I know there are v2 and v4 profile generation settings, but I can’t find settings for 8bit or 16bit profile to make?
Just about every program these days makes 16-bit profiles. Certainly ProfileMaker does. These are verified by looking at your profiles’ tags. Your LUTs should have mft2 tags which are described as “Double Precision(16 bit) Multi-Function Lookup Table.” If you come across an 8-bit profile, that will have an mft1 tag for a single precision profile.
If you are looking for greater precision in your profiles, you could try the Large Profile size option in ProfileMaker. That will make a grid resolution of 33 grid points instead of something like 25 grid points which you would get with the default size.
Thanks, I have this tag:
Double Precision(16 bit) Multi-Function Lookup Table
3 input channels, 3 output channels
33 CLUT grid points
Speaking about profile precision, I recently found out that to calculate the precision that test chart will provide is like this:
see attached image file:
Anyone knows how to know the number of patches that profilemaker generates as (service patches)?
For example IT8 9.18 has 936 patches, but 918real patches.
So a custom target 3000 patches has ??? real patches ?
It depends on what you mean as a difference between patches and real patches. Some target generators will add patches in order to fill out a complete row or column so that the final target arrangement will be a complete rectangle. We usually refer to these as filler patches.
You might be interested in the concept of duplicate patches. It is common for a target to be made that has duplicates of the same colors scattered in various places around the target. This can be used to determine measurement or printing consistency across a sheet if you have software that can analyze this (like Maxwell). For example, the 918 reference target actually has 7 or 8 identical patches on the neutral axis. So there might be only 910 unique patches in that patch set.
If you know how to use Excel, you can arrange the color columns by sort order to make it fairly easy to pick out the duplicates.
Well I don’t know how I have to sort RGB values with excel because once I sort by R the GB values are left unsorted.
What formula I should use?
I however read the actual chart files and it seems to have only 911unique colors, so you were close when said 910 
But if the extra patches are used in building the profile then I think it’s irrelevant that there are only 910unique patches.
Chromix color think has so many bugs that version 3 is unusable, only beta has some bugs solved. I wouldn’t touch the maxwell software with a ten yard pole. Global color management system, no no unless you have a money stashed for spectrophotometer not a colorimeter.
So it seems like for average Joe it’s better to pay 35-45$ for professional profile made with professional equipment and knowledge, than to buy cat in the bag.
Speaking about Chromix color ColorValet prices at $99 for RGB profile its a bit expensive to say the least.