Ice Tutorials

109 views
Skip to first unread message

Sam J. Bowling

unread,
Apr 12, 2011, 1:36:41 AM4/12/11
to soft...@listproc.autodesk.com
I recently ran across some great ICE tutorials from Paul Smith. The "Zombie
outbreak" set are great and very well done and thinking of them as Zombies
and Humans instead of particles seems to really help keep me interested. I
can already think of a lot of uses for the things he covers in this set and
it's all a lot easier than I would have imagined.


--
Sam J. Bowling

Rob Wuijster

unread,
Apr 12, 2011, 3:22:00 AM4/12/11
to soft...@listproc.autodesk.com
Yes,

Paul's ICE videos are excellent, and his way of explaining things make ICE less 'complicated'...

Not sure if this was posted before, but the direct link is http://vimeo.com/user4895541/videos if people are interested.

Rob

\/-------------\/----------------\/
-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 10.0.1209 / Virus Database: 1500/3565 - Release Date: 04/11/11


Tim Leydecker

unread,
Apr 12, 2011, 4:23:30 AM4/12/11
to soft...@listproc.autodesk.com
His ICE raycast/raytrace in viewport images look more natural
than a great many images I�ve produced sofar...

No gamma issues, an intuitive way to create a fisheye, soft lighting, .etc

Awesome... absolutely awesome.

Brilliant and fun.

Cheers

tim

Tim Crowson

unread,
Apr 12, 2011, 9:03:41 AM4/12/11
to ro...@casema.nl, soft...@listproc.autodesk.com
I'm of the opinion of that although ICE has that word "creative" in it, it really is for TDs, and not very artist-friendly (at least not without the help of a TD first!). Paul's videos make things so easy to understand. I wish there was more content like that out there!

+1 for Pooby!


Tim Crowson   
Asset Artist

http://www.magneticdreams.com/md_logo.gif


Magnetic Dreams Animation Studio, Inc.
2525 Lebanon Pike, Building C. Nashville, TN 37214
Ph  615.885.6801 | Fx  615.889.4768 | www.magneticdreams.com
t...@magneticdreams.com



From: "Rob Wuijster" <ro...@casema.nl>
Sent: Tuesday, April 12, 2011 2:23 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

Guillaume Laforge

unread,
Apr 12, 2011, 9:08:56 AM4/12/11
to soft...@listproc.autodesk.com
You meant that TD's can't be creative ?

;)

Guillaume

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Tim Crowson
Sent: Tuesday, April 12, 2011 9:04 AM
To: ro...@casema.nl; soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

I'm of the opinion of that although ICE has that word "creative" in it, it really is for TDs, and not very artist-friendly (at least not without the help of a TD first!). Paul's videos make things so easy to understand. I wish there was more content like that out there!

+1 for Pooby!

Tim Crowson
Asset Artist

[http://www.magneticdreams.com/md_logo.gif]


Magnetic Dreams Animation Studio, Inc.
2525 Lebanon Pike, Building C. Nashville, TN 37214

Ph 615.885.6801 | Fx 615.889.4768 | www.magneticdreams.com<http://www.magneticdreams.com/>
t...@magneticdreams.com<http://mail.magneticdreams.com/Main/t...@magneticdreams.com>

________________________________
From: "Rob Wuijster" <ro...@casema.nl>
Sent: Tuesday, April 12, 2011 2:23 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

Yes,

Paul's ICE videos are excellent, and his way of explaining things make ICE less 'complicated'...

Not sure if this was posted before, but the direct link is http://vimeo.com/user4895541/videos if people are interested.


Rob

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

On 4/12/2011 07:36 AM, Sam J. Bowling wrote:
I recently ran across some great ICE tutorials from Paul Smith. The "Zombie outbreak" set are great and very well done and thinking of them as Zombies and Humans instead of particles seems to really help keep me interested. I can already think of a lot of uses for the things he covers in this set and it's all a lot easier than I would have imagined.


--
Sam J. Bowling

-----
No virus found in this message.

Checked by AVG - www.avg.com<http://www.avg.com>

winmail.dat

cte...@comcast.net

unread,
Apr 12, 2011, 10:04:19 AM4/12/11
to Tim Crowson, ro...@casema.nl, soft...@listproc.autodesk.com
Hear hear!

Christopher Tedin
Creative Director
Dahlstrom


----- Reply message -----
From: "Tim Crowson" <t...@magneticdreams.com>
Date: Tue, Apr 12, 2011 8:03 am
Subject: Ice Tutorials
To: <ro...@casema.nl>, <soft...@listproc.autodesk.com>

I'm of the opinion of that although ICE has that word "creative" in it, it really is for TDs, and not very artist-friendly (at least not without the help of a TD first!). Paul's videos make things so easy to understand. I wish there was more content like that out there!

+1 for Pooby!


Tim Crowson   
Asset Artist

http://www.magneticdreams.com/md_logo.gif


Magnetic Dreams Animation Studio, Inc.
2525 Lebanon Pike, Building C. Nashville, TN 37214
Ph  615.885.6801 | Fx  615.889.4768 | www.magneticdreams.com



From: "Rob Wuijster" <ro...@casema.nl>
Sent: Tuesday, April 12, 2011 2:23 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials


Yes,

Paul's ICE videos are excellent, and his way of explaining things make ICE less 'complicated'...

Not sure if this was posted before, but the direct link is http://vimeo.com/user4895541/videos if people are interested.
Rob

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

On 4/12/2011 07:36 AM, Sam J. Bowling wrote:
I recently ran across some great ICE tutorials from Paul Smith. The "Zombie outbreak" set are great and very well done and thinking of them as Zombies and Humans instead of particles seems to really help keep me interested.  I can already think of a lot of uses for the things he covers in this set and it's all a lot easier than I would have imagined.




--
Sam J. Bowling



-----
No virus found in this message.
Checked by AVG - www.avg.com

Tim Crowson

unread,
Apr 12, 2011, 10:18:06 AM4/12/11
to soft...@listproc.autodesk.com
Ah sorry! I didn't mean to imply that! Actually...it's quite the opposite...it takes a very creative mind to come up with some of the ICE solutions I've seen. I'm just jealous!

-Tim
--

 



 

Bradley Gabe

unread,
Apr 12, 2011, 10:45:27 AM4/12/11
to soft...@listproc.autodesk.com
I'd reply to that with some kind of phrasing of protest. But unfortunately I can't think of what to say.

Alan Fregtman

unread,
Apr 12, 2011, 11:08:56 AM4/12/11
to soft...@listproc.autodesk.com
Perhaps we can string some ICE nodes together to compound a retort? :p

Guillaume Laforge

unread,
Apr 12, 2011, 11:10:36 AM4/12/11
to soft...@listproc.autodesk.com
No problem :), it was just too easy to be sarcastic on this one and I couldn't resist ;).

Cheers !

Guillaume

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Tim Crowson
Sent: Tuesday, April 12, 2011 10:18 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

Ah sorry! I didn't mean to imply that! Actually...it's quite the opposite...it takes a very creative mind to come up with some of the ICE solutions I've seen. I'm just jealous!

-Tim

On 4/12/2011 8:08 AM, Guillaume Laforge wrote:

You meant that TD's can't be creative ?

;)

Guillaume

Sent: Tuesday, April 12, 2011 9:04 AM

To: ro...@casema.nl<mailto:ro...@casema.nl>; soft...@listproc.autodesk.com<mailto:soft...@listproc.autodesk.com>

Subject: Re: Ice Tutorials

I'm of the opinion of that although ICE has that word "creative" in it, it really is for TDs, and not very artist-friendly (at least not without the help of a TD first!). Paul's videos make things so easy to understand. I wish there was more content like that out there!

+1 for Pooby!

Tim Crowson

Asset Artist

[http://www.magneticdreams.com/md_logo.gif]

Magnetic Dreams Animation Studio, Inc.

