Page 650 of 739 FirstFirst ... 150 550 600 640648649650651652 660 700 ... LastLast
Results 12,981 to 13,000 of 14766
  1. #12981
    Haha, you did it (i man separate and cleaned core volume code from my fat almost-obfuscated-patch), conglatualtions!. At this state, i barely can understand my patch parts and fear to touch code that not lookes 2 weeks, it become so unrealistically complex to maintain and fix.

    You even do some one step sampling from light direction (limited bidirectional), i have no that parts, it is Stuart invention, it help to clean noise around light a lot.

    About "g" parameter, it is like Phong specular power, but related to volume particles. of 0, it do uniform sphere distribution (smoke?), if 0.6-0.8, more close to human tissue like skin/blood, and at extreme cases close to 1 it go to transparent material like colored glass (not recommended, big computation overflow errors). Must note, that water vapor and small droplets in air have very complex Phase function, with many peaks, and related to wave lenght, current Henyey-Greenstein model too simple to make that, so do not expect to see rainbow. I think that "g" around 0.7-0.8 make fog looks more like real street fog.

    Beware, real media scattering is multi bounced, and you need 2-3 bounces to get light that "bend" around objects, it very slow, for example this is 15 hour render on i3770K (my patch used, but sure DingTo variant can do same w/o limitation). One bounce (or 0) scenes can ve rendered on reasonable speed, ofc.
    Click image for larger version. 

Name:	Снимок экрана и&#1.jpg 
Views:	144 
Size:	428.4 KB 
ID:	258445
    Last edited by storm_st; 08-Sep-13 at 05:44.



  2. #12982
    Originally Posted by storm_st View Post
    Beware, real media scattering is multi bounced, and you need 2-3 bounces to get light that "bend" around objects, it very slow, for example this is 15 hour render
    but how much can a scene like that expect to be optimized?, for cycles, which is meant for animation.

    I don't know how long it takes in other render engines, so I have nothing to compare too. I know its early, just curious.
    MY BLENDER SKETCHBOOK (updated 1st september)
    Computer Specs WIN7 / I7 3930 / 32gb Ram / Gtx 560 TI 2GB



  3. #12983
    Unfortunatly. no, correct multi-scattering integral that hard to compute, it is principal. More hard in complex setup, that bike is "cheating", light paths are very simple. Most ppl will stick to single scatter, it work nice for many scenes. About other engines, they just show you single scattering in gallery, and usually avoid render time/hardware info. I think Luxrender can do same in compare time in bidir mode, but MLT samler (Lux default to MLT) sure will be much faster in average scene.

    For example, same scene, 0 bounce (single scattering), 64 samples, 5 min

    Click image for larger version. 

Name:	Снимок экрана и&#1.jpg 
Views:	62 
Size:	611.0 KB 
ID:	258472

    256 samples, 20 min
    Click image for larger version. 

Name:	Снимок экрана и&#1.jpg 
Views:	83 
Size:	448.2 KB 
ID:	258473

    you can notice how important is secondary bounce in volume rendering, it have big difference, and only 1, must be 8+ but i have no patience to do it now .

    Edit:

    Must note as well, i am still not sure about correctness of light measure in my patch, it can be some bug lurking, but pictures looks cose to what i expect to see.
    Last edited by storm_st; 08-Sep-13 at 07:59.



  4. #12984
    Member
    Join Date
    Dec 2011
    Location
    Oxford
    Posts
    173
    The volumetric patch is still at quite an early stage. We have loads of cleaning and optimizing still to do before considering adding new functionality. For example I am sure the Equiangular sampler is wrong, although it gives nice looking results. The (exp) functions are just rewrites I made of the sampling algorithms. There are subtle differences but they were more for my testing.

    The hair shader is on its way. Only a very basic one for now though. I sent a patch to Brecht yesterday and if its ok I will commit it.



