PDA

View Full Version : APELS (Advanced Parametric EdgeLoop Split)


passwr
09-03-2003, 05:26 PM
APELS is a parametric loopSplit algorithm which aims at bringing interpolating Bspline behaviour to the world of polygon modeling. It's main use is for splitting edgeloops by, after performing a traditional loopsplit, placing the newly created edgePath CVs to flow with the surrounding topology, a bit like 'insert isosparms' function on NURBs, but for polygons. This second function can also be performed on any existing edges, or edgepath on the mesh.

http://www.swanimator.com/melscripts/APELS/description.htm

http://www.swanimator.com/melscripts/APELS/APELSplane.gif

TheWraith
09-03-2003, 06:17 PM
yay! i've been wanting something like this for awhile. not sure if theres any other free plugz out there that do the same thing, i thought there were, but in any event, i don't have them installed! so i'll hafta grab this tonight and give it a whirl!

KidderD
09-03-2003, 06:41 PM
Maybe I'm missing the purpose of this, but is this to be used as a directional smooth. I guess I'm trying to determine what this would be useful for. I apologize ahead of time if this insults anyone's intelligence, in truth it only insults mine.

bombay_potato
09-03-2003, 07:26 PM
I have given it a whirl and I think this is great! I have been thinking along these lines but have not gone headfirst into scripting or the math involved, thank you for writing this!

As far as the script is concerened my first suggestion is to allow for the smoothsplit command to be added to the repeat last stack.

and looking under future development on the download site, I see talk of future implementation as a node. This could make a big difference in removing repetitive actions while tweaking/polishing a shape.

I have been writing down some ideas for a while now thinking of how 2d vector drawing behavior could be implemented in 3d to help in modeling. vector programs like illustrator and freehand have some options to "beautify" existing lines, where it looks at the number of knots and determines which ones should be moved based on the relationshop between all of the control points (decides what points look out of place and moves them). I have been thinking that a 3d (or maybe 2.5d based on screen) beautify operation could be powerfull.

having used the script I am sold that APELS operations that could be performed on existing loop selections would be extremely usefull.

thanks for making this available and seeking input, very fun!

bb_potato

wgreenlee1
09-03-2003, 07:35 PM
I'm confused on the installation.....
Hints anyone?

trthing
09-03-2003, 08:56 PM
execute apels.mel (have MJpolytools on you script dir; copy available on his site)

a panel will come up.

enjoy;)

Very nice script. I wonder if it will work in conjunction with BPT or take advantage of it and get to another level.

Will keep testing it here...

Thnx

rebo
09-04-2003, 12:01 AM
Im confused. What is the difference between this and a traditional edge loop split.


EDIT::

Oh i see, it adds a new edge whilst keeping the curvature of the existing transverse edges?

This is quite nifty, i would love to see this in BPT.

wgreenlee1
09-04-2003, 12:25 AM
@trthing...thanks,got it.
:thumbsup:

gundog
09-04-2003, 03:15 AM
nice. this looks like a great script. i'll give it a run tomorrow for doing some beveling.

thanks.

rebo
09-04-2003, 04:27 AM
I find it quite slow on large splits, I think a compiled plugin solution would be much much faster.

ceql
09-04-2003, 10:14 AM
Whoa, Thanks!! I've been looking for something like this too :D :applause:

/me will dl now :D

1000101
09-04-2003, 11:19 PM
Great stuff Shyal

I haven't had much time to play with it yet; but so far I like what I see.

haha so when are you going to start digging into the API?
muahha

APELS
09-05-2003, 04:01 PM
Hi Folks,
It's great to see interest in this script.

OK the newest version (APELS v0.2 Alpha) (http://www.swanimator.com/melscripts/APELS/download.htm) has new features:


-Interactivity: After performing a split the cursor changes to a crosshair, and you can interactively set the height of the split (drag towards the left to make it a traditional split, or towards the right to make it etched out. 6 different levels of etching).

-Bypasses UI: The UI was just for the primer development stage, it is still there but if you want it you'll have to call it, else you can call the commands from your shelf, or from Maya's main menu and Edit Polygons. (make sure you add the userSetup.mel code into your userSetup file)




From Bombay_Patato
As far as the script is concerened my first suggestion is to allow for the smoothsplit command to be added to the repeat last stack.



