vrayvelocity max velocity

[pailhead]

Hi,

I can't seem to find any reference about the parameters for this element. I need to output a velocity map for AE. I left the settings at default and tried to output to an .exr. I definitely got some data in it, but the results were not quite as expected. I figure the max velocity parameter is important, but i have no idea what it is.

[instinct]

If your target app properly supports float data (and you are saving to a float supporting format) you can leave max at 1.0 and get correct absolute results.

In this case the pixel value means "pixels per frame". So if your velocity element contains -4 in the red channel than this pixel is supposed to move 4 pixels to the left in that frame.

If the application (or the plugin doing the motion blur?) does expect values encoded differently you might have to change that. Setting the Maximum value means, that the values
will be mapped to that range. If you set max velocity to 4 then a velocity value of 4 will be saved as 1.0. This way you can make sure your velocities fit into 0-1 for non float formats
or applications.

The non-editable "max velocity in last frame" spinner can be used to determine the max velocity in the last frame. So basically you would choose the frame with the fastest movement,
render it, take that value, add a bit of margin and use that for max. Note that this will mostly lead to eyeballing, whereas when saved as absolute values you can actually work with
shuttertimes and settings correctly if your target app supports that. (e.g. nuke)

Regards,
Thorsten

[wxyz]

Anyone know if baseline AE can use float data correctly?

[Morbid Angel]

Anyone know if baseline AE can use float data correctly?

I don't see why it wouldn't, your project has to be in 32 bit float though. Another tip is that if you wish to use float data you must uncheck clamp, and leave max vel at 1.0

[3LP]

Hey

I did that research couple of months ago, and if turn out that the Developpers of RSMB recommend to put the max at 2048 and unclamp the velocity. You don't need to worry about any "max velocity in last frame" anymore, the floating point information, the 2048 and unclamp is taking care of that, so much room for pixel information with 32bit...

in AFX, you need to be setup in 32bit sRGB and linear.

once you drop your velocity pass in your comp, just add a level, and the spikes should be really tight in the center of the level, if it's the case, then your have done everything right. Delete the level

add a rsmb pro vector, route it to the velocity pass and put the max to 2048 in rsmb as well.

I'm using it all the time since with that workflow and it's working great.
The only thing is that since your are ticking the linear workflow in AFX, looks is not going to work properly anymore, we already talked to the support guys from red giant and there is nothing they can do for the time being :-/

but velocity work great


Stan

[tammo]

Are you using it to interpolate ever second frame on slow Archvis walkthroughs
in order to save half the rendertime?