Why It Looks Like That
Article 25

How to Read a TV Calibration Report

Read the target, signal, measurement system, patch set, grayscale, EOTF, color, capability, and limitations before trusting one average Delta E number.

By · Published · Updated

Direct answer

In brief

Read the setup block before the charts. Identify the report type, target, signal format, picture mode, meter correction, pattern source, patch set, and conditions. Then read grayscale chromaticity and luminance separately, use the correct SDR, PQ, or HLG target, combine color chromaticity with luminance, inspect maximum and individual errors before the average, and evaluate gamut coverage, color volume, uniformity, stability, and limitations separately. A low average Delta E is evidence only under the declared formula and measurement conditions.

A display-measurement report is a record of:

The charts are not self-explanatory.

Read the setup information before the error numbers.

The practical answer

Use this order:

  1. Identify whether the report is pre-calibration, post-calibration, validation, or product evaluation.
  2. Identify the SDR, PQ, or HLG target.
  3. Identify the white point, color space, EOTF, and luminance target.
  4. Identify the meter, meter correction, pattern source, and signal format.
  5. Identify the measured patch set.
  6. Read grayscale chromaticity and luminance separately.
  7. Read color chromaticity and color luminance together.
  8. Inspect individual and maximum errors before trusting the average.
  9. Read gamut coverage and color volume separately from color accuracy.
  10. Look for repeatability, stability, uniformity, and limitations.

A low average Delta E can be useful.

It is not a complete verdict.

A ten-step vertical flow shows the recommended order for reading a television measurement report: report type, target, signal, meter and generator, patch set, grayscale, EOTF, color, capability, and limitations. The final note says error numbers should be read only after the measurement conditions are known.
A measurement result is interpretable only after report type, target, signal, system, and patch set are known.

Identify the report type

Pre-calibration

This records the television before adjustment.

It can show:

Post-calibration

This records the result after adjustment.

It should be measured under the same conditions as the pre-calibration report.

Validation

Validation uses a measurement set intended to test the final result.

A stronger validation can use:

Product evaluation

A review can measure:

It is not necessarily evidence that the display was calibrated.

Three sequential report cards show pre-calibration, post-calibration, and independent validation. Pre-calibration records the starting display state. Post-calibration measures the adjusted result under the same target and conditions. Independent validation changes patch order or density and can add another meter, correction, generator, or software system while keeping the target consistent. A note says changed conditions can create a false improvement.
Pre, post, and validation reports must be distinguished; fair comparison requires matching conditions.

Read the setup block first

A useful report identifies:

Without this information, the charts are difficult to reproduce or compare. [16]

Confirm the target

SDR

The report should state:

BT.1886 is not simply a label for pure gamma 2.4.

Its equation includes measured black and white luminance. [7]

PQ

The report should state:

HLG

The report should state:

PQ and HLG are separate systems and should not be mixed. [8]

Confirm the measurement system

Ask:

A report with no meter correction can be less trustworthy on a narrow-spectrum display.

A report using the same incorrect meter correction before and after can look internally consistent while remaining inaccurate.

An eight-stage trust chain runs from meter through correction, generator, signal, patch, display, software, and report. A coral fault marker at any stage points to the final report, showing that a smooth chart can be consistently wrong when the measurement system or signal path is invalid. A final verification box lists repeat readings, independent checks, and documented limitations.
Meter, correction, generator, signal, patch, display state, software, and report conventions can each affect the result.

Confirm the patch set

Record:

An average over six colors is not comparable with an average over hundreds of volumetric patches.

Delta E is a family of formulas

Delta E is a calculated difference between a measured value and a target.

The symbol can be written:

The formula must be named. [11]

CIE 1976

CIE 1976 Delta E uses Euclidean distance in CIELAB.

It is easy to calculate.

The visual uniformity is limited.

CIE94

CIE94 adds application-dependent weighting.

The parametric choices matter.

CIEDE2000

ISO/CIE 11664-6:2022 defines CIEDE2000.

It adjusts for variation in perceived differences involving:

It has reference conditions.

It is not a universal television visibility scale. [4]

Delta E ITP

ITU-R BT.2124-0 defines Delta E ITP for television using display-referred ICtCp.

