CGTalk > Software > Autodesk 3ds max
Login register
Thread Closed share thread « Previous Thread | Next Thread »
 
Thread Tools Search this Thread Display Modes
Old 12-15-2013, 01:39 AM   #16
Graeck
CAT User
 
Graeck's Avatar
portfolio
Gregor Weiß
Character TD & Animator
Berlin, Germany
 
Join Date: Aug 2003
Posts: 1,052
Send a message via ICQ to Graeck
Well, the point is, if I'm not sure about what to answer because I probably have too many other things to do, it's a better idea to skip the question, rather than producing confusion. By the way, I'm also not happy about the questions being asked in that interview. It doesn't matter much to the overall Max development how someone would animate a loopable flag. But if you come up with an answer like "Try something different like Waveform or try to solve it using Morpher", it's just an incompetent answer, and I'm allowed to criticize that. I'm sure Chris never tried to use Morpher to create a loopable animation, because if, he would have realized it's not possible, as you will always see the loop. So why he is suggesting something he does not know about? And this is the point. Answering without knowing, that is the problem. It's supposed to be an interview with experts, so I'm expecting good answers.

And I think it's an important point. If you keep answering questions like that, saying "Hey we have that feature already, you just didn't realize", it means the feature is being ignored, although there is actually not such a feature, or the feature is not satisfying. That's actually what I want to make clear.

Quote:
The fish is stinking from the head.

Not always true. Decisions made by the responsibles is dependent on the feedback they get. And if a technical expert in their team is saying we have that feature already, it's fine, although it's not, we have a problem. I even think this might be one of the big reasons why certain things in Max development are being ignored like that.
__________________
My Demoreel 2011:
http://www.youtube.com/watch?v=Gfq-4VoeAPU
 
Old 12-15-2013, 01:19 PM   #17
Erka2
Lord of the posts
 
Erka2's Avatar
 
Join Date: Aug 2002
Posts: 520
Quote:
But if you come up with an answer like "Try something different like Waveform or try to solve it using Morpher", it's just an incompetent answer, and I'm allowed to criticize that. I'm sure Chris never tried to use Morpher to create a loopable animation, because if, he would have realized it's not possible, as you will always see the loop.

http://vitsly.wordpress.com/2009/01/23/loopable-flag/
 
Old 12-16-2013, 12:01 AM   #18
Graeck
CAT User
 
Graeck's Avatar
portfolio
Gregor Weiß
Character TD & Animator
Berlin, Germany
 
Join Date: Aug 2003
Posts: 1,052
Send a message via ICQ to Graeck
Quote:
Originally Posted by Erka2

The tutorial basically describes the Point Cache method, the Morpher is just being used to create the blend between both. It's not necessary to use the Morpher as you can always do the blending within the Point Caches themselves. The main idea is to use the Point Cache and create the time offset. A Morpher alone wouldn't work, as pointed out in that tutorial as well.

A little thing that is missing in the tutorial though is a proper weight-balancing as two different animations tend to cancel each other out in the middle of the blending.

I'm not saying a demo artist needs to know every little trick in every field of a 3d software. I guess nobody is able to do that. The question about the loopable flag is dispensable and a bit unfair too as it is a specific question about animation. What really bothers me is the way Chris is answering, saying we need to discuss the question from the ground up and bla bla. It's obvious he had no idea how to solve it, so why he is answering like that? I've watched another Expert Challenge from March this year (http://www.youtube.com/watch?v=Ys5dr9OQNbk) where Chris tries to explain the differences between Edit Mesh and Edit Poly, saying Edit Mesh is more like a triangle-based workflow, where you can see and edit triangles better than in Edit Poly. Holy crap? And he says Edit Poly is used for subdivision workflows. Does that mean subdivisions do not work on Edit Mesh? He is just producing confusion rather than answering questions. I was sitting there throwing up my hands in despair. Never seen a demo artist talking so much rubbish, sorry. Glad Joe Gunn summed in up in just a few words.

Ok, this is now getting very personal against Chris Murray, I'm sorry about that. My actual point however is that Max development, whether XBR is still alive or not, can only go the right way if the responsibles listen to the right people. Having people in their team who have no clear vision or who don't know well about workflows will no be good for the decisions being made in the future.
__________________
My Demoreel 2011:
http://www.youtube.com/watch?v=Gfq-4VoeAPU

Last edited by Gräck : 12-16-2013 at 12:14 AM.
 