2525 Lebanon Pike, Building C. Nashville, TN 37214

Ph 615.885.6801 | Fx 615.889.4768 | www.magneticdreams.com<http://www.magneticdreams.com><http://www.magneticdreams.com/><http://www.magneticdreams.com/>

t...@magneticdreams.com<mailto:t...@magneticdreams.com><http://mail.magneticdreams.com/Main/t...@magneticdreams.com><http://mail.magneticdreams.com/Main/t...@magneticdreams.com>

________________________________

From: "Rob Wuijster" <ro...@casema.nl><mailto:ro...@casema.nl>

Sent: Tuesday, April 12, 2011 2:23 AM

To: soft...@listproc.autodesk.com<mailto:soft...@listproc.autodesk.com>

Subject: Re: Ice Tutorials

Yes,

Paul's ICE videos are excellent, and his way of explaining things make ICE less 'complicated'...

Not sure if this was posted before, but the direct link is http://vimeo.com/user4895541/videos if people are interested.

Rob

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

On 4/12/2011 07:36 AM, Sam J. Bowling wrote:

I recently ran across some great ICE tutorials from Paul Smith. The "Zombie outbreak" set are great and very well done and thinking of them as Zombies and Humans instead of particles seems to really help keep me interested. I can already think of a lot of uses for the things he covers in this set and it's all a lot easier than I would have imagined.

--

Sam J. Bowling

-----

No virus found in this message.

Checked by AVG - www.avg.com<http://www.avg.com><http://www.avg.com><http://www.avg.com>

winmail.dat

Eric Thivierge

unread,
Apr 12, 2011, 11:13:56 AM4/12/11
to soft...@listproc.autodesk.com
I see what you did there!

--------------------------------------------
Eric Thivierge
Technical Director
http://www.ethivierge.com

Christopher Tedin

unread,
Apr 12, 2011, 11:39:28 AM4/12/11
to soft...@listproc.autodesk.com
As Jim Gaffigan would say "That's a nerdy-ass joke!" ;-)

Sam J. Bowling

unread,
Apr 13, 2011, 12:21:44 AM4/13/11
to soft...@listproc.autodesk.com
I see this complaint about so many time, but there are tons of artistic things done with ICE. I see people complaining about 3d programs not being artist friendly, yet there are tons of artistic renderings produced on a daily basis. Are the people who create these not artists? I think what most of these people are trying to say it that it's not easy, but by that definition, most art work would not be artist friendly.

I would consider a hammer and a chisel and a slab of stone, or working with clay much less artist friendly. After all it you make one wrong cut with your chisel, you have to start over, or if you don't wrap you clay project up well enough it will dry out, but it I make a wrong change all I have to do is hit undo. Most artists didn't master their art over night, of even in a few weeks, and I bet if anyone put as much effort into learning ICE as they did into learning the other aspects of art, they would be able to use ICE as proficiently as they use any of their other tools.

--
Sam J. Bowling

Matt Lowery

unread,
Apr 13, 2011, 8:15:14 AM4/13/11
to soft...@listproc.autodesk.com
I here what you saying and for the most part I agree.
However with the exception of a few truely exceptional people, most
artist are not great logical thinkers. A logical mind is what's really
needed to understand the maths and programming logic in ICE, that's the
conundrum. To use your chisel and stone analogy, ICE is a bit like
showing a sculptor a fantastic new 'wonder' chisel that makes sculpting
way easier. But then instead of giving him the new 'wonder' chisel, you
give him the component parts of the machine that makes the 'wonder'
chisel and ask him to figure out how to put the machine together to make
the chisel....

m@

Sam J. Bowling wrote:
> I see this complaint about so many time, but there are tons of
> artistic things done with ICE. I see people complaining about 3d
> programs not being artist friendly, yet there are tons of artistic
> renderings produced on a daily basis. Are the people who create these
> not artists? I think what most of these people are trying to say it
> that it's not easy, but by that definition, most art work would not be
> artist friendly.
>
> I would consider a hammer and a chisel and a slab of stone, or working
> with clay much less artist friendly. After all it you make one wrong
> cut with your chisel, you have to start over, or if you don't wrap you
> clay project up well enough it will dry out, but it I make a wrong
> change all I have to do is hit undo. Most artists didn't master their
> art over night, of even in a few weeks, and I bet if anyone put as
> much effort into learning ICE as they did into learning the other
> aspects of art, they would be able to use ICE as proficiently as they
> use any of their other tools.
>
>
> At 06:03 AM 4/12/2011, you wrote:
>> I'm of the opinion of that although ICE has that word "creative" in
>> it, it really is for TDs, and not very artist-friendly (at least not
>> without the help of a TD first!). Paul's videos make things so easy
>> to understand. I wish there was more content like that out there!
>>
>> +1 for Pooby!
>>
>>

>> *Tim Crowson
>> */Asset Artist
>> /
>> http://www.magneticdreams.com/md_logo.gif
>> *


>> Magnetic Dreams Animation Studio, Inc.

>> *2525 Lebanon Pike, Building C. Nashville, TN 37214
>> *Ph* 615.885.6801 | *Fx* 615.889.4768 | www.magneticdreams.com

>> *From*: "Rob Wuijster" <ro...@casema.nl>
>> *Sent*: Tuesday, April 12, 2011 2:23 AM
>> *To*: soft...@listproc.autodesk.com
>> *Subject*: Re: Ice Tutorials


>>
>> Yes,
>>
>> Paul's ICE videos are excellent, and his way of explaining things
>> make ICE less 'complicated'...
>>
>> Not sure if this was posted before, but the direct link is
>> http://vimeo.com/user4895541/videos if people are interested.
>>
>> Rob
>>
>> \/-------------\/----------------\/
>>
>> On 4/12/2011 07:36 AM, Sam J. Bowling wrote:
>>> I recently ran across some great ICE tutorials from Paul Smith. The
>>> "Zombie outbreak" set are great and very well done and thinking of
>>> them as Zombies and Humans instead of particles seems to really help
>>> keep me interested. I can already think of a lot of uses for the
>>> things he covers in this set and it's all a lot easier than I would
>>> have imagined.
>>>
>>>
>>>
>>>
>>> --
>>> Sam J. Bowling
>>>
>>>
>>>
>>> -----
>>> No virus found in this message.

>>> Checked by AVG - www.avg.com <http://www.avg.com>

Olivier Jeannel

unread,
Apr 13, 2011, 9:05:01 AM4/13/11
to soft...@listproc.autodesk.com
Hello list :)

I'm a little stuck here :

I'm working on some cartoon art direction. I make the char & set
modeling, as well as the textures.
The company I'm working with uses Maya for animation and rendering.
I used to give them an XSI scene from which they export the geometry
into some Maya format.

The big problem I'm facing is that most of the set is made of tall
cartoonish grass.
I made that grass using the classic xsi hair. These are several textured
objects instanciated on the hair strand (Enable Instance Group).

Now I'm asked if it is possible to generate some polymesh from these hair ?
I've tried via AndyHayes FXnuts SaveRender mesh, but with no luck.
I don't know if is because the Addon doesn't work with hair objects, or
if it is because I'm on 2011.5

Does any one have a good idea of how I could tranform those hair
instancied objects into some solid geometry ?

Thank's a lot,

Olivier


Tim Crowson

unread,
Apr 13, 2011, 9:25:23 AM4/13/11
to soft...@listproc.autodesk.com
I really didn't mean to sound like I was complaining. I fully realize that the problem is more with me than with ICE, simply because my mind doesn't work "programmatically" the way ICE needs it to. And as I said in another reply, I fully agree that it takes very creative people to come up ICE solutions. There's no hating going on here at all. In fact, ICE continues to impress, and the latest modeling tools it offers in 2012 are going to suck me right in!

