Here's how things look in the 'Pose Room' - pretty well identical, as you can see from these screen shots.

Can others with both P7 and PPro run these sorts of tests to see if it's just my system, or a 'feature' of Poser Pro?

Just lower gamma to, say, 1.8
Try a bit to find the best parameter for your renders.

It does, IF you set the gamma properly, and I believe this is even covered over at poserpro.net

http://www.poserpro.net/New_Index_9000.html

New version, new things to learn, plain and simple.

After reading it, it's hard to teel if I'm seeing what you are actually seeing.. for instance, if you have a crt and are rendering on that, on a lcd monitor, that could look quite different.

When I got my lcd a few months back, all my prevous stuff looked quite bad in some cases.

My lcd is about 800 times brighter then my old crt monitor. I turned not only it's brightnes settings down, but then also even turned dwon the r/g/b settings as well, so it wasn;t so bright. And odds are, someone else with a new lcd is going to have thier turned up very bright compared with my settings.

Basically, they've just given us another "in app" tool to adjust things that we might have done in post with poser 7.

Quote - Just lower gamma to, say, 1.8
Try a bit to find the best parameter for your renders.

Thanks for the suggestion, but no amount of messing about with gamma settings will allow me to match Poser 7.

I understand what gamma is and what it means within the context of Poser, and I can see uses for it in the future. However, for now it is incredibly frustrating that Poser Pro can't render exactly like Poser 7.

I'm not asking for tips on achieving good renders in PPro: I can do that. What I am asking is, does anyone else have this issue? It would be really useful to get input from others so that I can determine whether this is something to do with my system or something that SmithMicro have (in my opinion) got wrong.

Whenever you add features and improve a render engine, you can't really duplicate the old one exactly, because it does NOT have the same limitations as the previous render engine does.

Notice that ther's no "poser 4,5,6,7" render options? They've improved things over time, and why someone would want poorer quality renders is beyond me.

And I wasn't offering tips on doing better poser pro renderings. Turn gamma on, and do some test renders until you get very close to the old poser 7 renders, and then just use that setting when you render an old scene. Since it's going to vary from monitor to monitor, and system gamma settings, and video cards, there's no "magic bullet" number for all systems for the best gamma setting.

I ran into the same thing when I was beta testing poser pro, and was told the same thing, and when I asked why it wasn't the same, they asked why I would want inferior render quality in the new version?

With every new poser version, there have been new tools to learn to get our renders how we want them.

Or just use poser 7 when you want renders with the older render engine.

Quote - It does, IF you set the gamma properly, and I believe this is even covered over at poserpro.net

http://www.poserpro.net/New_Index_9000.html

New version, new things to learn, plain and simple.

The Gamma thing: in PoserPro you can turn it off. The point being that if it is off, it will not influence the render and that means it should be exactly the same as Poser 7.

I'm running lots of tests while I'm on here, but I'd really like some kind soul to quickly set something up in P7, render it, save the PZ3, open that up in P-Pro, render it with gamma correction turned OFF, and post both results here.

of course in the real world I expect some differences, it's just that this seems to be so very very different. I don't recall, for example, a similar expereince when switching from Poser 6 to Poser 7.
That clearly involved a major improvement in the Firefly engine, yet the renders from P6 and P7 were (to all intents and purposes) identical when using the same scene. I've probably got my tests backed up on DVD somewhere; the point is, I ran them, and had no issues.

Surely you'd accept that with gamma off, P-Pro and P7 should produce similar results?

EDIT: just did these histograms from the renders. The P-Pro one is generally much better than the P7, but with the fatal flaw that it's clipping the shadows!

Top is P7, bottom is PPro.

Also taking a look at your renders, the main difference seems to be in shadow density. Have you tried turning shadows off to see if you get identical results ?

EDIT : i'll try running my renders through the histogram as well

ghonma: thanks!

Well, yours look near-as-dammit identical [just did a quick analysis - they are 100% identical], and your comments have prompted me to look into the way the lights are set up.

Thanks again.

EDIT: Not sure i like these 'improvements.' After some more testing i also noticed the crazy amount of banding in the PPro render (compare the gradient on the thin purple box on the ground) And i tried disabling AO and still get differences. So the changes are most probably in how IBL is being done

Meanwhile I went the other way, creating a really basic scene, no IBL, AO, fancy shaders or anything. No point in posting an image, both renders were identical, pixel-for-pixel.

Having run a few more tests, it doesn't seem to be light intensity, shadow stettings or AO that affect the render, so my next set of tests are going to be with IBL But that'll be tomorrow.

