Show Us Your Bryce Renders! Part 3

mermaid010 · January 2013

David Brinnen said:
Just to let you know, I've not been idling my time away (not entirely), I've been working on more video tutorials. And I thought I might share with you an little discovery I made during the course of a recent tutorial.
On the same test scene.

Soft shadows rendered with Super Fine art AA - 16 Rays Per Pixel, 24 minutes
Soft shadows rendered with Premium effects - 16 Rays Per Pixel, 14 seconds

Thanks for the tip and looking forward to your new videos :-)

@electro-elvis - I like your little grove, the ground is not bad.

@ eireann- thanks for sharing your experiment.

GussNemo · January 2013

@electro: Thank you very much. That grove really look nice, better than I sometimes see in real life. The image itself is really great, very well done. Should I add, as usual? :lol:

I took dwsel's suggestions and remapped the pillars of the temple I made. While I was at it, I decided to have another look at the mapping for the rest of the objects. Remapping the pillars made them look much better, even the dome and the dais. But when it came to the dome support, that was a different story.

I ran through all the mapping modes, checked material setting with the original material, played with the DTE, Mat editor, but still got black splotches on the underside of the dome support. NOT on the top side of the dome support. The images posted are, and maybe not in this order, another closeup image of the temple with everything remapped; an image of the top and bottom of the temple dome support from the temple file before it was placed into the canyon; the other two are again top and the underside of the dome support from the temple closeup and with sizing set only in one direction. I did this to see what colors I was dealing with. With these two images, you can see quite a bit of difference between top and underside.

If anyone has any idea why the splotches are happening I'd appreciate knowing.

Rashad Carter · January 2013

David Brinnen said:
Just to let you know, I've not been idling my time away (not entirely), I've been working on more video tutorials. And I thought I might share with you an little discovery I made during the course of a recent tutorial.
On the same test scene.

Soft shadows rendered with Super Fine art AA - 16 Rays Per Pixel, 24 minutes
Soft shadows rendered with Premium effects - 16 Rays Per Pixel, 14 seconds

That's quite a big difference. So, on the strength of this observation, I would strongly recommend that if you are thinking of using Super Fine art AA - don't - use Premium effects instead and save yourself a lot of render time.

I redid the test twice, I was so surprised at the difference in render times.

Worth knowing?

In this case, Premium effects was over 100x faster than Super Fine art AA, for the same RPP settings!

Funny, I came to this realization quite a while back. I think I even mentioned it in one of the Committee threads, about the uselessness of Superfine AA. We simply doent need it, Premium works much better.

Eireann, I think you are correct, Superfine doesn't take rays per pixel into consideration, it has a pre-progammed setting for that. Generally, if the software makes too many choices for you it will lead to longer render times. Len was one of the first to mention that it really is essential to give the user control over as many parameters as possible if one is to speed render times. That's why we have so many new AA controls.

eireann.sg · January 2013

GussNemo said:
@electro: Thank you very much. That grove really look nice, better than I sometimes see in real life. The image itself is really great, very well done. Should I add, as usual? :lol:
I took dwsel's suggestions and remapped the pillars of the temple I made. While I was at it, I decided to have another look at the mapping for the rest of the objects. Remapping the pillars made them look much better, even the dome and the dais. But when it came to the dome support, that was a different story.

I ran through all the mapping modes, checked material setting with the original material, played with the DTE, Mat editor, but still got black splotches on the underside of the dome support. NOT on the top side of the dome support. The images posted are, and maybe not in this order, another closeup image of the temple with everything remapped; an image of the top and bottom of the temple dome support from the temple file before it was placed into the canyon; the other two are again top and the underside of the dome support from the temple closeup and with sizing set only in one direction. I did this to see what colors I was dealing with. With these two images, you can see quite a bit of difference between top and underside.

If anyone has any idea why the splotches are happening I'd appreciate knowing.

Are you using bump or displacement?
Try true ambiance.
If you are using bumps then the black splotches might be shadows.

David Brinnen · January 2013

Good points Eireann, Rashad! More investigation needed then...

Speaking of investigation.

Jamie and your splotches. Two things suggest themselves, if the black is not part of the texture - it could be a very thin vein of black that is being sliced out of the texture? Then... if you are using TA there is a funny bug that only afflicts itself on perfectly horizontal surfaces (there is a potentially related mapping issue that also afflicts spheres - but I digress.