  5. #12985
    well if single scatter is the only way to speed up renders, I guess we don't have choice (in animation)

    15hours per frame render is not an option haha. I know its early and it probably shouldn't even be "tested", but from what storm has said, it seems with scatter set to 2 or above, very slow renders are unavoidable for something like volumes. ( I can't imagine its possible to bring 15hrs down to 2-3hrs)

    I think the single scatter results are still good. Really i'm interested in smoke, we'll see how this gsoc goes in coming weeks
    MY BLENDER SKETCHBOOK (updated 1st september)
    Computer Specs WIN7 / I7 3930 / 32gb Ram / Gtx 560 TI 2GB



  6. #12986
    Member Bao2's Avatar
    Join Date
    Aug 2009
    Location
    Galicia - Spain
    Posts
    4,832
    Originally Posted by storm_st View Post
    you can notice how important is secondary bounce in volume rendering, it have big difference,
    First image in 12996 blurred and screen blend mode above original and we have for free a similar look to post 12994:



    Originally Posted by ng-material View Post
    ( I can't imagine its possible to bring 15hrs down to 2-3hrs)
    Use post production thingies
    Last edited by Bao2; 08-Sep-13 at 11:12.
    >>> Neytiri WIP <<<
    When explanations are complex you know there is no Truth on it.
    Tangled Fan Art



  7. #12987
    Originally Posted by Bao2 View Post



    Use post production thingies
    Yes ofcourse


    Brecht just committed the thingy node. Its solves just about every problem.

    good result btw.
    MY BLENDER SKETCHBOOK (updated 1st september)
    Computer Specs WIN7 / I7 3930 / 32gb Ram / Gtx 560 TI 2GB



  8. #12988
    Member J_the_Ninja's Avatar
    Join Date
    Sep 2010
    Location
    Portland, OR, USA
    Posts
    1,185
    Originally Posted by ng-material View Post
    Yes ofcourse


    Brecht just committed the thingy node. Its solves just about every problem.

    good result btw.
    Nah, we've had that node for awhile. :P

    Name:  Screen Shot 2013-09-08 at 9.01.12 AM.png
Views: 1241
Size:  37.7 KB



  9. #12989
    Member Ace Dragon's Avatar
    Join Date
    Feb 2006
    Location
    Wichita Kansas (USA)
    Posts
    22,049
    Originally Posted by ng-material View Post
    but how much can a scene like that expect to be optimized?, for cycles, which is meant for animation.
    It would seem to me like it would indeed be possible for the volumetrics to be optimized a bit more if we follow the information given by the Arnold developers at SIGGRAPH in terms of their fast volumetric shading, as I believe Brecht has stated that they revealed some info on how they did it.

    Another big thing that would speed up the volume shading (and many other lighting scenarios), would be bidirectional sampling (which would be a serious speedup even without metropolis sampling added on top).
    Sweet Dragon dreams, lovely Dragon kisses, gorgeous Dragon hugs. How sweet would life be to romp with Dragons, teasing you with their fire and you being in their games, perhaps they can even turn you into one as well.
    Need Cycles materials / node setups: Get them here



  10. #12990

    clouds

    A test with clouds, 400 samples took around 4 hours . The new work in volumetrics is looking great, gotta say thanks to to Dingto, Stuart, and Storm. Also the new sky model is looking great too.
    Click image for larger version. 

Name:	clouds.png 
Views:	204 
Size:	1.51 MB 
ID:	258601



  11. #12991
    Member
    Join Date
    Dec 2009
    Location
    Vancouver Canada
    Posts
    118
    Any chance of a node settings screengrab from Indy_Logic or dailerob? I can't seem to get any volume areas to be fully transparent or get a sharp separation between dense and less dense areas.



  12. #12992
    Member DingTo's Avatar
    Join Date
    Mar 2008
    Location
    Southern Germany
    Posts
    1,508
    @dailerob: Great render, thanks for sharing!

    Really excited what will be possible with Volumetrics. We plan to commit the patch to my branch soon, and will continue development there.
    Blender developer (Cycles / User Interface)
    Blog -- Blender Podcast -- OpenShading -- BlenderDay -- Blender Network profile



  13. #12993
    what is actually required to render realistic good looking clouds?

    is it just instanced volume objects?

    for something like a cumulus clouds, you'd probably need a smoke volume?. Right now blender can't really do a convincing "large scale" things like pyroclastic style smoke, especially when it comes to the movement. (google pyroclastic if you need too)
    MY BLENDER SKETCHBOOK (updated 1st september)
    Computer Specs WIN7 / I7 3930 / 32gb Ram / Gtx 560 TI 2GB



  14. #12994
    Volumetrics inching ever closer, great work!

