Jump to content


Photo

Genetica 3.5 Beta 1


  • Please log in to reply
62 replies to this topic

#31 MonsterMind

MonsterMind

    Guru

  • Full Members
  • PipPipPipPipPipPipPip
  • 3,602 posts

Posted 10 November 2009 - 03:27 AM

This is very handy. Thanks.
I love swimming but more than that I love Genetica!


#32 Music3000

Music3000

    Grandmaster

  • Full Members
  • PipPipPipPipPipPip
  • 1,187 posts

Posted 10 November 2009 - 08:11 AM

Nice this will be great :thumbsup:

#33 Mr_Lamppost

Mr_Lamppost

    Forum Master

  • Full Members
  • PipPipPipPipPip
  • 468 posts

Posted 10 November 2009 - 11:50 AM

The little details are always good to know. :thumbsup:

#34 Atlas

Atlas

    Staff

  • Admin
  • PipPipPipPipPipPipPipPip
  • 4,508 posts

Posted 11 November 2009 - 03:15 PM

We also reworked the upsampling used by the Make Environment node to make 2:1 panoramas square. The improvement is most visible with non-AA previews.

LatLon_Smoothing.JPG

#35 MonsterMind

MonsterMind

    Guru

  • Full Members
  • PipPipPipPipPipPipPip
  • 3,602 posts

Posted 11 November 2009 - 10:08 PM

Nice.
I love swimming but more than that I love Genetica!


#36 Music3000

Music3000

    Grandmaster

  • Full Members
  • PipPipPipPipPipPip
  • 1,187 posts

Posted 12 November 2009 - 03:03 AM

Nice, these will be very useful thanks.

Is it customizable though? on the screenshots i can't see where the option changed but im guessing it's done through another panel/menu?

For me though apps where i can adjust things to find just the right settings/style i need are the ones i like most which is why i like genetica and other modular/nodal apps usually far more than others as you can do so much stuff. :)

#37 WillBellJr

WillBellJr

    Active Member

  • Full Members
  • PipPipPip
  • 57 posts

Posted 12 November 2009 - 11:13 AM

I've loving the Synthesis node and what can be achieved so quickly!

Perhaps I'm missing out on some settings but one of the main uses I want to get out of this node is perspective removal.

I've seen others here mention they've come from TextureMaker - I have also and that was one of the main tools I used within that program.

I like how you can mask off and isolate features out of your images with Synthesis, but a lot of times I don't need the tiling features, so I have to go through various pages to remove the jittered edges and dial down repeats to 1, and turn off stuff, etc.


I mostly need to just isolate an object out of an image, remove (or add) perspective and then use that as say input into something else.


One thing I've been trying to achieve recently is drawing with blocks (2.5D artwork) for creating concept drawings.


Say for example I want to create an image of a wall with the details (nurnies) as shown on the wall here (middle picture - ignore the spaceship for now):
http://www.meshweave...ctsnurn1top.htm

And I have a library of nurnie images like these (click the panesl to zoom into the library, click again to go back):
http://www.meshweave...iecatalogue.htm


What I'm trying to do is capture each shape into an image object I can then rotate/perspective distort, position, scale etc., into another image (again, the wall in the 2nd picture of 1st link above would be the result)


Instead of me having to fire up my 3D app and actually build all this stuff, if I could some how create an image based nurnie object library, I should be able to recreate the same image in 2D - at least to some degree as a pre-viz step before actual modeling.

Depending on the images (say like hull panels) I may be able to use the synthesized mock-up as the actual textures themselves; throw in a vent panel here, a port hatch etc render a texture (with normals??) - done!


With v3.5, I'm able to isolate the panels (sheet 2 panels #011 - #015 for example) within the Synthesis node but it's not as straight forward as I had hoped - Genetica is mostly trying to tile and blend/merge whatever I isolate which is not what I'm trying to do in this case.


Hopefully I've described what I'm trying to accomplish with the links I've provided - to generate textures or final images by placing image based objects - using SRT and perspective functions to position and scale each object within a final texture or image.


Is there a way to do this using v3.5 with the existing nodes? (I'm not all that versed in Genetica as most of you are here unfortunately)

Somehow I can envision using the synthesis AND canvas nodes in some fashion to accomplish something really grand creating images in this fashion...


