trying out the Arnold renderer

[Andreas Bystrom]

Recently I have been betatesting the arnold renderer for softimage. I
have been testing various scenes in both mental ray and arnold and so
far arnold has proven to be extremely impressive.

The first test I'd like to show is an interior:

this is the arnold version:
http://ericknelson.com/wurp/arnold_scifi_01.png

and this is the mental ray version:
http://ericknelson.com/wurp/mentalray_scifi_01.png

As you can see they look pretty much the same except arnold has a lot
more detail, look at the wall on the left for example, arnold catches
every little detail in the model while the mental ray
version looks pretty washed out. Note that in the mental ray version
I'm using the architectural shader with the ambient occlusion mode
enabled (detail enhancement set to 2), without that it would lack even
more detail.

These two examples rendered in about the same time (27 minutes or so
on a quadcore). The arnold version however would be completely stable
in animation, the mental ray version
would flicker a lot due to the interpolation it uses.

The second test is a simple occlusion render (the scene is around 500k
polys and the grass has some 350k hairs):
http://ericknelson.com/wurp/mr_vs_arnold.jpg

Both renderers uses an occlusion shader with depth of field enabled,
as you can see the arnold version has a bit less noise while still
rendering way faster than the mental ray version.
Enabling either depth of field or motionblur in arnold has a very
small impact on rendering speeds, even combining the two is very fast,
in mental ray however it slows down a LOT.

Another example I'd like to show is rendering hair with gi/fg in both
renderers, here is a version without motionblur:
http://ericknelson.com/wurp/mr_vs_arnold_hair_no_mblur.jpg

The two versions looks quite different and I guess I could match the
look more closely spending a bit more time on it, however the
important thing here is that they both use an identical setup:
Both versions has one arealight with exactly the same radius as well
as gi/fg enabled that also illuminates the scene using an environment
map.

Here is the same scene, but this time I've enabled motionblur (that is
with deformation blur, two motion steps), I've also switched the
mental ray version to use the rasterizer mode which is supposed to be
best way to render hair with motionblur in mental ray:

http://ericknelson.com/wurp/mr_vs_arnold_hair_with_mblur.jpg

The difference in renderspeed is even more obvious, the arnold version
has hardly slowed down at all while the mental ray version is taking
several times as long to render. Again these versions looks quite
different but
they still use an identical lighting setup (one arealight as well as gi/fg).

Here are a few more examples that i've rendered in arnold:

http://ericknelson.com/wurp/mr_vs_arnold_mblur.png (clothsim with
deformation blur and displacement)

http://ericknelson.com/wurp/arnold_light_falloff.jpg (this is with GI,
one arealight, again this renders in just over 5 minutes)

http://ericknelson.com/wurp/tank_glossy_new_0.jpg (this is with GI,
one arealight as well as lots of glossy reflections, took around 10
minutes.)

http://ericknelson.com/wurp/Arnold_classroom_01.png

Sony imageworks has also recently completed their second movie all
rendered in arnold: http://www.youtube.com/watch?v=BPH0ct2oXBg (first
one was monster house:
http://www.apple.com/trailers/sony_pictures/monster_house/)

I hope you enjoy these examples, stay tuned for more.
and by the way, it's not an open beta but I have been given permission
to post these examples on the list...

[David Saber]

Thanks Andreas it's very impressive.
Does Arnold for Softimage work in the render region and is it compatible
with the render tree nodes?
See you,
David

[Sebastian Kowalski]

jeah its looks impressive, so is looking 3delight in side by side
comparsion to mental ray.
at the end ist all about integration into xsi. 3delight studio sucks
(yet), in a studio pipeline. we ended up writing our own exporter for
3delight (renderman compilant render engines).

would like to see some 3delight xsi to arnold xsi comparsions. i could
render some examples in 3delight. do you mind to post some test scenes?

bests

sebastian

[Matt Morris]

I've always thought that mental ray's motion blur sucked balls - in this case, big grassy ones ;)
Not just that it's slower but just doesn't look physically correct either - the arnold one looks far superior. Any ideas when it might be available as a finished product?

2009/7/5 Sebastian Kowalski <li...@sekow.com>

--
www.mattmos.com | www.xsilondon.com

[Rob Wuijster]

Hi Andreas,

That looks really impressive, esp. the render times.

If all beta's will become products, we could look at MR, Arnold, Vray,
3DLight and maybe finalrender for picking a renderer of choice :-)

Rob

\/-------------\/----------------\/

Matt Morris wrote:
> I've always thought that mental ray's motion blur sucked balls - in this
> case, big grassy ones ;)
> Not just that it's slower but just doesn't look physically correct
> either - the arnold one looks far superior. Any ideas when it might be
> available as a finished product?
>
>
>

> 2009/7/5 Sebastian Kowalski <li...@sekow.com <mailto:li...@sekow.com>>

> www.mattmos.com <http://www.mattmos.com> | www.xsilondon.com
> <http://www.xsilondon.com>
>
>
> ------------------------------------------------------------------------
>
>
> No virus found in this incoming message.
> Checked by AVG - www.avg.com
> Version: 8.5.375 / Virus Database: 270.13.3/2217 - Release Date: 07/03/09 18:11:00
>

[Andreas Bystrom]

Glad you like the examples so far.

David, yes, it's fully integrated, it even has support for some ICE things.

Sebastian, comparing with 3delight can be interesting for sure but a
lot of the examples are hard to reproduce, all the interiors are
pretty much impossible to recreate in 3delight.(at least using the
same lighting setup) You could try the grass spheres yourself if you
like (it uses 100k hairs by the way) but last time I checked 3delights
pointbased GI does not support illumination from an envmap (at least
not in xsi) which is what I'm using here, also if you were to use
arealights on hair in 3delight it would grind to a halt, so you'd have
to use deep shadowmaps.
I already tried the lighthouse scene in 3delight, using its pointbased
occlusion with DOF, and from what I remember it took around 10 minutes
(not including baking out the pointclouds), but ofcourse one has to
bear in mind that there is no noise in that render.
http://ericknelson.com/wurp/3delight_pointbased_ao_white_bg.jpg

Another cool thing to mention about arnold is that it's interactive,
the first time you render the scene it can take some time for the
render to start just like with mental ray but after that, no matter
how heavy the scene is, the render will start instantly as long as you
don't add any objects or change the topology on them. Anything to do
with lighting, adding lights, moving them around, tweaking shaders,
moving the camera around etc is all interactive, you can even
hide/unhide objects while still having it interactive...

[Joe Williamsen]

Thanks very much for the samples, Adreas :)

A few questions come to mind: 1) Any idea when it might be available? 2) Any idea what the price rainge is going to be? and 3) Do you foresee any rendering compliance issues?

