ISO 15739:2003

INTERNATIONAL ISO
STANDARD 15739
First edition
2003-05-01


Photography — Electronic still-picture
imaging — Noise measurements
Photographie — Imagerie des prises de vue électroniques —
Mesurages du bruit




Reference number
ISO 15739:2003(E)
©
ISO 2003

---------------------- Page: 1 ----------------------
ISO 15739:2003(E)
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.


©  ISO 2003
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Published in Switzerland

ii © ISO 2003 — All rights reserved

---------------------- Page: 2 ----------------------
ISO 15739:2003(E)
Contents Page
Foreword. iv
Introduction . v
1 Scope. 1
2 Normative references. 1
3 Terms and definitions. 1
4 Test conditions. 3
4.1 General. 3
4.2 Illumination. 3
4.3 Temperature and relative humidity . 4
4.4 White balance. 4
4.5 Infrared (IR) blocking filter . 4
4.6 Photosite integration time. 4
4.7 Colour noise weighting . 4
4.8 Compression. 4
5 Noise measurement procedures . 5
5.1 General. 5
5.2 Uniform field noise measurement methods. 5
5.3 Cameras with removable lenses . 5
5.4 Whole camera. 6
5.5 Density measurements for methods 5.3 and 5.4 . 7
5.6 Test chart noise measurements . 7
6 Calculation and reporting of results . 8
6.1 General. 8
6.2 Signal-to-noise ratios — large area. 9
6.3 ISO DSC dynamic range. 10
6.4 Visual noise measurements. 11
6.5 Evaluation of noise measurements with software . 11
Annex A (normative) Test chart . 12
Annex B (normative) Noise component analysis. 15
Annex C (informative) Visual noise measurements. 18
Annex D (informative) Method for measuring edge noise. 22
Annex E (informative) Incremental gain calculation . 24
Bibliography . 26


© ISO 2003 — All rights reserved iii

---------------------- Page: 3 ----------------------
ISO 15739:2003(E)
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 15739 was prepared by Technical Committee ISO/TC 42, Photography.
iv © ISO 2003 — All rights reserved

---------------------- Page: 4 ----------------------
ISO 15739:2003(E)
Introduction
Noise is an important attribute of electronic photographic systems. Standardization assists users and
manufacturers in determining the quality of images that may be obtained from an electronic still-picture
camera. The camera noise measurements described in this International Standard are performed in the digital
domain, using digital analysis techniques. For electronic cameras that include only analogue outputs, the
analogue signal has to be digitized, so that the digital measurement can be performed. The digitizing
equipment has to be characterized, so that the effects of the digitization can be removed from the
measurement results. When this is not possible, the type of digitizing equipment used is to be reported along
with the measurement results.
Since the noise performance of an image sensor may vary significantly with exposure time and operating
temperature, these operating conditions are specified. The visibility of noise to human observers depends on
the magnitude of the noise, the apparent tone of the area containing the noise and the spatial frequency of the
noise. The magnitude of the noise present in an output representation depends on the noise present in the
stored image data and the contrast amplification or gain applied to the data in producing the output. The noise
visibility is different for the luminance (or monochrome) channel and the colour (or colour difference) channels.
Therefore, this International Standard accounts for these factors in measuring and reporting the camera noise
measurements. Annex A gives the construction of an ISO standard camera noise test chart. Annex B specifies
the method for determining the components of the digital camera noise from a number of samples. Annex C
describes a procedure using a human visual model as a method for weighting the spectral components of the
noise.
Many electronic still-picture cameras use extensive signal processing to reduce noise in uniform areas, and
the noise levels measured in the large area test defined in this International Standard may not be
representative of the noise levels found in the pictures taken by the camera. Therefore, new methods of
measuring edge noise have been investigated. One method is described in Annex D. The incremental gain
calculation is given in Annex E.