It includes color and luminance differences over HDR and wide-color-gamut ranges. [6]

A central measured-versus-target comparison branches to four Delta E formulas: CIE 1976, CIE94, CIEDE2000, and Delta E ITP. Each branch lists a different space or purpose. A footer states that formula, scale, conversion, and reference conditions must be identified before comparing values.
Delta E is not one universal metric; each formula has its own space, scale, assumptions, and intended use.

What a Delta E number means

A smaller value generally means the measured result is closer to the target under the selected formula.

A value of zero means that the values supplied to that formula match.

It does not prove:

Do not reuse one threshold for every formula

The source article used fixed bands such as below 1, below 3, and above 5.

Do not treat those as universal.

Visibility depends on:

Delta E ITP scale

BT.2124 scales 1 Delta E ITP to indicate the potential of a just noticeable color difference under the most critical adaptation state.

The Recommendation says the metric can overpredict the difference in other conditions. [6]

That statement does not apply automatically to CIEDE2000.

A flow diagram shows target and measured display-referred values converted into ICtCp, then separated into intensity and chroma-related differences, combined by the ITU-R BT.2124 Delta E ITP calculation, and reported on a scale where 1 indicates the potential of a just noticeable difference under the most critical adaptation state. A caution says the metric may overpredict differences in other conditions and is not numerically interchangeable with CIEDE2000.
BT.2124 defines Delta E ITP from display-referred ICtCp and scales 1 to potential JND under critical adaptation, with an overprediction warning.

Report tolerance

A report should state its pass-fail threshold and why that threshold was selected.

A professional workflow can have a tighter requirement than a home preference workflow.

Average Delta E

The average can summarize one patch set.

It can also hide a failure.

Consider this example:

The average is about 0.81.

The one large error remains important.

Read these values

When available, read:

Find the outlier

Ask:

A conceptual bar chart contains 24 patch errors. Twenty-three bars are 0.5 and one bar is 8.0. A calculation shows the average is about 0.81 even though one large error remains. The figure tells the reader to inspect the maximum and patch identity, not only the average.
Average error is one statistic and can hide an important outlier.

CIE chromaticity charts

A CIE chart places chromaticity points on a two-dimensional diagram.

Common choices include:

The CIE 1931 x,y diagram is familiar but not perceptually uniform. [1]

ISO/CIE 11664-5:2024 defines the CIE 1976 u′,v′ uniform-chromaticity-scale diagram. [2]

Target markers

A report can plot targets for:

Measured markers

Measured points show where the display produced the patch chromaticity.

The plot can show:

A conceptual CIE x,y diagram shows two pairs of points with equal plotted line lengths in different regions. The pairs are labeled equal plot distance but not necessarily equal perceptual difference. A separate box says Delta E requires a defined color space and formula.
Equal x,y distances do not guarantee equal perceptual differences, and x,y omits luminance.

Do not read plot distance as Delta E

Equal distances on the x,y chart do not represent equal perceptual differences.

A large plot offset in one region can have a smaller calculated perceptual error than a shorter offset elsewhere.

Use the chart for location.

Use the selected Delta E formula for the report’s calculated difference.

Chromaticity is not luminance

A point can sit directly on target x,y and still be too bright or too dim.

Read color data as:

or another complete representation.

Common error components

Some reports separate:

The exact definitions depend on the formula and software.

Read the legend.

Two panels compare incomplete and complete color reporting. The left panel shows measured x and y landing on the target while a luminance bar is too high. The right panel lists x, y, and Y together and marks chromaticity matched but luminance failed. A note says color luminance must be read separately from the CIE point.
x and y locate chromaticity; Y reports luminance. Both are needed for a complete color measurement.

Reading target direction

Hue clue

A point rotated around the white point relative to the target can suggest hue error.

Saturation clue

A point closer to or farther from the white point along the target direction can suggest saturation error.

Luminance

The two-dimensional diagram does not show it.

Read the table or luminance chart.

These are clues.

The full diagnosis depends on the selected color model and target.

White point

The report should identify:

CCT

Correlated color temperature describes the nearest Planckian-radiator chromaticity under a defined calculation.

