ICE: For TDs only?

[Chris_TC]

Okay, but I refuse to believe that you or Max (or most everybody else) can't wrap your minds around that with a bit of initial effort. Could you give an example maybe? What was the last thing you wanted to do in ICE but couldn't?

[Hirazi Blue]

We're veering dangerously close to a "yes it is"/"no it isn't" argument here.

For the record (and I have implied as much in an earlier post in this thread, but should probably have stated it explicitly): it's taken me a long time to grasp ICE, I think I get it now for the most part, but the time it took me made me understand other people's struggle perhaps a little better and I am not afraid to admit my own struggle in this respect...
;)

[Pooby]

@Chris_TC - Sorry, but IMHO you and Pooby just do not seem to get it: it's true you CAN learn ICE and it might even turn out to be easier than you initially thought, but if you don't see it (the "workflow") and/or don't get it (the "mindset"), you will be lost forever. This has nothing to do with intelligence and/or computer skills IMHO. It's a visual way of coding (yes, it's coding!) and some people just can't get their minds around that (through NO fault of their own...).

Yes. That was me for my initial scattered attempts I had when ICE came out, I know EXACTLY what its like to feel totally lost in ICE. I also know that what is stopping people is a lack of Getting it.. And it's something that can be learnt very fast with decent tuition. I didn't think there was much around and none for free that dealt with deforms. Thats why I did my series pf tutorials.
I throw down the gauntlet. Anyone who doesnt think they can learn ICE, but would like to use it? I guarantee that, as long as you don't deliberately decide to prove your point by not paying attention and making no effort, I can get you up to a good standard in 2 days. No more arguing needed. I'm prepared to prove it can be done to anyone who thinks otherwise. The real point is that a lot of people have just closed their mind. Made a decision about it based on frustrated failed efforts and decided not to revisit it.

The bottom line is. Human nature will dictate that most will prefer to defend their point and maintain their stance of ICE being inaccessible rather than make an effort and have to admit they might have been wrong. Its the path of least resistance thing in action.
I have had, however, loads of people contact me about my ICE videos, saying that they have taken up ICE and its starting to make sense to them, so yes, while of course, you cant teach someone who doesnt try. I know for a fact that iCE is a lot more straightfoward than those who don't try it, seem to think.

[Hirazi Blue]

The bottom line is. Human nature will dictate that most will prefer to defend their point and maintain their stance of ICE being inaccessible rather than make an effort and have to admit they might have been wrong.

The same argument could be made for the opposite:

The bottom line is. Human nature will dictate that most will prefer to defend their point and maintain their stance of ICE being easy rather than make an effort to accept some people have severe difficulties with it.

Sorry, I couldn't resist... ;)

[Maximus]

Thanks Pooby and Chris for answer, first i'm kinda glad people like you guys have fun and find ICE usefull in your daily work, personally my ICE knowledge is a bit above basic, i studied it a freakton of time since it came out and i'm not completely new to it.

@Chris, regarding the example you posted before that is really basic stuff you could learn just reading or browisng vimeo, i probably wasnt enough clear, we might end up into a "show me an example" thread, wich i wouldnt like to.

in general the comparison you guys are doing, that learning ICE is like learnin Modeling it doesnt really fit well at all, you have the tools for modeling, and you start model, with ICE you dont have the tools you have something called ICE that is a thing that creates tools for you to work with.

You should compare ICE to the code behind the modeling operators, not to the modeling itself.
Of course every human can learn everything unless he/she lacks some really basic brain (no offense) but that is not the point.
The point is too personal to be general, the point is in what % ICE helps you in your daily work and how much effort you have to put to use it efficiently.

My % of ICE usage is near 1% to 0. I dont need it thats why i complain about an heavy development on something that i cant benefit.
You say i should learn it because i could benefit because i could create my own modeling tools.
This is not true, because creating a sweep modifier like Eugen Sares (if i didnt mix his name, dont kill me pls) is something that requires a massive amount of knowledge and skill.

We arent speaking about moving 1 point or assignin set/data, get data, set particle size, randomize and thos factory basic compound.
Are you really reducing ICE to something like this? I dont think so, ICE is way more than this, way more complex and way more powerful.
I'm personally frustrated with ICE because since devs seem to not update what i need for my daily work i would love to build myself, but here comes the pain.

1) the effort and the time you have to spend into learning an environment to create your tool
2) the difficulty to create tools, wich varies depending on tools, but the tools i need arent that easy to create
3) the purpose of a software, i'm really supposed to create my own tools? i mean..dont you find this a bit awkward?
Why then dont start a new 3D program and code yourself everything from scratch.
Would you like to create a spotlights connecting 20 nodes, or would you like to find a nice button that says "Create Spotlights?"
Becase being an artist what i would expect to find is the button, not the ground to create myself the button, if i wanted to be a developer i wouldnt be an Artist.

