*** Paul's Playground ***, WIPs and WotNot Options

[martin]

Paul, try the Hair "jitter" feature in Multipass.

[Paul Forwood]

Jitter is already at 30%. I'll try turning it up.

[Paul Forwood]

I have done a number of test renders with different jitter settings which I will compile for comparison.
Still fiddling with the hair and not quite getting what I am looking for yet. I will be adding a map to drive the hair colour. Haven't painted the new textures yet either but have tweaked the model a little:

 PictureView_A00_pass3_jit30_noskinshader_e.mov ( 4.94mb ) Number of downloads: 59

[robcat2075]

the hair just seems way too bright for gray hair.

[Shelton]

Paul

Very nice model. Great Job!

Steve

[Paul Forwood]

the hair just seems way too bright for gray hair.

Agreed. I am trying everything but I'm still not getting soft grey and white hair.

Very nice model. Great Job!

Thanks, Shelton.


--------------
Edit: (Removed this movie to conserve a little space)

[robcat2075]

perhaps the muhair shader is just too shiny for old-person gray hair.

[Fuchur]

perhaps the muhair shader is just too shiny for old-person gray hair.

(You might want to play with the "transparency" over length-property...

The new A:M avatare for 15 has something like that applied to its fury cloth.
It looks blurred and very soft. Dont know if that can help you there.)

Sorry, didnt read all posts before...

*Fuchur*

PS: Nice character... like him much!

[Paul Forwood]

I couldn't get transparency along length to work in A:M 15e, Fuchur. Thanks for the thumbs up.

I think that I might be getting somewhere with Muh Hair. Slow but sur.....not too sure. I found Kevin's, (Cosmonaught), description of the Muh Hair properties and realised that I had the secondary highlight wiping out the primary highlight. After some experimenting with MuHair I hope to get closer to finding a flexible workflow for hair. Sometimes I would just like to capture a slightly enhanced version of hair as it appears in the realtime view though.

This is muHair after tweaking the speculars a bit. Also reduced the amount of taper in the hair so it is now thinner overall. Increased the number of passes to 5. Somehow I think that really soft fluffy hair with a good range of shades is going to require lots of passes.
 FaceTest_A00p.mov ( 6.1mb ) Number of downloads: 49


Oh. Jitter doesn't appear to be working or maybe it requires more passes before it becomes apparent.

[largento]

The hair looks pretty good, Paul... and the model looks GREAT!

Now you can have Clint tell the Academy what he thinks about them snubbing Gran Torino. :-)

[Kelley]

You might not be getting exactly what you want, but what you've got is really good.

[Paul Forwood]

Thanks, guys.

This is just a little spriticle effect that I put together as an example in another thread:

 SpriteTest_A00l.mov ( 3.85mb ) Number of downloads: 73

[Paul Forwood]

A little update on my particle experiments:

(Removed these two examples and posted a slightly better version in my next post)

[steve392]

Oh thats good , your gonna have to show us how to do that Paul ,its magic

[johnl3d]

Nice, Paul - between your and Vern's postings, it looks like I can retire.

[agep]

Very cool effects Paul

[robcat2075]

Very impressive result!

I see a movie in this... "Attack of the 50 Foot Flower Man"

[Paul Forwood]

Steve, you know what the Magic Circle would say about that!

Johnl3d, you can never retire! We all need you for inspiration. Keep tinkering!

Thanks, Stian!

Robert, I may throw together a 50 ft flower man, just for you, but this effect has inspired me to try something else. Oh dear. Yet another project...

I've taken those other two movies down and replaced them with this slightly better version, which includes sound:

 Kick_comp_A00a_12fps_7.mov ( 3.27mb ) Number of downloads: 83

[steve392]

Ha ha ha thats graet ,sound makes a differance I must say

[Paul Forwood]

Just more particle testing:

[steve392]

This one looks good ,but looks like she needs some yorkshire pudding down her,.Is that from a normal charector or a skelleton Paul ?

[martin]

That's a great experiment, Paul! I'm thinking "story" as I watch it... (My measure of an animation's appeal.)

[robcat2075]

That would be a great character in the land of Oz.

[Rodney]

That is outstanding Paul.
I'd call that a very successful test.
I like where you are going with this.

File the following in its proper place as I'm sure you already know it:
That last walk seems to be more of a heavy set manly walk than that of a woman.
That aspect may work against the lightness of the particles effect and character based interests in some instances (not necessarily in this test).