At the very least, compounds make ICE more approachable for idiots like me. When I open up a compound and see half a dozen arrays and a bunch of math nodes with noodles crisscrossing like a web, I can't help but feel a bit intimidated. But that's fully my problem and shortcoming. I realize that there are those who have full control over that sort of thing and have no real problem with it. And statistically, those people seem to be in the TD category.



Tim Crowson   
Asset Artist

http://www.magneticdreams.com/md_logo.gif

Magnetic Dreams Animation Studio, Inc.
2525 Lebanon Pike, Building C. Nashville, TN 37214



From: "Matt Lowery" <ma...@glassworks.co.uk>
Sent: Wednesday, April 13, 2011 7:17 AM

To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

Guillaume Laforge

unread,
Apr 13, 2011, 9:54:38 AM4/13/11
to soft...@listproc.autodesk.com
>. I realize that there are those who have full control over that sort of thing and have no real problem with it. And statistically, those people seem to be in the TD category.

Don't worry, I don't think humans are comfortable at all with those ICE spaghettis graph. In fact ICE programming is facing the same problem that classic coding. If you don't organize your graph using compounds with good names, comments etc..., it will be very difficult to understand how things work. The creator of those spaghettis compounds is not smarter than you. It is just that as ICE is an interactive tool, he can "code" and see the results at the same time. This is good in absolute but it is also very bad as most of the time the "ICE designer" will dive in its graph without planning things too much. When things start to give nice results, the compound will be saved on disk and when its creator will re-open it two months later, the only thing he will want to do is run very far from its workstation :).

I'm the first one doing this mistake on lot of compounds and I'm trying to improve this area. I'm thinking about writing something on how to plane and organize your work when doing ICE programming, but it is something that need time, so can't say when I will be able to do it.


Cheers,

Guillaume

Sent: Wednesday, April 13, 2011 9:25 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

I really didn't mean to sound like I was complaining. I fully realize that the problem is more with me than with ICE, simply because my mind doesn't work "programmatically" the way ICE needs it to. And as I said in another reply, I fully agree that it takes very creative people to come up ICE solutions. There's no hating going on here at all. In fact, ICE continues to impress, and the latest modeling tools it offers in 2012 are going to suck me right in!

At the very least, compounds make ICE more approachable for idiots like me. When I open up a compound and see half a dozen arrays and a bunch of math nodes with noodles crisscrossing like a web, I can't help but feel a bit intimidated. But that's fully my problem and shortcoming. I realize that there are those who have full control over that sort of thing and have no real problem with it. And statistically, those people seem to be in the TD category.


Tim Crowson
Asset Artist

[http://www.magneticdreams.com/md_logo.gif]


Magnetic Dreams Animation Studio, Inc.
2525 Lebanon Pike, Building C. Nashville, TN 37214

Ph 615.885.6801 | Fx 615.889.4768 | www.magneticdreams.com<http://www.magneticdreams.com/>

winmail.dat

Christopher Tedin

unread,
Apr 13, 2011, 10:30:05 AM4/13/11
to soft...@listproc.autodesk.com
Well, at the very least, it allows some generous folks to post their
compounds and lets others like me putz around with them, reverse
engineer them and use them for their evil plans to take over the world,
so that's good.

At the very best, it lets someone with a slightly logical mind, but no
memory for coding syntax (I've learned and forgotten Fortran, Cobol,
Basic, ActionScript, etc...) be able to work an idea out with some
simple math and some visual logic. At least, it's a lot more fun than
coding and you can comment out the parts to keep it all organized in
your head-bone. :-)

Thanks for your help with all of this, Guillaume. And, thanks to Pooby
and Thiago for their work as well trying to spread the word about ICE's
concepts.

-Chris

Stephen Davidson

unread,
Apr 13, 2011, 11:04:16 AM4/13/11
to soft...@listproc.autodesk.com
I am not a programmer, although I have played one in the past. :)
I have played with ICE to see if I could even figure out how to do simple tasks.
I have tried to follow "basic" ICE tutorials.
It seems I have been missing some very basic concepts.
It has to do with connecting inputs and outputs...
Just when I thought I have that simple task under my belt, I see
some output being connected in a way that I didn't expect.

I keep thinking there is some more basic level of understanding that I am missing.
I keep reading ICE posts, hoping that I will some day absorb some basic concepts.

Does anyone else, on this list feel this lost, or am I just giving everyone a good laugh.

I see the potential, I just can't seem to wrap my head around the basics.

Best Regards,
  Stephen P. Davidson 
       (954) 552-7956



Check My BLOG

My Website is GREEN, Is yours?

affiliate_link


Christopher Tedin

unread,
Apr 13, 2011, 11:28:44 AM4/13/11
to soft...@listproc.autodesk.com
The part that seems to be the hangup for me is "thinking backwards", that this node expects this input, and that comes from there, and you work backwards as well as forwards with your thinking, so there is no way to think in a linear fashion, that you think up and down the data streams to get your results. You can't just "visually" look at an ice graph and "get it", but you need to diagram it all out first to get the results you want. I think that's why they invented those plastic logic template things, so you can diagram out the flow of information:

http://www.cset.sp.utoledo.edu/sample/engt1050/engt1050_flowchart.html

Just so you feel better, I am having the same hangup. It's just getting your mind to work in this fashion that takes some time and practice. This comes more naturally to some, but the rest of us have to struggle to get through it. I think that's why Pooby's videos are so great. You can see someone like us, who admits he is not a programmer, make it work through just by wading through it.

Olivier Jeannel

unread,
Apr 13, 2011, 11:30:04 AM4/13/11
to soft...@listproc.autodesk.com
I'm just like you. I keep doing Ice projects other and other again. I found the idea of using Ice higly exiting. I mean setting up procedural parametric stuff that you can play with once operationnal is just so pleasing way of working compared to a bunch of keyframes everywhere that are a disaster to modify to the client desire.
I love all that turbulence noisy natural mouvements stuff. Enhance the quality of the particle animation a lot.

But hell, it takes time to setup when you start. I rarely start from scratch. Usually I'm starting with complex compound made by other experienced users and tear them appart until I find what I want. That's because the global ice logic involves (I think) a lot of different concepts.
I think I could't work without an internet access. There are so many questions to ask and so many tutorials to watch other again.

But, when it works, it is IMHO a new and different pleasure in our dayly job.


Olivier

Robert Chapman

unread,
Apr 13, 2011, 11:37:10 AM4/13/11
to soft...@listproc.autodesk.com
imho it helps to 'do' stuff rather than only read stuff. I'm no math / programming wizard either but by simply exposing myself to the process , doing as many tutorials' (Thiago's on cmivfx & Brad's on Vimeo helped the most), after a few weeks struggling it starts to sink in and becomes natural, its now very 'normal' , I can read others trees and get an understanding of whats going on and I couldn't approach any FX task without things through in an ICE way. yeah it may be a struggle initially but its well worth it.

Graham Bell

unread,
Apr 13, 2011, 11:38:05 AM4/13/11
to soft...@listproc.autodesk.com
Yep, I get what you're saying and this is something I'm hoping to try and do for my blog.
I want to try and go to a very low level in explaining ICE, as its often the only way Maya and especially Max guys understand it. ICE has been around for a while not, but I still come across people who don't, for want of a better term, 'get it'


From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Stephen Davidson
Sent: 13 April 2011 16:04
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

I am not a programmer, although I have played one in the past. :)
I have played with ICE to see if I could even figure out how to do simple tasks.
I have tried to follow "basic" ICE tutorials.
It seems I have been missing some very basic concepts.
It has to do with connecting inputs and outputs...
Just when I thought I have that simple task under my belt, I see
some output being connected in a way that I didn't expect.

I keep thinking there is some more basic level of understanding that I am missing.
I keep reading ICE posts, hoping that I will some day absorb some basic concepts.

Does anyone else, on this list feel this lost, or am I just giving everyone a good laugh.

I see the potential, I just can't seem to wrap my head around the basics.