Again i'm glad people like you guys have fun and use of ICE, i personally dont, and not because i didnt try, i try everyday since v7 came out, to friends home, with friends help, and i never once was able to produce something good if not basic things, like this

viewtopic.php?f=19&t=1435
wich was a simple drag and drop properties into ICE and connect them, agan this is basic level.

To be honest one of the most cool things i've had the chance to test with ice was that Eugen Sares extrude along curve compound, since Softimage lacks that command (for some shocking reason), and still is not perfect and produces some errors, but i tried to dig that compound and open it to look inside, i just closed and lost hopes into learning ICE, i admire Eugen and people like him because they are great skilled and talented, but i'm far from this kind of skill, i didnt have a math background it simply doesnt come to my mind that i should use an Array in that particular case, i just dont think i should use a "Get distance between" node, this is the real ICE problem, when i say it is hard, i mean this, i dont have the mindset to think on how to achieve things.

I hope i explained myself a bit better, sorry for english.
My work requires a lot of rendering and quality, and this things are made from modeling/texturing/rendering department, where Softimage is behind every other 3d package, even blender. I couldnt care less about ICE because what i need desperately is those things. IF AD will release a version of softimage where i got all new tools that connect ICE nodes under the hood, i couldnt be more happy, i dont care what ICE does, i care about tools, and in this same moment Softimage lacks a lot of things.

Max

[Bellsey]

This is a very interesting thread with some valid points. For me there are a few 'things' surrounding ICE and how/what people think of it. Many people refer to Softimage and then ICE as if they are literally two different packages. And in some ways you could argue that they almost feel like it, but I like and prefer to think on them as the same thing, Softimage is ICE and ICE is Softimage. That's how I view them.

Which kinda leads to another thing and one that has always been tricky, but it's the same with any kind of 'technology', and that's explaining what ICE actually is, and then how you work with it. And even now after a few years, there are still people who don't quite 'get it'. And I can see where they are coming from, it can look scary and daunting.
Many people point to the lack of tutorials, presets and general help about ICE and in some ways, I agree there has been a shortfall here, but it has been getting better, we still have some way to go, but we're getting there.

When we first launched ICE, we said it was for artists as well as the TDs, and I still think there's still truth to that. But its hard to show and explain, because both those people will look and any tools/feature build with ICE, in different ways. I think Paul's right, anyone can learn ICE, but perhaps one person might need abit more handholding than the other.
Once you sit down with someone and show them ICE in the right way and what its doing, then it can all fall into place. That's easier said than done though, which is why I liked Paul's videos, because they start quite basic.
I've been planning some videos myself, which follow Paul's example. I want to try and re-introduce people to ICE, but not just Soft users, Max and Maya ones as well.
But I'm going even more granular and low level to explain the absolute basics. And I mean really basic. I think the trap we (including myself) fall into when showing/teaching ICE, is that we kinda assume someone will quickly grasp and understand the principles after we explain them, when often they might will sit there and not understand what's actually going on in the graph and how data flows.

[Chris_TC]

@Chris, regarding the example you posted before that is really basic stuff you could learn just reading or browisng vimeo,

If this was basic stuff then why were you earlier saying that you couldn't possibly make Mograph functionality in ICE because it's way too complicated for you? You made it sound as if you're completely clueless about anything ICE-related.

extrude along curve compound, since Softimage lacks that command (for some shocking reason)

Modify->Poly Mesh->Extrude along curve

from modeling/texturing/rendering department, where Softimage is behind every other 3d package

You can't possibly be serious. If you are, then why are you still using Softimage at all?

[Chris_TC]

I've been planning some videos myself, which follow Paul's example. I want to try and re-introduce people to ICE, but not just Soft users, Max and Maya ones as well.
But I'm going even more granular and low level to explain the absolute basics. And I mean really basic.

That sounds excellent. I think there's too much misinformation about what ICE is, especially among non-Softimage users. It would be great if you could explain the "scope" of ICE and show that it's not just a glorified particle simulator. And that it can be used by artists.
I recently made a spot light flicker by driving its brightness with a Turbulize Around Value ICE node. I have no idea what kind of expression I'd have to write to make this happen. A single ICE node did the trick. If that's not artist-friendly, I don't know what is.

[Pooby]

For me. I do mainly performance capture character /creature work. I constantly try to improve the way I can describe the skin material, in the way it moves, slides, rolls over itself and interacts with underlying structures.
Shapes and 'bones' just don't cut it. Something like cmuscle in Maya wont cut it. With ICE I can design my own deformation qualities and do things that simply cant be done any other way without a severe and tedious amount of coding.
I find this extremely exciting. I know that I have a toolset in ICE that is leagues ahead of my 'competitors' using Max and Maya. And thus have access to all kinds of deformers that don't even have names because I've just made them up. When some new impressive Maya deformer comes out, I can often replicate it in ice within minutes.

So, whilst I don't NEED ICE. ( I could still work like everyone else using same old techniques) It gives me a huge advantage And has opened up all kinds of possibilities and made doing CGI enormously entertaining for me.