It does not fully locate the measured white. [5]

Two whites can have similar CCT and different green-magenta displacement.

Duv

A Duv-like value can describe signed distance from a locus.

The report must identify:

Do not interpret an unlabeled Delta uv value.

A conceptual u-prime v-prime-style diagram shows a curved Planckian locus, an iso-temperature line, D65 near but not defined only by a Kelvin label, and two white points with similar correlated color temperature but opposite green and magenta offsets. A footer says Duv-like values require a stated coordinate system, locus, sign convention, and method.
CCT is tied to the Planckian locus and does not fully specify D65 or green-magenta offset.

Grayscale tracking

A grayscale series measures achromatic signal levels from near black to white.

It can contain:

The count describes measured patches.

It does not guarantee the number or location of the television’s controls.

RGB balance charts

An RGB balance chart compares measured red, green, and blue contributions with the selected white target.

The chart can use different reference modes.

Relative reference

Portrait Displays says a relative Calman RGB chart shows the relative color balance and normally centers the target at 100. [12]

A good relative result means:

It does not prove the luminance is correct.

Absolute reference

An absolute Calman RGB chart includes luminance error relative to the selected EOTF and normally centers the target at zero. [12]

It can show:

Other software can use different normalization.

Two conceptual RGB balance charts use the same grayscale data. The relative chart centers on 100 and shows red, green, and blue close together, hiding luminance error. The absolute chart centers on zero and shows all three channels below target at several levels because luminance is low. A note says the chart reference and legend control the interpretation.
Relative Calman RGB charts emphasize channel balance; absolute charts include luminance error against the EOTF target.

Equal chart values do not mean equal spectral power

D65 white does not require the panel’s red, green, and blue emitters to produce equal radiometric power.

The chart shows each channel relative to its required contribution for the target.

Read the level labels

A report can label the horizontal axis as:

Do not assume all 0 to 100 axes are IRE.

Six horizontal-axis label cards show percent signal, digital code value, PQ percentage, target luminance in cd per square meter, HLG percentage, and IRE. Each card states what must be known before interpreting it. A warning says a zero-to-one-hundred axis should not automatically be called IRE.
Digital and HDR reports can use several axis quantities; IRE is not a universal label for every 0 to 100 scale.

Near-black grayscale

Near black can be affected by:

Look for repeatability.

Do not judge a 1 percent or 2 percent point from one unstable reading.

SDR EOTF

An SDR EOTF chart compares measured luminance with the selected target.

Possible targets include:

BT.1886

BT.1886 uses:

If black luminance is not zero, the target is not a single pure 2.4 power function.

A conceptual SDR EOTF plot compares the BT.1886 zero-black case with a nonzero-black case, both normalized to the same white. The zero-black case approaches a pure 2.4 power function. The nonzero-black case departs near black because the complete BT.1886 equation includes measured black and white luminance. A measured line is shown as a separate dotted example.
BT.1886 uses exponent 2.4 plus measured black and white luminance; the zero-black case approaches a pure power law.

Best chart

The clearest chart shows:

A Gamma Point chart can be useful.

It is less direct near black.

Reading a Gamma Point chart

Check what the vertical axis means.

It can show:

On a pure-power chart:

Near black, the exponent can become unstable.

Read the luminance chart before diagnosing crush or lift.

Two conceptual charts show the same luminance response in different forms. The left chart plots effective gamma point, where a higher exponent can correspond to darker midtones. The right chart plots measured minus target luminance, where positive and negative signs have a different meaning. A warning says curve direction cannot be interpreted without the vertical-axis definition.
Gamma charts can plot exponent, fitted response, or luminance error, so sign and direction depend on chart type.

PQ EOTF

PQ is an absolute display-light EOTF.

A report can plot:

Portrait’s chart system supports several distinct luminance-response views. [13]

Trackable region

A consumer display can follow the reference PQ relationship through part of the range.

Display mapping

When requested luminance exceeds the display’s capability, the display can:

The mapping usually begins before the target demand would exceed the usable display peak.

Tone-mapping onset

The report should identify:

An early roll-off is not automatically wrong.

It can be part of a deliberate mapping.

It should match the declared mode objective.

