Japanese document

Description of Exif file format

Currently, most of new digicam use Exif file format to store images. This specification was made by JEIDA, but there is no open document in internet. So I made tiny description about Exif from some of open documents that can get from internet.

Add: Now we can get official document of Exif2.1 from PIMA's website.

ISO is now working to build DCF (Design rule for Camera File system) Specification. The entire digicam manufacturer is moving to adopt it and their newest digicam already uses DCF. DCF defines whole file-system of digicam; directory structure, file naming method, character set and file format etc. The file format of DCF is based on Exif2.1 specification.

I believe this document is basically based on Exif2.1/DCF specification, if you have information about 'unknown' item or find errata, please e-mail me, TsuruZoh Tachibanaya , t s u r u z o h @ b a . w a k w a k . c o m


This is a FREE document, you may use this document for any purpose (commercial/non-commercial) of all/part of this document. All trade names mentioned in this document are trademarks or registered trademarks of their respective holders.

TsuruZoh Tachibanaya, t s u r u z o h @ b a . w a k w a k . c o m

http://park2.wakwak.com/~tsuruzoh/

rev. 1.4 Feb.03,2001
rev. 1.3 Sep.09,2000
rev. 1.2 Jul.19,2000
rev. 1.1 Dec.19,1999
rev. 1.0 May.28,1999

Referenced materials

Exif file format written by itojun (Japanese language document)
Exif file format written by Mamoru Ohno (Japanese language document)
TIFF6.0 Specification written by Adobe
TIFF/EP Specification written by ISO TC42 WG18
exifdump program written by Thierry Boush
DCF Specification written by ISO TC42 WG18
Exif2.1 Specification written by JEIDA


What is Exif file format?

Basically, Exif file format is the same as JPEG file format. Exif inserts some of image/digicam information data and thumbnail image to JPEG in conformity to JPEG specification. Therefore you can view Exif format image files by JPEG compliant Internet browser/Picture viewer/Photo retouch software etc. as a usual JPEG image files.

JPEG format and Marker

Every JPEG file starts from binary value '0xFFD8', ends by binary value '0xFFD9'. There are several binary 0xFFXX data in JPEG data, they are called as "Marker", and it means the period of JPEG information data. 0xFFD8 means SOI(Start of image), 0xFFD9 means EOI(End of image). These two special Markers have no data following, the other Markers have data with it. Basic format of Marker is below.

0xFF+Marker Number(1 byte)+Data size(2 bytes)+Data(n bytes)

Data size(2 Bytes) has "Motorola" byte align, starts from bigger digits. Please notice that "Data" contains Data size descriptor, if there is a Marker like this;

FF C1 00 0C

It means this Marker(0xFFC1) has 0x000C(equal 12)bytes of data. But the data size '12' includes "Data size" descriptor, it follows only 10 bytes of data after 0x000C.

In JPEG format, some of Markers describe data, then SOS(Start of stream) Marker placed. After the SOS marker, JPEG image stream starts and terminated by EOI Marker.

SOI Marker Marker XX size=SSSS Marker YY size=TTTT SOS Marker size=UUUU Image stream EOI Marker
FFD8 FFXX SSSS DDDD...... FFYY TTTT DDDD...... FFDA UUUU DDDD.... I I I I.... FFD9

Marker used by Exif

The marker 0xFFE0~0xFFEF is named "Application Marker", not necessary for decoding JPEG image. They are used by user application. For example, older olympus/canon/casio/agfa digicam use JFIF(JPEG File Interchange Format) for storing images. JFIF uses APP0(0xFFE0) Marker for inserting digicam configuration data and thumbnail image.

Also Exif uses an Application Marker for inserting data, but Exif uses APP1(0xFFE1) Marker to avoid a conflict with JFIF format. Every Exif file formats starts from this format;

SOI Marker APP1 Marker APP1 Data Other Marker
FFD8 FFE1 SSSS 457869660000 TTTT...... FFXX SSSS DDDD......

It starts from SOI(0xFFD8) Marker, so it's a JPEG file. Then APP1 Marker follows immediately. All the data of Exif are stored in this APP1 data area. The part of "SSSS" on upper table means the size of APP1 data area (Exif data area). Please notice that the size "SSSS" includes the size of descriptor itself also.

After the "SSSS", APP1 data starts. The first part is a special data to identify whether Exif or not, ASCII character "Exif" and 2bytes of 0x00 are used.

After the APP1 Marker area, the other JPEG Markers follows.