Hmmm. Looking over my past correspondence with SM, there is no explicit guarantee that Poser Pro will render identically with P7.

However, having observed that gamma correction, despite all its merits for realism, might interfere with existing scenes and result in loss of artistic control, I was told "That's the reason why there will be a checkbox to turn on/off all gamma correction." That is a quote.

Now I'm no lawyer, but it seems pretty obvious the intent was to maintain the existing results for existing scenes, when you turn off GC.

You may very well have found a bug. Might I suggest you disable gamma, and save the scene as a new file, exit Poser, restart, and load it fresh.

See, when you switch it on and off, Poser has to re-evaluate all its image files. It may be that the IBL image is erroneously not refreshed.

Another thing to try as a workaround, is to enable gamma, but set it to 1.0. This should make no difference, but it might.

Restart didn't change anything and this is with gamma=1.0. Seems identical to the one with gamma disabled.

ciao

Bopper.

I want to post a picture but it does not show up, why is there a maximum size or so?

There's a maximum size of, I believe, 200kb.

And just leave me alone with this Gamma Correction ... no matter what settings the result is always crappy. But yay, we've got something new but we could not fix that transparency bug that always occurs when the trans map has not the exact same size as the texture map: then Poser 7 and Poser Pro will replace the trans map with the texture map. And I've reported that bug last year in June and was told they thought is was fixed with SR2 but surely they will reopen the bug and now fix it ... bla bla ...

I think I've found the general area where Poser Pro differs from Poser 7. I built a scene (of questionable artistic merit) and lit it using some of Poser's standard lighting setups. I started to see differences when using IBL lights. Subtle, but they are there.

EDIT: image didn't load, but it's not important.

EDIT: forgot I can't upload zips, to right click the link and 'save as', then change the extension to .zip.

I do recall some said about improvements to IBL lighting, but since I never use it, I wouldn't really see the improvements.

It might be interesting to see a simlar ibl light in something like, say Vue6I, and see it compared with a poser pro render.

I've spent around half an hour digging around in the Material Room and I can't see any reason for the inconsistency described in previous posts, nor any way to fix the problem.

My conclusion is that SmithMicro have made a simple, probably unintended change to something in the way shaders are calculated that has a drastic effect on renders.

In other words, they screwed this up, and for reasons I can't begin to fathom this wasn't picked up during testing.

I am desperately disappointed.

I think we've noticed it since we loaded scenes into PoserPro that we've already made and rendered in P7.

Btw, not only do the renders look different, the filesize of the PPro renders is smaller than the P7 render although it's exactly the same resolution and size.  The images I posted above are both rendered at 3000x4000, 300 dpi and both saved as bitmap: the P7 render is 50.5 MB and the PPro render has 38.1 MB. How that?

The saga continues...

Some parts of the shader nodes / render engine work together very nearly exactly the way they did before.

Other parts work together very differently.

If the whole thing was different, that would be a bleeding nuisance, but acceptable if renders were a lot better.

What seems to be happening - that some small part is working totally differently - can not possibly be intentional. If it is, I'd love to hear the justification.

Well, I can't waste any more time on this issue. I'll look back in a few days.

Surely this must have been tested by the beta testers, comparing renders & setups from P7 to P7 Pro?
Unless they changed something at the last moment.

You'd think so, wouldn't you?

This image shows P7 render on the left, P-Pro right, the big one being the 'difference map'. This is created by putting the P-Pro image on a layer on top of the P7 layer and changing blending mode to 'difference'. I've used levels to emphasise the differences.

But I think we could live with these differences.

This may be totally off base, but going back to Poser 6 I've had issues with IBL not doing what I expect, particularly it is always bright and washed out, unable to produce darker areas.

The "Contrast" parameter affects this - perhaps the contrast interpretation has changed?

While I understand the nuisance that existing scenes don't obey your careful tuning of the past - it may just be that we have to re-tune everything, even if Gamma is not involved. So far - yes some of those Pro renders are darker - but maybe they are more correct?

That was my thinking at well, BB.. that IBL rendering has been improved, accounting for difference with the older P7 renders.

And it could be that in the next P7 patch, that the same render improvements are implimented, rendering all this research a moot point.

I'd agree with BB and Gareee that PoserPro renders are, in general, better than those in previous versions. But there is an issue here. I've looked at my screenshots of the Material Room settings for my IBL light posted earlier, and the results have convinced me that 'something' is wrong.