Any thoughts, pointers or opinions are really appreciated!


I guess my one request would be to add one more mode within the Synthesis module to lets you isolate parts of an image to different layers, SRT (and perspective distort) the layers (parts) into a final image instead of just creating tiles.

In effect I would then be creating 2.5D images.

For creating textures - if normal maps could be created - I could place panels, grates, vents, portholes etc., and then texture that onto my 3D model pretty much without having to always add the nurnies themselves - cool! :)

-Will

PS - Thinking again, I guess I could just fire up Adobe AfterEffects and use that since I'd have full control over isolating sources, moving/placing layers, SRT / perspective distorting etc., but then I'd be missing out on all the other lovely Genetica features!

Edited by WillBellJr, 12 November 2009 - 11:21 AM.


#38 manleystanley

manleystanley

    Grandmaster

  • Full Members
  • PipPipPipPipPipPip
  • 1,328 posts

Posted 12 November 2009 - 01:03 PM

nurnies: I had to look that up, which is wild because I have a rather outrageous understanding vocabulary. Greebles I know what are, I've used Genetica to make quite a few maps to use in the displacement node of my Carrara shaders; which worked quite nicely :)

I wont say I fully comprehended what you were asking but I don't think you can do it without starting with an orthographic shot of the object.
"An eye for an eye leads to a blind world" R.T.

#39 WillBellJr

WillBellJr

    Active Member

  • Full Members
  • PipPipPip
  • 57 posts

Posted 12 November 2009 - 01:22 PM

nurnies: I had to look that up, which is wild because I have a rather outrageous understanding vocabulary. Greebles I know what are, I've used Genetica to make quite a few maps to use in the displacement node of my Carrara shaders; which worked quite nicely :)

I wont say I fully comprehended what you were asking but I don't think you can do it without starting with an orthographic shot of the object.



Hi, yes, Greebles, Nurnies, same thing - artificial high-tech looking bits...

My ultimate goal is to create the middle image on this page:
http://www.meshweave...ctsnurn1top.htm

But not in 3D but using 2D greeble cutouts, rotated, scaled, positioned (SRT - scale/rotate/translate) and perspected to look the same way - like they're placed on the wall in the final image.

Using a library of 2D greebles I should be able to achieve the same image - in theory. Basically a 2.5D image.


I agree, that I would most likely have to make orthographic renders of the greebles I make, which I could then SRT/Perspect in Genetica, but I was able to remove a good amount of the perspective from a couple of the images in the 2nd link I posted previously using the Synthesis node - I just didn't need any tiling or everything else the node was trying to do with the cutout.

That's why I'm wondering if the Synthesis node can be used to just isolate parts of images and remove perspective distortion - or is that coming (or available) in perhaps another image related node?

-Will

#40 Allen 1

Allen 1

    Alien Genius

  • Full Members
  • PipPipPipPipPipPipPipPip
  • 4,258 posts

Posted 12 November 2009 - 02:10 PM

I tried to create a basic greeble setting within the canvas and instanced it using the synthesis node. There are many ways to situate these shapes using the synthesis nose using both auto placement and manual placement of the patches. I noticed many of the greebles are cylindrical in nature on that link you have and i have not figured how to get the bevels to do that but i think it is impossible currently but squares, circles, rectangles, hexagons, octagons, pentagons, stars and triangles are possible.

Attached Images

  • greebles.jpg

Attached Files



#41 Atlas

Atlas

    Staff

  • Admin
  • PipPipPipPipPipPipPipPip
  • 4,508 posts

Posted 12 November 2009 - 02:37 PM

As you noticed, the perspective-correcting functionality of the Repeating Pattern layer type (within the Synthesis node) is specifically attuned for stitching the result into seamless textures. There's currently no feature that was specifically created for extracting part of an image and correcting its perspective without trying to make a seamless texture out of it.

Once you have head-on renders of your nurnies, it's easy to places them using the manual patch placement functionality of the Synthesis node. I can show you that part if needed.

I love those nurny images! I didn't know collections like that existed.

