here are a couple of images of probably a better test,Model is from Zbrush.
I rendered the base proxy, the clamped Dmap, the unclamped with negative and positives dmap and the sculpted hi res model together to compare.
There seem to be a bug with the scale value in the Dmap node, it doesnt like values around 1,so the scale of the model is affected which is a bummer,had to scale it down to 0.1 and still some artifacts around the ear and face , values around (0.001) is all good.
let me know if anyone is interested and i will upload the files to the cloud.
displacement question - which color is 'neutral' ?
Forum rules
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
hey Gordonrobb, yes,the values at render time are raw values from Zbrush, so there is no trial and error.gordonrobb wrote:Also, is your point more about a work flow that makes sure you are getting the right amount of displacement without it being trial and error?
miko3d,
You are of course absolutely correct, I really don't know where my brain was.
I apologise.
Gordonrobb,
Sorry for jumping in and confusing the matter.
T.
You are of course absolutely correct, I really don't know where my brain was.

Gordonrobb,
Sorry for jumping in and confusing the matter.
T.
Win10 x64|i7-9750H 2.6 GHz|32 GB RAM | RTX2080 max Q 8GB
- gordonrobb
- Posts: 1247
- Joined: Sat Mar 16, 2013 9:08 am
But....
Your offset parameter (I thought it was called shift) is minus a half of the amount parameter. Isn't that what I said? My head hurts.
Your offset parameter (I thought it was called shift) is minus a half of the amount parameter. Isn't that what I said? My head hurts.
Windows 8 Pro | i7 3770 OC | 32 GB Ram | Single Titan (plus Black Edition on Order) | Octane Lightwave |
- itsallgoode9
- Posts: 894
- Joined: Thu Apr 03, 2014 9:04 am
- Location: New York City
- Contact:
based on this it sounds like there's two definitions of what "neutral" is?gordonrobb wrote:But....
Your offset parameter (I thought it was called shift) is minus a half of the amount parameter. Isn't that what I said? My head hurts.
Definition 1: . value halfway between the lightest and darkest values of your displacement map (this is dependent on the values in your displacement map)
Definition 2: value of 128, 128, 128 (this value is always considered middle grey)
Did I understand this correctly and is this what is causing the misunderstanding between everybody?
I think the main confusion is that there are so many ways to handle this, that and not all programs support all options. For displacement you can use a float image (both signed and unsigned), a integer image (both signed and unsigned), and all the bit depths from 8 bits to 32 bits (maybe higher? 48 bits). Displacement can affect the size of an object, sometimes significantly so usually you want to limit how it's changing the size. One way is to have both positive and negative displacement, you pick a mid-range as zero displacement and it minimizes the size change. You need a signed image format (for negative numbers) to do that and I'm not sure that Octane supports it. From my experiments it appears that Octane only supports unsigned images, so you need the shift offset to move the low values negative. This effectively moves the neutral (zero/mid-range) value of the image so the displacement looks better and doesn't change the size as much.
I have asked before how Octane handles this but no one has responded yet (this was during the beta testing phase). Sorry for the rant, at least this is how I understand it.
Jason
I have asked before how Octane handles this but no one has responded yet (this was during the beta testing phase). Sorry for the rant, at least this is how I understand it.

Jason
Linux Mint 21.3 x64 | Nvidia GTX 980 4GB (displays) RTX 2070 8GB| Intel I7 5820K 3.8 Ghz | 32Gb Memory | Nvidia Driver 535.171
- gordonrobb
- Posts: 1247
- Joined: Sat Mar 16, 2013 9:08 am
Miko3d's comments I now believe are all about having a reliable way to create a displacement map that takes away the guesswork of how big to make the Amount parameter in Octane. That is good. However, my comments are all about the need, and there episode a need in almost every circumstance of using a displacment created for a model in ZB, to use the Offset parameter. It seems I was wrong about it always being negative a half, but only when you've changed the default.?
My issue is and always has been that if you displace a relative complex but of geometry out the way, it looks great, but if you have to use the Offset, it invariably creates tears somewhere or other. Again, if people are not finding this, then I'd love to know how.
My issue is and always has been that if you displace a relative complex but of geometry out the way, it looks great, but if you have to use the Offset, it invariably creates tears somewhere or other. Again, if people are not finding this, then I'd love to know how.
Windows 8 Pro | i7 3770 OC | 32 GB Ram | Single Titan (plus Black Edition on Order) | Octane Lightwave |
According to the release notes for v2.12.1 "Fixed/improved displacement mapping with negative offsets."gordonrobb wrote:My issue is and always has been that if you displace a relative complex but of geometry out the way, it looks great, but if you have to use the Offset, it invariably creates tears somewhere or other. Again, if people are not finding this, then I'd love to know how.
I have not tested displacement in this version myself though, if you are still getting the tearing in this version or above it may be a good idea to post it to the development build forum.
T.
Win10 x64|i7-9750H 2.6 GHz|32 GB RAM | RTX2080 max Q 8GB