[Bf-funboard] Some skinning feature ideas

[Michael Crawford]

On Apr 3, 2007, at 8:04 PM, Benjamin Tolputt wrote:

> Michael Crawford wrote:
>> But what If I want to add influence to just one bone, without  
>> influencing the weights any other bone.
> Simply don't toggle on the "normalized paint" button. This would  
> then work as it does currently

I'm cool with that.  Sounds good.

>> What I do is consider light green to be my "max influence"..
> Yes, but this is counter-intuitive (which is the point I am  
> arguing, not that it is a better/worse workflow).

Yeah...  I dunno.  I would suspect it's more intuitive for new  
(non-"convert") users.

> When one first starts painting weights, you would expect that  
> painting "full bright" (i.e. weight 1.0) would mean that the  
> selected bone/influence is going to fully affect the painted area.  
> Just like when painting red with full opacity in the texture image  
> (either in Blender, PhtoShop, or GIMP), you would expect that the  
> colour would be red on the image, not green or orange depending on  
> layers you cannot see/influence (this is ignoring special brushes  
> as the analogy starts breaking down between images/weights there).

I would think most people would find it easier to judge weight by  
color, being able to remember a particular hue after switching from  
group to group.  Try remembering a precise luma value and doing the  
same thing.  Embrace subjectivity and feel the color.

> The point is, you have had to bend your mind around an  
> implementation detail in the system (Blender) to get things to work  
> as you would like; just as one has to do with Maya's weight  
> painting. The only difference is what you bend you mind around. In  
> Blender you bend it around  the fact that the weight you see is not  
> necessarily what you get

see above: really easy to remember color as opposed to value.

> (I think this is a bad thing). In Maya you bend it around the fact  
> that you can only add weight vertices, as subtracting or smoothing  
> (equally bad in my opinion).

There is one other way to do it...  Paint extremities first, flooding  
to one, smoothing, locking that joint down, and moving one up the  
chain, repeating the process, etc etc.  It "squeezes the toothpaste  
inwards" (you paint the torso last).  Since stray influence can only  
go in one direction (due to the lock) all stray influence ends up in  
one section(torso).   But don't expect any students (except one thank  
god) to ever take your advice and actually follow that advice.

> My contention is the way it SHOWS you the weights is more intuitive  
> than the way Blender does.

silly idea perhaps.. what about multiplying the value and color for  
this hypothetical display mode.

> Hence the suggestion I made about having a settable "auto-truncate"  
> level, below which weights are completely removed. To be honest, I  
> cannot see how having a weight of 0.004 is going to help something.

Yeah, but whatever you set the "auto-truncate" value to becomes the  
effective new zero point (if my brain is still working, i need  
coffee).  If you have stray weights now, they're always going to be  
bigger than this.

> The only difference I am asking is that *I* be allowed (not that  
> *you* be forced) to see and paint the weight colours as they will  
> be applied.

If it's a preference, it sounds good, but I still don't think it can  
work.  I can "see" why but i can't quite find the worlds to verbalize  
it.  Oh well.  You'll hit that roadblock, and either trip or jump.

> However, I have confidence that Blender will evolve when needed.  
> Nigh on all Maya users know of the issues with weight painting, yet  
> the development team will not change it.

My guess is it's management that's the problem, rather than the techs  
(as is the usual in corporate nepotistic hierarchies where morons and  
every boss's cousin "Bob" get promoted out of the way, and hence tend  
to rise quickly to reek havoc at the top).

> Blender has already proven many times that the "current  
> implementation" of a feature is not always the "best  
> implementation" and has evolved to newer and (in general) better  
> methods. In alot of cases, the old way was not removed, just a new  
> method added to make things simpler.

Oh shit.  Where did mesh>tube go? Simpler?  eeh.  I'll patch my  
source and shut up about it. some obnoxious user is probably going to  
raise a stink about it at some point anyway :P

> And right here you admit that you need to bend you mind around  
> something that is not immediately obvious. Which is exactly the  
> reason I would like the new method of being able to do things.

Oh heck no it's not obvious at first, but it makes a lot more sense  
once you "get it" (at least to me).  I would imagine the concept of  
blender's weight painting would have been easier for me to grasp had  
I not already had a concept of what weight painting "should work like"

>> I've "converted" several maya users at my university to blender,  
>> and not a single one wants the old system of "paint weights" back.
> And I know alot of Maya users that will not touch Blender due to  
> many of it's "quirks" as they see it.

Blender is lacking in some areas ... but in my experience, whenever I  
say  "well i kinda miss that feature", it magically pops up in the  
next release.  Like AO baking for example.  That's the cool thing  
about open source.  Yeah Maya is "open source" as well.. given that  
it's 500Mb of mel scripts on top of a teeny tiny binary.. but that's  
not exactly what I had in mind.  Point is blender is evolving at such  
a rate that in my estimation it will start to surpass commercial  
packages within the next two years or so.  Alias/discreet can only  
keep up it's "upgrade now.. we added *three more paint FX brushes*"  
for so long before they run out of useless things to tack on (or  
users get inundated and decide they like blender's "k.i.s.s." approach.)

So many users expect bloat to the point where i would almost suggest  
adding a few megs of padding to the binary just to make it feel more  
complete.  If you're marketing to the government, pad a few hundred  
megs and charge licensing fee for the "propritary data" :)  Make em  
feel "special"!

I don't know if I would aggree totally with the transportation  
analogy... Id say some people are faster with horses and some are  
faster riding...  well something similar to horses.  I think it's a  
mostly matter of a users familiarity with a given tool when it comes  
to efficiency.

>
> Regards,
> B.J.Tolputt
>
> _______________________________________________
> Bf-funboard mailing list
> Bf-funboard at projects.blender.org
> http://projects.blender.org/mailman/listinfo/bf-funboard