m:studio 2.4d in use on SA Sports Hall of Fame

#301

It appears that your ribbon has expanded slightly with the displacement and may contribute to noise if the ridges are overlapping slightly one on another (swollen). To maintain the size of the object but still have the same look, use the MinMax filter inbetween the Texture Map and the main material node. Use -.5 and .5 :slight_smile: Also increase the translucency to lighten the noise - and since you’re probably going to be using Unilimit, the ribbon won’t appear so blown out.

#302

Thanks for the tips Gary :thumbsup: I knew you knew more than you were letting on :slight_smile:

here’s your suggestion with Unilimit at 20 (no bump displace)

here’s the same with bump displace ON and subdivision = 30 which increases render time but drops the noise since the polys are much smaller (that’s my understanding anyway - please correct me if I’m mistaken). It may be more evident on closeups which I’ll test later.

#303

I’m using MinMax, but not centred around zero - good tip. I’ll also play with translucency.

Thanks Gary :thumbsup:

#304

Hi Paul, I´m following your work all the time, I´ts a great resource!!, and … can I do a petition? :rolleyes:

Can you post the 2 images between?, I can´t see the differences of your renders with and without SSS effect :shrug:

Thanks!

#305

Since you’re using Unilimit, take that bad boy back up to .5 translucency! :bounce:

#306

(A) no SSS

(B) subD Level 15, translucency 0.5, UniLimit 20
Also bump map now -0.1 to 0.1 instead of 0.0 to 0.1.

This is too much translucency IMO with too much fill in the bumps and a waxy look developing. But it most definitely takes care of bump noise.

© translucency at 0.35

Taking your votes now…A, B or C? or should I tweak some more for an X, Y and Z? :slight_smile:

#307

I like B … but the waxy look can be moderated by either

  1. Decreasing Surf Thickness
  2. Increasing Falloff

It’s more predictable to play with Falloff. Try increasing it to 40. This will tighten up the translucent transition area that you see along the edge … thereby making the ribbon appear more dense and less waxy.

#308

Great images Paul, each one have this particular beauty but I vote for the C (Z) :applause:
I like to much shadow contrast of the first of this, but the SSS effect of the third makes the composition more realistic.

Thanks!!

#309

(D) translucency 0.5, falloff 40

(E) translucency 0.5, falloff 80

Note how the shadow density is increasing.

#310

Very nice! What I like about SSS is that it’s definitely softening the edges making it appear more rounded and it let’s light through because I’m sure it’s made of some type of nylon polymer.

Side note: if you were to render this with a Point or Distant Light Source you’ll see something special in the shadow. :slight_smile:

#311

This was also a concern I had since I started the SSS renders. Without SSS the ribbon is more grounded with it’s dark shadow. The later renders with higher falloff do show more dense shadows but still not as much.

Perhaps two renders comped to have the best of both worlds?

Thanks for your crits :thumbsup:

#312

Also consider reducing the Surf Thickness. This will make the shadow more evident. What is happening is that the Surf Thickness is actually creating a “subshadow”, an additional shadow that is, for this example, at 50% translucency. You can see this with a Point light source - appears as double shadows. When you decrease the Surf Thickness, this portion of the shadow is reduced. The sphere and panel lights blend the main shadows and subshadow.

#313

no SSS

settings as before, falloff = 80

falloff = 2 (typo, wanted to type 20)

falloff = 20

#314

(no SSS)
Translucency 0.5
UniLimit 20
Falloff 40
Surf Thickness 0.0002

More dense shadows and less light penetrating from the edge = less nylon polymer :slight_smile:

It still gives a nice soft fabric feel. I’m quite pleased with it. Playing with SSS can be endless . . . and so,many different settings look good, each with its own merits . . . hard to decide.

#315

Yes it is rather addictive - I’ve spent so much time doing test renders after test renders. Whatever you decide on, I think it’s a nice subtle touch that will lend an air of realism when everything is said and done.

Side note: Surf Thickness of .0002 is very tiny - 2 micrometers. Consider that .002 is 2 millimeters.

Here’s a nice conversion utility page for the mathematically challenged.

#316

Thanks for your input Gary. It is indeed much appreciated. I agree with the benefit of SSS in this scene. I’ve learned a lot and still far to go…

Yes, I’m aware of the thickness. I just took it smaller and smaller as I did some tests. I’ll still decide on a final setting. Perhaps animate through them all then I’ve got them all . . . just kidding. Deciding can be hard with such subtle changes. I’m most impressed with the speed at which messiah renders SSS. I’m so used to doing the basics because anything more makes a scene unrenderable for animation. But this is a pleasure. Another setting which added almost no render time (at least on this scene) was ray depth. I went from 2 to 4 with negligible delay.

So I’m enjoying messiah and since I’ve worked through all aspects of this scene without a show stopper, I can announce that it is a success. Yes, there were delays, near misses, frustrations, etc. Each project has that. Working with any package has that, perhaps in a lesser degree compared to where m:studio is now, but that is changing with each release.

I can just again say that you guys on this forum have made it worth it :thumbsup: I’ll still do a proper credit roll when I’m done :slight_smile:

The project is soon due for a 2nd client preview. Then final render. So I’m not done yet. My Digital Fusion comps are set up. I’m happy that everything is working out.

#317

After tweaking my coordinate system to get rid of as much Gimbal madness as possible the remaining Gimbal lock was “solved” by:
[ol]
[li]baking the object’s rotation channels that was being affected by a Target expression[/li][li]deactivating the Target expression[/li][li]manually editing the gimbal flipping problem directly on the baked keyframes in the motion graph[/li][/ol]This whole Gimbal problem really sucked a lot of the pleasure out of setting up the animation. It was a struggle because the animation was required to move through a range of roughly 160 degrees from one extreme to the other and no matter how I adjusted the initial coord system, the animation crossed into illegal range somewhere. I just had to choose where I’d better be able to do damage control as outlined above.

Thanks again to Thomas for feeding me Euler and Gimbal theory and for all you guys listening…it’s very therapeutic for me to post and consolidate. Plus, if I don’t post, this knowledge will be lost 2 weeks from now as I move on again.

#318

Does messiah’s render have a pre-roll for dynamics computation before rendering? For instance if I want to render from frame 480 but would like my cloth to “settle” for 20 frames before that without explicitly rendering from frame 460.

Or should I rather “bake” my cloth sim using SaveMorphSequence and then import that sequence prior to render?

Any other advice?

#319

Baking.
Everything Dynamic.
Always.

I do it with PO, but that internal stuff should work too.

Preroll? LOL ROFL gasp…

Cheers!

#320

Thanks I’ll do that.

Hey, isn’t messiah a non-linear 3D animation package? Pre-roll is part of this business, even if heavily used on the old analodue video editing side… OK, OK, I see your point. :scream::slight_smile:

If messiah’s dynamics was rock-solid and baking was not needed except perhaps for network rendering, then a pre-roll would have been a good idea, no? Anyway, we’re talking theory here. I’ll bake.