Exif data structure

Roughly structure of Exif data (APP1) is shown as below. This is a case of "Intel" byte aligns and it contains JPEG format thumbnail. As described above, Exif data is starts from ASCII character "Exif" and 2bytes of 0x00, and then Exif data follows. Exif uses TIFF format to store data. For more details of TIFF format, please refer to "TIFF6.0 specification".


FFE1 APP1 Marker
SSSS APP1 Data APP1 Data Size
45786966 0000 Exif Header
49492A00 08000000 TIFF Header
XXXX. . . . IFD0 (main image) Directory
LLLLLLLL Link to IFD1
XXXX. . . . Data area of IFD0
XXXX. . . .    Exif SubIFD Directory
00000000 End of Link
XXXX. . . . Data area of Exif SubIFD
XXXX. . . .    Interoperability IFD Directory
00000000 End of Link
XXXX. . . . Data area of Interoperability IFD
XXXX. . . . Makernote IFD Directory
00000000 End of Link
XXXX. . . . Data area of Makernote IFD
XXXX. . . . IFD1(thumbnail image) Directory
00000000 End of Link
XXXX. . . . Data area of IFD1
FFD8XXXX. . . XXXXFFD9 Thumbnail image

Structure of TIFF header

First 8bytes of TIFF format are TIFF header. First 2bytes defines byte align of TIFF data. If it is 0x4949="I I", it means "Intel" type byte align. If it is 0x4d4d="MM", it means "Motorola" type byte align. For example, value '305,419,896' is noted as 0x12345678 by sixteenth system. At the Motorola align, it is stored as 0x12,0x34,0x56,0x78. If it's Intel align, it is stored as 0x78,0x56,0x34,0x12. It seems that most of digicam uses Intel align. Ricoh uses Motorola align. Sony uses Intel align except D700. Kodak DC200/210/240 use Motorola aligns, but DC220/260 use Intel aligns though they are using PowerPC! Therefore when we need the value of Exif data, we MUST check byte align every time. Though JPEG data uses Motorola align only, Exif allows both alignments. I can't understand why Exif didn't fix a byte align to Motorola.

Next 2bytes are always 2bytes-length value of 0x002A. If the data uses Intel align, next 2bytes are "0x2a,0x00". If it uses Motorola, they are "0x00,0x2a". The last 4bytes of TIFF header are an offset to the first IFD(Image File Directory, described in next chapter). Includes this offset, all the offset value used in TIFF format counts offset bytes from the first of TIFF header("I I" or "MM"). Usually the first IFD starts immediately next to TIFF header, so this offset has value '0x00000008'.

Byte align TAG Mark Offset to first IFD
"I I" or "MM" 0x002a 0x00000008

IFD : Image file directory

Next to TIFF header, there is the first IFD:Image File Directory. It contains image information data. At the chart below, the first 2bytes('EEEE') means the number of directory entry contains in this IFD. Then directory entry (12bytes per entry) follows. After last directory entry, there is a 4bytes of data('LLLLLLLL' at the chart), it means an offset to next IFD. If its value is '0x00000000', it means this is the last IFD and there is no linked IFD.

EEEE No. of directory entry
TTTT ffff NNNNNNNN DDDDDDDD Entry 0
TTTT ffff NNNNNNNN DDDDDDDD Entry 1
. . . . . . . . . . . . . . .
TTTT ffff NNNNNNNN DDDDDDDD Entry EEEE-1
LLLLLLLL Offset to next IFD

'TTTT'(2bytes) of above chart is Tag number, this shows a kind of data. 'ffff'(2bytes) is data format, 'NNNNNNNN'(4bytes) is number of components. 'DDDDDDDD'(4bytes) contains a data value or offset to data value.

Data format

Data format ('ffff' at the above chart) is defined as below. "rational" means a fractional value, it contains 2-signed/unsigned long integer values, and the first represents the numerator, and the second, the denominator.
Value 1 2 3 4 5 6
Format unsigned byte ascii strings unsigned short unsigned long unsigned rational signed byte
Bytes/component 1 1 2 4 8 1
Value 7 8 9 10 11 12
Format undefined signed short signed long signed rational single float double float
Bytes/component 1 2 4 8 4 8

You can get the total data byte length by multiplies a 'bytes/components' value (see above chart) by number of components stored 'NNNNNNNN' area. If total data length is less than 4bytes, 'DDDDDDDD' contains the value of that Tag. If its size is over 4bytes, 'DDDDDDDD' contains the offset to data stored address.