A conceptual PQ report plots target luminance and measured display luminance. The measured line follows the PQ reference in the lower range, passes a marked HDR Reference White point, begins to depart at a tone-mapping onset, rolls off toward the measured display peak, and finally reaches a clipping region. Labels also call for metadata, pattern area, duration, and tone-mapping mode.
PQ is an absolute reference relationship; consumer display mapping can begin before requested output exceeds usable capability.

Smooth is not automatically correct

A smooth curve can still:

A hard clip can be:

Read the intended behavior.

HDR Reference White

BT.2100-3 identifies HDR Reference White as 203 cd/m² on a PQ display or a 1,000 cd/m² nominal-peak HLG display under the stated reference conditions. [8]

BT.2408-9 describes approximately:

A report should show whether this region is:

Peak luminance is a separate measurement.

Peak-luminance report

Peak output should include:

One 10 percent window does not describe the complete television.

HLG report

HLG is a relative HDR system.

A useful report identifies:

Do not compare an HLG measured line with a PQ target.

Two report requirement panels compare PQ and HLG. PQ is labeled absolute display-light EOTF and requires target luminance, metadata, HDR Reference White, tone-mapping state, area, and duration. HLG is labeled relative HDR system and requires nominal peak, system gamma, black condition, room assumption, user gain, and signal level. A warning says the targets must not be interchanged.
PQ and HLG are separate HDR systems with different report requirements.

Delta E ITP

BT.2124 defines:

Delta E ITP = 720 × distance in scaled ITP

The metric uses display-referred ICtCp and includes luminance and color. [6]

Interpretation

A value of 1 indicates potential just-noticeable difference in the most critical adaptation state.

The metric may overpredict under other conditions.

Do not compare directly

Do not compare a Delta E ITP value directly with:

The formulas and scales differ.

HLG caution

BT.2124 says display-referred Delta E ITP requires an assumed nominal peak for relative HLG.

It also defines Delta ITP R for some relative scene-referred comparisons.

Read the metric name carefully.

Primary and secondary colors

A six-color set measures:

Read for each:

These points are useful.

They are not the complete gamut interior.

Saturation sweeps

A saturation sweep measures colors between white and a primary or secondary endpoint.

It can reveal:

A display can hit the 100 percent primary and miss the 25, 50, and 75 percent points.

A split diagram compares six endpoint measurements with saturation sweeps. The endpoint panel shows red, green, blue, cyan, magenta, and yellow at 100 percent. The sweep panel shows 25, 50, 75, and 100 percent steps for each color direction, revealing interior tracking, hue drift, luminance error, and gamut mapping. A note says endpoints are useful but do not describe the complete gamut interior.
Endpoint checks do not test intermediate saturation, interior mapping, or luminance-dependent errors.

Expanded patch sets

Software can use:

The report should name:

Do not assume every 24-patch set uses identical display targets.

Error patterns

One region fails

This can suggest:

Many regions fail

This can suggest:

The pattern suggests where to investigate.

It does not prove that one control can fix the problem.

Gamut coverage

Gamut coverage describes how much of a target color space the display can reach under the selected calculation.

A two-dimensional method often uses the primary triangle area.

Coverage is not accuracy

A display can cover 100 percent of a target and map the colors inaccurately.

A display can cover less than 100 percent and reproduce many interior colors accurately.

Target matters

State:

Do not compare percentages from different diagrams or methods.

Two conceptual gamut diagrams compare capability and calibration accuracy. The left display covers the full target triangle but measured interior points are misplaced. The right display has a smaller reachable boundary but the interior points it can reproduce sit close to their targets. A note says coverage asks how much space can be reached, while accuracy asks how close measured output is to target.
Gamut reach and measured target conformity are separate questions.

Native gamut and calibrated gamut

A report can show:

The result can identify hardware capability and mapping behavior separately.

Color volume

Color volume adds luminance to hue and saturation.