I'd love to have an alternative to MentalRay.  While (IMO) MentalRay might have been a great renderer 5 years ago, compared to other renderers available in other packages, it's getting pretty long in the tooth.  It really seems like Mental Images is just sitting around doing nothing - and still expecting a premium for render seats, etc.....

[Steven Caron]

"but ofcourse one has to bear in mind that there is no noise in that render."

it also looks like the 3delight one has more effects than the others, ie. water ripples and detail on the ground.

regardless your testing does look extremely impressive!

steven

[Arnar Gunnarsson]

very interesting and impressive

looks like arnold has been under the radar for a while, who is actually developing it now? finding it hard do google.

although it seems to have been used more extensively then I thought and even in development for xsi since 2007, would be interesting to know what time frame they are thinking about releasing it.

been looking into replacing mr for 3delight, there are always pros and cons, I see it more as a companion for now.

.arg

[Gene Crucean]

Interesting. I haven't heard anything about Arnold for a lonnnnng time.

--
Gene Crucean
CG/VFX Sup
** Freelance for hire **

Main email: genecrucean [at] gmail [dot] com

[Luc-Eric Rousseau]

On Sun, Jul 5, 2009 at 7:20 AM, Andreas
Bystrom<andreas...@gmail.com> wrote:
> Recently I have been betatesting the arnold renderer for softimage

>[...]

> Sony imageworks has also recently completed their second movie all
> rendered in arnold: http://www.youtube.com/watch?v=BPH0ct2oXBg (first
> one was monster house:
> http://www.apple.com/trailers/sony_pictures/monster_house/)

Can you clarify what company is making and selling this Arnold
renderer you are beta testing?
Is it Imageworks' renderer, messiah render, or somethign else?

[piotrek marczak]

These are really impresive!

Is it the same Arnold that was beta tested like 10 years ago by Daneil
martinez Lara from pepeland.com ?
and then incorporated in messiah ??

Somebody please tell me why release cycles of all renderers take sooooo long ?

2009/7/5 Andreas Bystrom <andreas...@gmail.com>:

[james.d...@gmail.com]

arnold is a fair whack older than the pepeland stuff,

originally it was written by one guy, marcos something...forget his name now...started around the same time Max got MR 2.1 or about that. i got some things rendered in it before textures were supported around 7 - 8 years ago i think, will have to check but its certainly been quite some time.

he used to frequent #3dsmax irc channel many years ago...I still remember the cornell box renders he would post.

started out life in max...so dont know why it went to messiah commercially first...


james,

[Andreas Bystrom]

here is another example comparing it with 3delight:

http://ericknelson.com/wurp/arnold_vs_3delight_ao.jpg

I had the same scene rendered with dof in Arnold and as you can see
that one actually rendered slightly faster even..

As I said it's a bit hard comparing a reyes renderer like 3delight to
a raytracer like arnold as you would set things up in a different way
(using more shadowmaps etc in 3delight) so an occlusion test will have
to do for now..

I will try to post some animated GI examples with motionblur/dof later on..

[guillaume laforge]

Hi Andreas,

So, when is it released ?

:)

--
Guillaume Laforge
modeling, texturing, compositing, animation, effects, lighting artist and/or TD (not always in this order) at la maison.
http://www.alamaison.fr
my blog : frenchdog.wordpress.com/

[Sebastian Kowalski]

keep'em coming. these are much appreciate tests.
some animation test would be great. if you could spare some time.

[peter boeykens]

Surprising, how clean the renders are in such short rendering times.

The Arnold examples I remember from way back when were poststamp sized,
extremely grainy and supposedly took ages to render.
I'm recalling this dark room, a door opening with light pouring in and some
simple stickman figure walking inside. Nice GI for the time, but nothing
you'd want to build a production around.
If this is the same one (how many Arnold renderers can there be?) they must
have come a long way.

[Andreas Bystrom]

Peter, it is the same renderer yes. I also remember seeing those
pepeland examples back then and even though they were very noisy I
liked the look of them a lot. No doubt has both hardware and the
actual renderer improved a lot since then.

Guillaume, I can't speak about pricing or release dates, not that I
know anyway. I'm still under NDA but as I said I have been given
permission to post some examples here.

[Gene Crucean]

I just want to know WHO is working on it now?



http://www.daniel3d.com/pepeland/gallery/imas/pepeland-com_01War.jpg ;)

[guillaume laforge]

Marcos Fajardo ?

[Andreas Bystrom]

It's co-developed between Imageworks and Solid Angle (http://www.solidangle.com)

[guillaume laforge]

Then, lets hope to learn more after SIGGRAPH as it looks really promising !

But maybe there are looking for other beta testers ?

Cheers

Guillaume

[tak...@earthlink.net]

Yes. 

-----Original Message-----
From: guillaume laforge
Sent: Jul 6, 2009 10:02 AM
To: soft...@listproc.autodesk.com
Subject: Re: Re: trying out the Arnold renderer

Marcos Fajardo ?

On Mon, Jul 6, 2009 at 3:53 PM, Gene Crucean <emailgene...@gmail.com> wrote:

I just want to know WHO is working on it now?



http://www.daniel3d.com/pepeland/gallery/imas/pepeland-com_01War.jpg ;)

[Joe Williamsen]

I'm very impressed at how the AO isnt' blown out in the bright areas.
Very nice...

[Steven Caron]

hey andreas

i have noticed on the examples where you put mental ray and arnold side by side, arnold only shows 3 render settings... AA, Hemi, and Glossy samples. are these the only render settings you need to adjust? if so... thank god

steven

[Andreas Bystrom]

Steven, yes, arnold is very easy to tweak, and way more predictable
than mental ray (arnold works very well in animation, after all that's
what it was designed for)

Tomorrow I will post a simple animation rendered in arnold, stay tuned ;)

[Andreas Bystrom]

Heres' a GI animation, no prebaking or anything.
http://ericknelson.com/wurp/arnold_gi_anim_rbd.mov

This one has gi, dof, mblur, glossy reflection and one arealight.
Rendertime is around 2.5-3.5 minutes per frame on an 8-core box.

Might try the same thing in mental ray but I suspect a single frame
would take about as long to render as the whole animation did in
arnold.

On Tue, Jul 7, 2009 at 10:23 AM, Andreas

[Antonin Messier-Turcotte]

Very impressive. Have you come across any issue that would prevent you from using Arnold in a production context?

2009/7/8 Andreas Bystrom <andreas...@gmail.com>

[John Payne]

Phenomenal!  Especially the speed of the glossy reflections and area light.  What level of
anti-aliasing are we looking at?

Cheers,

- John

[Ciaran Moloney]

Very nice. Can you talk about workflow at all? e.g. how well Arnold works with the Render Tree nodes etc.

[piotrek marczak]

holy shit..
how come GI doesn't flicker...???

