Ugly non uniformal bisier splines – DIE !!!!!

(click below to see video (a MUST SEE)

Rhonda is here :

ok, ok .. we are all quick to announce deaths when new things show up. painting was dead when photography was invented. softimage was dead when it was bought by autodesk, i think maya too :)

Ok Ok rahonda is probably not threatening rhino.  Or nubs in general. But still i think it could be very nice tool for two things.  three might be.

1. sketching. the speed u can (it seems) create is incredible (gooooogles sketch up – look out :)

2. creating models for animation films, in very distinct style. (well i wonder how exploitable models will be, and will it be only lines or will we be able to make surfaces out of them, and how.)

3. making 3d more accessable to people who do not belong to this industry. again look at sketch up.

imagine something as rhonda as interface for drawing nurbs… that would make me create some models with nurbs instead of polys for sure.

So i had this idea about how to make clouds in 3d.

I think it works quite well, but as all things in life have draw backs. First – the image:

clouds

And here is how i did this.

The foreground clouds are done using a blob mesh rendered in mental ray using a depth of field. The background clouds are simple 3d Max volume fogs.

My goal  here was to create clouds that could be quickly rendered and quite easy to do. I used a standard volume fog in the background which was computationally effective, and in the foreground ablob mesh (also known as implicit surfaces) was used to create foreground clouds. The foreground clouds were rendered in mental ray, with a depth of field.

Depth of field adds this cloud like apearence.  Softness, which is quite tricky to do with simplle material, and fallofs in opacity chanal.

So even though i like results of this experiment, unfortunately its not the fastest, i meen rendering time. but oh well :)

and a screenshot:

one other good thing .  your mech topology can be as ugly as it gets, depth of field hids it :)

so its one way of using inplicit surfaces practicaly.

—————–

further readings at veterrain.org

(look at section research)

Lately I was working with some textures. I had a model, and for texture work in zbruch and photoshop,
I had to use unwrap uwv. I guess that’s a term in 3d max.
Not sure how you call it in other software, but I guess name shuld be similar as well as function.
So i selected parts of my 3d model, corrected seems, and used pelt mapping in most cases.
Again actually it does not mater. the main idea is that at the end i had something like this:
So when it comes to zbrush naturally you don’t see any seems where two texture blocks connect.
You work without knowing that they exist. (more or less)
But when it comes to photoshop, You yourself must make sure that all pixels corresponding to other texture
blocks pixels (at the edges of texture blocks) Are either same or very similar color, so that when you use
your texture back in your 3d program, no one would see where texture blocks connect.

But how on earth do you do that?
how do you manipulate image so that each edge pixel in two separated texture blocks would be the same?
I am sure there must be solutions, or work arounds. I would be very happy to hear how you do it.

But I was thinking, what if we would have a plug in for photoshop ( gimp? corel photo paint? …? )
that would help doing that?

the idea is to create two “guidelines” for both texture blocks, they could resemble curves in photoshop.
So first here is an image to better understand what do i talk about:

so if, lets say i use brush tool and draw in one block of texture, and at some point brush moves out
of that texture block, crossing one seem, and  “magically” it appears in other texture block, crossing second seem.
it would be like working with tilling textures. or better example, like this OLD game “snakes”, where u have to
eat stuff, and as a result you get longer, but you can not bump in to yourself. but you can hit the right side of
your screen, and you will end up in left side of screen. The idea is the same.
You hit green seem line from one side, and you appear crossing also read seem in other texture block.

OK and now lets imagine a bit more complex situation.
well might be its not more complex, just for a sake of this example.

so if, lets say i use brush tool and draw in one block of texture, and at some point brush moves out
of that texture block, crossing one seem, and  “magically” it appears in other texture block, crossing second seem.
it would be like working with tilling textures. or better example, like this OLD game “snakes”, where u have to
eat stuff, and as a result you get longer, but you can not bump in to yourself. but you can hit the right side of
your screen, and you will end up in left side of screen. The idea is the same.
You hit green seem line from one side, and you appear crossing also read seem in other texture block.

OK and now lets imagine a bit more complex situation.
well might be its not more complex, just for a sake of this example.

so we see a situation where same edge is shared in 3 texture blocks, and how the same edge represented in
each block has different length and shape.

OK my representation of “free transform” tool is not very accurate, but i hope its good enough for understanding.
I guess “Free transform” tool is quite limited, therefore you would be able to control it only on one texture block,
and other one (on next texture blog) would be automatically generated…  or we write free transform tool too :)
Now in a case of brush (look at second image) it might at some points appear in 3 different locations on screen, on 3
different texture blocks.

well, any one needs an idea for computer graphicks degree work? That could be it :)

p.s.
If such tool wuld exist…
the author could go further, and write another plugin for max/maya/xsi to eport seems….. or?
it would take years to define all seems in photoshop for more coplex objects..