Taking the brownish tone in Color_Math_2, in the P7 version the RGB values are 143/130/114. These values are indentical in Color_Math_4 and _5, as we'd expect. The values in the final bit of the shader are 93/85/74 which, since we have intensity set to 0.65, is exactly what we'd expect.

Compare this with the same set of values in the P-Pro version. In Color_Math_2 the same colors have been selected. Yet the RGB values in Color_Math_4 and _5 read 73/57/42, which are just plain wrong. I can't figure out what is modifying them, but they shouldn't be getting modified at all.

Well, this exercise has convinced me that there is a glitch in the software, somewhere, but I seem to be having a hard time convincing people whose opinions I'd normally defer to.

I've spent way too long on this already. I'll post again if I get anywhere or if any interesting replies get posted.

Quote -
And it could be that in the next P7 patch, that the same render improvements are implimented, rendering all this research a moot point.

Isn't this what Pro should have been really, a patch?

Quote - That was my thinking at well, BB.. that IBL rendering has been improved, accounting for difference with the older P7 renders.

Gareee, I did not use IBL or AO in my Poser 7 scene and still it rendered darker in PoserPro. It's not the IBL whatever thing, there is something wrong in general.

Quote - And it could be that in the next P7 patch, that the same render improvements are implimented, rendering all this research a moot point.

Well, I hope they leave P7 alone and don't try to implement this shit into it too. Before they start adding stuff no one wants they might want to fix the still existing bugs first.

Punt, when was the last time you contacted them abotu that transparency bug you keep going on about?

cspear: Can you post a larger version of the material room screencap ? If rosity doesn't let you post a big version, you could use something like imageshack.us or similar.

I'd like to try duplicating the problem but i can't make head or tail of the small pic. Thanks.

Quote - Punt, when was the last time you contacted them abotu that transparency bug you keep going on about?

Today :P And I even told them why it happens. Maybe this helps them to create a fix a bit faster as when I rely on them to find out why it does that  ;-) LOL

Good ya poked them in the ribs again. I poke them every time I find something that seems odd, or something that could possibly be a better workflow improvement

A quick update on this:

I've presented the issue to SmithMicro and they've asked for more info. I'll let everyone know how this pans out.

Hello cspear

Are there any news from SmithMicro on the issue.

A had the same problem from the first Day using Poser Pro.
My own IBL lights produced far to dark images in Poser Pro.

For the moment I found a fix for some scenes including customized IBL lights.
If you double the IBL Light intensity. the results are much closer to the Poser 7 output.

e.g. Poser 7 intensity 120% -> Poser Pro 240%.
(1,2 -> 2,4 settings in material Room)

Maybe this helps someone.

Well, the new Poser 7 sr3 release updates poser 7, so what you see in it is the same as Poser pro now.. BUT the downside is all your old P7 stuff needs to be updated.

They did corrections to a number of render issues, and the option was to still do them "wrong"  and keep things consistant with P7, or update P7 to do things properly and update P7 to the new proper Poser Pro standard, and they chose the later.

How was it "wrong" before ?

Poser has always done IBL incorrectly - too bright. This drove me crazy as I would try to create an IBL probe with very specific amounts of light and it never worked right, always over-lighting my scene.

Here is a very simple test you can do to see what your version of Poser is doing. I'm not sure what's going on with SR3 or PPro - at the moment I have not installed SR3 on the machine I'm on. I have P7 SR2 and a beta of PPro. They are behaving the same for me, and they are both wrong.

Create a scene with only one light - set it to IBL.

Do not use an image - just set the light Color to white RGB(255, 255, 255). Set the light intensity to 1 (100%). Set the IBL Contrast to 1. This should produce uniform light in all directions that is numerically 1. The result is that any color you put onto a shader for Diffuse_Color should be reproduced exactly.

Now - Add a primitive one-sided square or a sphere or a box - what kind really doesn't matter. Edit the shader for this prop in the advanced material room. Starting from the default shader, make the following settings:

Diffuse_Color = RGB(127, 127, 127)  (roughly 50% gray)
Diffuse_Value = 1
Specular_Value = 0

Render the prop. Now we want to measure the color of the render. Click on the Specular_Color. This will bring up the eye-dropper color grabbing tool. Click on the rendered prop. This will copy the rendered color into the Specular_Color. Now Alt-Click the specular color to bring up the Color dialog - from which you can read the resulting RGB value. On Poser 7 SR2 and my PPro beta, the value is RGB(222, 222, 222). It should be 127, not 222.