W dniu 2009-07-08 20:15, Andreas Bystrom pisze:

[Andreas Bystrom]

Quick replies here :)

Ciaran: yes, I can say that it supports the rendertree, just like for
example 3delight it's all integrated.

Antoin: Arnold is already being used in production by Imageworks, as
for the plugin I have already tried it on a few production scenes.

John, this one used 8 AA samples.

[Steven Caron]

outstanding!

steven

[Votch]

Wow, I'm very impressed.

I've been battle the demons within MR for the last couple of years. Seeing your GI animation test is a breath of fresh air.

Votch-

[Francois Lord]

What about the displacement? How does it compare to mr?

[Rui Feliciano]

(speechless....)

o u t s t a n d i n g !

Does it support softimage's render passes as well?

Ok, who do I have to kill to get my hands on it?

Rui Feliciano

[Gene Crucean]

I'd like to see more production type examples if you don't mind.

RGB Mattes, Holdouts, No Primary Rays on something, Object ID's, Overrides etc... Lets see some passes. Also, does it support buffers (channels)? If yes, is the custom setup similar to MR's color in channel workflow?

[Andreas Bystrom]

Francois:

http://ericknelson.com/wurp/mr_vs_arnold_mblur.png

its a very simple test, but having more polys in arnold doesnt slow it
down much at all, so displacement is not an issue either, so far
everything I've tried mental ray just doesn't have anything over it..

Piotrek: the GI never flickers in arnold, the only thing you will get
is a light buzzing effect which you can get rid of by increasing the
AA samples, but there is never any "popping" or weird things happening
like in mental ray.

[Marc-Andre Carbonneau]

What about particles and ICE strands? ;) ya ok, you all got us excited like little girls in front of the Jonas Brothers now.

 

Please get the developer on board this mailing list if he isn’t already!

[Jason Brynford-Jones]

On behalf of Phil Taylor

http://area.autodesk.com/downloads/plugins/phil_s_ice_pack

100 ICE Compounds

-Compounds-
--Arrays--
Build Array from Per Point Data
Build Index Array
Build Linearly Interpolated Array
Find Closest in Array
Get Array Maximum 2
Get Array Minimum 2
Move in Array
Rebuild Array
Reverse Array
Select SubArray in Array
------------------------

--2D arrays--
Build 2D Array
Calc Sub Array Size
Get 2D Array Size
Get 2D SubArray Size
Initialize 2d Array
Pop SubArray from 2D Array
Push SubArray on 2D Array
Resize Array in 2D Array
Select Array in 2D Array
Select SubArray in 2D Array
Set SubArray in 2D Array
------------------------

--Curves--
Bezier4
Bezier5
Curve Distance to Curve Location
Curve Location to Curve Distance
Curve Location to Normalized U
Fitted Bezier Curve
Normalized U to Curve Location
------------------------------

--Debugging--
Build Arrow
Build Capsule
Build Circle
Build Cross
Build Debug Grid
Build Rectangle
Build Sphere
Build Square
Display Arrow
Display Line
Display Point
-------------

-Deformation-

--Effects--
Transform Mesh
Turbulize Mesh
--------------

--Hull Deformer--
Deform by Hull
Find Hull Surface Point
-----------------------

--Skinning--
Calculate Deformer Weights
Cleanup Painted Weights
Dual Quaternion Deformation
----------------------------

--Verlet--
Apply Surface Stiffness
Blend Verlet Effect
Correct Edge Lengths
Inflate Volume
Init Verlet Geometry Data
Move Points Outside Geometry
Pin Point to Position
Simulate Mesh using Verlet Integration
Verlet Framework
----------------

--Forces--
Add Verlet Force
Add Verlet Forces
Restitution Force
Verlet Drag Force
Verlet Wind Force
-----------------

--Execution--
Execute Once
Repeat with Counter
While with Counter
------------------

--Geometry Queries--
Reinterpret Location on New Geometry
------------------------------------

--Math--
LawOfCos
Limit Vector Length
Quaternion nlerp
WorleyNoise
-----------

-Particles-

--Emission control--
Filter by Particle IDs
Offset Particle From Emitter
----------------------------

--Getters--
Get Location by Raycast
Get Next Particle Position
Get Next Particle
Get Particle Speed
Get Particle Transform
Get Prev Particle Position
Get Prev Particle
-----------------

--Setters--
Set Particle Transform
----------------------

--Simulation--
Simulate Particles
------------------

--Strands--
Curl Strand
Display Strand Shape
Get Strand Length
-----------------

--Strand Dynamics--
Add Strand Force
Add Strand Forces
Constrain To Guide Hairs
Curl Strand
Emit Filler Strands
Emit Strands
Init Strand Dynamics
Shape Strand With Curve
Simulate Strand using Verlet Integration
Stick Particle to Emit Location
Strand Drag Force
Strand Dynamics Framework
Strand Gravity Force
Strand Length Bidirectional Constraint
Strand Length Constraint
Strand Stiffness
----------------

--Tests--
Test Animated Value
Test Particle ID
----------------

-Scenes-
Accessing Curves by Distance
Display Circle
Display Strand Shape
Fit Hermite Curve
Hair
Hull Deformer
Worley Noise
------------

[Andreas Bystrom]

Gene:
http://ericknelson.com/wurp/arnold_simple_production_examples.jpg

1.) no primary rays
2.) shadow caster disabled
3.) ray switch type material
4.) constant black
5.) constant red that emits GI
6.) constant green that does not emit GI
7.) normal object

and yes, all this works with passes/partitions ofcourse.

ice support?
here's a very simple example I made some time ago:
http://ericknelson.com/wurp/arnold_gi_anim.mov
If I have time I should try something a bit more complex ;)

[Bradley Gabe]

Many Thanks!

[Marc-Andre Carbonneau]

Me jaw just dropped. Thanks Phil for your generosity and Chinny(for posting)

MAC

[Steven Caron]

yes, thanks phil, chinny, and rest of softimage/autodesk!

s

[Andy Moorer]

Wahooo! Very cool!

[Nils Engler]

Wow, Thank you :)

[Rens Heeren]

Wow, absolutely amazing.... many thanks!

[Gene Crucean]

Buffers/Channels?

What is the worst part of the renderer itself so far? Worst part of it's workflow? Slowest? A couple things that you really think need to be worked on before release? Instanced Geo in ICE?

Sorry I'm not a ooh and ahh kinda guy with renderers anymore. Gimme the meat! :)

-g

[Gene Crucean]

Thanks Phil!

And thanks for schooling me on ICE a while back haha. I'm still blown away by your massive noggin power (look at the sieeeze of that boys heed! It's like an orange on a tooth pick! It's got it's own weather system!) :)

-gene

[Ciaran Moloney]

How about a shadow pass? It's been the biggest pain in the arse to get anything remotely correct from 3Delight.