© ISO 2003 — All rights reserved v

---------------------- Page: 5 ----------------------
INTERNATIONAL STANDARD ISO 15739:2003(E)

Photography — Electronic still-picture imaging — Noise
measurements
1 Scope
This International Standard specifies methods for measuring and reporting the noise versus signal level and
dynamic range of electronic still-picture cameras. It applies to both monochrome and colour electronic still-
picture cameras.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
ISO 7589:2002, Photography — Illuminants for sensitometry — Specifications for daylight, incandescent
tungsten and printer
ISO 12232:1998, Photography — Electronic still-picture cameras — Determination of ISO speed
ISO 14524:1999, Photography — Electronic still-picture cameras — Methods for measuring opto-electronic
conversion functions (OECFs)
ITU-R BT.709-5, Parameter values for the HDTV Standards for production and International programme
exchange
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
3.1
camera opto-electronic conversion function
camera OECF
opto-electronic conversion function
focal plane opto-electronic conversion function
relationship between the input scene log luminances and the digital output levels for an opto-electronic digital
capture system
NOTE The units of measurement for this function are log candelas per square metre.
10
3.2
digital output level
digital code value
numerical value assigned to a particular output level
3.3
digital still camera
DSC
camera incorporating an image sensor which outputs a digital signal representing a still picture, or records a
digital signal representing a still picture on a removable medium, such as a memory card or magnetic disk
© ISO 2003 — All rights reserved 1

---------------------- Page: 6 ----------------------
ISO 15739:2003(E)
3.4
electronic still-picture camera
camera incorporating an image sensor which outputs an analogue or digital signal representing a still picture,
and/or records an analogue or digital signal representing a still picture on a removable medium, such as a
memory card or magnetic disk
3.5
image sensor
electronic device which converts incident electromagnetic radiation into an electronic signal
NOTE A charge coupled device (CCD) array is an example of an electronic signal.
3.6
incremental gain function
change in the output level (digital code value) divided by the change in the input level (luminance or exposure)
as a function of the light level
NOTE 1 For the determination of incremental gain values, log input values are not used.
NOTE 2 If the input exposure points are very finely spaced and the output noise is small compared to the quantization
interval, the incremental gain function can have a jagged shape. Such behaviour is an artefact of the quantization process
and is removed by using an appropriate smoothing algorithm, or by fitting a smooth curve to the data. In some cases, it
may be desirable to fit a curve to the input-output data and then determine the incremental gain function by taking the first
derivative of the function used for the curve fit.
3.7
incremental output signal
input level (luminance or exposure, not logged) multiplied by the system incremental gain at that level
3.8
incremental signal-to-noise ratio
ratio of the incremental output signal to the root mean square (rms) noise level, at a particular signal level
NOTE This is typically expressed as a graph or table showing the rms noise level versus output signal level for the
full range of output signal levels.
3.9
ISO DSC dynamic range
ratio of the maximum luminance level which appears unclipped to the minimum luminance level which can be
reproduced with an incremental signal-to-temporal-noise ratio of at least 1, as determined in accordance with
ISO 15739
3.10
noise
unwanted variations in the response of an imaging system
3.10.1
total noise
all the unwanted variations captured by a single exposure
3.10.2
fixed pattern noise
unwanted variations which are consistent for every exposure
3.10.3
temporally varying noise
random noise due to sensor dark current, photon shot noise, analogue processing and quantization, which
varies from one image to the next
2 © ISO 2003 — All rights reserved