IFD data structure

At Exif format, the first IFD is IFD0(IFD of main image), then it links to IFD1(IFD of thumbnail image) and IFD link is terminated. But IFD0/IFD1 doesn't contain any digicam's information such as shutter speed, focal length etc. IFD0 always contains special Tag Exif Offset (0x8769), it shows an offset to Exif SubIFD. Exif SubIFD is IFD formatted data also, it contains digicam's information.

In case of Extended Exif format (Exif2.1/DCF), Exif SubIFD contains special Tag Exif Interoperability Offset (0xa005). It also points to the Interoperability IFD. By the DCF specification, this tag is mandatory and both SubIFD (main image IFD) and IFD1 (thumbnail image IFD) may have Interoperability IFD. In usual, only main image have this tag.

And some of digicam uses IFD data format for Makernote; Maker-specific magic number area. It's very hard to decision whether makernote is IFD format or not, be careful to coding program. See Appendix for information of Makernote. 
0000: 49 49 2A 00 08 00 00 00-02 00 1A 01 05 00 01 00
0010: 00 00 26 00 00 00 69 87-04 00 01 00 00 00 11 02
0020: 00 00 40 00 00 00 48 00-00 00 01 00 00 00
If the first part of TIFF data is above, it can read as;

Thumbnail image

Exif format contains thumbnail of image (except Ricoh RDC-300Z). Usually it is located next to the IFD1. There are 3 formats for thumbnails; JPEG format(JPEG uses YCbCr), RGB TIFF format, YCbCr TIFF format. It seems that JPEG format and 160x120 pixels of size are recommended thumbnail format for Exif2.1 or later. By the DCF specification, thumbnail image MUST use JPEG format and image size is fixed to 160x120 pixels.

JPEG format thumbnail

If the value of Compression(0x0103) Tag in IFD1 is '6', thumbnail image format is JPEG. Most of Exif image uses JPEG format for thumbnail. In that case, you can get offset of thumbnail from JpegIFOffset(0x0201) Tag in IFD1, size of thumbnail from JpegIFByteCount(0x0202) Tag. Data format is ordinary JPEG format, starts from 0xFFD8 and ends by 0xFFD9.

TIFF format thumbnail

If the value of Compression(0x0103) Tag in IFD1 is '1', thumbnail image format is no compression(called TIFF image). Start point of thumbnail data is StripOffset(0x0111) Tag, size of thumbnail is the sum of StripByteCounts(0x0117) Tag.

If thumbnail uses no compression and PhotometricInterpretation(0x0106)Tag in IFD1 has a value '2', thumbnail uses RGB format. In that case, you can see thumbnail image by simply copy data to computer's RGB format(such as BMP format, or copy to VRAM directory). Kodak DC-210/220/260 uses this format. Be mention that at TIFF stores pixel data as 'RGB' order, but BMP stores 'BGR' order.
If that tag has a value '6', thumbnail uses YCbCr format. If you want to see thumbnail, you must convert it to RGB. Ricoh RDC4200/4300, Fuji DS-7/300 and DX-5/7/9 use this format(newer RDC5000/MX-X00 series use JPEG). Next section is brief description to conversion of Fuji DS's thumbnail. For more details, refer to TIFF6.0 specification.

At DX-5/7/9, YCbCrSubsampling(0x0212) has values of '2,1', PlanarConfiguration(0x011c) has a value '1'. So the data align of this image is below.

Y(0,0),Y(1,0),Cb(0,0),Cr(0,0), Y(2,0),Y(3,0),Cb(2,0),Cr(3.0), Y(4,0),Y(5,0),Cb(4,0),Cr(4,0). . . .

The numeric in parenthesis is pixel coordinates. DX series' YCbCrCoefficients(0x0211) has values '0.299/0.587/0.114', ReferenceBlackWhite(0x0214) has values '0,255,128,255,128,255'. Therefore to convert from Y/Cb/Cr to RGB is;

B(0,0)=(Cb-128)*(2-0.114*2)+Y(0,0)
R(0,0)=(Cr-128)*(2-0.299*2)+Y(0,0)
G(0,0)=(Y(0,0)-0.114*B(0,0)-0.299*R(0,0))/0.587

Horizontal subsampling is a value '2', so you can calculate B(1,0)/R(1,0)/G(1,0) by using the Y(1,0) and Cr(0,0)/Cb(0,0). Repeat this conversion by value of ImageWidth(0x0100) and ImageLength(0x0101).