[rob...@texturelighting.com]

Thank you indeed!

-Robert

-- 
Render Theory: It either renders or it doesn't

[Francois Lord]

Wow! That seems nice. Now if only I could log on to the Area...

[Alan Fregtman]

Looks fantastic. Thank you!

[tak...@earthlink.net]

Thanks, Chinny, Phil & co. for sharing - very cool to see this stuff!

-T

[Robert Chapman]

/bows

many thanks! :D

2009/7/8 <tak...@earthlink.net>

[Dan Yargici]

I found clicking the link asked me to logon, but if I just went to the Area I was already logged on and the link on the front page took me straight there... Grrrrrrrrrrr.

Thanks Chinny / Phil!

DAN

[Jason Brynford-Jones]

Download
http://www.xsibase.com/Phils_ICE_Pack.zip

http://www.xsibase.com/news.php?detail=2350

[Andreas Bystrom]

That's fair enough Gene.

This is again something I've had to ask permission to post, but here goes:
The actual arnold API is very mature and has been used in production
for quite a long time, as for the plugin it has only been in
development for about 4 months now, however it's progressing very
fast, new things are fixed/added every single day, it still needs work
but the goal is to have a fully integrated renderer that supports as
much as possible.

As for the workflow it's pretty much the same as in mental ray, you
work with the same rendertree, you use the region etc, can't really
say what the worst part of it is except that it could have support for
a few more shaders, but that will come I'm sure.
The best part of it, besides the render speed is that it's
interactive, meaning when tweaking lighting and shading even if you
have millions of polygons the render starts instantly after rendering
the first time...

About ICE/instances and buffers, both are supported but still needs
things added, a few more buffers could be supported and ICE doesnt
work with volumes yet.

Also as you can see from the examples I have tried a lot of different
things in it, I guess full interiors/arch viz slows it down the most.
Other than that it has all been fast so far, usually with a raytracer
motionblur and dof is what kills it, arnold has performance pretty
close to a reyes renderer when it comes to that while still kicking
ass at raytracing, that is something I haven't really seen in any
other renderer I've tried so far.

[David Barosin]

Truly generous.  Thanks Phil and Chinny!

[Matt]

Any SSS tests? Glossy refractions? Caustics? I just want to see more of this! what shaders are not supported yet. Any MR ones? Architectural, SSS, Glossy relection, etc. Dielectric? What if you override say a mix8 colors in the render tree, or a setting in a shader. Does it behave as expected? Hair examples?

So many questions... Can't wait to test it out here!

--
Matt

CG artist
The Embassy Visual Effects

[William Pleydon]

Thanks in advance...I'm not playing in ICE yet, but I'm sure I'll appreciate
it later!

William

----- Original Message -----
From: "Jason Brynford-Jones" <Jason.Bryn...@autodesk.com>
To: <soft...@listproc.autodesk.com>
Sent: Wednesday, July 08, 2009 3:58 PM
Subject: Phil's ICE Pack

[Adam Sale]

great stuff. Thanks for the heads up chin

On Wed, Jul 8, 2009 at 2:33 PM, Jason Brynford-Jones <Jason.Bryn...@autodesk.com> wrote:

Download
http://www.xsibase.com/Phils_ICE_Pack.zip

http://www.xsibase.com/news.php?detail=2350

[Nick]

I have a few questions too, I hope you don't mind, it looks all very exciting and I'd love to know more.
So along those lines, how about a programmable shading language?
What's the API like?
Custom geometry file format support (dso or geometry shader equivalent)?
Volume shaders, Caustics, Refractions, SSS (dipole, multipole support?), hair with transparency and displacement - what hair shaders are implemented (Kajiya, Marschner?), Geometry Instances, LOD support?

What kind of (local and network) caching mechanisms does it have for geometry, textures, shaders etc?
What is the support like for subdivision surfaces? Are they real subdivision or do you have to, like in Mental Ray, specify the number of recursions?
What mip-mapped file format does it use for texures?
Distributed rendering support?

The renders look really nice, I'd really like to see much heavier production renders with it, FX heavy stuff too. Any chance you might be able to throw some fluid renders at it? It'd be great to see that with everything thrown at it, volumes, scattering, GI, DOF, mo-blur, etc...  

And as a completely side note, the bokeh looks nice, do you have control over the shape of it?

Cheers,

Nick

2009/7/9 Andreas Bystrom <andreas...@gmail.com>

[Sebastian Kowalski]

now i am really impressed.

[Morten Bartholdy]

That looks very impressive considering quality and rendertime.

Morten Bartholdy
3D & VFX Supervisor

----- Original Message -----
From: "Andreas Bystrom" <andreas...@gmail.com>
To: <soft...@listproc.autodesk.com>

Sent: Wednesday, July 08, 2009 8:15 PM
Subject: Re: trying out the Arnold renderer

> Heres' a GI animation, no prebaking or anything.
> http://ericknelson.com/wurp/arnold_gi_anim_rbd.mov
>
> This one has gi, dof, mblur, glossy reflection and one arealight.
> Rendertime is around 2.5-3.5 minutes per frame on an 8-core box.
>
> Might try the same thing in mental ray but I suspect a single frame
> would take about as long to render as the whole animation did in
> arnold.
>
>
>