---------------------- Page: 7 ----------------------
ISO 15739:2003(E)
3.11
noise spectral power distribution
curve or equation which expresses the camera output noise as a function of two-dimensional image spatial
frequencies
3.12
photosite integration time
total time period during which the photosites of an image sensor are able to integrate the light from the scene
to form an image
3.13
signal processing
operations performed by electronic circuits or algorithms which convert or modify the output of an image
sensor
3.14
video signal-to-noise ratio
ratio of the maximum (peak) output signal level to the root mean square (rms) noise level in video systems
NOTE 1 This value is typically expressed in decibels (dB).
NOTE 2 This term is not used to express the noise in an electronic still-picture imaging system.
4 Test conditions
4.1 General
The following measurement conditions should be used as nominal conditions when measuring the noise of an
electronic still-picture camera. If it is not possible or appropriate to achieve these nominal operating conditions,
the actual operating conditions shall be listed along with the reported results.
4.2 Illumination
4.2.1 Characteristics
The noise measurements shall indicate whether a source conforming to the daylight or tungsten illuminant
was used. ISO 7589 describes the procedures for determining if the characteristics of the illumination used in
a specific speed rating determination test are an acceptable match to the daylight and tungsten sensitometric
illuminants.
4.2.2 Daylight illumination
For daylight measurements without the camera lens, a source conforming to the ISO sensitometric daylight
illuminant specified in Table 1 of ISO 7589:2002 shall be used. This illuminant is defined as the product of the
spectral power distribution of CIE illuminant D55 and the spectral transmittance of the ISO standard camera
lens. For measurements with the camera lens in place, the spectral characteristics of the illuminant shall be
equivalent to CIE illuminant D55.
4.2.3 Tungsten illumination
For tungsten measurements without the camera lens, a source conforming to the ISO sensitometric tungsten
illuminant specified in Table 2 of ISO 7589:2002 shall be used. This illuminant is defined as the product of the
average spectral power distribution of experimentally measured sources having a colour temperature of
approximately 3 050 K and the spectral transmittance of the ISO standard camera lens. For measurements
with the camera lens in place, the spectral characteristics of the illuminant shall be equivalent to the product of
the average spectral power distribution of experimentally measured sources having a colour temperature of
approximately 3 050 K.
© ISO 2003 — All rights reserved 3

---------------------- Page: 8 ----------------------
ISO 15739:2003(E)
4.2.4 Reflection illumination
For measurements using a reflection test chart, the illumination should meet the uniformity requirements of the
measurement procedures described in Clause 5. The sources are positioned so that the angular distribution of
influx radiation is at its maximum at 45° to the test chart normal, and is negligible at angles of less than 40° or
more than 50° to the normal, at any point on the test chart.
4.3 Temperature and relative humidity
The ambient room temperature during the acquisition of the test data shall be 23 °C ± 2 °C, as specified in
ISO 554, and the relative humidity shall be 50 % ± 20 %. Additional measurements at 0 °C and 40 °C are
recommended. The normal camera operating temperature (internal rise above ambient) shall be achieved
before beginning the tests. If the ambient temperature varies throughout the room, for example as a result of
heat generated by light sources, the ambient room temperature shall be measured at a distance of between
0,1 m and 0,2 m from the camera under test.
4.4 White balance
For a colour camera, the camera white balance shall be adjusted, if possible, to provide proper white balance
(equal RGB signal levels) for the illumination light source, as specified in ISO 14524.
4.5 Infrared (IR) blocking filter
If required, an infrared blocking filter shall be used, as specified in ISO 14524.
4.6 Photosite integration time
The photosite integration time should not be longer than 1/30 s.
4.7 Colour noise weighting
For colour cameras using a single exposure process, the camera noise may be determined using a weighted
sum of the colour outputs to derive the luminance. If the proper luminance weighting values for the RGB
channel spectral sensitivities are known, they shall be used to calculate the luminance channel data. If these
values are not known, the following weighting, given in ITU-R BT.709, shall be used:
Y = 0,2125 R + 0,7154 G + 0,0721 B (1)
For colour cameras with luminance and colour-difference outputs, the standard deviation of the camera noise
may be computed from the luminance channel standard deviation σ (Y), the red minus luminance channel
standard deviation σ (R-Y) and the blue minus luminance channel standard deviation σ (B-Y). The following
Equation (2), as specified in 6.2.3 of ISO 12232:1998 shall be used:
2 2 2 1/2 (2)
σ (D ) or σ (D ) = [σ (Y) + 0,64 σ (R-Y) + 0,16 σ (B-Y) ]
H L
4.8 Compression
If the electronic still-picture camera includes any form of lossy compression, the compression shall be
disabled, if possible, during the noise measurements. If the compression cannot be turned off, then
measurements should be taken and the compression level reported with the noise measurement result, for
example, the actual camera switch setting (fine, standard, etc.) and the approximate average number of bits
per pixel.
4 © ISO 2003 — All rights reserved

---------------------- Page: 9 ----------------------
ISO 15739:2003(E)
5 Noise measurement procedures
5.1 General
These measurement procedures shall be used to determine the camera noise, the camera signal-to-noise
ratio and the camera dynamic range. The minimum requirement is to specify the signal-to-total-noise ratio and
the dynamic range of the digital camera under test. In addition, the fixed pattern and temporal noise
components can be expressed individually and the total noise weighted to match a known expression for
human visual response.
5.2 Uniform field noise measurement methods
The method of measuring the uniform field noise will be dependent on the type of camera and its level of
exposure automation. If the camera lens can be removed, then the sensor noise level can be measured
without any shading effects from the lens.
On automatic exposure cameras having through the lens (TTL) exposure control and no manual exposure
control override capability, the test chart and measurement methods described in 5.6 shall be used.
5.3 Cameras with removable lenses
5.3.1 General
This method involves the exposure of the electronic still-picture camera sensor directly to specific quantities of
uniform illumination with the lens removed. The illumination shall have the spectral characteristics specified in
4.2 and shall be produced by a small source at a distance, such that the largest dimensions of the source and
the sensor are no greater than one twentieth of the distance between them, as shown in Figure 1. Reflective
surfaces shall not be placed where they could cause additional illumination to be incident on the sensor.

Key
1 light source
2 test density
3 camera under test
4 lens removed
Figure 1 — Illumination for cameras with removable lenses
5.3.2 The camera OECF shall first be measured according to ISO 14524.
5.3.3 The light source shall be fixed level with combined short term and supply amplitude variations of less
than ± 2 %.
© ISO 2003 — All rights reserved 5

---------------------- Page: 10 ----------------------
ISO 15739:2003(E)
5.3.4 The light source shall be adjusted to give the maximum unclipped level from the camera. If necessary,
an appropriate neutral density filter should be used to cover the camera exposure control sensor, in order to
adjust the signal level to provide the maximum unclipped level from the camera.
5.3.5 Test densities shall completely cover the area exposed, when the camera lens was removed.
5.4 Whole camera
5.4.1 General
These measurements shall be used for all cameras that use manual exposure control, or exposure control
based on a separate exposure control sensor.
5.4.2 The camera OECF shall first be measured according to ISO 14524.
5.4.3 The light source shall be fixed level with combined short term and supply amplitude variations of less
than ± 2 %.
5.4.4 If the camera lens focus is adjustable, it shall be set to infinity.
5.4.5 The diffuser shall be uniform and close to the camera, preferably less than one tenth of the minimum
focus distance of the camera under test, to prevent diffuser blemishes from influencing the noise
measurements. The diffuser may be illuminated by either transmissive or reflective light (see Figure 2).

Key
1 uniform fixed level light source
2 diffuser
3 test density
4 camera lens
5 camera under test
6 camera exposure control sensor
Figure 2 — Uniform field noise measurements
5.4.6 The light source and diffuser shall be adjusted to give the maximum unclipped signal level from the
camera. If necessary, an appropriate spectrally neutral density filter shall be used to cover the camera
exposure control sensor to adjust the signal level, in order to provide the maximum unclipped signal level from
the camera.
5.4.7 Test densities shall only cover the camera lens.
6 © ISO 2003 — All rights reserved