Thanks for the feedback. This is now possible if you use either the shelf button or launch the procs through the menu. I don't think it's possible via the UI though, but I have the feeling we won't be using the UI any longer anyway.



From Bombay_Patato
having used the script I am sold that APELS operations that could be performed on existing loop selections would be extremely usefull.


This is possible, and is the command "Smooth Selection" ( APESmoothSelection ). It's quite helpful to "beautify" existing meshes as you called it, ideally if used in conjunction with edgePath, selecting one every two edgesPaths (evens), smoothing, then the odds, and smoothing again, this evenly spreads the polys and usually yields good results.


From Rebo
I find it quite slow on large splits, I think a compiled plugin solution would be much much faster.

From Russ
haha so when are you going to start digging into the API?


First problem is MJPT is quite slow at performing splits on large models, I'm hoping to bypass this problem by either writing some simplified loopSplit algorithm, or using Sebastian's BPT, the latter is probably the most likely scenario.

And secondly, I accede, the speed would increase manifold were I to implement an API version of this, which I'm most probably going to do. But first you'll notice as you keep on using it there are still a few bugs such as vertices moving where and/or when they shouldn't, and a few features I still want to implement, primarily for a more heuristic selection of pin vertices for the algorithm to perform better with border-edge CVs, border-edgeloops, and non-quad areas of the mesh.


Here's the link to the newest version. (http://www.swanimator.com/melscripts/APELS/download.htm). As requested I've created a list (http://www.swanimator.com/melscripts/APELS/APELSlistSubscribe.htm) to keep you folks notified when updates are released. And many thanks for the feedback, very useful in terms of development, very much appreciated.

From where do you think the commands should be available, shelf? main menu? edit polygons? UI? combination of?

I'll post a video shortly of what it can do for you on a more complex mesh.

Peace & Groove.

rebo
09-05-2003, 06:11 PM
Thanks for the update.

trthing
09-05-2003, 06:19 PM
Just downloaded the new version.

Will give it a check later but I bet I will love it!

Thanks for your effort!:buttrock:

bombay_potato
09-05-2003, 08:03 PM
the new tool behavior is very cool. Nowadays I use a custom marking menu for modeling and tend not to use the standard dropdowns or shelf for the tools I use constantly.

I would suggest not to add it to the edit polygons menu as that is already very full and changes to the core menus of maya can be confusing especially if the scripts fail to load on startup.

I discovered that smooth selection worked shortly after my last post, I think its great but now with this new version I can't get it to work:
Error: Line 1.29: Wrong number of arguments on call to edge2Vertex.

glad to see smoothSplit is still included as in many cases I find that I prefer to make another operation elsewhere rather than adjust the height of the split as is the case with APEslpit. the default that was listed at "3" before works very well. -maybe left click can still be used to select for another operation and mm click could be reserved to adjust the curve height....

I tend to use the "g" repeat last command hotkey allot. MJP connect and loopcut operations are added to the repeat stack and I end up using them allot over and over again. if APELs operations could be added to the stack it would be a big plus.

thanks for the new toys, bb_potato

DesignDawg
09-05-2003, 08:15 PM
Hey all,

I don't really get this part:

You can make the commands available from the main menu and Edit Polygons by inserting the userSetup code into your userSetup.mel file.

First, what is the "Main Menu", and where is the "userSetup.mel" into which we are supposed to insert the code? Personally, before downloading this file, I had no userSetup.mel file on my computer. ?

Ricky

APELS
09-05-2003, 09:31 PM
Thanks heaps for mentioning about the bug, it's fixed now, it still occurs from time to time (especially if you select multiple edgepaths and then hit smoothSelection). These minor bugs are what I'm hoping to have totally fixed by next release, so I can start looking into making this node based.

About using the MMB for tweaking the split rather than the tool (draggerContext) coming on straight after executing a command, I hear ya, this is a feature which I'll release at the same time as making it node based, because they are mutually exclusive (dragAttrContext only works with object.attribute).

About being able to call the commands from Edit Polygons, well that's pretty much what I thought, so it's good we concur, too messy. I think another advantage with using a menu of it's own is that it's compact so can be torn off and float around without crowding the screen too bad. Spose I'll stick to that.

Thirdly and finally, "g" will repeat the last command, given you use the APELS custom menu as opposed to the UI, I don't think it's possible adding a commands launched from a window button to the repeast last action stack, at least to my knowledge, I haven't come accross one either yet.