As its rather hard to tell the gender of the character at this point in testing thats not going to be a problem but it may be more of a factor as you settle into specific characters. Would the acting and personality have to be a bit more stereotypical? Perhaps. The more diverse you want to make these characters the more importnant that aspect.

Apologies where I'm just stating the obvious.

[TheSpleen]

impressive
A wood nymph perhaps?
Guardian of the forest?

[Paul Forwood]

Is that from a normal charector or a skelleton Paul ?

Not sure what you mean, Steve. This is one of my faerie models. The skeleton will be formed by the gnarled and twisted trunk and braches. It will be quite surreal.

That's a great experiment, Paul!

Thanks!

I'm thinking "story" as I watch it... (My measure of an animation's appeal.)

These are just tests, Martin. The story is just outlined at the moment, (just some notes and 16 thumbnail sketches so far), and will not be revealed until I have an animatic at the very least. I'm imagining it as a 20 - 25 minute short but it could end up being much shorter. With the way I jump from project to project it may be a while in coming.

That would be a great character in the land of Oz.

Indeed, but not this time.

That last walk seems to be more of a heavy set manly walk than that of a woman.

Agreed. The walk can't be finalised until the character's size and weight have been fully established. At the moment she can go either direction. Eventually each character will have motion suited to their physical makeup and their personality. (That's the plan anyway).

A wood nymph perhaps?

Yes. Sort of. She is a supporting actress. One of several who will enhance some of the scenes but at the moment she is primarily a test model for a number of effects.

[Paul Forwood]

I tried a render with a volumetric light and found that it doesn't play well with this model. One for A:M Reports I guess.

Particles, Volumetric light and Fog:

Particles and Volumetric light:

(movie removed to conserve space)

The back faces of the transparent body seem to be reacting with the volumetric light, showing up as visible geometry and acting like a filter on the scene immediately behind.

[Paul Forwood]

Getting back to Clint for a minute, here is a test in A:M 14c where I tried the skin shader for the first time:
 Shortlegs_A00B.mov ( 2.18mb ) Number of downloads: 96

[steve392]

I think that looks good ,the light gives a nice shadow and the geometry looks good like that ,sorta differant but wierd.The skinn shader is very good ,I have to try it out ,you can still see all the bump (face lines) so thats working better than I would have thought .This guy is looking real good now

[KenH]

That old guy has a really distinctive face. I like it.
How are the flowers falling off the Plant character done? Hand animated or automated?

[Paul Forwood]

Thanks, Steve and Ken.

How are the flowers falling off the Plant character done? Hand animated or automated?

Particle emitters with very low velocity and emission rates. Sprite's gravitational effect set low, to suit.

[Paul Forwood]

I have no idea what is happening here, in A:M15e:

Somehow I managed to get a volumetric light to work with a transparent surface but I get these dark streaks appearing for no apparent reason. There is nothing that could be casting a shadow above the balls.

I have tried everything that I can think of to get volumetrics to work with transparent objects in this other project but without success. I'm still getting the white-out. I'm wondering if this might be caused by groups, higher up the tree, which share the same patches as the group that has been set to 100% transparent. I doesn't make sense but I will try stripping out all the groups except for the Transparency group
and see if that has any effect.

[KenH]

I wonder what flipping the patches would do?

[Paul Forwood]

I wonder what flipping the patches would do?

I stripped out all groups except for a bark colour group and the skin/particle emitter. Same story.
I then tried your suggestion, Ken, and flipped the normals:

No change and, of course, the particles are forced into the body shape, which is not really what I want.
I guess I'll have to forget about creating any moody lighting or leave it until sometime in the future.

[KenH]

If you strip down the model to just an arm or something and post the project file here, perhaps someone might be able to tweak something. I'd be interested in having a look anyway.

[Rodney]

If you strip down the model to just an arm or something and post the project file here, perhaps someone might be able to tweak something. I'd be interested in having a look anyway.

I'd be very interested as well.

If there is no other solution the thought that comes to mind would be to test a completely transparent patch image applied to all of the body mesh. That should vanish the mesh for you. I'm not sure what elements are there that might confict with that method however. Having a small test project would tell us.

