Regarding multiresolution output settings, what factor(s) effect or contribute to a smooth rotation and relatively quick load time? Would adding more cube size levels help smooth rotation, even though it adds more bandwidth? Would a smaller base cube face help increase load speed, particularly for rural America
Recent threads have clearly covered how to calculate the Multiresolution settings, but I'm wondering if you would expand on the above.
Jim Watters wrote this good summary:
[quote="JimWatters"]No need to have levels other than 1/4 the previous size by halving the width. Creating extra levels for every factor of the tile size is a waste.
if 8000 is your ideal cube width, then use:
8000, 4000, 2000, and 1000
and cube faces of 500
Making cube sizes of 500, 1500, 3000, and 6000 is just a waste of bandwidth for the user, and storage space for the content creator.
No need to downsample to a nice round equirectangular size before letting Pano2VR create cube faces.
The calculations are pretty simple.
Estimate cube size based on a value of Equirectangular width divided by 4 or Pi
divide by 2 until less than or equal to 512
to accommodate preview track of 1/4 width divide by 4,
round to nearest whole number
multiply by 2 until you are close to the starting estimate.