Design Dawg, I should have added (if you don't have a userSetup.mel file simply copy this one in your scripts directory). The userSetup.mel file is used to launch whatever MEL code you wish when Maya starts up, in our case, creating an extra menu ("APELS", top right of the screen, right next to "Help").

many thanks.

bombay_potato
09-05-2003, 10:54 PM
ah I didnt understand at first regarding the repeat last operations but I loaded up the menu and got it to work as you said.

some other things I noticed, may or may not be helpful:
when testing the results on a shape that has an interior, like a volcano, the outer surface calculates fine but the interior surface especially near the lip of the opening seems to shear left or right rather than out or in. this is most noticeable when adjusting the split at the extremes.

and for a feature request, if its possible to add APEL functions to a connect edges command. This would provide more control over where new edges are created. I find that I will connect a selected group of edges much more often than just using a loop split, as it can be hard to predict just where the new edges will go across a complex surface. I understand that the poles of the resulting cut will not move due to the nature of the smoothing but if the vertices inbetween the poles can still be reliably smoothed that would be worthwhile I think...

bb_potato

APELS
09-06-2003, 12:18 AM
well sorry about this. Sometimes I thought I could see an edge go towards the right or the left but then I'd rotate around and would turn out to be just an illusion.. but maybe you really have hit a particular scenario I didn't. It would be quite helpful if you'd send me your volcano model on which the culprit occurs (just mail it to APELS @ swanimator.com)

If there is a culprit, it most probably is because of my algorithm being a substitute for de boor interpolation, which I'm still working on translating into MEL.

This script is only about 5 days old, so thanks for your tolerance. making it available this early has it's inconvenience (though I didn't post it here myself), but it also has it's advantages, such as your comments.

Thanks again for your feedback : )

APELS
09-06-2003, 01:15 AM
Well, the loops on your volcano seem a bit skewed themselve. The algorithm is performing correctly, now is that algorithm correct for what we want to do, that's another question... but anyway I'll keep this object, to compare the current results with the ones I'll get with the proper de boor polynomial interpolation working on the polys (when I'll get that working in MEL).

Oh, and having had a quick play, it looks like getting APELS to work with connect edges should work no probs.

Take care.

Per-Anders
09-07-2003, 11:17 PM
maybe you guys should check out this older thread. the actual technique is called superflange.

http://www.cgtalk.com/showthread.php?s=&threadid=83921

and here's a couple of movies of it from that thread

http://www.peranders.com/c4d8/samples/mov/superflange01.mov

http://www.peranders.com/c4d8/samples/mov/superflange01.mov

this is primitive early version though. the look further on in the thread to see it in action properly.

JIII
09-08-2003, 04:46 AM
umm is this like hand in the cookie jar or what....

APELS
09-08-2003, 11:07 AM
Pure coincidence : ) Which I won't try to defend at length because even though I came up with the idea of smooth loopsplits on my own, the idea of dragging in loopSplits like isosparms arose by discussing with other Maya users (in fact I didn't even come up with that one).

And even if it were a hand in the cookie jar, so what? People bounce ideas off each other all the time.

The real tragedy to me here is to be developing this script whilst a Maya plug will be released for it.. oh well.

APELS
09-08-2003, 03:15 PM
Oh, sorry I only just realised that plug is for Cinema 4D, and it's not freeware...

Hum, must I conclude Maya developers are more generous, and develop plugs for free? : )

Sebastian Thiel told me he wanted such a feature as APELS in BPT although he didn't know the maths involved which is why it's not got it.. (Yeah you can make it a PURE coincidence)and, I suppose many people had the same idea too but were not into scripting, or didn't have time on their hands to develop it..

The poly tools which we are going to see emerge within the next year or two are the logical evolutionary step in polygon modeling, and it's normal many people have the same ideas at the same time.. if you try to suss out who copied who you'll still be at it by Christmas.

I just hope you'll manage to enjoy and appreciate APELS as it comes, and the long hours that went into it.

Peace.

TheWraith
09-08-2003, 03:43 PM
i don't think it really matters who copies who anymore. now it's just about who does it best! and a tool like yours is taking maya one step closer to the poly functionality that max and mirai have.

rebo
09-08-2003, 04:01 PM
I would disregard that "cookie jar" comment APELS, i think we all appreciate your efforts. Whilst its true that many people have had the idea , the fact is you are the only one who has decided to put it into practise. So kudos to you.