Portrait describes CIELAB and ICtCp color-volume methods with different sampling and target options. [14]

Light Illusion documents volumetric profiling and warns that two-dimensional primary coverage can miss luminance-dependent gamut reduction. [15]

Read the method

Identify:

Do not compare unlike methods

A CIELAB volume percentage is not directly equivalent to:

A two-dimensional primary triangle is compared with a three-dimensional color-volume stack. The 2D view shows primary chromaticity at one projection. The 3D view shows a smaller saturated boundary at high luminance than at low luminance. A checklist states that color-volume results depend on color space, target peak, measured peak, patch count, sampling grid, and absolute or relative method.
Color volume adds luminance and is method-dependent; 2D coverage does not show brightness-dependent gamut.

Millions of distinguishable colors

Some software estimates the number of distinguishable colors from an ICtCp model.

Treat this as a calculated metric.

It depends on:

It is not a count of colors shown individually to observers.

Uniformity

A center-screen calibration report does not show the complete panel.

A uniformity report measures several positions.

It can reveal:

Read:

A television screen grid shows a low-error center measurement surrounded by corners and edges with conceptual luminance and tint variation. The center cell is teal and labeled center report. Several edge cells are amber or coral and labeled dimmer, brighter, warm tint, or cool tint. A note says center-screen accuracy does not establish spatial uniformity.
A center measurement cannot establish spatial luminance or chromaticity uniformity.

Stability

A stability test repeats measurements over time.

It can reveal:

A report without stability data is a snapshot.

Repeatability

Repeat selected patches without changing settings.

A trustworthy result should be stable within the measurement system’s repeatability.

Large variation can make small Delta E changes meaningless.

Measurement uncertainty

A report can include:

When measured differences are smaller than the combined uncertainty, do not claim a meaningful improvement.

Before and after comparison

Use identical:

Read more than the average

Compare:

An adjustment can reduce average Delta E while creating banding or clipping.

Independent validation

A stronger validation can change:

It can also use:

The validation should still use the same target.

Green, yellow, and red chart indicators

Software can assign colors based on configured thresholds.

The threshold can be:

The colors are not self-validating.

Read the threshold and formula.

Recognizing a trustworthy report

A strong report answers these questions.

Target

Measurement system

Test

Results

Limitations

Quick reading sequence

When time is limited:

  1. Read the setup page.
  2. Confirm the target.
  3. Check D65 grayscale.
  4. Check luminance or EOTF.
  5. Check maximum Delta E.
  6. Find failed patches.
  7. Check saturation sweeps.
  8. Check gamut coverage.
  9. Check color volume for HDR.
  10. Read limitations.

Do not begin with the average.

A twenty-item checklist groups the information needed for a trustworthy television measurement report into Target, Measurement system, Test, Results, and Limitations. The checklist includes formula, signal, meter, correction, generator, patch count, maximum error, repeatability, uniformity, stability, capability limits, and uncertainty. A final box says a clean chart without this context is incomplete.
A strong report identifies target, measurement system, test design, results beyond averages, and limitations.

What a good measured result looks like

A useful result shows:

No one universal Delta E cutoff defines all of these.

What a report can prove

A well-designed report can show that the display:

under the stated conditions.

What a report cannot prove

The report cannot by itself prove:

A note about capability

Calibration accuracy and hardware capability are separate.

A broad evaluation report can measure both.

Capability measurements can include:

Calibration measurements show how closely the selected output follows the target within those capabilities.

The next article

The next measurement-sidebar article explains the calibration process.

It covers:

Read the report before changing the television.

A chart tells you where the difference is.

It does not automatically tell you which control should move.

Sources and standards

CIE and ITU standards plus current Portrait Displays and Light Illusion documentation used in this article. Last verified August 6, 2026.

  1. International Commission on Illumination. Colorimetry. CIE 015:2018, 4th Edition.
    https://cie.co.at/publications/colorimetry-4th-edition

    ↩

  2. International Commission on Illumination and International Organization for Standardization. Colorimetry, Part 5: CIE 1976 Luv Colour Space and u′,v′ Uniform Chromaticity Scale Diagram.* ISO/CIE 11664-5:2024.
    https://cie.co.at/publications/colorimetry-part-5-cie-1976-luv-colour-space-and-u-v-uniform-chromaticity-scale-diagram

