I don't understand IT8 reference files

[racecannon]

When I try to calibrate my IT8 scan, I'm told to choose a reference file and am presented with files with names that provide no indication of what the file is for, like E130104.txt. All the files are like this. How do I know what to choose?

[LSI_Ketelhohn]

Dear user,

generally SilverFast shows exactly the file name it requires. Each target has its own code and reference file.
The name of the file is also shown as the code below the barcode on SilverFast targets.
You can download the corresponding reference file from our website's download area.

Kind regards,
Arne

[racecannon]

Ahh, I see. I don't recall it ever being mentioned that the target name was written underneath the barcode on the physical target. I received two targets when I bought my scanner. A transparent and a reflective. I've scanned them both and got the delta message. But I scanned the transparent again because I go an error about patch homogeneity (no idea what this means) and I got two different Deltas, a .5 and a .8. What should I do? And once I've done my IT8 calibration on each, do I need to do anything else related to IT8 regularly? - like load the IT8 results somehow?

[LSI_Ketelhohn]

The delta E value gives the quality of the calibration. Values below 3 are fully acceptable. Values lower then 1 are excellent.
After the calibration is done you do not need to take any further steps.
The calibration should be repeated about every three months.

An error in patch homogeneity means that the calibration process did get wrong values for some of the fields. This can have various reasons like the mask not being placed correctly, a wrong reference file, overaged targets or the scanner not being at warmed up.

Kind regards,
Arne

[LSI_Ketelhohn]

Dear user,

there is no certain way to do this because it also greatly depends on your device.
If you have a flat bed device make sure it is clean.

You can also use a hand measured high-precision target.

But it also is not necessary to actively try to get the the delta E below 1.

The Delta E is often misunderstood.
A low value means only few values differentiated between target values and reference values.

For example you could switch from a normal target to a high precision target the value could even get higher because same values got a more precise measurement. This can result in a higher deviation from the reference value and result in a higher total delta E. Which in this case would be a good thing because it would mean that the color is represented more precisely.

Kind regards,
Arne