On Wed, Apr 13, 2011 at 10:30 AM, Christopher Tedin <cte...@comcast.net<mailto:cte...@comcast.net>> wrote:
Well, at the very least, it allows some generous folks to post their compounds and lets others like me putz around with them, reverse engineer them and use them for their evil plans to take over the world, so that's good.

At the very best, it lets someone with a slightly logical mind, but no memory for coding syntax (I've learned and forgotten Fortran, Cobol, Basic, ActionScript, etc...) be able to work an idea out with some simple math and some visual logic. At least, it's a lot more fun than coding and you can comment out the parts to keep it all organized in your head-bone. :-)

Thanks for your help with all of this, Guillaume. And, thanks to Pooby and Thiago for their work as well trying to spread the word about ICE's concepts.

-Chris


On 4/13/11 8:54 AM, Guillaume Laforge wrote:
. I realize that there are those who have full control over that sort of thing and have no real problem with it. And statistically, those people seem to be in the TD category.
Don't worry, I don't think humans are comfortable at all with those ICE spaghettis graph. In fact ICE programming is facing the same problem that classic coding. If you don't organize your graph using compounds with good names, comments etc..., it will be very difficult to understand how things work. The creator of those spaghettis compounds is not smarter than you. It is just that as ICE is an interactive tool, he can "code" and see the results at the same time. This is good in absolute but it is also very bad as most of the time the "ICE designer" will dive in its graph without planning things too much. When things start to give nice results, the compound will be saved on disk and when its creator will re-open it two months later, the only thing he will want to do is run very far from its workstation :).

I'm the first one doing this mistake on lot of compounds and I'm trying to improve this area. I'm thinking about writing something on how to plane and organize your work when doing ICE programming, but it is something that need time, so can't say when I will be able to do it.


Cheers,

Guillaume

From: softimag...@listproc.autodesk.com<mailto:softimag...@listproc.autodesk.com> [mailto:softimag...@listproc.autodesk.com<mailto:softimag...@listproc.autodesk.com>] On Behalf Of Tim Crowson
Sent: Wednesday, April 13, 2011 9:25 AM
To: soft...@listproc.autodesk.com<mailto:soft...@listproc.autodesk.com>
Subject: Re: Ice Tutorials

I really didn't mean to sound like I was complaining. I fully realize that the problem is more with me than with ICE, simply because my mind doesn't work "programmatically" the way ICE needs it to. And as I said in another reply, I fully agree that it takes very creative people to come up ICE solutions. There's no hating going on here at all. In fact, ICE continues to impress, and the latest modeling tools it offers in 2012 are going to suck me right in!

At the very least, compounds make ICE more approachable for idiots like me. When I open up a compound and see half a dozen arrays and a bunch of math nodes with noodles crisscrossing like a web, I can't help but feel a bit intimidated. But that's fully my problem and shortcoming. I realize that there are those who have full control over that sort of thing and have no real problem with it. And statistically, those people seem to be in the TD category.


Tim Crowson
Asset Artist

[http://www.magneticdreams.com/md_logo.gif]


Magnetic Dreams Animation Studio, Inc.
2525 Lebanon Pike, Building C. Nashville, TN 37214

Ph 615.885.6801<tel:615.885.6801> | Fx 615.889.4768<tel:615.889.4768> | www.magneticdreams.com<http://www.magneticdreams.com><http://www.magneticdreams.com/>
t...@magneticdreams.com<mailto:t...@magneticdreams.com><http://mail.magneticdreams.com/Main/t...@magneticdreams.com>

________________________________
From: "Matt Lowery"<ma...@glassworks.co.uk<mailto:ma...@glassworks.co.uk>>


Sent: Wednesday, April 13, 2011 7:17 AM

To: soft...@listproc.autodesk.com<mailto:soft...@listproc.autodesk.com>
Subject: Re: Ice Tutorials

m@

+1 for Pooby!

*Ph* 615.885.6801<tel:615.885.6801> | *Fx* 615.889.4768<tel:615.889.4768> | www.magneticdreams.com<http://www.magneticdreams.com>
<http://www.magneticdreams.com/>
t...@magneticdreams.com<mailto:t...@magneticdreams.com>
<http://mail.magneticdreams.com/Main/t...@magneticdreams.com>


*From*: "Rob Wuijster"<ro...@casema.nl<mailto:ro...@casema.nl>>


*Sent*: Tuesday, April 12, 2011 2:23 AM

*To*: soft...@listproc.autodesk.com<mailto:soft...@listproc.autodesk.com>
*Subject*: Re: Ice Tutorials

Yes,

Paul's ICE videos are excellent, and his way of explaining things
make ICE less 'complicated'...

Not sure if this was posted before, but the direct link is
http://vimeo.com/user4895541/videos if people are interested.

Rob

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

On 4/12/2011 07:36 AM, Sam J. Bowling wrote:
I recently ran across some great ICE tutorials from Paul Smith. The
"Zombie outbreak" set are great and very well done and thinking of
them as Zombies and Humans instead of particles seems to really help
keep me interested. I can already think of a lot of uses for the
things he covers in this set and it's all a lot easier than I would
have imagined.


--
Sam J. Bowling

-----
No virus found in this message.

Checked by AVG - www.avg.com<http://www.avg.com><http://www.avg.com>


Version: 10.0.1209 / Virus Database: 1500/3565 - Release Date: 04/11/11
--
Sam J. Bowling

Best Regards,
Stephen P. Davidson
(954) 552-7956

[http://www.3danimationmagic.com/3Danimation_magic_logo_sign.jpg]

<http://www.3danimationmagic.com>Check My BLOG<http://3danimationmagic.blogspot.com>

My Website is GREEN, Is yours?

[affiliate_link]<http://www.fatcow.com/join/index.bml?AffID=626478>

winmail.dat

Christopher Tedin

unread,
Apr 13, 2011, 11:59:06 AM4/13/11
to soft...@listproc.autodesk.com
Oh, forgot to thank Bradley Gabe as well. Thanks all for your tutorials. Also, Aaron Kent has some very simple tutorials too, like this one called "Simple use of ICE". http://vimeo.com/19511854

Chris Johnson

unread,
Apr 13, 2011, 1:19:57 PM4/13/11
to soft...@listproc.autodesk.com
I'm very much in the same light....without the existing database of
tutorials and compunds I'd be lost. I usual have to take apart a couple
cars just to make one new car work....I keep waiting for that moment
when my head gets it....keep the examples and tutorials coming!

Andy Nicholas

unread,
Apr 13, 2011, 1:47:39 PM4/13/11
to soft...@listproc.autodesk.com
> I'm thinking about writing something on how to plane
> and organize your work

Yep, me too. I've been planning on writing a tutorial or do a video on it
for a while. There are some very simple rules that you can follow.

The most useful one I've found so far is to never share values between
separate branches. The temptation is to try to optimise by reusing the
result from one node multiple times. But it rarely results in any
optimisation, and usually results in misunderstandings. People think the
results are cached, but they're not, and the second time you evaluate it's
entirely possible that it'll give you a new value that is technically
correct, but which you won't be expecting.

The other thing that goes hand in hand with this tip, is to cache any
values that are going to be reused at the top of your tree in attributes.
That way you can reuse the results as much as you like.

If you follow both of these tips, you automatically optimise the tree, and
you also make it more readable with fewer nodes and no spaghetti.

Admittedly, it's not particularly clear writing this down, but I'll try to
do a blog post on this soon.

As far as learning about ICE goes, it's just about maths and logic. Most
people are logical, but it's making that work properly with the numbers
that can be tricky. Getting a good feel for numbers is essential, and that
only comes through practice.

