Hello guys.
We actually work on a film, on wich some shots are made on C4D, and others on 3DsMAX.
Except scales differences (on noises, triplanars ...), everything roughly match between the 2 softwares.
But at compositing time, we have issues with all the Z Depth that come from 3DsMAX.
C4D one works pretty good, but all max ones are bugy.
When we compute the same shot and compare the compositing with the 2 differents Z, that give us this :
And when we look at the two Z pass, we can notice that the 3DsMAX one shrinks all objects.
Can you compare the actual pixel values of the C4D and Max Z AOV output? Could be a scale issue. The default export scale in C4DtoA is 1 Arnold unit = 1 m. I'm not sure about Max.
I thought about scale issue too, but the scene isn't so big, neither in C4D, nor in MAX.
Moreover, we always place a black sphere on background to avoid alpha, z, and others stuffs problems.
Resulting white value from the 2 software still are not the same.
C4D : White value is above 7
MAX : White value is more than 1e+5
Frischluft doesn't understand this overexposed Z and in After Effects, we need to pre-process it with a minus exposure exposure.
C4D Expo = -5
MAX Expo = -11
In Nuke, I believe remember that Frischluft has the same default, but I'm sure that Native plugin and PG Bokeh doesn't care.
I suspect that could be related to the filter. In MAXtoA, the filter defined in the rendering settings applies to all the AOVs, and the default (Gaussian) does not make sense on numeric AOVs such as Z, P, N.
Can you try setting the filter to Closest and re-render the Z ?
We are aware of the limitation, and we're working on allowing per-AOV filters.
In MAX, in closest, the exact same scene give us a NAN background instead of white.