Tag number used by Exif/TIFF

Tag numbers used by Exif/TIFF are shown as below. If the Tag has upper limit of components number, CompoNo column has numeric value. If it has no value, there is no limitation.
Tags used by IFD0 (main image)
Tag No. Tag Name Format CompoNo Desc.
0x010e ImageDescription ascii string
Describes image. Two-byte character code such as Chinese/Korean/Japanese cannot be used.
0x010f Make ascii string
Shows manufacturer of digicam. In the Exif standard, this tag is optional, but it is mandatory for DCF.
0x0110 Model ascii string
Shows model number of digicam. In the Exif standard, this tag is optional, but it is mandatory for DCF.
0x0112 Orientation unsigned short 1
Value 0th Row 0th Column
1 top left side
2 top right side
3 bottom right side
4 bottom left side
5 left side top
6 right side top
7 right side bottom
8 left side bottom
The orientation of the camera relative to the scene, when the image was captured. The relation of the '0th row' and '0th column' to visual position is shown as right.
0x011a XResolution unsigned rational 1 Display/Print resolution of image. Default value is 1/72inch, but it has no mean because personal computer doesn't use this value to display/print out.
0x011b YResolution unsigned rational 1
0x0128 ResolutionUnit unsigned short 1 Unit of XResolution(0x011a)/YResolution(0x011b). '1' means no-unit, '2' means inch, '3' means centimeter. Default value is '2'(inch).
0x0131 Software ascii string
Shows firmware(internal software of digicam) version number.
0x0132 DateTime ascii string 20 Date/Time of image was last modified. Data format is "YYYY:MM:DD HH:MM:SS"+0x00, total 20bytes. If clock has not set or digicam doesn't have clock, the field may be filled with spaces. In usual, it has the same value of DateTimeOriginal(0x9003)
0x013e WhitePoint unsigned rational 2 Defines chromaticity of white point of the image. If the image uses CIE Standard Illumination D65(known as international standard of 'daylight'), the values are '3127/10000,3290/10000'.
0x013f PrimaryChromaticities unsigned rational 6 Defines chromaticity of the primaries of the image. If the image uses CCIR Recommendation 709 primaries, values are '640/1000,330/1000,300/1000,600/1000,150/1000,0/1000'.
0x0211 YCbCrCoefficients unsigned rational 3 When image format is YCbCr, this value shows a constant to translate it to RGB format. In usual, values are '0.299/0.587/0.114'.
0x0213 YCbCrPositioning unsigned short 1 When image format is YCbCr and uses 'Subsampling'(cropping of chroma data, all the digicam do that), defines the chroma sample point of subsampling pixel array. '1' means the center of pixel array, '2' means the datum point.
0x0214 ReferenceBlackWhite unsigned rational 6 Shows reference value of black point/white point. In case of YCbCr format, first 2 show black/white of Y, next 2 are Cb, last 2 are Cr. In case of RGB format, first 2 show black/white of R, next 2 are G, last 2 are B.
0x8298 Copyright ascii string
Shows copyright information
0x8769 ExifOffset unsigned long 1 Offset to Exif Sub IFD
Tags used by Exif SubIFD
Tag No. Tag Name Format CompoNo Desc.
0x829a ExposureTime unsigned rational 1 Exposure time (reciprocal of shutter speed). Unit is second.
0x829d FNumber unsigned rational 1 The actual F-number(F-stop) of lens when the image was taken.
0x8822 ExposureProgram unsigned short 1 Exposure program that the camera used when image was taken. '1' means manual control, '2' program normal, '3' aperture priority, '4' shutter priority, '5' program creative (slow program), '6' program action(high-speed program), '7' portrait mode, '8' landscape mode.
0x8827 ISOSpeedRatings unsigned short 2 CCD sensitivity equivalent to Ag-Hr film speedrate.
0x9000 ExifVersion undefined 4 Exif version number. Stored as 4bytes of ASCII character. If the picture is based on Exif V2.1, value is "0210". Since the type is 'undefined', there is no NULL(0x00) for termination.
0x9003 DateTimeOriginal ascii string 20 Date/Time of original image taken. This value should not be modified by user program. Data format is "YYYY:MM:DD HH:MM:SS"+0x00, total 20bytes. If clock has not set or digicam doesn't have clock, the field may be filled with spaces. In the Exif standard, this tag is optional, but it is mandatory for DCF.
0x9004 DateTimeDigitized ascii string 20 Date/Time of image digitized. Usually, it contains the same value of DateTimeOriginal(0x9003). Data format is "YYYY:MM:DD HH:MM:SS"+0x00, total 20bytes. If clock has not set or digicam doesn't have clock, the field may be filled with spaces. In the Exif standard, this tag is optional, but it is mandatory for DCF.
