Validation Results
Standards Validation measures a captured image of one or more targets and rates the results against a quality standard such as FADGI or ISO 19264. The Calibrate window shows the same results table after it creates a profile.
When to Validate
After you create a color profile, check it against a second target, one of a different make and model from the profiling target. See The Importance of Evaluating Color Quality Using an Independent Target on the DT website.
- If two or more supported targets are in the frame when you create the profile, DT Nexus does this for you. It builds the profile from one target and shows results for the others in the same results table.
- If you capture one target at a time, capture the second target and use Validate.
Validation is also worth repeating at regular intervals, such as the start of each day, shift, or project.
Running a Validation
In Capture One, select one image or a Fusion pair that contains one or more supported targets. Calibrate needs a raw file, but Validate also accepts a TIFF or JPEG that has already been processed.
In the Image Quality panel on the Calibrate tab of DT Nexus, click Validate. The Standards Validation window opens. DT Nexus exports the image from Capture One, using the format set in Preferences > Validation, and detects the targets. A TIFF or JPEG is measured as it is, without an export.
DT Nexus measures the image as it is adjusted in Capture One. It does not change the color profile, curve, LCC or other settings, so apply the settings you want to check before you click Validate.
Choose the quality standard in Standard: and the material type in the menu next to it. The results update to show how the image rates against them.
Optional: Click Export PDF to save a validation report.
Results Table
Each row is rated in stars against the selected standard. Hover over the stars to see the measured value ("Measured: …") and the details for that row, including the limits for the selected standard.
Row | What it measures | FADGI section |
|---|---|---|
ΔE Mean | Average color difference (ΔE 2000) between the patches and their reference values | Color accuracy |
ΔE 90th | The 90th percentile of that color difference | Color accuracy |
White Bal. | White balance error of the neutral patches | 2.4.4 |
PPI | Sampling frequency, in pixels per inch, measured from the target | 2.4.1 |
Scale | Measured PPI compared with the PPI in the file header | 2.4.9 |
SFR10 | SFR10 sampling efficiency | 2.4.8 |
SFR50 | The frequency at which the SFR falls to 50% | 2.4.8 |
Sharpening | The maximum SFR, which shows oversharpening | 2.4.10 |
Nyquist | SFR at the Nyquist frequency (informative in FADGI) | 2.4.8 |
Misregistration | The largest offset between color channels, in pixels | 2.4.7 |
Noise | Variation of L* inside each neutral patch, rated on the worst patch | 2.4.11 |
The sharpness rows (SFR10, SFR50, Sharpening, Nyquist) and Misregistration have results only for targets with sharpness sites: the FADGI 19264 Device Target, the GoldenThread Device-Level Target and the ISA Object-Level Target 1x.
Informational Rows
Noise, Scale and Misregistration are shown for information only. They are not part of the overall result for the selected standard.
Scale compares the measured PPI with the PPI written in the file header. In Standards Validation the header is written by the export DT Nexus makes for the measurement, not by your own workflow, so the row is always informational there. When you validate a TIFF or JPEG, the PPI comes from the file's own header. If the file has no header PPI, the row says "no header PPI".
How SFR Is Measured
DT Nexus measures SFR with the sfrmat5 slanted-edge method of ISO 12233. The sharpness of the target's own printed edge is divided out of the result. On the FADGI 19264 Device Target, each measurement site is placed on its printed slanted edge.
SFR is a standard, but like many standards it leaves various choices to be made in its implementation, so different analysis engines report slightly different values for the same image. The SFR implementation in DT Nexus now more closely matches that of GoldenThread. Its results will be a little different from those of other engines such as ImageZebra or NimbusQA: expect a difference of 3 to 5 points of SFR50 and 1 or 2 points of SFR10.
If the Results Do Not Meet the Standard
See Color Profiling: Troubleshooting, correct the cause, and create a new profile.
PDF Report
Export PDF saves a report named DT_Nexus_Validation_<date and time>.pdf. It has up to three pages:
- Image Quality Validation Report: whether the image is compliant with the selected standard (and which metrics fail), followed by the color and spatial metrics with their star ratings and thresholds.
- OECF / Tone Response: the tone response of the neutral patches and a chart of their difference from the aim values. It is titled "OECF / Tone Response - ICC" or "OECF / Tone Response - Fusion".
- SFR / MTF Analysis: the horizontal and vertical SFR curves, for targets with sharpness sites.
When a GoldenThread Device-Level Target is scored against its generic reference, the first page carries a yellow banner explaining that the reference is a nominal placeholder, not a measurement.
Validation Preferences
Preferences > Validation sets how DT Nexus exports the image it measures:
Setting | Options | Default |
|---|---|---|
Format: | TIFF (16-bit), TIFF (8-bit), JPEG (quality 95), JPEG (quality 100) | TIFF (16-bit) |
Color Profile: | Adobe RGB (1998), ProPhoto RGB | Adobe RGB (1998) |
A higher bit depth gives more accurate patch readings, so keep TIFF (16-bit) unless you need to test a specific delivery format.