200K limit, one per reply. But you can use PNG. However, they are usually bigger then JPG.

If you store the files off somewhere else, you can embed as many pics as you like in one post.

Ok, I'll have to do a couple sets....

Whichway

I'm heading in the same direction - tiny Irradiance sample size. Refining, I ended up with 17 minutes and just about the same image as Paul posted using max quality settings.

We're "hill climbing" a multi-dimensional space. Each local maximum may end up producing the same peak, so it could be a dead end. The thing is, there are so many local peaks, we may have to search dozens of them before drawing any conclusions.

Yeah note that garbage on the backdrop.  You need to disable the Smooth Polygons property for that prop and re-render to get rid of that.

Quote - I'm heading in the same direction - tiny Irradiance sample size. Refining, I ended up with 17 minutes and just about the same image as Paul posted using max quality settings.

I had thought that only applied to full-scene AO.  I'll turn it down on my side and see what goes on.

I have pretty much every thing cranked down as far as it will go. Otherwise, the only difference between the two sets is that the first has 0 raytrace bounces and the second has 1 raytrace bounce. Only 1 IDL bounce in both cases. Note that tis violates pjz99's conditions, but gets light everywhere! Also, not apparent here, is that the red ants seriously invaded the second IDL. The first set had maybe six dots total, while the second was very high density (but not solid).

Interesting.

Whichway

There's a trap doing IDL-only renders. Some of the splotches only show up on the rendering pass. The IDL pass may look OK, but then the render doesn't. Do full renders.

Zero raytrace bounces is unacceptable in any circumstances with indirect lighting, as it prevents all indirect lighting (which is why your GI-only pass is 100% black).

So why does 1 bounce of each kind light up everything?

[Gotta feed the cat. I may be able to skip dinner, but he won't skip his.]

Whichway

Quote - I'm heading in the same direction - tiny Irradiance sample size. Refining, I ended up with 17 minutes and just about the same image as Paul posted using max quality settings.

This has a huge impact (positive impact).  Quite surprising.  Smith Micro really should have added an "Advanced Render Settings" interface for managing this stuff.  At least the Dimension3D script is there though.  It renders slower with this setting at 0.1 but "slower renders" is preferable to "unusable renders".

Quote - So why does 1 bounce of each kind light up everything?

One bounce is literally one bounce of light, which does apply some indirect lighting, but if you look at the pics I gave earlier you'll see that the underside of the smaller square is completely black.  This is unrealistic, 1 bounce is unacceptable.  At least 3 for this composition, and preferably more like 7+.

I'm trying inferior, 1-minute renders. None are good, but by studying the differences, I'm learning more quickly what the parameters do. I wasted a lot of time on 20 minute renders, while not learning anything at all.

I've learned a few things, mostly that my assumptions were wrong. Don't have verifiable assumptions yet. But I know how to get a 1 minute render looking better than before.

I'm going to go eat now. I can't stand not eating.

Ok, got it, I think.

Whichway

Same here, BB. Do any of us actually know what the number in the Irradience Caching slot actually means? I could maybe argue that the higher the number the more that is cached, meaning that fewer accurate values are actually used. If so, higher cahce means faster, not better. Hyothesis available for testing...

Whichway

It's a quality slider that in previous versions, mostly impacted AO smoothness.  Smaller = fewer samples = less accurate shading, at least the way I understand it.  It may be that it doesn't work the same now, Stefan would be the one to know for sure.

Oh course, now there are two such knobs.

Whichway

P.S. - Almost have another set ready.

In the interests of only changing one thing at a time, I've left the bounces the same, but raised the Indirect Samples to 1001. pjz99, how does this look to you re blotches, artifacts, whatever?

Whichway

Quote - Same here, BB. Do any of us actually know what the number in the Irradience Caching slot actually means? I could maybe argue that the higher the number the more that is cached, meaning that fewer accurate values are actually used. If so, higher cahce means faster, not better. Hyothesis available for testing...

To quote the Poser 8 manual:
"Irradiance Caching : Determines the ratio of shading points where indirect light will be calculated by raytracing vs. those that will be interpolated.  Low values result in shorter render times, with more interpolation. 