To extend the analogy already mentioned (and sound highly pretentious at
the same time): if stone is the medium used by sculptors, then numbers are
the medium of an ICE artist ;-)

Cheers!
Andy


Grahame Fuller

unread,
Apr 13, 2011, 2:19:48 PM4/13/11
to soft...@listproc.autodesk.com
I experimented with a technique for organizing and documenting trees in this tutorial: http://softimage.wiki.softimage.com/index.php/Water_Ripples_in_ICE

Instead of comments and group comments, I relied mostly on compounds so that I could name them and their ports -- even wrapping a single node in a compound if appropriate. I think it worked quite well and the tree was very readable -- it was originally something I shared on the v7 beta list but I didn't revisit that scene to make the tutorial until a long time later. Unfortunately I haven't kept up that habit as well as I should have.

gray

winmail.dat

Matt Lind

unread,
Apr 13, 2011, 2:28:36 PM4/13/11
to soft...@listproc.autodesk.com

Think of a compound as a functions or a class in a programming language.  If you’re organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality.

 

Matt

 

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
Sent: Wednesday, April 13, 2011 11:26 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

 

I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night.  He does the same thing.  When something gets to a certain size, put it in a compound & expose & name the ports.  Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

 

-Paul

Guillaume Laforge

unread,
Apr 13, 2011, 2:27:58 PM4/13/11
to soft...@listproc.autodesk.com
Completely agree with you Andy,

Except that I don't think that is it math who is hard to understand. The difficult part with ICE is to understand contexts.
All is context dependant, and it can be tricky to deal with at the beginning. For example in 2012 with have a node to convert a set of point positions to an array of positions. For new users, it looks like it is the same thing ! I think this level of abstraction is more complicate to understand than doing math in ICE (that can be simple or complicate like in every app using math).

Cheers

Guillaume

-----Original Message-----
From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Andy Nicholas
Sent: Wednesday, April 13, 2011 1:48 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

winmail.dat

Paul Griswold

unread,
Apr 13, 2011, 2:25:33 PM4/13/11
to soft...@listproc.autodesk.com
I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night.  He does the same thing.  When something gets to a certain size, put it in a compound & expose & name the ports.  Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

-Paul
On Wed, Apr 13, 2011 at 2:19 PM, Grahame Fuller <Grahame...@autodesk.com> wrote:

Guillaume Laforge

unread,
Apr 13, 2011, 2:35:59 PM4/13/11
to soft...@listproc.autodesk.com
>Think of a compound as a functions or a class in a programming language

A compound can just be used as a #define too ;).
For example using a compound named "Invalid Index" instead of an integer of value "-1" can help understanding a graph more quickly.

Guillaume

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 2:29 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

Think of a compound as a functions or a class in a programming language. If you're organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality.

Matt


From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
Sent: Wednesday, April 13, 2011 11:26 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night. He does the same thing. When something gets to a certain size, put it in a compound & expose & name the ports. Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

-Paul

winmail.dat

Grahame Fuller

unread,
Apr 13, 2011, 2:36:16 PM4/13/11
to soft...@listproc.autodesk.com
My point is that they can also be used to document a tree. For example, taking a simple Scalar node, turning it into a compound, and renaming it to "Radius" can greatly improve the readability - much better than comments in most cases.

gray

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 02:29 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

Think of a compound as a functions or a class in a programming language. If you're organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality.

Matt


From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
Sent: Wednesday, April 13, 2011 11:26 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night. He does the same thing. When something gets to a certain size, put it in a compound & expose & name the ports. Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

-Paul

winmail.dat

Matt Lind

unread,
Apr 13, 2011, 2:52:52 PM4/13/11
to soft...@listproc.autodesk.com

IMHO – I don’t agree with that approach to compounds.  It sets the wheels in motion for more problems downstream with lesser attentive users.

 

ICE Editor should be stressing ‘form follows function’.  When you use compounds for naming or providing comments and nothing else, it confuses the issue and lesser experienced users.  They’ll assume any compound is something they can cut n’ paste into another ICETree and expect it to do something different.  In the case of a ‘radius’ compound, if it only consists of a single node of a different name, then you’re just adding bloat to the ICEtree with not much gained.  Over time as these compounds are accumulated and mixed n’ match over many ICEtrees, you get a big mess that somebody, like me, has to dive in one day and untangle the mess because something exploded and nobody knows what went wrong or how to fix it.  No thank you.  Been there, done that too many times already.  I would strongly stress better attention to organizing via function as a good ounce of prevention.  A pound of cure isn’t always available.

 

Form follows function.  If anything, I’d rework the comment system to be a little better.  The group comment node should have the ability to associate itself with specific node(s) in the ICETree so when the group comment is moved around, the nodes travel along with it.  If a node inside the group comment is dragged around the editor, then the group comment node should resize itself to keep it contained within.  That’s my one gripe with the group comment node right now, click “rearrange” in the ICE tree editor and all the nodes associated with the group comment go elsewhere making the comment useless.  For that reason alone I tend to use the built-in comment of each node to keep track of what’s happening.  In some ways the group comment node should be packagable like a compound, but without hiding the internal details.

 

Matt

 

 

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Grahame Fuller
Sent: Wednesday, April 13, 2011 11:36 AM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

My point is that they can also be used to document a tree. For example, taking a simple Scalar node, turning it into a compound, and renaming it to “Radius” can greatly improve the readability – much better than comments in most cases.

 

gray

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 02:29 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

Think of a compound as a functions or a class in a programming language.  If you’re organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality.

 

Matt

 

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
Sent: Wednesday, April 13, 2011 11:26 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

 

I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night.  He does the same thing.  When something gets to a certain size, put it in a compound & expose & name the ports.  Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

 

-Paul

Guillaume Laforge

unread,
Apr 13, 2011, 3:06:27 PM4/13/11
to soft...@listproc.autodesk.com
> IMHO - I don't agree with that approach to compounds. It sets the wheels in motion for more problems downstream with lesser attentive users.

Using compounds as a constant is the style that we promote here . For example, take a look in the preset manager under Constant and you will find a compound named "Pi"...

But style is a personal choice, so use any way you like ;).

Guillaume

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 2:53 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

IMHO - I don't agree with that approach to compounds. It sets the wheels in motion for more problems downstream with lesser attentive users.

ICE Editor should be stressing 'form follows function'. When you use compounds for naming or providing comments and nothing else, it confuses the issue and lesser experienced users. They'll assume any compound is something they can cut n' paste into another ICETree and expect it to do something different. In the case of a 'radius' compound, if it only consists of a single node of a different name, then you're just adding bloat to the ICEtree with not much gained. Over time as these compounds are accumulated and mixed n' match over many ICEtrees, you get a big mess that somebody, like me, has to dive in one day and untangle the mess because something exploded and nobody knows what went wrong or how to fix it. No thank you. Been there, done that too many times already. I would strongly stress better attention to organizing via function as a good ounce of prevention. A pound of cure isn't always available.

Form follows function. If anything, I'd rework the comment system to be a little better. The group comment node should have the ability to associate itself with specific node(s) in the ICETree so when the group comment is moved around, the nodes travel along with it. If a node inside the group comment is dragged around the editor, then the group comment node should resize itself to keep it contained within. That's my one gripe with the group comment node right now, click "rearrange" in the ICE tree editor and all the nodes associated with the group comment go elsewhere making the comment useless. For that reason alone I tend to use the built-in comment of each node to keep track of what's happening. In some ways the group comment node should be packagable like a compound, but without hiding the internal details.

Matt

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Grahame Fuller
Sent: Wednesday, April 13, 2011 11:36 AM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

My point is that they can also be used to document a tree. For example, taking a simple Scalar node, turning it into a compound, and renaming it to "Radius" can greatly improve the readability - much better than comments in most cases.

gray

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 02:29 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

Think of a compound as a functions or a class in a programming language. If you're organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality.

Matt


From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
Sent: Wednesday, April 13, 2011 11:26 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night. He does the same thing. When something gets to a certain size, put it in a compound & expose & name the ports. Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

-Paul

winmail.dat

Matt Lind

unread,
Apr 13, 2011, 4:08:31 PM4/13/11
to soft...@listproc.autodesk.com

Why don’t you just create a node with a comment attached?  Doesn’t creating compounds add overhead?

