inputted textures will get its baked-it monitor compensation removed (the GC=Gamma Correction), then rendering is done the linear way, and at the end the monitor compensations can be baked in again to make the result look good for you (GC reversed).
   And while doing so, it can give different adjustments to different textures to balance them against each other (see one of my earlier posts), cater for the fact that some images might not need adjustment at all, or a different one, and so on.

So, this definitely is the way to go.

But their are issues which cloud the sky, since more ways exist to handle the effects of the monitors non-linear behaviour:
 - your system can have CM (Color Management) implemented
 - you an use Photoshop which has CM embedded in the user interface
 - you can save to JPG, which has GC build into the format definition

And then you run into the confusion I had in the beginning: what if all three occur and you're using PP2010 at the same time? Fuel for the next post(s). 

Color Management is the professional answer to handle the effects of non-linear behaviour of (outputting) devices in a production environment. How can we make different outputting devices make the same result, like printing logo's on paper, plastic, textile and painting metal, or like viewing the same image on different (kind of) displays.

For each device, the operating system works together with the hardware to compensate for the non-linearities. This brings the forementioned gamma-curve back again, but we also have to define which shade is exactly meant by Full Red, or (255,0,0) or whatever. All this is captured in the color profile. sRGB is one of them, and this one deals with colors that can be reproduced by about all displays, printers and  so on. You can have a far better one for your monitor, but at the risk of not being able to reproduce some colors on your printer. Nowadays, sRGB is the standard for webpage colors, and the default for PC equipment. It's meant for home and office use, not for high-end graphics.

Implementing CM effectively implies that you do not have to adjust the image itself for the non-linear behaviour of your monitor. So, when you exchange the image with someone who also have CM implemented, colors will show the same. Well, almost because when both of you approximate the monitor behaviour by using sRGB, some differences will remain. If both of you deploy color profiles which exactly match the respective monitors, all differences will vanish. And textures produced will be linear in themselves.

Textures like this will be handled by P8 the proper way as the supposed linearity is there indeed, while in PP2010 they should be assigned a gamma=1. Otherwise, PP2010 will take out a distortion which isn't there in the first place.

So the question is: is the texture at hand produced in a CM-aware environment? You might know, you might not. I don't know about Apple and Linux machines. For Windows, CM is there for Vista and later, XP had an optional applet and before that it had to be implemented explicitely, using the profiles which came with the device (the CD no-one uses). It looks a safe guess to me that CM is "in" for recent handpainted textures, and that portion will rapidly grow the forthcoming years.

And then, Photoshop and the JPG file format have to be taken into consideration. These have their own answer to CM.