For both suggestions the test is the same. Tilt the surface slightly (if the texture is world space mapped) - or if not, then the test varies, to tilting the surface slightly for investigating the potential TA issue, and tilting the texture slightly for testing the potential texture slice issue.

If all that sounds baffling. I apologise - if the file is small, post it to us and I will take a look and give you a definitive answer no problem.

Edit: a third options suggests itself, a bump issue - this is even harder to describe and is related indirectly to the function of reflection correction. To test this, swap the bump mode around and see if things change - so try legacy bump, depreciated bump and enhanced bump in turn and look for differences.

Cheers,

David.

cjreynolds · January 2013

@GussNemo & @mermaid010 - Thanks - always appreciate the encouragement!

Jamahoney said:
As always, thanks Guss and Mermaid...the best thing about abstracts is that they can be quite surprising, they're quick and easy, and you can apply any title to them :)
Cjrenolds....great architectural works. For my eye, there's too much marble of the same colour, but that's not a problem as you can always apply changes in materials to define and enhance areas more.

Jay

I was really diggin' the 'rose marble' back then, LOL! Went a little crazy with it (I can't even find that mat in Bryce anymore)...

I'd like to revisit this image with what I know now, but it looks like the source files are gone forever :( Shame, too - the wireframe was pretty - elegant despite all the booleans :)

GussNemo · January 2013

@eireann: Thank you for your comments. The entire image is rendered with TA, bump, and no displacement. I did think about the bump causing shadows but if I flip the entire support 180 degrees, so the top side is now the underside, there still should be shadows. And none appear. With the top as the underside, and the underside as the top, the pattern appears normal. But undo the flipping and the splotches return.

@David: As I told eireann, I flipped the dome support 180 degrees and the pattern appears normal; top and bottom. Tilting wasn't something I considered until your post so I gave it a try. In either world space or object space, tilting .5-1.00, X or Z, eliminates the splotches, the pattern looks normal. I even tried rotating the material itself but gained nothing. Only tilting the support, on either X or Z axis, eliminated the splotches.

The bump was already set to enhanced, so I tried legacy and depreciated bump. Selecting either made no difference. I tried other settings in the Material options with similar results. I even disabled some settings and saw the same results. The files aren't large, 2 and 3 Mb, so I may go ahead and post them to you if you need to see them in order to find a fix. And thanks for the help.

cjreynolds · January 2013

Here's a little more recent render - from about 3 years ago when I bought ver 6. It was done from a tutorial, but not sure who's...

eireann.sg · January 2013

GussNemo said:
@eireann: Thank you for your comments. The entire image is rendered with TA, bump, and no displacement. I did think about the bump causing shadows but if I flip the entire support 180 degrees, so the top side is now the underside, there still should be shadows. And none appear. With the top as the underside, and the underside as the top, the pattern appears normal. But undo the flipping and the splotches return.

As I see it, light is hitting the ring at a very small angle or close to 0 angle leaving very strong shadows. TA doesnt always shed enough light onto shapes when they dont have direct light falling onto them.
How about you put a small light underneath, no shadowing and very weak. Then you see if it is a material or lighting issue.

GussNemo · January 2013

@cj: I think that's a very interesting image. Nice job.

@eireann: That thought did cross my mind, and when David mentioned tilting the support, I did something even better. I rotated the entire dome support 180 degrees. In essence, I flipped to support so the top was on the bottom and the bottom was on top. When I rendered the image, there were no black splotches on the bottom--which was actually the top. And looking at the top--actually the bottom--there were no splotches either. The trouble should have stayed on the bottom if lighting was the problem. But because they disappeared when the support was flipped, I don't see lighting as the main problem. Why the splotches only occur when the dome support is correctly orientated is the puzzling part. If the color goes through the support, I'd think there'd be splotches top and bottom. Not just the bottom, and only when the bottom is orientated as the bottom.

I tried David's suggestion and tilted the support, first along the X axis, then the Z axis. Tilting the support .5-1.00 degrees, along either axis, eliminated the splotches. The pattern on the bottom looked like the pattern on the rest of the temple objects. This type problem is way above my pay grade.

eireann.sg · January 2013

GussNemo said:

@eireann: That thought did cross my mind, and when David mentioned tilting the support, I did something even better. I rotated the entire dome support 180 degrees. In essence, I flipped to support so the top was on the bottom and the bottom was on top. When I rendered the image, there were no black splotches on the bottom--which was actually the top. And looking at the top--actually the bottom--there were no splotches either. The trouble should have stayed on the bottom if lighting was the problem. But because they disappeared when the support was flipped, I don't see lighting as the main problem. Why the splotches only occur when the dome support is correctly orientated is the puzzling part. If the color goes through the support, I'd think there'd be splotches top and bottom. Not just the bottom, and only when the bottom is orientated as the bottom.