Lots of real time as well as non real time 3d solutions have same strategies dealing with intersecting meshes.  We find that in havok physics engine, any cloth or hair simulator an so on. Regardless of average and very good results these tools provided, we still sometimes find objects or meshes which are intersecting.

That might happen due to many reasons.  sometimes its imperfect physics simulation tools, but more likely,  simple computing restrictions (not all the objects, in their highest resolution,) can be added to physics engine – it would take years to calculate “everything”.  and also sometimes some objects are key framed, and do not obey physics engine, i dono if that is a real reason, but i imagine physics engine could, in some situations, have hard time dealing with “restricted” situations.  Also small spaces, different scales and so on….   all these things do not help to achieve best results.

There is a long history of research as well as commercial tools unwalable. realflow is in 4 version, syflex is in 3,9 version.

But I have not heard yet of a research on “hiding” problems instead of solving them.  what does rendering engine do if we have two polygons in exactly same space?

From my experience they usually go mad. the shading becomes really fu$$$$ up.  therefore my question is, is there any research done on hiding problems in situations where we cannot solve them?  Imagine tree leaves moving in a wind.. it takes quite a bit of computing power to make sure that non of each individual leaves would intersect with each other.  would it be cool to mask these problems if they happen to exist?

Also we often use two or more intersecting planes for making trees. In many games we make trees like this:

its only two polygons, 4 triangles i think, and both are with texture of tree, plus alpha map, so we see only tree shape, we shuld not see shape of real geometry of our 3d tree.

what if renderer would automatically blend textures in intersection area, so we don’t see a line?

Is it already done? mightbe it used to happend only lon time ago?

anyone knows?

leave me a comment if u know something about the topic.

So I was rendering other day and a strange thing happened.

Suddenly mental ray decided thats its not enough to calculate 100 % of photon emission before starting rendering.  And it reads “rendering” instead of “photon emission”…

oh well, must be one of these Christmas miracles i guess.

So and then I remembered once I had my video card setup wrong in max, and it produced very nice results.  I would like to investigate visual qualities of this error. Not sure though what to do with them exactly.

You can still somehow see the 3d scene (with assistance of strong imagination).
So here are fiew screan grabs I did:

So, recently I was working a bit with architectural visualizations.
And I ran into a small problem.  So there are these people who are always very tight and do things in “right” order. But there are others who prefer “creative mess” so to speak.
And I am sure both ways can be good.  sometimes at least :)
In my case i have a tendency to get lost in modeling stage,  adding more and more small details.  So at the end of the day I have a scene full with detail, but no lights.  And some say it is a good idea to draft your scene from simple cubes first, then to make “ruff” lighting,  so later on you will need just to tweak minor details.  But when you have a billion of objects, and its hard to navigate in view port… you add a light and it takes years to render that small test.  So its kind of stupid way to go.
So I was thinking. (This is about 3d max only i guess)
So lets imagine situation where we have heavy scene with dense mesh,  something like this :

interior, lots of polys

and then lets turn on view port rendering mode “show as box”

interior, view as bounding boxes

So if max can generate bounding boxes, it would be nice to have a tool which could do the same but export a mesh of these boxes.

Even more advanced tool could somehow *(no idea how, still thinking) detect objects which resemble boxes (walls) and more complex shaped objects (trees) and by using multi res tool (which reduces number of polygons) automatically generate whole scene only for test renders. That would be time saver,  I guess,  or?

What do you think? might be there is similar tool already but I am not aware of it? let me know if thats the case!

and as usual link to my other website here where you can find interior used as example rendered. And dyrect link is here. By the way its pdf :)

Normally i don’t write such posts, but sins xsi was a long time dream to me…

i was looking in to xsi from version 4 i guess, gathering courage to move from max to xsi. At first i thought i should learn maya, but after autodesk bought it i was very serious to move to xsi. after ice release i downloaded demo and started playing around with it. Xsi always had this feeling of independence, and being not pro monopolist. Thinking about artist and so on.. And sins i am not a programmer (scripter) ice looked very apealing to me. But now???

What will it mean – if all biggest 3d aps belong to one company?

who dies first? well i am sure you will find much more about topic in base forums or wherever, but i just couldn’t resist…

on other hand: could it be a good thing?more money for developement, sins it seems avid was not very lucky with xsi lately….