    Is there a chance this will be in the main branch soon, at least in the experimental settings?



  15. #12995
    Member DingTo's Avatar
    Join Date
    Mar 2008
    Location
    Southern Germany
    Posts
    1,508
    It's too early for trunk, development will be done in the soc-2013-dingto branch soon.
    Blender developer (Cycles / User Interface)
    Blog -- Blender Podcast -- OpenShading -- BlenderDay -- Blender Network profile



  16. #12996
    Member JuanJosé Torres's Avatar
    Join Date
    Nov 2011
    Location
    Madrid, Spain
    Posts
    622
    Originally Posted by DingTo View Post
    It's too early for trunk, development will be done in the soc-2013-dingto branch soon.
    Oh my.. I must get me a build of your branch as soon as you start commiting volumetrics... These test look amazing...



  17. #12997
    Member
    Join Date
    Apr 2013
    Location
    In XYZ Space
    Posts
    10
    Originally Posted by dailerob View Post
    A test with clouds, 400 samples took around 4 hours . The new work in volumetrics is looking great, gotta say thanks to to Dingto, Stuart, and Storm. Also the new sky model is looking great too.
    Click image for larger version. 

Name:	clouds.png 
Views:	204 
Size:	1.51 MB 
ID:	258601
    Wow this looks amazing!! When I saw some first volumetric renders two days ago I thought we'd still be far from something like that.

    And 4 hours rendertime isn't bad at all for a very first render with working code I think. When this will eventually run on the GPU some time in the future the speed will increase dramatically



  18. #12998
    Two interesting things to try with volume rendering:

    • Displace the texture coordinate, so e.g. position + output color of a noise texture, and plug that into another texture. This way you're not just carving out a volume but displacing the surface of the volume too.
    • A relatively cheap multiple scattering approximation is to mix single scattering with different densities. So you would have one volume node with the normal density, and other ones with 1/2, 1/4, 1/8, .. of the density. That way you get some detail but also light that scatters from further away, so e.g. the backside of clouds isn't so dark. I'm not sure about the best mix weights to use for this.
    Last edited by brecht; 08-Sep-13 at 20:28.



  19. #12999
    Member Kai Kostack's Avatar
    Join Date
    Jun 2006
    Location
    Berlin, Germany
    Posts
    836
    Yay, rendering without geometry is fun.

    Click image for larger version. 

Name:	volumetric Lunar Surface.jpg 
Views:	97 
Size:	1.29 MB 
ID:	258628

    This image took 19 h for 40 samples (Woodcock).



  20. #13000
    Member Indy_logic's Avatar
    Join Date
    Nov 2008
    Location
    Portland, Oregon
    Posts
    552
    Originally Posted by jandress View Post
    Any chance of a node settings screengrab from Indy_Logic or dailerob? I can't seem to get any volume areas to be fully transparent or get a sharp separation between dense and less dense areas.
    Just just have to recompile and then I'll post a screen shot. But, don't despair! All you have to do is insert a brightness and contrast node just after the texture. This give you control over the sharpness of the texture. in most of the examples I posted all I did was to use a math node (set for subtract) to combine two textures. in the last two it was a Musgrave with a Voronoi (set to cells) subtracted from it.

    But I think the key to get what you are after is to use a brightness and contrast node. Don't forget to increase the overall volume density in the render settings to make it nice and dense. :-)

    Originally Posted by brecht View Post
    Two interesting things to try with volume rendering:

    • Displace the texture coordinate, so e.g. position + output color of a noise texture, and plug that into another texture. This way you're not just carving out a volume but displacing the surface of the volume too.
    • A relatively cheap multiple scattering approximation is to mix single scattering with different densities. So you would have one volume node with the normal density, and other ones with 1/2, 1/4, 1/8, .. of the density. That way you get some detail but also light that scatters from further away, so e.g. the backside of clouds isn't so dark. I'm not sure about the best mix weights to use for this.
    Sounds really cool but my brain can't connect the dots unless there are more dots to connect. ;-) I don't understand what you mean by position + output color? Position of what?

    Also, how does one set the volume for single scattering?

    Thanks for the tips!
    Last edited by Indy_logic; 08-Sep-13 at 21:19.
    Have a look at my Latest WIP



Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •