Tottes hideout

Mornin Everyone,

Yeah, a cool filter Totte. I will have to check it out later, had to force quit my computer as it locked up tight during processing, I think it must need some serious adjusting.

Have a GREAT day!

Ron

Quote
Totte wrote: I know it is my five chained density 3-5 bombers that causes the slowness...

yup, that might do it

nice fur!

Good , I'll see what I can do with beta 3
All previews renders an about 60 seconds on my 16 core

Yes, VERY nice fur Totte, though one source of the slowness might be that you're using about twice as many bombers as you probably need. One of the Particles you use is *always* transparent. Look at the 'Hair shadows' control, follow the connection upwards from it to the remapper. You'll find it remaps to 1 or 1 , which means transparent .
Please don't think I'm complaining, this is a great filter, but I'd love for the render time to come down . 2 hours and 50 on my machine . Sheeesh. But worth it.
Another Idea that may let you use less bombers and thus speed up rendering would be to have more than one hair in each particle. As many particles as you are using, likely noone will notice lots of them come in patterns .
Miavir

OK, I'm getting closer and still preserving acceptable render speed

I got my Sheaf looking good now;-)

Now its getting closer...

I've stumbled on this interesting problem where I think I can use one of those new math functions, just have to figure out the correct way, or maybe someone of you geniuses here already have solved that problem. It is this sheaf filter that is spoking me.

The situation:
One slider "waist control" is used to control the waist of the sheaf.
The slider actually control two things, the waist (a scale using voldermorts "pinch setup") and the width of the rope overlay (a gamma -> threshold of a curve).
Using remapping, I've gotten it to look good at min and max values, but the tricky part is that the two items controlled dont have the same linear curve, so at "50%", the rope is smaller that the waist.

// Totte

Craig - the problem is that the rope is "trimmed" using a gradient where I control the Treshold to the trimmer, and the sheaf itself is trimmed using a Scale, and they do not have the same linearity, I tweaked it "good enough, but it is an interesting problem. I can make a small sample filter to show you the problem.

Tiger - that what I was doing
This is the lasest version I've experimented with.
Check it and you see my problems
The slider is Waist.

sheaf.ffxml

You sort of were, Here's alternate solution that fixes your issue, and I made some changes so it keeps the roundness of the cutoff constant. There's still some odd horizontal offsetting issues (leftover form the original), which could easily solved by controlling it with another tone curve on "Width"

sheaf001.ffxml

Been playing with what I seems to do best, wood

The sideways move isnt done yet I shift them vertically right now, will also shift them Horizontally before the vertical shif, to break that obvious repeat.

very cool!

Yuck, I think I call this one "infected alien tissue" or something.

now you didnt have to do that. we liked your jetty. no reason to poop on it

nice infected whatever 'heart'(?).

hehehe

Been playing with some crates with iron rims.

Thank you It's in the library

This is rendered on my brand new Core i7 MacBookPro with FF 2.005, same filter, default settings
They really have different shades and metallic look, I really wonder why? Is there a setting I can have fiddled with on my MacPro (gamma is set to the same, I've checked), but some other "global setting"?

After some more testing (where I got the SAME result in other filters), I quit FilterForge and started it again on the MacPro, then it looked as on the Core-i7. I know I've had this before, where I've been tweaking reflections and metallic inputs and got a great effect, and the next day when I start FF again it looks different, so I guess there is some kind of bug in FF app where something sticks, overloading the metallic/reflection to stay in some kind of weird overload state until quit, and that is inside the "loaded filter", not in the saved to disk filter XML.

So, solved, but still a mystery

cool

- the bolt heads need a bit of texture or a bit darker