> On Tue, Jul 7, 2009 at 10:23 AM, Andreas
> Bystrom<andreas...@gmail.com> wrote:
>> Steven, yes, arnold is very easy to tweak, and way more predictable
>> than mental ray (arnold works very well in animation, after all that's
>> what it was designed for)
>>
>> Tomorrow I will post a simple animation rendered in arnold, stay tuned ;)
>>
>>
>>
>>
>>
>> On Tue, Jul 7, 2009 at 4:38 AM, Steven Caron<car...@gmail.com> wrote:
>>> hey andreas
>>>
>>> i have noticed on the examples where you put mental ray and arnold side
>>> by
>>> side, arnold only shows 3 render settings... AA, Hemi, and Glossy
>>> samples.
>>> are these the only render settings you need to adjust? if so... thank
>>> god
>>>
>>> steven
>>>
>>> On Sun, Jul 5, 2009 at 4:20 AM, Andreas Bystrom
>>> <andreas...@gmail.com>
>>> wrote:
>>>>
>>>> Recently I have been betatesting the arnold renderer for softimage. I
>>>> have been testing various scenes in both mental ray and arnold and so
>>>> far arnold has proven to be extremely impressive.
>>>>
>>>> The first test I'd like to show is an interior:
>>>>
>>>> this is the arnold version:
>>>> http://ericknelson.com/wurp/arnold_scifi_01.png
>>>>
>>>> and this is the mental ray version:
>>>> http://ericknelson.com/wurp/mentalray_scifi_01.png
>>>>
>>>> As you can see they look pretty much the same except arnold has a lot
>>>> more detail, look at the wall on the left for example, arnold catches
>>>> every little detail in the model while the mental ray
>>>> version looks pretty washed out. Note that in the mental ray version
>>>> I'm using the architectural shader with the ambient occlusion mode
>>>> enabled (detail enhancement set to 2), without that it would lack even
>>>> more detail.
>>>>
>>>> These two examples rendered in about the same time (27 minutes or so
>>>> on a quadcore). The arnold version however would be completely stable
>>>> in animation, the mental ray version
>>>> would flicker a lot due to the interpolation it uses.
>>>>
>>>>
>>>> The second test is a simple occlusion render (the scene is around 500k
>>>> polys and the grass has some 350k hairs):
>>>> http://ericknelson.com/wurp/mr_vs_arnold.jpg
>>>>
>>>>
>>>> Both renderers uses an occlusion shader with depth of field enabled,
>>>> as you can see the arnold version has a bit less noise while still
>>>> rendering way faster than the mental ray version.
>>>> Enabling either depth of field or motionblur in arnold has a very
>>>> small impact on rendering speeds, even combining the two is very fast,
>>>> in mental ray however it slows down a LOT.
>>>>
>>>>
>>>>
>>>> Another example I'd like to show is rendering hair with gi/fg in both
>>>> renderers, here is a version without motionblur:
>>>> http://ericknelson.com/wurp/mr_vs_arnold_hair_no_mblur.jpg
>>>>
>>>> The two versions looks quite different and I guess I could match the
>>>> look more closely spending a bit more time on it, however the
>>>> important thing here is that they both use an identical setup:
>>>> Both versions has one arealight with exactly the same radius as well
>>>> as gi/fg enabled that also illuminates the scene using an environment
>>>> map.
>>>>
>>>>
>>>> Here is the same scene, but this time I've enabled motionblur (that is
>>>> with deformation blur, two motion steps), I've also switched the
>>>> mental ray version to use the rasterizer mode which is supposed to be
>>>> best way to render hair with motionblur in mental ray:
>>>>
>>>> http://ericknelson.com/wurp/mr_vs_arnold_hair_with_mblur.jpg
>>>>
>>>> The difference in renderspeed is even more obvious, the arnold version
>>>> has hardly slowed down at all while the mental ray version is taking
>>>> several times as long to render. Again these versions looks quite
>>>> different but
>>>> they still use an identical lighting setup (one arealight as well as
>>>> gi/fg).
>>>>
>>>>
>>>> Here are a few more examples that i've rendered in arnold:
>>>>
>>>>
>>>> http://ericknelson.com/wurp/mr_vs_arnold_mblur.png (clothsim with
>>>> deformation blur and displacement)
>>>>
>>>> http://ericknelson.com/wurp/arnold_light_falloff.jpg (this is with GI,
>>>> one arealight, again this renders in just over 5 minutes)
>>>>
>>>> http://ericknelson.com/wurp/tank_glossy_new_0.jpg (this is with GI,
>>>> one arealight as well as lots of glossy reflections, took around 10
>>>> minutes.)
>>>>
>>>> http://ericknelson.com/wurp/Arnold_classroom_01.png
>>>>
>>>>
>>>> Sony imageworks has also recently completed their second movie all
>>>> rendered in arnold: http://www.youtube.com/watch?v=BPH0ct2oXBg (first
>>>> one was monster house:
>>>> http://www.apple.com/trailers/sony_pictures/monster_house/)
>>>>
>>>>
>>>>
>>>> I hope you enjoy these examples, stay tuned for more.
>>>> and by the way, it's not an open beta but I have been given permission
>>>> to post these examples on the list...
>>>
>>>
>>
>

[Morten Bartholdy]

I have a feeling they will get to sell a seat or two of this - I was
contemplating adding 3Delight to our MR pipeline but by the looks of it
Arnold could replace both....

Morten Bartholdy
3D & VFX Supervisor

----- Original Message -----
From: "Andreas Bystrom" <andreas...@gmail.com>
To: <soft...@listproc.autodesk.com>

[Frank Lenhard]

awesome! thanks!

ciao
franky

Wednesday, July 8, 2009, 9:58:19 PM, you wrote:

JBJ> On behalf of Phil Taylor

JBJ> http://area.autodesk.com/downloads/plugins/phil_s_ice_pack

JBJ> 100 ICE Compounds

JBJ> -Compounds-
JBJ> --Arrays--
JBJ> Build Array from Per Point Data
JBJ> Build Index Array
JBJ> Build Linearly Interpolated Array
JBJ> Find Closest in Array
JBJ> Get Array Maximum 2
JBJ> Get Array Minimum 2
JBJ> Move in Array
JBJ> Rebuild Array
JBJ> Reverse Array
JBJ> Select SubArray in Array
JBJ> ------------------------

JBJ> --2D arrays--
JBJ> Build 2D Array
JBJ> Calc Sub Array Size
JBJ> Get 2D Array Size
JBJ> Get 2D SubArray Size
JBJ> Initialize 2d Array
JBJ> Pop SubArray from 2D Array
JBJ> Push SubArray on 2D Array
JBJ> Resize Array in 2D Array
JBJ> Select Array in 2D Array
JBJ> Select SubArray in 2D Array
JBJ> Set SubArray in 2D Array
JBJ> ------------------------

JBJ> --Curves--
JBJ> Bezier4
JBJ> Bezier5
JBJ> Curve Distance to Curve Location
JBJ> Curve Location to Curve Distance
JBJ> Curve Location to Normalized U
JBJ> Fitted Bezier Curve
JBJ> Normalized U to Curve Location
JBJ> ------------------------------

JBJ> --Debugging--
JBJ> Build Arrow
JBJ> Build Capsule
JBJ> Build Circle
JBJ> Build Cross
JBJ> Build Debug Grid
JBJ> Build Rectangle
JBJ> Build Sphere
JBJ> Build Square
JBJ> Display Arrow
JBJ> Display Line
JBJ> Display Point
JBJ> -------------

JBJ> -Deformation-

JBJ> --Effects--
JBJ> Transform Mesh
JBJ> Turbulize Mesh
JBJ> --------------

JBJ> --Hull Deformer--
JBJ> Deform by Hull
JBJ> Find Hull Surface Point
JBJ> -----------------------

JBJ> --Skinning--
JBJ> Calculate Deformer Weights
JBJ> Cleanup Painted Weights
JBJ> Dual Quaternion Deformation
JBJ> ----------------------------

JBJ> --Verlet--
JBJ> Apply Surface Stiffness
JBJ> Blend Verlet Effect
JBJ> Correct Edge Lengths
JBJ> Inflate Volume
JBJ> Init Verlet Geometry Data
JBJ> Move Points Outside Geometry
JBJ> Pin Point to Position
JBJ> Simulate Mesh using Verlet Integration
JBJ> Verlet Framework
JBJ> ----------------