Old 12-16-2013, 07:13 AM   #19
Noren
Expert
 
Noren's Avatar
 
Join Date: Jan 2003
Posts: 492
I do agree with some of your points but I think you're projecting all your frustration with Max on one guy, which isn't really fair.
As for the flag question, I have to disagree.
He might have a convoluted way of saying things, but technically he wasn't wrong.
Point Cache is the go-to answer, obviously, and he mentioned it, as well as offering different approaches which might very well be sensible, depending on the situation.
Since he knew he was going to mention Point Cache, I'd say give him some credit and assume he had a reason to mention the other stuff anyway, even if your experience tells you differently.
If you have a very short loop, simulation is not necessarily the best and quickest way to solve it and if you have a long loop and find some roughly matching segments in your simulation, a short, blended, live Morpher might do the trick just as well and in some cases better than a long transition where you then have to compensate for interference, like you mentioned, or use tricks like in the linked tutorial.
Same goes for Point Cache of course and for longer transitions it's clearly the better tool, but the Morpher can handle shorter ones just fine, when you are dealing with a limited amount of keyframes.

But I do agree it would be better to actually show things like that, instead of just mentioning them in passing, which also begs the question why they have to do things like that "live".
But then what might be confusing to some, others can use as starting point for further research.


What I liked about the presentation was the fact that there's apparently still some love for Scanline in the team and that they plan to rely heavily on user feedback.
How this will turn out we'll have to see.
In a way Autodesk has been feeding us hopes for years now, though, and they can't play that card much longer.
Ironically 2014 is quite a reasonable release for my needs, but boy has it taken them long to get there. Always eager to take our money while stumbling along.
They better step up the pace now and take care of the neglected areas in 3ds Max instead of nudging us towards their other products.
I don't really buy the story about the newfound spirit quite yet, either, especially when it comes to the higher-ups.
My impression is that they want to sell Maya and their suites and they have been working hard to eliminate immediate customer feedback in the only language they understand: Money. New customers seem much more important to them than the ones they already have.

Last edited by Noren : 12-16-2013 at 07:23 AM.
 
Old 12-16-2013, 10:37 AM   #20
Graeck
CAT User
 
Graeck's Avatar
portfolio
Gregor Weiß
Character TD & Animator
Berlin, Germany
 
Join Date: Aug 2003
Posts: 1,052
Send a message via ICQ to Graeck
Quote:
Originally Posted by Noren
I do agree with some of your points but I think you're projecting all your frustration with Max on one guy, which isn't really fair.
As for the flag question, I have to disagree.
He might have a convoluted way of saying things, but technically he wasn't wrong.
Point Cache is the go-to answer, obviously, and he mentioned it, as well as offering different approaches which might very well be sensible, depending on the situation.
Since he knew he was going to mention Point Cache, I'd say give him some credit and assume he had a reason to mention the other stuff anyway, even if your experience tells you differently.
If you have a very short loop, simulation is not necessarily the best and quickest way to solve it and if you have a long loop and find some roughly matching segments in your simulation, a short, blended, live Morpher might do the trick just as well and in some cases better than a long transition where you then have to compensate for interference, like you mentioned, or use tricks like in the linked tutorial.
Same goes for Point Cache of course and for longer transitions it's clearly the better tool, but the Morpher can handle shorter ones just fine, when you are dealing with a limited amount of keyframes.

But I do agree it would be better to actually show things like that, instead of just mentioning them in passing, which also begs the question why they have to do things like that "live".
But then what might be confusing to some, others can use as starting point for further research.


What I liked about the presentation was the fact that there's apparently still some love for Scanline in the team and that they plan to rely heavily on user feedback.
How this will turn out we'll have to see.
In a way Autodesk has been feeding us hopes for years now, though, and they can't play that card much longer.
Ironically 2014 is quite a reasonable release for my needs, but boy has it taken them long to get there. Always eager to take our money while stumbling along.
They better step up the pace now and take care of the neglected areas in 3ds Max instead of nudging us towards their other products.
I don't really buy the story about the newfound spirit quite yet, either, especially when it comes to the higher-ups.
My impression is that they want to sell Maya and their suites and they have been working hard to eliminate immediate customer feedback in the only language they understand: Money. New customers seem much more important to them than the ones they already have.