[Paul Forwood]

There are still a few glitches present in volumetrics:

(movie removed to conserve space)
Note the black triangle where the volumetric lights cross and the way that it affects the particle hair that is infront of it. There is also some very unnatural distortion of the ground plane texture occuring but that doesn't seem to be related to the volumetrics. (Visible in the other images also.)

[Paul Forwood]

Aha! I suddenly realised that the only place that the black triangle occured was where the lights crossed one another, infront of an empty background, (Alpha). I added a skydome and the problem went away. Transparency is working well too. Now if I can just find out what is causing the black streaks I will be a happy bunny. Maybe volumetrics need many passes. The following example uses just 3:
 BallTest_B00__NoFog_dome.mov ( 1.29mb ) Number of downloads: 37

[steve392]

Thats looking a lot better Paul and I don't see any streeks anywhere

[Paul Forwood]

Thanks, guys!

After a whole load of tests I finally managed to get this transparent model to work with a volumetric light, apart from the streaks.

I gradually deleted all groups and eventually all decals and materials too. After that I could create new groups and add materials and hopefully everything is working now, as far as I can tell. I'll see what happens when I add the other groups, materials, lights and fog.

So the rule of thumb for A:M15e users, at the moment, is: When importing models from earlier versions of A:M be prepared to start afresh with your mapping.

[Paul Forwood]

Volumetric lights are working well with my transparent geometry now but there are still a few things that I need to understand, such as...

(movie removed to conserve space)
... those streaks! There must be some way of controlling the streaks. Some simple volumetric tests should shed some light.
Do volumetric lights actually have some kind of mass? Those particles look as if they have become stuck on the light beam. Strange!

[animas3D]

When you mean "streaks" are you talking about that black stuff in the middle of the beam such as in the last image, or are you talking about the darker areas of the volumetrics that emanate from the character after being hit by the light?

If you are talking about the latter, I think those "streaks" are natural shadows caused by the model within the volumetric rays.

[Paul Forwood]

Well of course I'm talking about the dark, grainy streaks that run down the whole length of the light beam/cone. Those other soft streaks of light and shade is what I am after.

[animas3D]

Right. That is weird. Is it a bug?

[Paul Forwood]

Is it a bug?

I'm not sure yet but I vaguely recall a post on the subject from way back so I'll see if I can find that.

[KenH]

The black streaks look suspiciously like they are occurring in the hotspot of the light. Do they still appear if you widen/shrink that middle part of the light? Or do you have shadow softness on?

PS He's got a tad too many flowers falling off him now. I think they should be subtle.....something to look out for.

[Paul Forwood]

The black streaks look suspiciously like they are occurring in the hotspot of the light. Do they still appear if you widen/shrink that middle part of the light? Or do you have shadow softness on?

That is what I am testing at the moment, Ken, in a new project, but I am getting perfect cones with none of those black streaks:
 LightTest_A00b.mov ( 89.98k ) Number of downloads: 41

There are three lights in a closed box and none of them are displaying the problem so this is a good place to start adding things in until something breaks.

He's got a tad too many flowers falling off him now.

Yes. I wanted plenty of particles visible for this test just so I had plenty of examples of how the flowers/petals would behave all along the path. Just as well or I might not have known that particles can become suspended on the surface of a volumetric light cone. These tests may not look very pretty but they answer alot of questions for me so that I can better judge what will be possible, or not.

[Paul Forwood]

I have been testing A:M 15e's volumetrics pretty intensively today and I haven't been able to reproduce those destructive streaks to any significant degree, just a few black specks in the core of the light beam. . I have yet to introduce a model from A:M 14 so maybe that will turn out to be the trigger.

The black areas, that appear when two volumetric lights overlap, always occur when there is no background behind them so it must just be the alpha mix from the two lights. I will do a proper render, with alpha, and try compositing the results with a background.

[Paul Forwood]

An example of the testing that I have been doing:

 Aligned15deg_Box_Hair_Sprites_LR_angle_gobos_.mov ( 1.22mb ) Number of downloads: 35

This includes most of the ingredients that I thought might have been causing the problem.
The troublesome black particles are less obvious but they are still there if you look closely.

[jakerupert]

I vaguely recall, that this has something to do with a percentage setting somewhere at the shadows casting of the lights.
You will have to increase that.
Sorry, can`t look into that closer at the moment, since I donT have AM with me right now.

But maybe somebody else can tell you it`s exact location.