JBJ> --Forces--
JBJ> Add Verlet Force
JBJ> Add Verlet Forces
JBJ> Restitution Force
JBJ> Verlet Drag Force
JBJ> Verlet Wind Force
JBJ> -----------------

JBJ> --Execution--
JBJ> Execute Once
JBJ> Repeat with Counter
JBJ> While with Counter
JBJ> ------------------

JBJ> --Geometry Queries--
JBJ> Reinterpret Location on New Geometry
JBJ> ------------------------------------

JBJ> --Math--
JBJ> LawOfCos
JBJ> Limit Vector Length
JBJ> Quaternion nlerp
JBJ> WorleyNoise
JBJ> -----------

JBJ> -Particles-

JBJ> --Emission control--
JBJ> Filter by Particle IDs
JBJ> Offset Particle From Emitter
JBJ> ----------------------------

JBJ> --Getters--
JBJ> Get Location by Raycast
JBJ> Get Next Particle Position
JBJ> Get Next Particle
JBJ> Get Particle Speed
JBJ> Get Particle Transform
JBJ> Get Prev Particle Position
JBJ> Get Prev Particle
JBJ> -----------------

JBJ> --Setters--
JBJ> Set Particle Transform
JBJ> ----------------------

JBJ> --Simulation--
JBJ> Simulate Particles
JBJ> ------------------

JBJ> --Strands--
JBJ> Curl Strand
JBJ> Display Strand Shape
JBJ> Get Strand Length
JBJ> -----------------

JBJ> --Strand Dynamics--
JBJ> Add Strand Force
JBJ> Add Strand Forces
JBJ> Constrain To Guide Hairs
JBJ> Curl Strand
JBJ> Emit Filler Strands
JBJ> Emit Strands
JBJ> Init Strand Dynamics
JBJ> Shape Strand With Curve
JBJ> Simulate Strand using Verlet Integration
JBJ> Stick Particle to Emit Location
JBJ> Strand Drag Force
JBJ> Strand Dynamics Framework
JBJ> Strand Gravity Force
JBJ> Strand Length Bidirectional Constraint
JBJ> Strand Length Constraint
JBJ> Strand Stiffness
JBJ> ----------------

JBJ> --Tests--
JBJ> Test Animated Value
JBJ> Test Particle ID
JBJ> ----------------

JBJ> -Scenes-
JBJ> Accessing Curves by Distance
JBJ> Display Circle
JBJ> Display Strand Shape
JBJ> Fit Hermite Curve
JBJ> Hair
JBJ> Hull Deformer
JBJ> Worley Noise
JBJ> ------------

[Phil Williams]

Wow – much appreciated! Thanks Phil J

 

---

From: Jason Brynford-Jones [mailto:Jason.Bryn...@autodesk.com]
Sent: 08 July 2009 20:58
To: soft...@listproc.autodesk.com
Subject: Phil's ICE Pack

 

On behalf of Phil Taylor

 

http://area.autodesk.com/downloads/plugins/phil_s_ice_pack

 

100 ICE Compounds

 

 

-Compounds-

            --Arrays--

            Build Array from Per Point Data

            Build Index Array

            Build Linearly Interpolated Array

            Find Closest in Array

            Get Array Maximum 2

            Get Array Minimum 2

            Move in Array

            Rebuild Array

            Reverse Array

            Select SubArray in Array

            ------------------------     

           

                        --2D arrays--

                        Build 2D Array

                        Calc Sub Array Size

                        Get 2D Array Size

                        Get 2D SubArray Size

                        Initialize 2d Array

                        Pop SubArray from 2D Array

                        Push SubArray on 2D Array

                        Resize Array in 2D Array

                        Select Array in 2D Array

                        Select SubArray in 2D Array

                        Set SubArray in 2D Array

                        ------------------------

           

            --Curves--

            Bezier4

            Bezier5

            Curve Distance to Curve Location

            Curve Location to Curve Distance

            Curve Location to Normalized U

            Fitted Bezier Curve

            Normalized U to Curve Location

            ------------------------------

                        Simulate Mesh using Verlet Integration

                        Verlet Framework

                        ----------------

 

                                    --Forces--

                                    Add Verlet Force

                                    Add Verlet Forces

                                    Restitution Force

                                    Verlet Drag Force

                                    Verlet Wind Force

                                    -----------------

           

            --Execution--

            Execute Once

            Repeat with Counter

            While with Counter

            ------------------

 

            --Geometry Queries--

            Reinterpret Location on New Geometry

            ------------------------------------

                                    Simulate Strand using Verlet Integration

                                    Stick Particle to Emit Location

                                    Strand Drag Force

                                    Strand Dynamics Framework

                                    Strand Gravity Force

                                    Strand Length Bidirectional Constraint

                                    Strand Length Constraint

                                    Strand Stiffness

                                    ----------------

 

                        --Tests--

                        Test Animated Value

                        Test Particle ID

                        ----------------

 

-Scenes-

Accessing Curves by Distance

Display Circle

Display Strand Shape

Fit Hermite Curve

Hair

Hull Deformer

Worley Noise

------------

 

           

 

Tinopolis
Tinopolis Centre, Park Street, Llanelli  SA15 3YE
UK Company Reg. 03832383
___________________________________________________
This email and any files attached are strictly confidential and
intended solely for the person or entity to whom they are
addressed. They may contain legally privileged and/or
confidential information. If you are not the intended recipient
you must not copy, distribute or take any action in reliance
on them. If you have received this email in error, please notify
us as soon as possible and delete it. The views and opinions
expressed in this email message and any attachments are
the author's own and they do not reflect necessarily the views
and opinions of Tinopolis.  

Mae'r ebost hwn ac unrhyw ffeiliau sydd ynghlwm yn
gyfrinachol ac at ddefnydd y person neu'r corff y cyfeiriwyd
hwy atynt yn unig. Gallant gynnwys gwybodaeth sy'n
freintiedig yn gyfreithiol a/neu'n gyfrinachol ac os nad chi yw'r
sawl y cyfeiriwyd y neges ato, yna peidiwch a chopio,
dosbarthu na chymryd unrhyw gamau o ganlyniad a gan
ddibynnu arni.   Os ydych wedi derbyn yr ebost hwn drwy
gamgymeriad, dylech ein hysbysu o hyn cyn gynted a phosib
a'i dileu os gwelwch yn dda.  Barn neu safbwyntiau'r awdur
yw'r rhai a fynegir yn y neges ebost hon ac unrhyw atodiadau
ac nid ydynt yn adlewyrchu o anghenraid barn neu
safbwyntiau Tinopolis.
___________________________________________________

[Juan Brockhaus]

very much appreciated!
thanks!

