Darby Edelen
Forum Replies Created
-
You can place a Correction deformer at the bottom of the HyperNURBs hierarchy to check sizes/locations of points on the smoothed surface. You just need to have the Correction deformer selected in point or edge mode. In this case you shouldn’t do any of your edits to the correction deformer, it’s just there to show you the structure you’d get if you made the HyperNURBs editable.
Update: the Correction deformer will show you the points based on the “Subdivision Editor” value on the HyperNURBs while making the object editable will use the “Subdivision Renderer” value. For this reason I recommend increasing your Subdivision Editor value to match your Subdivision Renderer value unless that becomes unwieldy.
Darby Edelen
-
Adding the vector mask to the layer group is as simple as selecting the group and going to the Layer > Vector Mask > Reveal All menu option.
Then you can use the pen tool (p) to draw a path. Make sure that you have the vector mask selected and that the pen tool is set to “Path” instead of “Shape” in the Options toolbar.
Darby Edelen
-
It looks like most of the images here are some combination of drawings and fractals/generative art (maybe photographs too?).
Here’s one of the applications the artist mentioned using:
https://www.andreas-maschke.com/?page_id=42
Darby Edelen
-
[George Ahlmeyer] “I am aware that all clients go through the GI pass individually and which ever computer finishes first, the iamge sequence and multi-pass render takes over.”
This should be the behavior when using the IR + QMC (Net Render) GI Mode and not providing a pre-rendered irradiance cache.
My suggestions would be to either A) wait it out or B) cancel and submit the same scene with much reduced GI quality to test the pre-pass process. If you’re a couple of days in then (A) might be preferable at this point 🙂
Darby Edelen
-
You can drag your long composition onto the comp button at the bottom of the project panel to create a new composition with the long comp as the only footage item, then apply time remapping and pull in the last keyframe to 2 seconds.
Time remapping by this method in the case of keyframed animation is exactly the same as stretching the keyframes in the source (long) composition, but I find much easier.
Is there any specific reason you didn’t want to use time remapping? It’s really the easiest way to remap keyframed animations and you lose nothing.
Darby Edelen
-
[Adam Barenblat] “However, the motion is jerky.”
Have you edited the spatial path to your liking? Aside from temporal interpolation you might want to adjust the motion path itself using the pen tool to pull out bezier curves.
If by “jerky” you really mean that there’s noise in the animation (as there may be in a hand held shot) then you could try the smooth() expression.
Darby Edelen
-
I meant clipped. If your full range of data is being mapped into 0-1 then you have your high dynamic range, it’s just compressed and fugly and needs to be expanded out.
At that point Effects > Utility > Color Profile Convert or Effects > Utility > Apply Color LUT would probably get the job done provided you have the proper color profile or LUT.
Darby Edelen
-
Deriving Z position from scale requires two bits of information:
1) Camera field of view/focal length/zoom
2) A reference point, e.g. [100,100] scale = 1000 pixels away
Since you described your shot as a “zoom” this means that #1 is a moving target. This makes it impossible to do without knowing the zoom at every frame.
Most camera trackers will allow you to solve for a variable field of view which would give you this information, but my guess is that any camera tracker that could calculate that information accurately should also be able to solve the camera and therefore render the exercise completed.
However, if your camera only zooms and you don’t get any closer/farther from your subject then you’re set with a scale. Zooming = scaling.
How large would you need to render your flagpole from Blender in order for this to work? AE is much more capable of handling large image layers since the transition to 64-bit .
Darby Edelen
-
[Eric Chard] “IS duplication of a keyed layer a valid workflow?”
Yes, but usually you’d do a separate key with different parameters on the duplicated layer to try and preserve the legs (or whatever part you happen to be working on) and then use a junk matte/mask to eliminate any issues that causes.
Most difficult keying takes some back and forth between keying and using junk mattes/masks on several layers.
Darby Edelen
-
Haven’t worked with RED files in AE in a while and it sounds like things may have changed a bit since the last time I did.
It’s quite possible that the data isn’t being quantized into 8bpc but rather clipped in the 0-1 range. In that case if you have AE set to 32bpc you could try lowering the exposure in the interpret dialog until you have no clipping, then raise the exposure an equal amount in AE with the Color Correction > Exposure effect.
I can’t test this (we don’t have our RED Epic yet) so I’m just throwing ideas out here but it might be worth a shot.
Darby Edelen