I have to disagree with you about the Morpher. Morpher will always be the worse solution compared to Point Cache, simply because it does not contain animation. It will just blend into a pose and that means you will most probably see the loop, even if you use very short transitions. Also there is no "pre-pose" to get a good transition, very important for a really smooth loop, so most of the time finding a pose that looks similar will also not help much. Additionally, using Point Cache on long loops does not necessarily mean to use long blendings too. You can always use short blendings, whether the loop is long or not. That's also what Chris said wrong. He was pointing to the length of the loop for no reason.

I don't want to talk about Chris Murray too much anymore, enough being said from my side. I just want to add that my opinion about him is not going to change. He did not mention Point Cache at the first place, he forwarded the question to Kelcey who was obviously introducing the Point Cache workflow to him before the interview. Again, I don't care much about whether he knows about that workflow or not as it might be a bit too specific for him. I do care about the way he is answering though. First thing he says is putting the question into another context like it is a strength of Max that there are different approaches to solve problems (although those approaches do not help answering the question). It's not a strength of Max, it is what you can do in any 3d package. My impression is they don't really know what to say or what to do with Max in the future and rather say how proud they are, how great the community is and trying to put things into another context so it sounds nicer.
__________________
My Demoreel 2011:
http://www.youtube.com/watch?v=Gfq-4VoeAPU
 
Old 12-16-2013, 11:25 AM   #21
Rawalanche
Senior Noob
 
Rawalanche's Avatar
portfolio
Ludvik Koutny
3D Generalist
Prague, Czech Republic
 
Join Date: Oct 2008
Posts: 324
Yeah, i do not think it is necessary to dig into the specifics either. The overall problem was that it was called an Expert Challenge and they claimed they will answer the toughest question thrown at them, yet they did manage to dodge specific answers for EVERY SINGLE DAMN QUESTION and just kept repeating salesman BS all over and over again. They do not deserve to be defended by anyone, because the entire session was a big slap into face of nearly every 3ds Max user.
__________________
http://raw.bluefile.cz/
 
Old 12-16-2013, 12:28 PM   #22
Noren
Expert
 
Noren's Avatar
 
Join Date: Jan 2003
Posts: 492
Quote:
Originally Posted by Gräck
I have to disagree with you about the Morpher. Morpher will always be the worse solution compared to Point Cache, simply because it does not contain animation. It will just blend into a pose


And that's were you're not quite right. A morpher with automatic target reloading enabled will do so for every frame, so it's perfectly possible to blend between animations.
 
Old 12-16-2013, 01:49 PM   #23
Graeck
CAT User
 
Graeck's Avatar
portfolio
Gregor Weiß
Character TD & Animator
Berlin, Germany
 
Join Date: Aug 2003
Posts: 1,052
Send a message via ICQ to Graeck
Quote:
Originally Posted by Noren
And that's were you're not quite right. A morpher with automatic target reloading enabled will do so for every frame, so it's perfectly possible to blend between animations.

You do not understand correctly. Using that option means using the Point Cache method. What is shown in the tutorial ist just a on-top-solution that is not necessarily needed. You can live without using the Morpher, and the Morpher is not the actual idea for solving the loop problem. The tutorial is all about using the Point Cache method. Using Morpher alone means blending into a pose only, as you have no animated offseted additional target mesh which you can only produce by using the Point Cache method and so actually make use of that option you are talking about.
__________________
My Demoreel 2011:
http://www.youtube.com/watch?v=Gfq-4VoeAPU
 
Old 12-16-2013, 03:44 PM   #24
Noren
Expert
 
Noren's Avatar
 
Join Date: Jan 2003
Posts: 492
I wasn't referring to the tutorial in particular and this is getting old.
Point Cache is as replaceable as the Morpher in the mentioned situations, especially for smaller transitions. (And that only for ease of use reasons. Theoretically you could do that whole tutorial without Point Cache in reasonable time if you ever needed to for any reason or happen to be a masochist. )

With the morpher you can add additional transitional animations without having to deal with keyframes and fading in and out and can see parametrical changes you make to those targets directly on your mesh if you ever needed to.
You probably don't, but it doesn't hurt to know about it.