I'd like to know why people say softimage's modelling , texturing rendering is behind other packages. I find it almost comprehensively superb, having used lightwave and modo previously. I also study rigging in all packages for research and the standard SI toolset for this seems leagues ahead too in ease of use and application.

[Maximus]

@Chris, regarding the example you posted before that is really basic stuff you could learn just reading or browisng vimeo,

If this was basic stuff then why were you earlier saying that you couldn't possibly make Mograph functionality in ICE because it's way too complicated for you? You made it sound as if you're completely clueless about anything ICE-related.

extrude along curve compound, since Softimage lacks that command (for some shocking reason)

Modify->Poly Mesh->Extrude along curve

from modeling/texturing/rendering department, where Softimage is behind every other 3d package

You can't possibly be serious. If you are, then why are you still using Softimage at all?

I meant Mograph modules, not mograph itself, stuff like dynamic arrays, tracer, cloner, all deformers, things that you control with just sliders that are premade and that quickly give you a lot of nice stuff by simply activating them.

Extrude along curve...i once again failed to write what i meant, i know Soft has that, but it doesnt take into account corner mitering, you cant do a precise profile sweep, if you have a profile curve and you want to sweep it into another curve path, good luck doing it in soft, thats why Eugen Sares compound was so amazing.

I'm using softimage for the same reason you are using it, because i like it, that doesnt mean that it doesnt have shiny and beautiful aspect as well as flaws.

I might just drop the thread here i guess
cheers, Max

[tomalc]

It's true Softimage should include more high level tools to keep and capture artistic mindset users...

I guess those work with trial/error until it works, like all of us, BUT without taking (or having) the time to look back and understand what they have done.

But if an artist knows what he is doing, in what order, and can write it down to paper, he is more than half way up to the technical way of doing.. he just made a procedure..

Technically, my own learning path with ICE was :

- grasp the context thing (but can you select both points, edges and objects with the ui ? No, same in Ice)
- (re)learn some basic math (ice is a nice teacher who draws very fast on the board.)

What I really don't understand is the focus on type (vector, scalar...) I don't believe any CG artist is lost in math to the point he thinks a 3D position can be a single number... Same with an array, it's just a ordered list...

Thomas Le Calvé

[Pooby]

http://www.xsibase.com/forum/index.php? ... #msg257015

I found this old thread from 2009 and re-read my posts. I can very clearly remember the frustration I felt.
The reasons I did my tutorials was because thats what I would had wanted to know then.

[Pooby]

The bottom line is. Human nature will dictate that most will prefer to defend their point and maintain their stance of ICE being inaccessible rather than make an effort and have to admit they might have been wrong.

The same argument could be made for the opposite:

The bottom line is. Human nature will dictate that most will prefer to defend their point and maintain their stance of ICE being easy rather than make an effort to accept some people have severe difficulties with it.

Sorry, I couldn't resist... ;)

I would surely accept the view after someone had made an effort to train and failed.

My point is that it is VERY easy to give the impression that you can't do something but this is only true if you can demonstrate that a concerted effort has been made and correct training material has been provided and STILL you have severe difficulties.
As far as I'm concerned, nobody should be in that position now, as, if there are any ICE concepts that people dont get, I would be more than happy to make a new video to thoroughly explain them until they do.

[Bellsey]

http://www.xsibase.com/forum/index.php? ... #msg257015

I found this old thread from 2009 and re-read my posts. I can very clearly remember the frustration I felt.
The reasons I did my tutorials was because thats what I would had wanted to know then.

The point made in that thread are the things that have niggled me for a while. There are some very good ICE tutorials out there, but its the absolute basic where I think we maybe fell short. And its these initial small steps in learning ICE basics that many can struggle with, and what I would like to do in some videos. That's not taking anything away from you Paul, your vids were great, and part of the reason for me trying to get back on track with this.

[Chris_TC]

if there are any ICE concepts that people dont get, I would be more than happy to make a new video to thoroughly explain them until they do.

Hi Paul,
I have a suggestion. The one thing that helped me the most was understand contexts. A year ago I had no idea what contexts are, and I was getting context errors left and right. It was very frustrating. It took me a long time to figure out how this stuff works.

I think a video would be good in which you show object and per-element contexts. Show that an object context gives you one value (or an array), while something like self.PolygonIndex gives you one value per element, and something like self.PointNeighbors gives you an array per element. Stuff like that.
One situation in which this comes up quite a bit is when you want to use the Randomize compounds. By default, these compounds are per-point, so you can't plug them into, say, an object's SRT to randomize its translation. When you know about context, you just open up the compound and see that a "Get Point ID" is forcing the per-point context, and that removing it makes the Randomize compound output only a single value in object context. If you don't know about contexts, you'd probably just conclude that ICE sucks.

[Pooby]

Good idea. I'll do one on contexts next.

One thing that I find slightly unintuitively named is 'switch context' . It might be technically accurate, but it gives no hint that its for allowing data from one object to be used in another identical copy.
One might assume from its name that object context data can be used in per point, or poly, but thats not what it does at all.