 

 

Matt

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Guillaume Laforge
Sent: Wednesday, April 13, 2011 12:06 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

> IMHO – I don’t agree with that approach to compounds.  It sets the wheels in motion for more problems downstream with lesser attentive users.

 

Using  compounds as a constant is the style that we promote here . For example, take a look in the preset manager under Constant and you will find a compound named “Pi”…

 

But style is a personal choice, so use any way you like ;).

 

Guillaume

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 2:53 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

IMHO – I don’t agree with that approach to compounds.  It sets the wheels in motion for more problems downstream with lesser attentive users.

 

ICE Editor should be stressing ‘form follows function’.  When you use compounds for naming or providing comments and nothing else, it confuses the issue and lesser experienced users.  They’ll assume any compound is something they can cut n’ paste into another ICETree and expect it to do something different.  In the case of a ‘radius’ compound, if it only consists of a single node of a different name, then you’re just adding bloat to the ICEtree with not much gained.  Over time as these compounds are accumulated and mixed n’ match over many ICEtrees, you get a big mess that somebody, like me, has to dive in one day and untangle the mess because something exploded and nobody knows what went wrong or how to fix it.  No thank you.  Been there, done that too many times already.  I would strongly stress better attention to organizing via function as a good ounce of prevention.  A pound of cure isn’t always available.

 

Form follows function.  If anything, I’d rework the comment system to be a little better.  The group comment node should have the ability to associate itself with specific node(s) in the ICETree so when the group comment is moved around, the nodes travel along with it.  If a node inside the group comment is dragged around the editor, then the group comment node should resize itself to keep it contained within.  That’s my one gripe with the group comment node right now, click “rearrange” in the ICE tree editor and all the nodes associated with the group comment go elsewhere making the comment useless.  For that reason alone I tend to use the built-in comment of each node to keep track of what’s happening.  In some ways the group comment node should be packagable like a compound, but without hiding the internal details.

 

Matt

 

 

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Grahame Fuller
Sent: Wednesday, April 13, 2011 11:36 AM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

My point is that they can also be used to document a tree. For example, taking a simple Scalar node, turning it into a compound, and renaming it to “Radius” can greatly improve the readability – much better than comments in most cases.

 

gray

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 02:29 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

Think of a compound as a functions or a class in a programming language.  If you’re organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality.

 

Matt

 

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
Sent: Wednesday, April 13, 2011 11:26 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

 

I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night.  He does the same thing.  When something gets to a certain size, put it in a compound & expose & name the ports.  Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

 

-Paul

Grahame Fuller

unread,
Apr 13, 2011, 4:29:11 PM4/13/11
to soft...@listproc.autodesk.com
There's a tiny overhead to the pre-evaluation phase but in most cases the effect would be swamped by the evaluation itself. I doubt it would be noticeable, even if you ran timer tests over and over.

To my mind the benefit of readability far outweighs any performance hit. For example,

[cid:image0...@01CBF9F7.A643A340] versus [cid:image0...@01CBF9F7.EB4977D0]

gray

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 04:09 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

Why don't you just create a node with a comment attached? Doesn't creating compounds add overhead?


Matt

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Guillaume Laforge
Sent: Wednesday, April 13, 2011 12:06 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

> IMHO - I don't agree with that approach to compounds. It sets the wheels in motion for more problems downstream with lesser attentive users.

Using compounds as a constant is the style that we promote here . For example, take a look in the preset manager under Constant and you will find a compound named "Pi"...

But style is a personal choice, so use any way you like ;).

Guillaume

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 2:53 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

IMHO - I don't agree with that approach to compounds. It sets the wheels in motion for more problems downstream with lesser attentive users.

ICE Editor should be stressing 'form follows function'. When you use compounds for naming or providing comments and nothing else, it confuses the issue and lesser experienced users. They'll assume any compound is something they can cut n' paste into another ICETree and expect it to do something different. In the case of a 'radius' compound, if it only consists of a single node of a different name, then you're just adding bloat to the ICEtree with not much gained. Over time as these compounds are accumulated and mixed n' match over many ICEtrees, you get a big mess that somebody, like me, has to dive in one day and untangle the mess because something exploded and nobody knows what went wrong or how to fix it. No thank you. Been there, done that too many times already. I would strongly stress better attention to organizing via function as a good ounce of prevention. A pound of cure isn't always available.

Form follows function. If anything, I'd rework the comment system to be a little better. The group comment node should have the ability to associate itself with specific node(s) in the ICETree so when the group comment is moved around, the nodes travel along with it. If a node inside the group comment is dragged around the editor, then the group comment node should resize itself to keep it contained within. That's my one gripe with the group comment node right now, click "rearrange" in the ICE tree editor and all the nodes associated with the group comment go elsewhere making the comment useless. For that reason alone I tend to use the built-in comment of each node to keep track of what's happening. In some ways the group comment node should be packagable like a compound, but without hiding the internal details.

Matt

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Grahame Fuller
Sent: Wednesday, April 13, 2011 11:36 AM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

My point is that they can also be used to document a tree. For example, taking a simple Scalar node, turning it into a compound, and renaming it to "Radius" can greatly improve the readability - much better than comments in most cases.

gray

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 02:29 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

Think of a compound as a functions or a class in a programming language. If you're organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality.

Matt


From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
Sent: Wednesday, April 13, 2011 11:26 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night. He does the same thing. When something gets to a certain size, put it in a compound & expose & name the ports. Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

-Paul

image001.png
image002.png

Guillaume Laforge

unread,
Apr 13, 2011, 4:39:54 PM4/13/11
to soft...@listproc.autodesk.com
It is maybe an overhead once stored on disk (two nodes instead of one). But really it should not be the reason to not use compounds !
This overhead is so negligible compared to the visual comfort of using nodes instead of comments.

It is exactly like reading a code like this :

Case A (like using comments):

// This value is there to describe the area of my triangle
float t

Case B (like using compounds):

float triangleArea

So compound are the way to go to describe a constant. Comments nodes are there to add ...some useful comments.

Cheers,

Guillaume

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 4:09 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

Why don't you just create a node with a comment attached? Doesn't creating compounds add overhead?


Matt

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Guillaume Laforge
Sent: Wednesday, April 13, 2011 12:06 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

> IMHO - I don't agree with that approach to compounds. It sets the wheels in motion for more problems downstream with lesser attentive users.

Using compounds as a constant is the style that we promote here . For example, take a look in the preset manager under Constant and you will find a compound named "Pi"...

But style is a personal choice, so use any way you like ;).

Guillaume

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 2:53 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

IMHO - I don't agree with that approach to compounds. It sets the wheels in motion for more problems downstream with lesser attentive users.

ICE Editor should be stressing 'form follows function'. When you use compounds for naming or providing comments and nothing else, it confuses the issue and lesser experienced users. They'll assume any compound is something they can cut n' paste into another ICETree and expect it to do something different. In the case of a 'radius' compound, if it only consists of a single node of a different name, then you're just adding bloat to the ICEtree with not much gained. Over time as these compounds are accumulated and mixed n' match over many ICEtrees, you get a big mess that somebody, like me, has to dive in one day and untangle the mess because something exploded and nobody knows what went wrong or how to fix it. No thank you. Been there, done that too many times already. I would strongly stress better attention to organizing via function as a good ounce of prevention. A pound of cure isn't always available.

Form follows function. If anything, I'd rework the comment system to be a little better. The group comment node should have the ability to associate itself with specific node(s) in the ICETree so when the group comment is moved around, the nodes travel along with it. If a node inside the group comment is dragged around the editor, then the group comment node should resize itself to keep it contained within. That's my one gripe with the group comment node right now, click "rearrange" in the ICE tree editor and all the nodes associated with the group comment go elsewhere making the comment useless. For that reason alone I tend to use the built-in comment of each node to keep track of what's happening. In some ways the group comment node should be packagable like a compound, but without hiding the internal details.