[Tim Borgmann]

I'm really impressed. All your tests are looking very promising and the
technical details are great so far. I really hope this renderer will be
released within a useful timeframe. Can't wait to test it by myself :)

[Alexander Samouridis]

Those all look awesome! I'm no render guru myself
but the guys here at work are pretty impressed by
your tests.

Is there any place on the web one can get more info
on the renderer, availability, pricing etc.?

--

Alexander Samouridis

Animator / Character TD
GROUND STUDIOS GmbH

[Andreas Bystrom]

Matt: as for SSS and glossy refractions, yes it does both, thought I
havent gotten around testing either much yet, but if you look at this
trailer you will find examples of both:
http://www.youtube.com/watch?v=BPH0ct2oXBg

afaik all the faces and a few other things uses SSS, there are also
some glossy refraction in there like that big jello thing.

Nick:
It has programmable shaders, all the shaders that are supported in xsi
was written this way, i believe you can also write custom shaders and
use them in the rendertree directly, just like with mental ray..
It also has its own format much like RIB files, archives etc. I'm not
a programmer so I dont know the API at all, perhaps Marcos will be
kind enough to step in here and explain a few things about it at some
point.
As for volume shader, I can't say, I've only tried volume lights and
those work like a charm, fast and accurate, and ofcourse it does
refraction.

As for LOD and dsos, I can't say, again something Marcos can answer.

Geometry instances I tried yesterday, this is through ICE also:

http://ericknelson.com/wurp/arnold_ice_instances.jpg

This is about 30 billion polygons in total all rendered in under 10
minutes at 2k (GI, glossy reflection, one arealight), I also tried
this with motionblur and as before with arnold it works without a
fuss. This all rendered with xsi using around 400mb of ram in total,
and xsi eats up around 150 of those by itself.
I also tried rendering this exact scene in mental ray, it ended up
using over 7gb of memory before I gave up, my box "only" has 8 gb of
ram so... using that much memory it didnt even render a single bucket
with the tanks in them, only the white background was rendering a bit.

In arnold I can also tweak the lighting/shaders without having to wait
for the render to start...

[Chris Marshall]

Super amazing!

Big thanks indeed!!

[Arnar Gunnarsson]

what about the interactive rendering, you say its extremely fast if your not adding or changing geometry.  so I guess its caching the data, hence the short render time, or...? 

 

so how much does a "fresh" frame take to render compared to a cached one?

would a ICE particle sim not count for a changing geometry and a constant "fresh" frame?

that did make me worry about memory usage but you actually answered that in your last post... very impressive indeed 

[Marc-Andre Carbonneau]

Ok stop it already Andreas! We can't take it anymore! ;)
I guess Siggraph is the time it will be announced officially?

Is it the only renderer on Meatballs movie?!!!
Thanks for your time answering all the questions...bet you didn't think
you'd get as much attention! ;)

-----Original Message-----
From: softimag...@listproc.autodesk.com
[mailto:softimag...@listproc.autodesk.com] On Behalf Of Andreas
Bystrom
Sent: July 9, 2009 5:32 AM
To: soft...@listproc.autodesk.com
Subject: Re: trying out the Arnold renderer

[adrian wyer]

considering the apparent lack of improvement in MR's image based lighting
and especially it's p1ss poor performance when using depth of field and/or
motion blur, it seems Arnold (assuming they don't price themselves out of
the park) could be very serious MR replacement.....

a

[Andreas Bystrom]

Arnar:

It's caching the data yes, a fresh frame would take about the same
time to export as with for example the 3delight plugin, it's still
fast. The tank scene took some 15-20 seconds to start, which was
actually quicker than with mental ray..

If you had a particle sim youd have to re-export for every frame the
animation changes, adding new particles and such, however on a single
frame you can still tweak things in realtime, just like on this tank
scene (it uses ice particles.)

And yes, memory usage has not been a problem, Im pretty sure I could
take it up to rendering 100 billion polygons using instances even on
32 bit, but it becomes a bit silly testing that I think :)

Marc-Andre: As for announcements it's not my place to answer such
questions, you will have to wait and see I guess :)

The meatball movie was all rendered in arnold yes, same with monster
house that came out a few years ago.

[Ciaran Moloney]

Great! Thanks for sharing.

[Gene Crucean]

The meatball movie was all rendered in arnold yes, same with monster
house that came out a few years ago.

This says almost enough about it right here, even though I'm assuming both of those were through Maya (boooooo). I had no idea of this.

[Rick Fuentealba]

Not sure how it was done for Monster House, but Cloudy was all lit,
rendered and comped on Imagework's proprietary software called Katana.

[Steven Caron]

yes but Katana is renderer agnostic. i have seen it demo'd using arnold.

steven

[Steven Caron]

sorry Rick, i think i misunderstood your comment. you were just suggesting they were using Katana instead of Maya but with Arnold?

s

[Scott Parrish]

Monster House was rendered from maya using a modified version of mArnie or mArnold plugin that was available at the time (cant remember the name exactly).
Katana was still in development at the time. I'm pretty sure Katana was designed to be renderer agnostic so they could be using it atop prman, arnold, mentalray, etc.

--
scott parrish
scot...@gmail.com

[Christopher]

It is nice to see that I can now get a Dual Q compound but how do you apply it to enveloped meshes, unless I'm wrong ?

Message: 1
Date: Wed, 8 Jul 2009 19:58:55 -0400
From: "William Pleydon" <will...@digimata.com>
Subject: Re: Phil's ICE Pack
To: <soft...@listproc.autodesk.com>
Message-ID: <FAFACDD334CC4D10BDF1AC942F9E4818@user51b7c7bbcc>
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
reply-type=original

[Christopher]

I'm sorry maybe I missed it but where can I read up on Arnold coming to SoftImage ? From the posts I've read it appears very promising.

[tak...@earthlink.net]

Nowhere yet apparently. :)

-T

-----Original Message-----
>From: Christopher <chris...@thecreativesheep.ca>
>Sent: Jul 9, 2009 9:01 PM
>To: soft...@listproc.autodesk.com
>Subject: RE: trying out the Arnold renderer
>

[Jason Brynford-Jones]

The DQ Compound OP drives the deformation not the weights, so:

Envelope and weight as normal.
Mute the envelope
Add ICE op
Move to Animation construction mode
Connect DQ compound (switch to DQ mode)
-now get all the deformers in the envelope, you might already have a group setup, if not...
Select Mesh
Envelope>select deformers from envelope
Create Group
D&D into ICE Tree
Connect to DQ compound - Volia

You can then add a weight map (connected to DQ compound) if you want to drive the areas you want to be DQ like the shoulders.

