OTOY Forums

This is an example of an optical coherence tomography photo acoustic data set of the blood vessels in mouse skin. Visualized using automatic iso-mesh segmentation and GPU-Physically based rendering. Keep in mind the patient was alive when this was done. The process was automatic. There was no mesh modeling required. This was rendered entirely on a GPU (Nvidea GTX 480) using CUDA technology using Cinema 4D and the Octane Render Plugin.

Nice visual style, but I highly recommend to switch off the autofocus. That bumping DOF is annoying as hell.
Try to center the camera focus manually.

I agree. which is why I really wish that it would be possible to do at least a minimal amout of KEYFRAMING inside Octane. Doing a variabel DOF inside Cinema and animating takes alot of time, because the mesh needs to be reloaded between every frame into octane. Which in this case triples the over all processing time per frame.
ny Ideas?

You can use the --cam-focaldepth command to focus to a specific length.
Or you can use a --cam-aperture value of 0.01 which should make all sharp.

face

This would give me a static DOF. But if I want a variable one, which focuses on areas of intrest without automatically jumping to the center, I have to do this inside Cinema. Which is VERY time consuming during the rendering phase, as the geometry needs to be reloaded every time. The same can be said for moving lights, and camera flight path. These things really need to be able to be KEYFRAMED inside of OCTANE. Or at least done so that the only thing that Octane loads between frames are parameters, and not the entire geometry.

I don´t know how Cinema that handles.
I think the focalDepth in the exporter has an alternative script name.
With that, it should be possible to calculate the distance between the camera and the interest and send the data to the script parameter.

Thats realy simple to calculate,
CV = cameraVector
iV = interestVector
dV = distanceVector

dV = cV - iV
distance = sqr(dV_x*dV_x + dV_y*dV_y + dV_z*dV_z)

Here is an old vidio, how we can handle it in Softimage...

I don´t know how Cinema that handles.
I think the focalDepth in the exporter has an alternative script name.
With that, it should be possible to calculate the distance between the camera and the interest and send the data to the script parameter.

Thats realy simple to calculate,
CV = cameraVector
iV = interestVector
dV = distanceVector

dV = cV - iV
distance = sqr(dV_x*dV_x + dV_y*dV_y + dV_z*dV_z)

Here is an old vidio, how we can handle it in Softimage...

[vimeo]http://www.vimeo.com/11891062[/vimeo]

face

Thanks. I do not know if this would in Cinema, but I will give it a try

Maybe you can use an expression on the parameter in the GUI...

face

Hi Carl,

The first thing you have to do, is to enable the export of the "Focal Depth" and tick the box "of Camera Object". When that is done the focus distance of the exported camera will be the the focus of the C4D camera.

Then just add a target object, which can be a null object and a target expression to your C4D camera object. Drag the target object into the link field "Target Object" of the expression tag and you are ready to go.

Now the camera is always pointing towards the target object and will automatically keep it in focus

Cheers,
Marcus

Was that animation actually done using the turn table animation feature of Octane? If it was, just disable the autofocus in Octane and use the focus of the C4D export.

Cheers,
Marcus