Shade maps? how does it work?

How does the new shade maps help to speed up DOF and motion blur?

regards.
/Tobias

It is quite simple:

(*) Add the stereo helper to your scene;
(*) Set the mode to “render shade map”
(*) Set some file name in the box below to hold the shade map;

Now when rendering, this will create a shade map for the current camera position. If you are rendering an animation, you must render the entire animation in this mode.

(*) Once you have the shade map(s), set the mode to “Use shade map”;
(*) If you don’t need the stereo features, set the stereo helper to render only the “left” view and set the “eye distance” to 0.0

Now when you render, V-Ray will attempt to use the saved shade map where possible. You can see the places where the shade map was used in a special “Shade map” render element if you are using the V-Ray VFB.

You can adjust DOF settings without recalculating the shade map(s), but for motion blur it needs to be recalculated.

Best regards,
Vlado

Very cool feature! Does this compare in quality and speed to post effects like Final DOF?

regards.
/Tobias

The scene is still raytraced so it will never be as fast as a post effect. But it is more accurate, especially when combining DOF with motion blur.

Best regards,
Vlado

The scene is still raytraced Thats exactly what I wanted to hear! :smile: Thanks!
regards.
/Tobias

Is there any speed improvement for anything else than DOF and MB ?

yes, stereo :stuck_out_tongue:

You mean even for a stereo image that as no DOF or MB ?
Have to do more tests as it is no yet very obvious for me :wink:

Well yes. You have a right view and a left view wich are only slightly different viewpoints, so the helper calculates the shademap in the first view and greatly speeds up the second.

Regards,
Thorsten

hum,
what is the workflow for camera animation ?
1) Render Gi prepass for my sequence. (no stereo)
2) Activate the vraystereoscopic helper.
3) render the shademap for all sequence.
4) finaly, reuse shademap to render both eye

does the render shademap pass needs fullrender BTW ?

Thanks for helping :slight_smile:

This is the workflow, yes. The shademap needs a full render (in terms of resolution and render settings quality).

You can also look at the “Shade map” render element in the V-Ray VFB - it will show you the areas where the shade map was reused. If you notice a lot of black areas, then you may need to increase the “Reuse threshold” parameter.

Also, be sure to specify a name for the shade map file - I often forget this one… :slight_smile:

Best regards,
Vlado

So when you want to render the initial shade map, you need to render the entire scene once (this is for a still job) just to get the shademap, or is it possible to just render a shade map some faster way?

b

It needs to render the entire scene, yes. There is no point to use any shortcuts as that would degrade the final quality afterwards. However dof and motion blur will not be calculated when the shade map is computed, so that part would go a bit faster.

Best regards,
Vlado

Thanks for the clarity. It certainly did render a cleaner DOF version faster!

b

I have to says SM are great for Stereo renderings !

Just to be sure as i didn’t tryed :
Can i disable all RE for SM prepass ?
I always have some occulsion pass and others that can be quite long to render !

Thanks

Anyone on this ?
I’ve trouble getting RE when using Shademaps !