I am deeply shocked.

what i would be interested is opinions from maya and max users who deal with autodesk.

any thoughts?

ps. “optimistick” article

“The World Stops Spinning – Autodesk Acquires Softimage” at  xsi user base

autodesk ’s oficial statement

and short article in 3d max underground, summarizing general user reaction here

thread on cgtalk’s 3dmax forum here

///////////////////////////////////////////////////////////////////

insiders jokes:

teapot

xsi_baseball_bat

standart icon

So as we can see in a previous post (if we read comments) there where some concerns about my proposals validity. Mainly Delt0r raised some valid questions. therefore i have been struggling to improve my initial idea, and here is what i came up with. Now, its still has some flows. and questions. but its for u to judge, sins i have no clue about maths, and here its quite important, i mean the math.

so the idea is this.

A. we have a mesh. *(base mesh) which we will use to to determine where the “blobs” appear.

B. we will get the normals of each vertex.

c. then the slicing planes should be generated. (here u can look at my old post or drawings beneath)

here comes the first problem. what i want to find out – is there a way to limit this “slice plane’s ” effect. here no one will understand me, so lets have a look at the picture:

screen shot of blob mesh section in max

Here we have a blob mesh, and a slice tool. the white plane is a slice plane, and green line is outline which is generated. the problem is that the green outline goes out of boundaries of slice plane, and correct me if iam wrong but iam sure algorithm behind this procedure works this way. so what we would like to have here would look like this:

blob mesh, fake intersection

Ok how to do it ? i have no clue. It might require rewriting all algorithms behind this operation, or doing it old way, and then subtracting unneeded parts. which is probably easier.

so now lets imagine our slicing works like we want it, and lets move on.

so in order to understand my “hand” drawings here is another diagram for explanation:

normals

so this diagram shows a mesh we will be thinking about. in next drawings i draw only 3 vertexis out of this mesh. In this image we can also see averaged normals (left)cos normally each plane is plane :) so all 4 vertexies should like in image on right. but we can get average values.

how to make planes

you can enlarge this image to see better. but the idea is to make planes for each vertex separately.

so here as well we have problems. first of all we would want to arrange planes so the would make a “continuity” i mean they would go like this: /\/\/\/\/\/\ shit i have no idea how to explane…. look at the drawings again. chm…. please tell me if anyone understands what i am talking about ? ah? anyone?

so and last step would be to “weld” vertexies who are very close to each other.

again i wonder how easy it would be to make so many “restricted” planes for slicing.

but from what i understand, if it would work it could improve mesh topology or?

any ideas?

Ok, to tell u the truth its not a review sins there is my name in the credits of film :)

so i am rather biased :) So as film dyrector states himself:

Alican Serbest:
I didn’t study cinema but educated myself and gathered a team of my friends, educated them as well :) Then shot short movies for 8 years. This was the latest of them. We are trying to get into the professional area :)

synopsis:

It is year 2092 and vegetation is almost completely perished all over the world.
A police team has just followed a suspect to a roof.
There, they will learn a bitter truth that will affect all human race.

so no biggy, but still for me = cool to see turkish movie.

good whaching THE LINK

Hello everyone. So I was (am) having this problem. I rendered some animation, but now i think i need to slow it down.
The problem is, rerendering everything and making everything twice as long is not an option, sins one frames rendering time is around an hour.
so i was thinking how to slow animation without rerendering… how to get more frames…
and then i remembered video morphing. So. first of all have to mention that the content of my animation is water surface.
Just water with waves. thats it. No characters or theyr faces.
So if anyone knows that this is already done, please let me know, i would like to try employ this technique in my work.
But the basick idea is, to write a script or separate program which tasks would be>
1. detect motion in frames. thats already possible in lots of software. think of motion tracking. the idea is we detect a part of video frame which is similar in next frame but its position changes. So this is already possible to do with lots of softwares. If i am not mistaken also it is part of mpeg video compression.
now step 2, would be to connect this automatic process with video morphing. You remember this “effect” when we have a photo of one person and it “nicely” “morphs”
to other person. So thats exactly what i was thinking to do.
To take two frames. detect “objects” using motion tracking.
and then user would need to say how many frames in between he wants to add. so lets say adding one frame would increase video time twice. two frames would make total time 3 times long.
and then in ideal case program or script would make inbetweens.
The problem here is that user usually has to “tell” program how the morphing is done. Usually user has to draw curves around object, then take that curve put it on next frame and  move points of to mach next object.
well if someone has a better descrotion, pls correct me :)
i hope its still kind of understandable.
ok ill try to find some links, which is hard if i dont whant to advertise any specifict software product :)

And other VIDEO
it shows how its done, i mean the curve drawing.

SHIIIIITTT.

so

couple of hints for bloging.

first chek wiki. i forgot. thats a link.

but thats not a mistake. mistake is DO YOUR RESEARCH BEFOREU WRITE!

so, ofcource i am not a first to come up with this techniqe.

read this article!