Like said I'd probably choose the Point Cache myself but that doesn't change the fact there are other options which might come in handy in certain situations.
And also like said, for very short loops I'd rather take an parametric approach before wrestling a simulation. (But that's surely different from person to person, target style and how experienced you are with both approaches. )

What you said was basically that he was an idiot for even mentioning other methods that according to you couldn't work at all and therefore he couldn't have even tried.
That's basically all I wanted to comment on and suggest you dial it down a bit.
 
Old 12-16-2013, 10:06 PM   #25
Graeck
CAT User
 
Graeck's Avatar
portfolio
Gregor Weiß
Character TD & Animator
Berlin, Germany
 
Join Date: Aug 2003
Posts: 1,052
Send a message via ICQ to Graeck
I guess we better stop here as you ignore to get the point correctly. You can't do the loop correctly without the Point Cache or without doing a time offset, if you disagree, you didn't understand the workflow.

I am close to setting up a scene to show the problem, but this really shouldn't be needed as it is actually not too hard to understand.

Quote:
Like said I'd probably choose the Point Cache myself but that doesn't change the fact there are other options which might come in handy in certain situations.
And also like said, for very short loops I'd rather take an parametric approach before wrestling a simulation. (But that's surely different from person to person, target style and how experienced you are with both approaches. )
What you said was basically that he was an idiot for even mentioning other methods that according to you couldn't work at all and therefore he couldn't have even tried.
That's basically all I wanted to comment on and suggest you dial it down a bit.

Again you are putting things wrong. I never said he was an idiot, just what he is saying is rubbish, and that is very obvious.
Suggesting different methods is fine, but without answering the question it's not. I also don't want to labour the point as there are plenty examples where Chris fails to give useful answers and rather produces a lot confusion, like the Animatable Pivot Point, like the Edit Mesh / Edit Poly question, etc, just watch the videos. He is even unable to actually show something reasonable in his demos. Most of the time he is just opening a scene and basically doing nothing.

But that's really it for me about Chris right now. I don't want to just blame a person all the time. But obviously I need to give more examples because you fail to understand my point.
__________________
My Demoreel 2011:
http://www.youtube.com/watch?v=Gfq-4VoeAPU

Last edited by Gräck : 12-17-2013 at 06:40 AM.
 
Old 12-17-2013, 07:19 AM   #26
Noren
Expert
 
Noren's Avatar
 
Join Date: Jan 2003
Posts: 492
Quote:
Originally Posted by Gräck
I guess we better stop here as you ignore to get the point correctly. You can't do the loop correctly without the Point Cache or without doing a time offset, if you disagree, you didn't understand the workflow.

I said "no Point Cache" not "no time offset." And again, I wasn't talking about that tutorial in specific, if you're aiming for that.

Baking and shifting a few keys is neither limited to Point Cache nor really that big a problem or anything obscure, but here's a file so we know what we are talking about:

https://www.dropbox.com/s/8zymdrwmr...orpher_send.zip

This is neither beautiful nor polished, just a quick proof of concept. There is some noise in the sim etc. and the sim is just a sim for a sim's sake.
There might be a slight jerk in the loop in the viewport sometimes, but I watched it in the RAM player at different speeds and it seems good enough to me to prove my point. A similar set up Point Cache would have the same problem and the same need of polishing.
It's not feasible for every case, but it works.

The advantage of using a Morpher to blend would be for example that you can deform your target over time to better match the parallel segments and see a live update without having to rebake a Point Cache. Of course you can use both in combination like in the tutorial as well and that's the preferred method but that's far from your claim that the Morpher is completely useless for blending loops.
(And noone ever mentioned the Morpher was to be used alone, you naturally would assume there is a morph target involved and in a case like that the step to an animated morph target seems logical, too. As long as you weren't completely oblivious to that possibility. )

And I'm surely not completely happy with the presentation and many of the answers either (though I partly blame the slightly weird concept and didn't know what to expect) and share some of your frustration with development in general, I think I made that clear. But if you are going to be that harsh and condescending (and coming from me this means something ), you better make sure you haven't a blind spot somewhere yourself.
 
Old 12-17-2013, 09:34 AM   #27
Graeck
CAT User
 
Graeck's Avatar
portfolio
Gregor Weiß
Character TD & Animator
Berlin, Germany
 
Join Date: Aug 2003
Posts: 1,052
Send a message via ICQ to Graeck
Oh hell, god, please, this discussion is driving me nuts. Whatever please use your Morpher. I don't care about it. Sure you can use Morpher if you have a proper target mesh. Whatever that is. Forget that freakin Point Cache please. You can also copy your simulated object and offset that, you can always find anything else to force prove me wrong, even if you suggest to hand-animate your mesh by hand for days, it doesn't matter. If Chris had suggested that, you would probably have shaked his hands for being such a genius.