:wavey:

I think its quite a shame and a bit off that two big Cinema4d posters decide to come here and insult someones hard work. Accusing somone of stealing is quite a serious accusation no matter how lightly its suggested. Poly modellers should be supportive of each other after all evolution of tools is in everyones best interest.

p.s. Superflange is a stupid name :p

Per-Anders
09-08-2003, 08:36 PM
rebo, i think you'll find we already had it in action if you look at the videos, if you don't believe me i can point you in the direction of several of our beta testers who frequent this board. and for a C++ plugin that does rather more than just split edges at their mid point uninteractively i see no reason for it to be free. the plugin was beta testing for several months, and can't be released till cinema 4d 8.2 comes out, superflange is just one of the tools in there, and so far we've had quite a lot of interest from maya users who want this ported to maya, xsi etc.

it's just a coincidence i'm sure, but it's odd seeing as we already posted displaying this feature last month, and then this turns up in the maya conference. and yes, it's good for ideas to be spread around, and mish mash of concepts (i do think this is true, unlike Alias), but sources matter, if you don't believe me look at the threads that have been about having your work ripped off in the general conference.

APELS
09-08-2003, 09:47 PM
What general conference? You're really coming out of the blue, and making lots of assumptions about me and what I do!

First off I'm a bit peeved reading this because:


1/ I do NOT frequent these boards

2/ I did NOT read the thread about superflange.

3/ I didn't even start this thread!!


You talk about ripping off of ideas.. but if you are a serious plugins developer you should know that if you want intellectual property over a piece of work, you have to either patent or copyright it. Approximately 100 countries are members of the Berne Convention which grants copyright protection for works which meet minimum standards, but quite frankly, even these precautions will NOT cover you for what you are moaning about since international intellectual property law only covers the source code, NOT the idea (Not that I think it's a good thing, but just to put this straight).

If you want to protect your source code then do so. "Work made for hire" copyrights last 75 years from publication or 100 years from creation whichever is shorter. Copyright infringement offenders will have to pay you up to $25,000 for first infringement and/or be imprisoned for a duration of a year.

If you own a copyright over your source code and think I ripped off your work, then get a lawyer and sue me. I have proof of having come up with this idea as a group-effort of discussion and exchanging ideas, with authorities within the world of 3D (For which I have multiple ICQ and Email logs).

Else, QUIT your derogatory unfounded accusations.

Oh, and really, if I try seeing things from your perspective, and I'd develloped this plug for cinema 4D, and found a Maya user was "copying".. well, to be perfectly honest, I'd be quite flattered. I think your plug looks great and you must be a star programmer, but trust me, the only one you are hurting here is yourself.

gundog
09-08-2003, 10:27 PM
the idea of this averaging is far from new. maya users have been talking it about as a possible solution for the maya's bevel since alias added the halfbaked tool years ago.

Per-Anders
09-08-2003, 10:27 PM
re-read my post

i said "It's a coincidence i'm sure".

what don't you understand about this statement?

i too have logs and posts from many months back, and betas fully working from months ago, but more importantly we published
publically last month the examples of a fully working model. but i'm not about to start legal proceedings. (despite your paranoid risible behaviour). i originally thought it was interesting nothing more. oh and i can claim the concept as mine simply because i did publish to the public before, not that i can do anything with that of course :D (well apart from say it as loud as possible).

APELS
09-08-2003, 10:56 PM
thanks for the precision gundog. I haven't been on the Maya boat long enough to know this, but here's hoping APELS will be the solution to this age-old problem.

mdme_sadie "it's just a coincidence I'm sure", "but it's odd seeing that X & Y...", sorry but your understatements are too strong to be overlooked!

Are you annoyed because you wanted to develop to release this as a commercial plug for Maya, and I'll be releasing a free script? If this is the case you should have contacted me.. you're going around this a very strange way, we most probably could have arranged something. had this been posted from your user base I could have understood, but coming directly from you I find this very odd and blunt.

> despite your paranoid risible behaviour

Glad to see we got more than just an idea about averaging vertices in common.

Peace out.

bentllama
09-09-2003, 02:20 AM
this thread is getting off track. I am going to close the thread before it gets too viscious. if you have personal problems with another user, please contact them in private

on a side note: there are many auto-riggers being developed. sometimes like features between 2 tools are coincidence...
...and in the odd occasion it is "borrowed", mimicry is the sincerest form of flattery.