---------------------- Page: 11 ----------------------
ISO 15739:2003(E)
5.5 Density measurements for methods 5.3 and 5.4
5.5.1 A test density of 0,9 (13 % transmittance) shall be used to provide an “18 % reference” signal level
equal to that of an 18 % reflectance card with a camera peak signal level of 140 %, corresponding to the
saturation signal level in ISO 12232.
5.5.2 A test density of 2,0 (1 % transmittance) shall be used to provide a “black reference” signal level to
determine the camera dynamic range.
5.5.3 Any optional additional test densities should correspond to the test chart densities listed in ISO 14524.
5.6 Test chart noise measurements
5.6.1 General
These measurements shall be used for TTL automatic exposure cameras having no manual exposure control
override.
5.6.2 The camera OECF shall first be measured in accordance with ISO 14524.
5.6.3 The light source shall be fixed level with combined short term and supply amplitude variations of less
than ± 2 %.
5.6.4 The light source shall be adjusted to give the maximum unclipped level from the camera, on the
lightest patch of the test chart, e.g. patch 12 on the OECF or noise test chart. If necessary, an appropriate
neutral density filter should cover the camera exposure control sensor, in order to adjust the lightest patch
signal level, to the maximum unclipped level.
5.6.5 The test chart shall be either the test chart specified in ISO 14524 or the test chart specified in
Annex A. The test chart can be either transmissive or reflective (see Figure 3).
5.6.6 If a transmissive test chart is used, the diffuser shall be uniform. If a reflective test chart is used, the
chart illumination shall be as described in 4.2.4. Additional shielding of the camera may be necessary to
prevent stray illumination from the light sources, or from other reflections, entering the camera lens.
5.6.7 Non-uniformity in the test chart density patches shall be less than one tenth of the expected camera
noise level, and any image structure spatial components shall be at a spatial frequency of at least 10 times
higher than the camera limiting resolution.
5.6.8 In order to overcome erroneous results due to gamma lookup tables in the digital still-picture cameras,
the noise measurement at the density of 0,9 shall be averaged with measurements at densities either side of
0,9 (measurements at densities of 1,05 and 0,77).
5.6.9 Shading from the illumination should be such that the effect of the shading does not affect the
standard deviation of the digital code values by more than one digital level across the sample patch.
5.6.10 The test target shall be correctly focused by the camera under test.
5.6.11 The whole apparatus shall be securely mounted to reduce movement between exposures to less than
one quarter of a pixel.
© ISO 2003 — All rights reserved 7

---------------------- Page: 12 ----------------------
ISO 15739:2003(E)

a) Test arrangement using a transmissive test chart

b) Test arrangement using a reflective test chart
Key
1 uniform fixed level light source
2 diffuser
3 test chart
4 camera under test
5 camera lens
6 45° uniform illumination
7 additional shielding
Figure 3 — Test chart noise measurements
6 Calculation and reporting of results
6.1 General
The measurements made in 5.5 and 5.6 are converted to reported noise values as follows.
For the measurements made according to 5.3 and 5.4, a minimum of eight images shall be captured for each
test density. The mean code value and the rms noise level shall be determined from an area of not less than
64 × 64 pixels in the centre of each of the images.
For the test chart case, a minimum of eight images shall be captured in a single session. The mean code
value and rms noise level shall be determined from an area of not less than 64 × 64 pixels in the centre of
each of the patches of the test chart specified in Annex A. The calculation can be conducted by using the
plug-in software described in 6.5.
8 © ISO 2003 — All rights reserved

---------------------- Page: 13 ----------------------
ISO 15739:2003(E)
6.2 Signal-to-noise ratios — large area
6.2.1 General
The signal-to-noise ratio is determined from data captured at the “18 % reference” signal level. In methods 5.3
and 5.4, this is accomplished by using the densities specified in 5.6.7, and in method 5.6, by using the density
patches in block 13 on the test chart specified in Annex A.
The total noise is converted to an input referred incremental signal-to-noise ratio for the test density, and
reported as the ISO standard camera signal-to-noise ratio.
6.2.2 The signal-to-total-noise ratio is determined by:
L ××0,18 incremental gain
sat
(3)
σ
total
for the case of a density of 0,9, where:
L is the target luminance which gives the maximum unclipped output from the camera, e.g. for an
sat
eight bit system, this is 255;
0,18 is the 18 % reflectance of the target of a density of 0,9 wi
...