The point is Chris did not mention anything close to that method (time offset). And I even don't care if he did. Please let's not talk about that example anymore. There are plenty of examples where Chris fails to explain things correctly or gives useful answers. And let's even not talk about Chris anymore, getting tired of that. The other guys in that interview are also not very much better.

Look at Duncan for example (Maya demo artist). This is how I think a good technical expert / demo artist should be like. He thinks before he says and specifically answers to questions, always keeping in mind possible production issues etc. Someone like him would be great to have in the Max discussion group. I don't know how much influence Chris really has to the Max development, but if they choose him for being the offical technical expert / demo artist, I begin to worry about if the responsibles are able to make the right decisions, whether it is in hiring people or deciding which features are developed in which way.

Quote:
And noone ever mentioned the Morpher was to be used alone, you naturally would assume there is a morph target involved and in a case like that the step to an animated morph target seems logical, too. As long as you weren't completely oblivious to that possibility.

Oh hell, yes, of course for using Morpher you need to use a target mesh. Good we agree about that! So you think Chris made this logical, natural step to the animated mesh that is offseted in time, the actual main point and initial idea the whole loop question is all about, but he just didn't mention it? You really think so? Haha. Nothing else to say, sorry. Just funny.
__________________
My Demoreel 2011:
http://www.youtube.com/watch?v=Gfq-4VoeAPU

Last edited by Gräck : 12-17-2013 at 09:42 AM.
 
Old 12-17-2013, 11:35 AM   #28
Noren
Expert
 
Noren's Avatar
 
Join Date: Jan 2003
Posts: 492
Quote:
Oh hell, yes, of course for using Morpher you need to use a target mesh. Good we agree about that! So you think Chris made this logical, natural step to the animated mesh that is offseted in time, the actual main point and initial idea the whole loop question is all about

Well, I know someone who clearly didn't. SCNR

In all seriousness, though: He probably had seen or done something like that in the past, so he mentioned it as a starting point. Not as good as demonstrating it or explaining it in detail, but also no harm done. I also find the whole concept of the presentation a bit silly and think they should have whatever time they need to prepare a demonstration.
I do share some of your concerns, make no mistake, but I don't think it was necessary to be such an ass about it.
 
Old 12-17-2013, 03:55 PM   #29
PiXeL_MoNKeY
Expert
 
PiXeL_MoNKeY's Avatar
Eric Craft
Dallas, USA
 
Join Date: Mar 2002
Posts: 5,654
Actually Duncan is not a demo artist, he is the creator/developer of what he is showing off. So he should have knowledge on what he is showing off.

Chris is a technical specialist/demo artist so his knowledge of what he needs to show has to cover the application, not specific features.

The real question is does the 3ds max team have those "white paper" types of developers like Duncan who can do the backend work, but also do some technical demonstrations as well. Also, do those on the 3ds max team have the time to step away from the development to do tech demos?

I personally would prefer the developers develop. On the demo side a lot can be done to improve, but the question is what do you focus on?

-Eric
__________________
"The Evil Monkey hiding in your closet."
 
Old 12-18-2013, 06:30 AM   #30
Noren
Expert
 
Noren's Avatar
 
Join Date: Jan 2003
Posts: 492
What I'm also a bit wary about is all the talk about "leapfrogging".
As far as I remember the questions answered with that phrase didn't concern me directly, but I'd much rather see solid development and creative solutions on small scale first, listening to the users like they promised.
If that's the case and they have free resources they can get creative with some of their CEO's more aspiring plans.
I have to say I'm still a bit grumpy about People Power and the resources it wasted. (Speaking from my current point of view. )
 
Thread Closed share thread


Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
CGSociety
Society of Digital Artists
www.cgsociety.org

Powered by vBulletin
Copyright ©2000 - 2006,
Jelsoft Enterprises Ltd.
Minimize Ads
Forum Jump
Miscellaneous

All times are GMT. The time now is 12:17 PM.


Powered by vBulletin
Copyright ©2000 - 2016, Jelsoft Enterprises Ltd.