Hi,
Why does my displacement map have jaggies?
I create this displacement follow this tutorial: Frank Gehry-style titanium scales , and save the map in 16bit TIFF file, but the result is not smooth, please help…
If it was the map you would be seeing jaggy squares, You are seeing a default screen display of mesh whose points have all been displaced, but the mesh is just showing on screen as a bunch of triangles. It might render fine. lots of possible solutions if it doesn’t. Try simply adding an ‘Edit Poly’,
select all polys, scroll down to ‘Polygon Smoothing Groups’ and autosmooth the deformed mesh,
instead of
HI, you’ve went too far with the amount. Try to use 3D mapping mode for the VRayDisplacementMod if you need that much amount. If you’d like send us the scene here or through our support contact page and we can help.
I have deleted the file.
Anyway, I have re-create a file with the same step and method, and rendering still pixelated like what I post here, I have archive and upload the file to the support contact page
Your map is causing the issue, you have extrapolated your displacement texture from a low density mesh. You can fix this in a couple ways, the fastest is to give a bit of blur to the map in PS, the second is to recreate the map from a mesh with more subdivisions to get a smoother result.
If you’re in a rush then post the scene here too, in case Chaos is away for the weekend 
This should work with no problems and is sure to be easy to fix.
EDIT: It’s very likely that Sirio correctly identified the problem, so probably you can adjust from that
Thank you for your suggestion, will give it a try
I try to upload the file here but it failed
here is my google drive link: https://drive.google.com/file/d/1bjkpHGgob7mfXeLLfQ82xlp8QHkX9LjM/view?usp=sharing
Thanks in advance
Well I did find something very unexpected.
You ‘can’ use that png but in order for it to work you need to set the resolution to
e.g. in this case around 140, using 2d subdivision on a 1 segment plane, which yields the attached result.
You can also push the values without much issue, as in the other image, which is also unexpected.
I can’t quite work out why this works actually, so would love to know, as it seems counterintuitive 
Apart from that it’s as suggested, you’d need to normally get your zdepth from a much higher resolution reference mesh
and still may need to blur it to avoid some stepping, depending on the particular geometry and differences in brightness.
thank you for your reply, will give it a try when i am free