Try a different RGB value - like this orange: RGB(255, 128, 64). Render, and read back the rendered color. I get 255, 224, 112. I should get 255, 128, 64. This is a HUGE error, and like it or not, letting Poser continue to behave this way means we cannot get accurate lighting from IBL. This drives me absolutely crazy. It is why I never had the proper amount of real-life contrast from IBL probe photos. I will be very happy if they change this.

We have to learn to stop compensating for the old renderer bugs, and deal with the correct results produced with the newest versions, if they're correct. If SR3 is still wrong, then I say fix it again.

Somebody who has SR3 should try this and report the results.

Forgot to mention, under the existing behavior I described above, to get the exact amount of IBL illumination to be right, you have to set its intensity to .575 (57.5%).

Hmmm. I just downloaded and installed P7 SR3.  I see no difference.

Quote - Well, the new Poser 7 sr3 release updates poser 7, so what you see in it is the same as Poser pro now.. BUT the downside is all your old P7 stuff needs to be updated.

As long as SR3 is that slow I'm not going to use it. I made some quick benchmarks.

a) Intel Q6600 Vista 64bit  8gb ram
b) Intel Q6600 Vista 32bit 4gb ram

Test Scene from http://renderfred.free.fr/p7benchmark.html using all 4 Cores

                                   poser process / seperate process
Poser 7 (SR2)          59 sec                  90 sec
Poser 7 (SR3)         180 sec                196 sec
Poser Pro  (SR1)     177 sec                50 sec  / 47 sec (64bit)

(only measured the actual rendering time not the loading times)
Those differences are more or less the same on other rendering jobs.
If I find the time I'll test SR3 on an AMD CPU.

Anyone made similar experiences ?

There is definitely something not right with Poser Pro.

I loaded two figures into a scene (one V4, one M3 with high-shader textures) in P7, used the Andre shader lights by Atlantistyle. Rendered, turned out as expected.

Loaded up PPro, and loaded the same two figures from the same runtime, and applied the same lights, and used the identical render settings. The shaders on the textures seemed to have been ignored, as did the displacement maps, even though the render settings were very high and Use Displacement Maps was checked (again, identical to the P7 settings.)

So, I went back to P7, and repeated my steps, then saved the scene. Closed P7, Opened PPro, loaded the scene, and rendered. It came out exactly as it did in P7.  The lights did not use IBL or AO.

So there's definitely a difference in how PPro is handling shaders and lights. I can't say it's "right" because it's not working as expected, and one DOES expect displacement maps to render when that option is checked. But it wasn't just displacement, it was most of the shaders (skin sheen, falloff, etc...) Big bug in my opinion.

And I'm basically seeing exactly what cspear originally reported with IBL/AO lights. The shaders are being ignored there too.

Ok, looks like the problem is completely with the Gamma setting.

In my first test above, I thought the settings were the same with the manually setup scene, but didn't realize Gamma was checked in the firefly auto settings. When I uncheck Gamma, The shaders and displacement maps are vivid and obvious. Correction also, the shaders and displacement maps aren't being ignored, but they are definitely subdued. The top image is PPro no gamma, the bottom is gamma check with default of 2.20.

So, something is clearly different. I would NOT say the gamma version is a better render.

Every shader you've every liked before Poser Pro + Gamma was a shader that was designed to compensate for the absence of gamma. Shaders that produce good looking results without gamma correction at the end will NOT EVER produce good results WITH gamma correction.

The GC should be applied to new shaders - not anything designed for non-gamma Poser. For example, all my old shaders look terrible in PPro with Gamma, and I think face_off's shaders don't work the same either. All the effects were carefully tuned to produce good looking results, using NON-LINEAR MATH. Now you're trying to use linear math with these shaders and it just isn't going to look good.

What you must learn to do is design shaders to work WITH GC, not instead of GC. You can't add GC on top. It's overkill.

Also, you can't just flip gamma on and off. You have to go into every last shader, and make sure that any image map that is being used for data, such as displacement maps, has a forced GC value of 1. You cannot let those images track the rendering gamma, as they are not visual - they are data. Changing the rendering gamma results in changes to that data, which alters the displacements.

I warned the SM guys this sort of thing would be a problem, and that I'd be explaining this over and over and over.

They really should hang a HUGE RED SIGN on every forum, a sticky, that shows how to use GC.

Yes, the default is too high and will wash out most of the detail from your rendering. 1.20 or lower values is more like it.

Seems like 1.0 is the neutral setting for  gamma.