Guess then there must be somewhere a division by zero or infinity to give such kind of "out of spec" results.

David Brinnen · January 2013

Jamie, OK looked at the files and it isn't your fault, it is actually a bug in the way TA renders. It crops up in these tutorials,

Bryce "Nuts and Bolts" - HDRI lighting project - red dragon in a white box - part 1 - a video tutorial by David Brinnen
Bryce "Nuts and Bolts" - HDRI lighting project - red dragon in a white box - part 2 - a video tutorial by David Brinnen

What it boils down to is this: The underside of perfectly level (and also maybe some bias with top sides too - but not as bad) polygons do not get correcty lit with TA. In real life, this would not be much of an issue, because nothing much is perfectly level (or even just plain level in Lincolnshire) but in render land, it is more or less the default state. So as I did in my tutorials, the answer is to just jog things slightly out of kilter and the problem will go away.

cjreynolds · January 2013

Thanks again, Guss! Really diggin the temple renders :)

Here's another ver. 6 render - done from a tutorial, I think one of Robin Wood's (I might be mistaken)...

Does anyone know how to get rid of that red moire-like artifact to the left of the God beams? Or would I even have that in ver. 7? I remember beating that render to death trying to get rid of that and still get the full effect of the volumetric light :(

GussNemo · January 2013

@David: Thanks David, but that problem begs another question. In my case, shouldn't the problem exist even if the support were rotated 180 degrees around either the X or Z axis? With the top surface now acting as the underside, the problem still should exist, right? And since the underside surface is now acting like the top surface light hits it correctly and the problem is gone--that I can understand. But I tested the support rotated 180 degrees around either the X or Z axis and the bottom pattern looks correct. I'll have to post a render of the dome support flipped 180 degrees around the X axis.

@cj: Love the look of that image. Would decreasing the light intensity eliminate the dust look particles of the red?

David Brinnen · January 2013

Well I see your point, but I don't know how rotation is handled in Bryce, any slight error introduced in the rotation and the bug will vanish because it is out of kilter. Instead test with something with less variables, use flip y. This will give you an exact 180 rotation (and mirror the model - but don't worry about that, we are just concerned with the replication of the upper surface as a perfectly level lower surface).

I haven't tested this yet - I'm rendering four scenes right now. So you may beat me to the results.

dwsel · January 2013

More interesting glass materials:

Chohole · January 2013

I did start playing around with some glass, but then I got distracted, cos I came across a set I made ages ago, to work in Poser, and decided to see what they would look like in Bryce.

Jamahoney · January 2013

cjrenolds...Re getting rid of the moire, I wonder if changing the wall material, or lowering its bump qualities (if any) might help?

Moire patterns usually results from misalignment of several lines or curves against other lines/curves that are at a different angle/orientation to the first. In your case, it may be that the topmost red rays (let's say that the rays are made up of hundreds of individual 'light lines') are misaligned against the wall lines where there orientation are slightly out to those of the rays. Else, if that doesn't work, would finer aliasing get rid of them for you. Personally, I like the moire effect in your render, but can see that if they were missing, too, the light rays would look more natural.

Dwsel...very nice glasses...all crystally ;)

Jay

Edit: Oooh..chohole...those are really nice...love the rightmost one.

Chohole · January 2013

Thanks Jay. Trouble is, cos the bottle is just one single mesh, the jpg I used is mapped for it, and I am not sure if it will work on anything else.

David Brinnen · January 2013

Both Dwsel's and Pam's glass look very good, but for different reasons. Dwsel's has her grainy look while Pam has her render set perpetually to "pastel-o-vision" - I'm sure she's found a hidden check box. Those groves in Dwsel's look convincing, but I suspect they are just bump.

Anyway, experiments continue into the obscure mechanics of the render engine. Today I found the point at which reflection switches off. It doesn't just fade out, as I expected. It just fails. The floor in this render is reflective 0.3 but by using metallicity and fully black diffuse I can grade that down towards the camera. Turns out 0.11 is the least reflective thing that reflects. 0.1 and below does not reflect. This would not normally be seen, but it can be exposed by using very intense HDRI light.