#42 WillBellJr

WillBellJr

    Active Member

  • Full Members
  • PipPipPip
  • 57 posts

Posted 12 November 2009 - 04:03 PM

As you noticed, the perspective-correcting functionality of the Repeating Pattern layer type (within the Synthesis node) is specifically attuned for stitching the result into seamless textures. There's currently no feature that was specifically created for extracting part of an image and correcting its perspective without trying to make a seamless texture out of it.

Once you have head-on renders of your nurnies, it's easy to places them using the manual patch placement functionality of the Synthesis node. I can show you that part if needed.

I love those nurny images! I didn't know collections like that existed.



Thanks Atlas - I'm looking forward to more documentation on v3.5's features - at least I believe you understand what I'm trying to do.

I purchased those 2 sets of Nurnies when I first discovered them myself - it's an excellent collection, especially if you're into Scifi / spaceships and all!

My thing is, the creativity comes with choosing which ones to use! - that's why I'd like to - pre-viz my designs by grabbing images of the nurnies and >positioning them< as I would have in my 3D modeler.

OR grab the images, place them as desired and then generate a texture / normal map to be used on a low-poly 3D model...




Chris, thanks for your effort but I'm not quite trying to create the nurnies using the Canvas node (yet!) - that would come in handy if I could then SRT and perspective distort those Canvas nurnies and place them at will within a target image / texture.)


What I'm speaking of is creating a "Nurnie Library" similar to the 2nd link I posted - this would contain 2D images of the individual parts.

I was originally hoping that the Synthesis node would allow me to grab those individual images, scale, rotate, position and perspective distort them to create images similar to the example images shown on the 1st list I provided.

It would in effect be like I was 3D modeling but only using the 2D images - basically drawing and creating textures in 2.5D if you will.


The Canvas node would be great for adding logos, warning chevrons, registry numbers and other paint patterns on my ship hulls etc.


Atlas, would it be hard to add image warping functionality to Genetica? The isolation stuff is pretty much there, but I'm looking to be able to pull pieces out of one image, perspective correct (or distort) the extraction and then place it (with perspective) within another image.


Hopefully this seems like a reasonable feature to have - I certainly would find it hella powerful and useful!


-Will

#43 Silkrooster

Silkrooster

    Grandmaster

  • Full Members
  • PipPipPipPipPipPip
  • 1,152 posts

Posted 12 November 2009 - 04:42 PM

If you guys are interested in nurnies/greebles or other 3rd party mathematical applications goto http://3drecursions....forum/index.php There is a lot of information pack with in that forum including links to some applications.

#44 Music3000

Music3000

    Grandmaster

  • Full Members
  • PipPipPipPipPipPip
  • 1,187 posts

Posted 13 November 2009 - 04:36 AM

Will i have requested a texture maker style extractor here before. The synthesis lab does extraction but you don't get a cropped end result like TM but rather it trys to make it seamless as said. You could do the crop outside of the lab though. I was originally thinking it would be a seperate lab but 3.5 packs it into the synthesis but there should be a seperate thing for extracting parts in my opinion.

A basic extract without repeating/spreading you do it like this -

Insert a synthesis lab, select use input yes, use a shape etc or click the edit full size and add new points etc basically just select the area wanted



Set auto to no, the display will go now



On manual tab click create new copy, you can position it and resize, i used the edit boxes to fill the hole area and center it



The difference/result



Note - Seems a value of 100 did left a visual gap so i used 101, you can render the area without transforming also and do it to a transparent background. If it is another shape you can rotate but it doesn't fill the area like TMs tool, if other tools work better though you can just export images from them and load in genetica to edit.

Edited by Music3000, 13 November 2009 - 05:00 AM.


#45 Atlas

Atlas

    Staff

  • Admin
  • PipPipPipPipPipPipPipPip
  • 4,508 posts

Posted 13 November 2009 - 11:12 AM

I was originally hoping that the Synthesis node would allow me to grab those individual images, scale, rotate, position and perspective distort them to create images similar to the example images shown on the 1st list I provided.

Yeah, I get you. The Synthesis node will do all of that except the perspective bits.

Atlas, would it be hard to add image warping functionality to Genetica?

I've added it to our to-do list. (Related requests by Music3000 and others were already there.)




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users