Matt

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Grahame Fuller
Sent: Wednesday, April 13, 2011 11:36 AM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

My point is that they can also be used to document a tree. For example, taking a simple Scalar node, turning it into a compound, and renaming it to "Radius" can greatly improve the readability - much better than comments in most cases.

gray

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 02:29 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

Think of a compound as a functions or a class in a programming language. If you're organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality.

Matt


From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
Sent: Wednesday, April 13, 2011 11:26 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night. He does the same thing. When something gets to a certain size, put it in a compound & expose & name the ports. Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

-Paul

winmail.dat

Matt Lind

unread,
Apr 13, 2011, 5:15:39 PM4/13/11
to soft...@listproc.autodesk.com

It’s subjective, not objective.  I still don’t agree with compounds being any better for this particular niche task.

 

The issue I’ve seen in the past is users have multiple compounds and names to represent the same thing.  Over time it confuses people trying to read the graph because one thing is referenced by a gazillion different names.  One day things go boom and it’s a real mess trying to figure it out and clean it up - And it always happens at the most stressful time before a deadline.

 

In practice it’s better to force a spade to be called a spade.

 

 

Matt

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Guillaume Laforge
Sent: Wednesday, April 13, 2011 1:40 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

It is maybe an overhead once stored on disk (two nodes instead of one). But really it should not be the reason to not use compounds !

This overhead is so negligible compared to the visual comfort of using nodes instead of comments.

 

It is exactly like reading a code like this :

 

Case A (like using comments):

 

// This value is there to describe the area of my triangle

float t

 

Case B (like using compounds):

 

float triangleArea

 

So compound are the way to go to describe a constant. Comments nodes are there to add …some useful comments.

 

Cheers,

 

Guillaume

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 4:09 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

Why don’t you just create a node with a comment attached?  Doesn’t creating compounds add overhead?

 

 

Matt

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Guillaume Laforge
Sent: Wednesday, April 13, 2011 12:06 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

> IMHO – I don’t agree with that approach to compounds.  It sets the wheels in motion for more problems downstream with lesser attentive users.

 

Using  compounds as a constant is the style that we promote here . For example, take a look in the preset manager under Constant and you will find a compound named “Pi”…

 

But style is a personal choice, so use any way you like ;).

 

Guillaume

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 2:53 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

IMHO – I don’t agree with that approach to compounds.  It sets the wheels in motion for more problems downstream with lesser attentive users.

 

ICE Editor should be stressing ‘form follows function’.  When you use compounds for naming or providing comments and nothing else, it confuses the issue and lesser experienced users.  They’ll assume any compound is something they can cut n’ paste into another ICETree and expect it to do something different.  In the case of a ‘radius’ compound, if it only consists of a single node of a different name, then you’re just adding bloat to the ICEtree with not much gained.  Over time as these compounds are accumulated and mixed n’ match over many ICEtrees, you get a big mess that somebody, like me, has to dive in one day and untangle the mess because something exploded and nobody knows what went wrong or how to fix it.  No thank you.  Been there, done that too many times already.  I would strongly stress better attention to organizing via function as a good ounce of prevention.  A pound of cure isn’t always available.

 

Form follows function.  If anything, I’d rework the comment system to be a little better.  The group comment node should have the ability to associate itself with specific node(s) in the ICETree so when the group comment is moved around, the nodes travel along with it.  If a node inside the group comment is dragged around the editor, then the group comment node should resize itself to keep it contained within.  That’s my one gripe with the group comment node right now, click “rearrange” in the ICE tree editor and all the nodes associated with the group comment go elsewhere making the comment useless.  For that reason alone I tend to use the built-in comment of each node to keep track of what’s happening.  In some ways the group comment node should be packagable like a compound, but without hiding the internal details.

 

Matt

 

 

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Grahame Fuller
Sent: Wednesday, April 13, 2011 11:36 AM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

My point is that they can also be used to document a tree. For example, taking a simple Scalar node, turning it into a compound, and renaming it to “Radius” can greatly improve the readability – much better than comments in most cases.

 

gray

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Wednesday, April 13, 2011 02:29 PM
To: soft...@listproc.autodesk.com
Subject: RE: Ice Tutorials

 

Think of a compound as a functions or a class in a programming language.  If you’re organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality.

 

Matt

 

 

 

 

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
Sent: Wednesday, April 13, 2011 11:26 AM
To: soft...@listproc.autodesk.com
Subject: Re: Ice Tutorials

 

I picked up Thiago's scalar tutorials on CMIVFX and was watching them the other night.  He does the same thing.  When something gets to a certain size, put it in a compound & expose & name the ports.  Until then I really hadn't thought about compounds as a way to organize ICE, but it really makes a lot of sense.

 

-Paul

Steven Caron

unread,
Apr 13, 2011, 5:22:40 PM4/13/11
to soft...@listproc.autodesk.com
i am not seeing the difference between what grahame and guillaume are suggestion and what you mentioned earlier...


"Think of a compound as a functions or a class in a programming language.  If you're organized, the compound will be a generic task that can be reused in any ICETree that calls for the functionality."

s

Matt Lind

unread,
Apr 13, 2011, 5:42:41 PM4/13/11
to soft...@listproc.autodesk.com

We’re discussing a case of a compound consisting of a single node.  What’s the point of that?

Steven Caron

unread,
Apr 13, 2011, 5:52:54 PM4/13/11
to soft...@listproc.autodesk.com
ok i see.. subjective yes. you prefer comments attached to nodes, they prefer wrapping nodes with compounds

s

Matt Lind

unread,
Apr 13, 2011, 6:00:39 PM4/13/11
to soft...@listproc.autodesk.com

I prefer a system where the labeling doesn’t obfuscate the tree.

Guillaume Laforge

unread,
Apr 13, 2011, 6:03:54 PM4/13/11
to soft...@listproc.autodesk.com
I prefer a system where the labeling doesn’t obfuscate the tree.

That's why I tend to avoid comments :)


//this is my name
GL

Thomas Helzle

unread,
Apr 13, 2011, 7:11:54 PM4/13/11
to soft...@listproc.autodesk.com
Well, IMO this discussion rather points to a general weakness of nodes in XSI:

- The current node comments get in the way, are flimsy and tedious to use.
- One-node-compounds go against my grain as well as creating compounds
just for naming.
- Rendernodes can be renamed, but then you no longer know what kind they are...

I personally would love to be able to give every node a name directly
with F2 and STILL be able to see what kind of node it is. So I'd give
every node a name-property that can be set with F2 and that would be
bold and on top. Directly below that in normal print I'd put the node
description.
I didn't test it in a mockup, but that would be the first direction
I'd look into.

The current comments could maybe get a "Flyout-Flag" so they are only
visible when you go over the node with the mouse (there's a comment
icon on commented nodes already, so you'd know it's there) or could
even just be visible when the node is double clicked inside the
standard ppg - scratch the separate window for comments AFAIC.

This combined with group comments which "know" what they contain and
grow and shrink and move with that content would make creating
readable trees much easier IMO.

Cheers,

Thomas


On 14 April 2011 00:03, Guillaume Laforge
<guillaume....@gmail.com> wrote:
>>�I prefer a system where the labeling doesn�t obfuscate the tree.


> That's why I tend to avoid comments :)
>
> //this is my name
> GL
>
>
> On Wed, Apr 13, 2011 at 6:00 PM, Matt Lind <ml...@carbinestudios.com> wrote:
>>

>> I prefer a system where the labeling doesn�t obfuscate the tree.


