[Half Life]

I'm still convinced that everything would be perfect if only Maxwell let us set the depth based on the U or V tile scale rather than on some random % of the biggest bounding box edge size. That makes no sense to me...

It already does exactly that -- RealScale/Normalized UV's on geometry that is to scale will work exactly as it is supposed to based on whatever amount you put in under "cm".

Using meters and putting in the correct units will only work if the UV's are 1m x 1m -- any other scale would obviously invalidate the simple designation of "meters"... as far as tiling goes.

The depth is (more or less)accurate, if it isn't coming out right on the geometry it's because it is not to scale.

If you don't post the files I can't help.

Best,
Jason.

[Burnum]

depth is still not working as expected.
3.23195cm is far to deep, but .32319cm is getting closer...

[Half Life]

I measured and double checked, your circumference is 0.06785840131753954m (or 6.78584cm if you prefer) -- which makes your diameter 0.0216m (or 2.16 cm if you prefer).

If you want to displace to 0.218cm deep, simply put that (0.218cm) into your displacement value. You can see the results of a test render below:



1) Tiling of the displacement map as set to relative 1 and 2.27149 (due to the nature of your UV's) with interpolation enabled.
2) Precision is set to adaptive.
3) I recalculated the normals to smooth at 45 degrees

All on my end works as I would expect given your UV setup... I think you will want to increase the density of your geometry for better results though -- 48-sided with more square shaped subdivisions will work better, right now the elongated large polygon rectangles are going to not give such great results in Maxwell.

Best,
Jason.

[Burnum]

Thanks, that looks much closer than my last attempt!
Yours seems to have little missing areas though?
I woke up this morning thinking I should go back to relative rather than meters for the reason you stated. my UVs are matched to the circumference, not 1 meter. the depth however is still frustrating the way it works. do you understand why anyone would want to use % they way it is now? I can't picture why...

[Half Life]

That's an old jpeg -- the server hosting it hasn't refreshed to the new version yet, but the problem there was the smoothing angle wasn't high enough.

I think % is there for people who want to adjust the UV's and displacement through the eyeball method... which is alot of people.

Best,
Jason.

[Burnum]

even better after server update!