0x9101 ComponentsConfiguration undefined
Shows the order of pixel data. Most of case '0x04,0x05,0x06,0x00' is used for RGB-format and '0x01,0x02,0x03,0x00' for YCbCr-format. 0x00:does not exist, 0x01:Y, 0x02:Cb, 0x03:Cr, 0x04:Red, 0x05:Green, 0x06:Bllue.
0x9102 CompressedBitsPerPixel unsigned rational 1 The average compression ratio of JPEG (rough estimate).
0x9201 ShutterSpeedValue signed rational 1 Shutter speed by APEX value. To convert this value to ordinary 'Shutter Speed'; calculate this value's power of 2, then reciprocal. For example, if the ShutterSpeedValue is '4', shutter speed is 1/(24)=1/16 second.
0x9202 ApertureValue unsigned rational 1 The actual aperture value of lens when the image was taken. Unit is APEX. To convert this value to ordinary F-number(F-stop), calculate this value's power of root 2 (=1.4142). For example, if the ApertureValue is '5', F-number is 1.41425 = F5.6.
0x9203 BrightnessValue signed rational 1 Brightness of taken subject, unit is APEX. To calculate Exposure(Ev) from BrigtnessValue(Bv), you must add SensitivityValue(Sv).
Ev=Bv+Sv   Sv=log2(ISOSpeedRating/3.125)
ISO100:Sv=5, ISO200:Sv=6, ISO400:Sv=7, ISO125:Sv=5.32.
0x9204 ExposureBiasValue signed rational 1 Exposure bias(compensation) value of taking picture. Unit is APEX(EV).
0x9205 MaxApertureValue unsigned rational 1 Maximum aperture value of lens. You can convert to F-number by calculating power of root 2 (same process of ApertureValue:0x9202).
0x9206 SubjectDistance signed rational 1 Distance to focus point, unit is meter.
0x9207 MeteringMode unsigned short 1 Exposure metering method. '0' means unknown, '1' average, '2' center weighted average, '3' spot, '4' multi-spot, '5' multi-segment, '6' partial, '255' other.
0x9208 LightSource unsigned short 1 Light source, actually this means white balance setting. '0' means unknown, '1' daylight, '2' fluorescent, '3' tungsten, '10' flash, '17' standard light A, '18' standard light B, '19' standard light C, '20' D55, '21' D65, '22' D75, '255' other.
0x9209 Flash unsigned short 1 '0' means flash did not fire, '1' flash fired, '5' flash fired but strobe return light not detected, '7' flash fired and strobe return light detected.
0x920a FocalLength unsigned rational 1 Focal length of lens used to take image. Unit is millimeter.
0x927c MakerNote undefined
Maker dependent internal data. Some of maker such as Olympus/Nikon/Sanyo etc. uses IFD format for this area.
0x9286 UserComment undefined
Stores user comment. This tag allows to use two-byte character code or unicode. First 8 bytes describe the character code. 'JIS' is a Japanese character code (known as Kanji).
'0x41,0x53,0x43,0x49,0x49,0x00,0x00,0x00':ASCII
'0x4a,0x49,0x53,0x00,0x00,0x00,0x00,0x00':JIS
'0x55,0x4e,0x49,0x43,0x4f,0x44,0x45,0x00':Unicode
'0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00':Undefined
0x9290 SubsecTime ascii string
Some of digicam can take 2~30 pictures per second, but DateTime/DateTimeOriginal/DateTimeDigitized tag can't record the sub-second time. SubsecTime tag is used to record it.
For example, DateTimeOriginal = "1996:09:01 09:15:30", SubSecTimeOriginal = "130", Combined original time is "1996:09:01 09:15:30.130"
0x9291 SubsecTimeOriginal ascii string
0x9292 SubsecTimeDigitized ascii string
0xa000 FlashPixVersion undefined 4 Stores FlashPix version. If the image data is based on FlashPix formar Ver.1.0, value is "0100". Since the type is 'undefined', there is no NULL(0x00) for termination.
0xa001 ColorSpace unsigned short 1 Defines Color Space. DCF image must use sRGB color space so value is always '1'. If the picture uses the other color space, value is '65535':Uncalibrated.
0xa002 ExifImageWidth unsigned short/long 1 Size of main image.
0xa003 ExifImageHeight unsigned short/long 1
0xa004 RelatedSoundFile ascii string
If this digicam can record audio data with image, shows name of audio data.
0xa005 ExifInteroperabilityOffset unsigned long 1 Extension of "ExifR98", detail is unknown. This value is offset to IFD format data. Currently there are 2 directory entries, first one is Tag0x0001, value is "R98", next is Tag0x0002, value is "0100".
0xa20e FocalPlaneXResolution unsigned rational 1 Pixel density at CCD's position. If you have MegaPixel digicam and take a picture by lower resolution(e.g.VGA mode), this value is re-sampled by picture resolution. In such case, FocalPlaneResolution is not same as CCD's actual resolution.
0xa20f FocalPlaneYResolution unsigned rational 1
0xa210 FocalPlaneResolutionUnit unsigned short 1 Unit of FocalPlaneXResoluton/FocalPlaneYResolution. '1' means no-unit, '2' inch, '3' centimeter.