>>
>>
>>
>>
>>
>> Matt
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> From: softimag...@listproc.autodesk.com
>> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Steven Caron
>>
>> Sent: Wednesday, April 13, 2011 2:53 PM
>> To: soft...@listproc.autodesk.com
>> Subject: Re: Ice Tutorials
>>
>>
>>
>> ok i see.. subjective yes. you prefer comments attached to nodes, they
>> prefer wrapping nodes with compounds
>>
>> s
>>
>> On Wed, Apr 13, 2011 at 2:42 PM, Matt Lind <ml...@carbinestudios.com>
>> wrote:
>>

>> We�re discussing a case of a compound consisting of a single node.� What�s


>> the point of that?
>>
>>
>>
>>
>>
>> Matt
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> From: softimag...@listproc.autodesk.com
>> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Steven Caron
>> Sent: Wednesday, April 13, 2011 2:23 PM
>>
>> To: soft...@listproc.autodesk.com
>>
>> Subject: Re: Ice Tutorials
>>
>>
>>
>> i am not seeing the difference between what grahame and guillaume are
>> suggestion and what you mentioned earlier...
>>
>> "Think of a compound as a functions or a class in a programming language.

>> �If you're organized, the compound will be a generic task that can be reused


>> in any ICETree that calls for the functionality."
>>
>> s
>>
>> On Wed, Apr 13, 2011 at 2:15 PM, Matt Lind <ml...@carbinestudios.com>
>> wrote:
>>

>> It�s subjective, not objective.� I still don�t agree with compounds being


>> any better for this particular niche task.
>>
>>
>>

>> The issue I�ve seen in the past is users have multiple compounds and names
>> to represent the same thing.� Over time it confuses people trying to read


>> the graph because one thing is referenced by a gazillion different names.

>> One day things go boom and it�s a real mess trying to figure it out and


>> clean it up - And it always happens at the most stressful time before a
>> deadline.
>>
>>
>>

>> In practice it�s better to force a spade to be called a spade.


>>
>>
>>
>>
>>
>> Matt
>>
>>
>>
>>
>>
>> From: softimag...@listproc.autodesk.com
>> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Guillaume
>> Laforge
>> Sent: Wednesday, April 13, 2011 1:40 PM
>>
>> To: soft...@listproc.autodesk.com
>> Subject: RE: Ice Tutorials
>>
>>
>>
>> It is maybe an overhead once stored on disk (two nodes instead of one).
>> But really it should not be the reason to not use compounds !
>>
>> This overhead is so negligible compared to the visual comfort of using
>> nodes instead of comments.
>>
>>
>>
>> It is exactly like reading a code like this :
>>
>>
>>
>> Case A (like using comments):
>>
>>
>>
>> // This value is there to describe the area of my triangle
>>
>> float t
>>
>>
>>
>> Case B (like using compounds):
>>
>>
>>
>> float triangleArea
>>
>>
>>
>> So compound are the way to go to describe a constant. Comments nodes are

>> there to add �some useful comments.


>>
>>
>>
>> Cheers,
>>
>>
>>
>> Guillaume
>>
>>
>>
>>
>>
>>
>>
>> From: softimag...@listproc.autodesk.com
>> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
>> Sent: Wednesday, April 13, 2011 4:09 PM
>> To: soft...@listproc.autodesk.com
>> Subject: RE: Ice Tutorials
>>
>>
>>

>> Why don�t you just create a node with a comment attached?� Doesn�t


>> creating compounds add overhead?
>>
>>
>>
>>
>>
>> Matt
>>
>>
>>
>>
>>
>>
>>
>> From: softimag...@listproc.autodesk.com
>> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Guillaume
>> Laforge
>> Sent: Wednesday, April 13, 2011 12:06 PM
>> To: soft...@listproc.autodesk.com
>> Subject: RE: Ice Tutorials
>>
>>
>>

>> > IMHO � I don�t agree with that approach to compounds.� It sets the


>> > wheels in motion for more problems downstream with lesser attentive users.
>>
>>
>>

>> Using �compounds as a constant is the style that we promote here . For


>> example, take a look in the preset manager under Constant and you will find

>> a compound named �Pi��


>>
>>
>>
>> But style is a personal choice, so use any way you like ;).
>>
>>
>>
>> Guillaume
>>
>>
>>
>> From: softimag...@listproc.autodesk.com
>> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
>> Sent: Wednesday, April 13, 2011 2:53 PM
>> To: soft...@listproc.autodesk.com
>> Subject: RE: Ice Tutorials
>>
>>
>>

>> IMHO � I don�t agree with that approach to compounds.� It sets the wheels


>> in motion for more problems downstream with lesser attentive users.
>>
>>
>>

>> ICE Editor should be stressing �form follows function�.� When you use


>> compounds for naming or providing comments and nothing else, it confuses the

>> issue and lesser experienced users.� They�ll assume any compound is
>> something they can cut n� paste into another ICETree and expect it to do
>> something different.� In the case of a �radius� compound, if it only
>> consists of a single node of a different name, then you�re just adding bloat
>> to the ICEtree with not much gained.� Over time as these compounds are
>> accumulated and mixed n� match over many ICEtrees, you get a big mess that


>> somebody, like me, has to dive in one day and untangle the mess because

>> something exploded and nobody knows what went wrong or how to fix it.� No
>> thank you.� Been there, done that too many times already.� I would strongly


>> stress better attention to organizing via function as a good ounce of

>> prevention.� A pound of cure isn�t always available.
>>
>>
>>
>> Form follows function.� If anything, I�d rework the comment system to be a
>> little better.� The group comment node should have the ability to associate


>> itself with specific node(s) in the ICETree so when the group comment is

>> moved around, the nodes travel along with it.� If a node inside the group


>> comment is dragged around the editor, then the group comment node should

>> resize itself to keep it contained within.� That�s my one gripe with the
>> group comment node right now, click �rearrange� in the ICE tree editor and


>> all the nodes associated with the group comment go elsewhere making the

>> comment useless.� For that reason alone I tend to use the built-in comment
>> of each node to keep track of what�s happening.� In some ways the group


>> comment node should be packagable like a compound, but without hiding the
>> internal details.
>>
>>
>>
>> Matt
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> From: softimag...@listproc.autodesk.com
>> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Grahame Fuller
>> Sent: Wednesday, April 13, 2011 11:36 AM
>> To: soft...@listproc.autodesk.com
>> Subject: RE: Ice Tutorials
>>
>>
>>
>> My point is that they can also be used to document a tree. For example,
>> taking a simple Scalar node, turning it into a compound, and renaming it to

>> �Radius� can greatly improve the readability � much better than comments in


>> most cases.
>>
>>
>>
>> gray
>>
>>
>>
>> From: softimag...@listproc.autodesk.com
>> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Matt Lind
>> Sent: Wednesday, April 13, 2011 02:29 PM
>> To: soft...@listproc.autodesk.com
>> Subject: RE: Ice Tutorials
>>
>>
>>
>> Think of a compound as a functions or a class in a programming language.

>> If you�re organized, the compound will be a generic task that can be reused


>> in any ICETree that calls for the functionality.
>>
>>
>>
>> Matt
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> From: softimag...@listproc.autodesk.com
>> [mailto:softimag...@listproc.autodesk.com] On Behalf Of Paul Griswold
>> Sent: Wednesday, April 13, 2011 11:26 AM
>> To: soft...@listproc.autodesk.com
>> Subject: Re: Ice Tutorials
>>
>>
>>
>> I picked up Thiago's scalar tutorials on CMIVFX and was watching them the

>> other night. �He does the same thing. �When something gets to a certain
>> size, put it in a compound & expose & name the ports. �Until then I really

>> That way you can reuse the results �as much as you like.

Matt Lind

unread,
Apr 13, 2011, 8:03:09 PM4/13/11
to soft...@listproc.autodesk.com

> This combined with group comments which "know" what they contain and
> grow and shrink and move with that content would make creating
> readable trees much easier IMO.

Funny you should say that, I had submitted a feature request for this very thing earlier today.

As for the 'flyout flag', I think you mean 'Tooltip'. Personally I'd rather see the comments in full view as I often need to see comments from multiple nodes at a time.


Matt