Chohole · January 2013

David Brinnen said:
Both Dwsel's and Pam's glass look very good, but for different reasons. Dwsel's has her grainy look while Pam has her render set perpetually to "pastel-o-vision" - I'm sure she's found a hidden check box. Those groves in Dwsel's look convincing, but I suspect they are just bump.
Anyway, experiments continue into the obscure mechanics of the render engine. Today I found the point at which reflection switches off. It doesn't just fade out, as I expected. It just fails. The floor in this render is reflective 0.3 but by using metallicity and fully black diffuse I can grade that down towards the camera. Turns out 0.11 is the least reflective thing that reflects. 0.1 and below does not reflect. This would not normally be seen, but it can be exposed by using very intense HDRI light.

LOL David I have been known to really push for fairytale Pastel. This one was done back in 2005

Rareth · January 2013

chohole said:
David Brinnen said:
Both Dwsel's and Pam's glass look very good, but for different reasons. Dwsel's has her grainy look while Pam has her render set perpetually to "pastel-o-vision" - I'm sure she's found a hidden check box. Those groves in Dwsel's look convincing, but I suspect they are just bump.
Anyway, experiments continue into the obscure mechanics of the render engine. Today I found the point at which reflection switches off. It doesn't just fade out, as I expected. It just fails. The floor in this render is reflective 0.3 but by using metallicity and fully black diffuse I can grade that down towards the camera. Turns out 0.11 is the least reflective thing that reflects. 0.1 and below does not reflect. This would not normally be seen, but it can be exposed by using very intense HDRI light.

LOL David I have been known to really push for fairytale Pastel. This one was done back in 2005

Bryce Fairytale Barbie?

Chohole · January 2013

Yeah well in 2005 my youngest grandaughter was just 5 and very much into fairies and Barbie dolls. Of course she also liked Spongebob Squarepants and Bob the Builder, but I wasn't inspired to try rendering them for her.

cjreynolds · January 2013

GussNemo said:
@cj: Love the look of that image. Would decreasing the light intensity eliminate the dust look particles of the red?
Thanks! Yes, I found that decreasing the intensity helped, but then the light rays were much less pronounced..

cjrenolds…Re getting rid of the moire, I wonder if changing the wall material, or lowering its bump qualities (if any) might help?

Moire patterns usually results from misalignment of several lines or curves against other lines/curves that are at a different angle/orientation to the first. In your case, it may be that the topmost red rays (let’s say that the rays are made up of hundreds of individual ‘light lines’) are misaligned against the wall lines where there orientation are slightly out to those of the rays. Else, if that doesn’t work, would finer aliasing get rid of them for you. Personally, I like the moire effect in your render, but can see that if they were missing, too, the light rays would look more natural.

Hadn't thought of that - I'll have to re-visit this image and experiment.. Thanks!

GussNemo · January 2013

@dwsel: Those looks really good, just wondering if they're battery operated or work on electricity?

@Pam: Really like the glass. And the other image is really pretty.

@David: Darn, should have thought about using that setting instead of trusting the Bryce counter in the lower left of the screen. But I'll give it a try and see if it differs from manually rotating 180 degrees.

Miss B · January 2013

chohole said:
I did start playing around with some glass, but then I got distracted, cos I came across a set I made ages ago, to work in Poser, and decided to see what they would look like in Bryce.

Those are gorgeous Pam, and I really like your fairy render as well. I like the gradient look of her dress.

GussNemo · January 2013

Have been playing around with David's suggestions and came up with the following images.

The first image, I hope, shows the dome support rotated manually, using the icon on the Edit pallet. The next image shows the dome support rotated using the Flip command, as David suggested. In the image I manually rotated 180 degrees, the splotches are absent. But using the Flip command to flip the support 180 degrees the splotches are present. Dare I draw the conclusion that perhaps using the numbers displayed at the bottom left of the screen, as a guide for the number of degrees rotated, aren't as accurate as using the Flip command?

The third image shows the underside as the top surface after rotation. I show this image because I noticed black specks which I don't see in the first image; which is a view of the top surface after rotation--it is now the underside in that image.

The last image is the entire scene with the dome support rotated 0.10 degrees. As you can see, this slight rotation has eliminated the black splotches. I'd like the dome support perpendicular to the ground plane, but I can live with this slight rotation if it eliminates the splotches.

Kerya · January 2013

chohole said:
I did start playing around with some glass, but then I got distracted, cos I came across a set I made ages ago, to work in Poser, and decided to see what they would look like in Bryce.

Lovely - just lovely!