High values achieve more precise results at the expense of increased render times."

I have been doing renders in Poser 8.  I ran the settings all the way up, the way I used to, with indirect lighting, and the sucker hung.  To use indirect lighting, you need some kind of set around the figure to get bounce from.  Otherwise it hangs.  After I put in a set, the render at highest settings took just about 48 hours.  Not going there.

So I found that if I lowered my settings to final and then imported the setting from auto, I could turn on the indirect lighting and get something decent in an acceptable amount of time.    We're talking three or four hours for a render, which I will take.  I'm used to Vue 7 Infinite taking a day or so to render, so I can live with that. 

Injection poses don't seem to work, which kinda sucks.

As I mentioned elsewhere, WHAT'S WITH THE TINY THUMBNAILS?!  ARGH!!!  I am not 20 anymore!  I can't seeee them!  

Feh.

I will say however, that the inbox people are the UGLIEST human figures I have ever seen in my entire life.  Jeeeez, they make the Poser Dork from P4 look like adonis.  WTF?  Have we just given up on competing with Daz?  Feh.

So far, I'm thinking I like Daz Studio Advanced better by far.  But I'm still playing with it, so we'll see if my mind changes.

Juliekitty

I'm completely uninterested in how anything looks with 1 bounce.  3 at the absolute minimum, preferably more.

That's what I thought it said, and far be it from me to argue with the render god, but it doesn't feel like it's working that way. Hmmm.

Whichway

Sorry, that's not the sample size I changed. I changed the other one. (Whoops, you changed both. I changed only the one labled Samples.)

1 raytrace, 7 IDL bounces rendering now.

Whichway

Quote - Now let's make the quality sliders maximum (actually 99%).  Now instead of a few artifacts of very large size, we have tons of artifacts of much smaller size.  Sort of an improvement, and the overall lighting intensity is all right, but wow, that looks like crap.  20 minutes of render time for this one, vs. 3:25 for the last one.  I expect render time to go up with high quality settings, but this is pretty much unusable.  I might as well go back to faking it with IBL (and until patching, I will).

My eyes aren't that great, and my contacts are a bit foggy today, but I honestly can't see anything wrong with this particular image. If it gets lost in the quote, see here: www.renderosity.com/mod/forumpro/media/folder_9/file_436537.jpg

What are you seeing that I'm missing?

The other sample size? There's only one parameter called sample size.

Some monitors with low dynamic range make this difficult to see. I can barely see it on my laptop, but on my $500 Samsung it is glaringly obvious.

I used Photoshop to increase the contrast in a narrow band of luminance values, so you can see just how the artifacts are shaped. If this was not such a perfectly flat surface, you'd have a hard time picking these out.

There is one called Irradience Sample Size before the Enable Indirect Light box and one called Samples after it.

Whichway

Quote - There is one called Irradience Sample Size before the Enable Indirect Light box and one called Samples after it.

Whichway

Of course there is, but I said "Irradiance sample size". I assumed you modified it as well, because I saw .1 in your settings.

"Samples" is a count, not a size. It is how many rays are used to sample from any given point to populate the irradiance cache.

Quote - If this was not such a perfectly flat surface, you'd have a hard time picking these out.

Oh so it's like that huh.  Okay, try this instead.
http://cid-b233dcaeefa9709c.skydrive.live.com/self.aspx/Public/Poser8%20samples/wibbly.zip

Quote - I did this test a week ago. This is all indirect light, no light source. Placing a light in the room masks the artifacts, but they're always there.

I had made the assumption (foolishly) that this was caused by lighting only through the Ambient channel, which would certainly be bad, but I didn't go on to consider that it might happen with regular lights and shadows.  A bit slow on the uptake I guess.

Your Wibbly looks pretty good to me. One render is certainly no proof, but I didn't run into any problems. This is in your room. 28 seconds on my craptop.

Standard dialog settings, not D3D:

Bounces = 6
IC = 30
ILQ = 0!!!
Pixel Samples = 4
Min shading rate = 1

Note: I had to turn the wibbly around. Were you rendering the backside? We already know that's a problem.