Announcement

Collapse
No announcement yet.

Use Camera Path Explanation

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Use Camera Path Explanation

    Hallo,

    this time a little less verbose:

    I am, currently rendering an animation with a moving camera and moving objects with "Brute Force" set as primary engine and "Light Cache" set as secondary engine. I simply followed Vlado's advise here who said this method was the best to avoid any flickering in these kind of animations.
    However, I don't understand what exactly the Light Cache and Irradiance Map option "Use Camera Path" is doing internally. It somewhat lacks a thorough documentation.


    How does this technique fit in between normal and "Fly Through" mode?

    In my head I somehow think that "Fly Through" and "Use Camera Path" do the same, but obviously this isn't the case. How do they differ?


    What technique/algorithm does it use that makes "Use Camera Path" able to even deal with moving objects in a scene?


    Also, how high should the SubDivs value be set for the Light Cache with "Use Camera Path" On?


    Please, do not shy away from a technical explanation if this is necessary.


    I thank you all very much for any input!


    Capture_MM

  • #2
    The "Use camera path" option simply reduces flickering introduced by the fact that the light cache samples might end up in different positions on each frame because of the camera movement. It doesn't do anything about the moving objects specifically.

    It works by shooting rays from the camera path for all frames rather than just the current frame, so light cache samples on static objects are more consistent between one frame and the next. They will still be different for moving objects though, but it's better than nothing.

    Does that make sense?

    Best regards,
    Vlado
    I only act like I know everything, Rogers.

    Comment


    • #3
      Thanks Vlado. So with "Use Camera Path" ON, the camera first goes through the entire animation sequence and shoots LC rays AT EVERY frame onto STATIC as well as DYNAMIC objects? At every new frame, LC samples on dynamic objects are discarded while samples for static objects are indrementally added to a map?

      So all in all, the flicker on dynamic objects won't be reduced, only on static?

      Also, how do I calculate the amount of Subdivs I need for the LC in "Use Camera path" mode? I guess the Subdivs value will have to be high enough to cover the entire animation path? Or does the Subdiv value still only pertain to a single frame?

      Comment


      • #4
        Originally posted by CAPTURE_MM View Post
        So all in all, the flicker on dynamic objects won't be reduced, only on static?
        Yes.

        Also, how do I calculate the amount of Subdivs I need for the LC in "Use Camera path" mode? I guess the Subdivs value will have to be high enough to cover the entire animation path? Or does the Subdiv value still only pertain to a single frame?
        Render one frame with light cache as both primary and secondary bounces and see how "noisy" the light cache is - if it is too noisy, you will have to increase the subdivs. This mostly depends on what distance the camera travels through the scene.

        However, it might be easier to just forget about this and render a new light cache for each frame. The docs page here suggests some suitable settings for animation in V-Ray 3.20; they work in a large number of cases with or without the "use camera path" mode:

        http://docs.chaosgroup.com/display/V...nginanimations

        (see the section "Avoiding flickering in animations").

        Best regards,
        Vlado
        I only act like I know everything, Rogers.

        Comment


        • #5
          Originally posted by vlado View Post
          However, it might be easier to just forget about this and render a new light cache for each frame.
          You mean with "Single Frame" mode? SO that way no LC map is ever saved to disk, right? All LC data is calculated, kept in memory during rendering and then discarded at the end of a frame?

          And with "Use Camera Path" we are generating one huge incremental LC map file?

          Comment


          • #6
            Originally posted by CAPTURE_MM View Post
            You mean with "Single Frame" mode? SO that way no LC map is ever saved to disk, right? All LC data is calculated, kept in memory during rendering and then discarded at the end of a frame?
            This is correct, yes.

            And with "Use Camera Path" we are generating one huge incremental LC map file?
            No, this option still calculates a separate light cache for each frame. You can think of it as an extreme case of camera motion blur.

            I see this is confusing for you, so just forget about this option. It will be fine without it.

            Best regards,
            Vlado
            Last edited by vlado; 13-07-2015, 06:18 AM.
            I only act like I know everything, Rogers.

            Comment

            Working...
            X