Check out this vid,
http://polimeno.blogspot.com/2009/02/softimage-ice.html
Phil does some extra clean up but it is not necessary for you

Chin

[Christopher]

Alright let me try it out, tomorrow. :)

Message: 6
Date: Thu, 9 Jul 2009 18:20:10 -0700
From: Jason Brynford-Jones <Jason.Bryn...@autodesk.com>
Subject: RE: Phil's ICE Pack
To: "soft...@listproc.autodesk.com"
<soft...@listproc.autodesk.com>
Message-ID:
<3FF2397445743C47B0507...@ADSK-NAMSG-01.MGDADSK.autodesk.com>

Content-Type: text/plain; charset="utf-8"

[Kris Rivel]

Jesus!  This looks incredible!  The cubes falling render sent a chill up my spine...fantastic.  So...anyone...how much will it cost and when will it be release!?  I so want to be on beta for this.  Anyone have a contact?  I'm shocked Luc asked what this was about...you would think everyone on the SI team would know what renderers are beta testing...or am I assuming too much?  I'm just growing do so tired of MR and its "suckiness".

Kris

[Steven Caron]

i think the rendering api is allowing people to connect to renderers without as much help from softimage. i seem to remember halfdan saying something like... its up to you guys now

s

[James G. Jones]

Are there any more standardized render time tests, like the sponza
atrium, for example? Just to get an idea how Arnold might fair
compared to renderers other than MR...

-Jim

[Steven Caron]

i think andreas posted this originally...

http://ericknelson.com/wurp/Arnold_classroom_01.png

i would consider this scene pretty popular for those 'other renderers'

steven

[Christopher]

- Where do you put ICE compounds to store as a collection ? This is sweet now ICE has DualQ, finally no more DualQ add-on working only on 32-bit systems. Especially when I upgrade my 32-bit system to Win7 !
- The only thing is what do the rest of these nodes do in Phil's ICE pack ?

Message: 2
Date: Fri, 10 Jul 2009 01:43:59 +0000
From: Christopher <chris...@thecreativesheep.ca>

Subject: RE: Phil's ICE Pack
To: soft...@listproc.autodesk.com

Message-ID:
<4a569cdfbb07e_122...@app02.zenbe.com.tmail>
Content-Type: text/plain; charset=utf-8

[Chris Marshall]

hey,

When do we get Phil's 100 ICE tutorials?

;-)

...no really, some support would be good.

[Christopher]

I second that, I don't know really on how to use these and one has to has those who are pretty good
at ICE can any of these nodes in Phil's pack be reconstructed with the default nodes in ICE ?

Message: 2
Date: Mon, 13 Jul 2009 09:05:08 +0100
From: Chris Marshall <chris.m...@eclipsecreative.co.uk>
Subject: Re: Phil's ICE Pack
To: soft...@listproc.autodesk.com
Message-ID:
<1f902a650907130105kcf...@mail.gmail.com>
Content-Type: text/plain; charset="iso-8859-1"

hey,When do we get Phil's 100 ICE tutorials?

[Chris Marshall]

As far as I know, these compounds are simply made of the current base nodes in ICE, which is why they're called compounds. New core nodes aren't likely to be added until we get an upgrade to XSI.

[Christoph Schinko]

Holy smokes! Thanks very much!! :)

-----Original Message-----
From: softimag...@listproc.autodesk.com
[mailto:softimag...@listproc.autodesk.com] On Behalf Of Jason
Brynford-Jones
Sent: 08 July 2009 20:58
To: soft...@listproc.autodesk.com
Subject: Phil's ICE Pack

On behalf of Phil Taylor

http://area.autodesk.com/downloads/plugins/phil_s_ice_pack

100 ICE Compounds

-Compounds-
--Arrays--
Build Array from Per Point Data
Build Index Array
Build Linearly Interpolated Array
Find Closest in Array
Get Array Maximum 2
Get Array Minimum 2
Move in Array
Rebuild Array
Reverse Array
Select SubArray in Array
------------------------

--2D arrays--
Build 2D Array
Calc Sub Array Size
Get 2D Array Size
Get 2D SubArray Size
Initialize 2d Array
Pop SubArray from 2D Array
Push SubArray on 2D Array
Resize Array in 2D Array
Select Array in 2D Array
Select SubArray in 2D Array
Set SubArray in 2D Array
------------------------

--Curves--
Bezier4
Bezier5
Curve Distance to Curve Location
Curve Location to Curve Distance
Curve Location to Normalized U
Fitted Bezier Curve
Normalized U to Curve Location
------------------------------

--Debugging--
Build Arrow
Build Capsule
Build Circle
Build Cross
Build Debug Grid
Build Rectangle
Build Sphere
Build Square
Display Arrow
Display Line
Display Point
-------------

-Deformation-

--Effects--
Transform Mesh
Turbulize Mesh
--------------

--Hull Deformer--
Deform by Hull
Find Hull Surface Point
-----------------------

--Skinning--
Calculate Deformer Weights
Cleanup Painted Weights
Dual Quaternion Deformation
----------------------------

--Verlet--
Apply Surface Stiffness
Blend Verlet Effect
Correct Edge Lengths
Inflate Volume
Init Verlet Geometry Data
Move Points Outside Geometry
Pin Point to Position
Simulate Mesh using Verlet Integration
Verlet Framework
----------------

--Forces--
Add Verlet Force
Add Verlet Forces
Restitution Force
Verlet Drag Force
Verlet Wind Force
-----------------

--Execution--
Execute Once
Repeat with Counter
While with Counter
------------------

--Geometry Queries--
Reinterpret Location on New Geometry
------------------------------------

--Math--
LawOfCos
Limit Vector Length
Quaternion nlerp
WorleyNoise
-----------

-Particles-

--Emission control--
Filter by Particle IDs
Offset Particle From Emitter
----------------------------

--Getters--
Get Location by Raycast
Get Next Particle Position
Get Next Particle
Get Particle Speed
Get Particle Transform
Get Prev Particle Position
Get Prev Particle
-----------------

--Setters--
Set Particle Transform
----------------------

--Simulation--
Simulate Particles
------------------

--Strands--
Curl Strand
Display Strand Shape
Get Strand Length
-----------------

--Strand Dynamics--
Add Strand Force
Add Strand Forces
Constrain To Guide Hairs
Curl Strand
Emit Filler Strands
Emit Strands
Init Strand Dynamics
Shape Strand With Curve
Simulate Strand using Verlet Integration
Stick Particle to Emit Location
Strand Drag Force
Strand Dynamics Framework
Strand Gravity Force
Strand Length Bidirectional Constraint
Strand Length Constraint
Strand Stiffness
----------------

--Tests--
Test Animated Value
Test Particle ID
----------------

-Scenes-
Accessing Curves by Distance
Display Circle
Display Strand Shape
Fit Hermite Curve
Hair
Hull Deformer
Worley Noise
------------