    ↩

  3. International Commission on Illumination and International Organization for Standardization. Colorimetry, Part 6: CIEDE2000 Colour-Difference Formula. ISO/CIE 11664-6:2022.
    https://cie.co.at/publications/colorimetry-part-6-ciede2000-colour-difference-formula

    ↩

  4. International Commission on Illumination. Validity of Formulae for Predicting Small Colour Differences. CIE 230:2019.
    https://cie.co.at/publications/validity-formulae-predicting-small-colour-differences

    ↩

  5. International Commission on Illumination. Terms Related to Planckian Radiation Temperature for Light Sources. CIE TN 013:2022.
    https://files.cie.co.at/Pubs/TN/CIE_TN_013_2022.pdf

    ↩

  6. International Telecommunication Union Radiocommunication Sector. Objective Metric for the Assessment of the Potential Visibility of Colour Differences in Television. Recommendation ITU-R BT.2124-0, January 2019.
    https://www.itu.int/dms_pubrec/itu-r/rec/bt/R-REC-BT.2124-0-201901-I%21%21PDF-E.pdf

    ↩↩↩

  7. International Telecommunication Union Radiocommunication Sector. Reference Electro-Optical Transfer Function for Flat Panel Displays Used in HDTV Studio Production. Recommendation ITU-R BT.1886-0, March 2011.
    https://www.itu.int/dms_pubrec/itu-r/rec/bt/R-REC-BT.1886-0-201103-I%21%21PDF-E.pdf

    ↩↩

  8. International Telecommunication Union Radiocommunication Sector. Image Parameter Values for High Dynamic Range Television for Use in Production and International Programme Exchange. Recommendation ITU-R BT.2100-3, February 2025.
    https://www.itu.int/dms_pubrec/itu-r/rec/bt/R-REC-BT.2100-3-202502-I%21%21PDF-E.pdf

    ↩↩

  9. International Telecommunication Union Radiocommunication Sector. Guidelines for operational practices in high dynamic range television production. Report ITU-R BT.2408-9, March 2026.
    https://www.itu.int/pub/R-REP-BT.2408-9-2026

    ↩

  10. International Telecommunication Union Radiocommunication Sector. Parameter Values for the HDTV Standards for Production and International Programme Exchange. Recommendation ITU-R BT.709-6, June 2015.
    https://www.itu.int/dms_pubrec/itu-r/rec/bt/R-REC-BT.709-6-201506-I%21%21PDF-E.pdf

    ↩

  11. Portrait Displays. About DeltaE (ΔE). Current page reviewed August 1, 2026.
    https://www.portrait.com/resource-center/about-deltae-e/

    ↩

  12. Portrait Displays. Absolute vs Relative RGB Charts in Calman. August 3, 2020.
    https://www.portrait.com/resource-center/absolute-vs-relative-rgb-charts-in-calman/

    ↩↩

  13. Portrait Displays. Design Mode in Calman Ultimate. Current page reviewed August 1, 2026.
    https://www.portrait.com/resource-center/design-mode-in-calman-ultimate/

    ↩

  14. Portrait Displays. Evaluating Displays Using the Calman Toolkit Workflow. Current page reviewed August 1, 2026.
    https://www.portrait.com/resource-center/evaluating-displays-using-the-calman-toolkit-workflow/

    ↩

  15. Light Illusion. Calibration Reports, Reading ColourSpace Graphs, and Volumetric Gamut Coverage Analysis. Current pages reviewed August 1, 2026.
    https://lightillusion.com/reports_guide.html

    ↩

  16. International Telecommunication Union Radiocommunication Sector. A Reference Viewing Environment for Evaluation of HDTV Program Material or Completed Programmes. Recommendation ITU-R BT.2035-0, July 2013.
    https://www.itu.int/dms_pubrec/itu-r/rec/bt/R-REC-BT.2035-0-201307-I%21%21PDF-E.pdf

    ↩
Related reading: Choosing a TV Colorimeter: Build a Compatible Measurement System Review how meter choice, correction, software, and pattern generation affect the measurements in a report. Next step: TV Calibration Workflow Complete the measurement sidebar with a controlled workflow for baseline measurements, grayscale and EOTF adjustment, CMS decisions, Samsung AutoCal, LUTs, HDR states, validation, and stopping rules.