Note:Some of Fujifilm's digicam(e.g.FX2700,FX2900,Finepix4700Z/40i etc) uses value '3' so it must be 'centimeter', but it seems that they use a '8.3mm?'(1/3in.?) to their ResolutionUnit. Fuji's BUG? Finepix4900Z has been changed to use value '2' but it doesn't match to actual value also.
0xa215 ExposureIndex unsigned rational 1 Same as ISOSpeedRatings(0x8827) but data type is unsigned rational. Only Kodak's digicam uses this tag instead of ISOSpeedRating, I don't know why(historical reason?).
0xa217 SensingMethod unsigned short 1 Shows type of image sensor unit. '2' means 1 chip color area sensor, most of all digicam use this type.
0xa300 FileSource undefined 1 Indicates the image source. Value '0x03' means the image source is digital still camera.
0xa301 SceneType undefined 1 Indicates the type of scene. Value '0x01' means that the image was directly photographed.
0xa302 CFAPattern undefined
Indicates the Color filter array(CFA) geometric pattern.
Length Type Meaning
2 short Horizontal repeat pixel unit = n
2 short Vertical repeat pixel unit = m
1 byte CFA value[0,0]
:
:
:
1 byte CFA value[n-1,0]
1 byte CFA value[0,1]
:
:
:
1 byte CFA value[n-1,m-1]

The relation of filter color to CFA value is shown below.
Filter Color Red Green Blue Cyan Magenta Yellow White
CFA value 0 1 2 3 4 5 6

R G
G B
For example, ordinary RGB filter uses the repetition of left chart, the value is '0x0002,0x0002,0x00,0x01,0x01,0x02'.



Tags used by Interoperability IFD
Tag No. Tag Name Format CompoNo Desc.
0x0001 InteroperabilityIndex Ascii string 4 If this IFD is main image's IFD and the file content is equivalent to ExifR98 v1.0, the value is "R98". If thumbnail image's, value is "THM".
0x0002 InteroperabilityVersion Undefined 4 Records the interoperability version. "0100" means version 1.00.
0x1000 RelatedImageFileFormat Ascii string any Records the file format of image file. Value is ascii string (e.g. "Exif JPEG Ver. 2.1").
0x1001 RelatedImageWidth Short or Long 1 Records the image size.
0x1001 RelatedImageLength Short or Long 1



Tags used by IFD1 (thumbnail image)
Tag No. Tag Name Format CompoNo Desc.
0x0100 ImageWidth unsigned short/long 1 Shows size of thumbnail image.
0x0101 ImageLength unsigned short/long 1
0x0102 BitsPerSample unsigned short 3 When image format is no compression, this value shows the number of bits per component for each pixel. Usually this value is '8,8,8'
0x0103 Compression unsigned short 1 Shows compression method. '1' means no compression, '6' means JPEG compression.
0x0106 PhotometricInterpretation unsigned short 1 Shows the color space of the image data components. '1' means monochrome, '2' means RGB, '6' means YCbCr.
0x0111 StripOffsets unsigned short/long
When image format is no compression, this value shows offset to image data. In some case image data is striped and this value is plural.
0x0115 SamplesPerPixel unsigned short 1 When image format is no compression, this value shows the number of components stored for each pixel. At color image, this value is '3'.
0x0116 RowsPerStrip unsigned short/long 1 When image format is no compression and image has stored as strip, this value shows how many rows stored to each strip. If image has not striped, this value is the same as ImageLength(0x0101).
0x0117 StripByteConunts unsigned short/long
When image format is no compression and stored as strip, this value shows how many bytes used for each strip and this value is plural. If image has not stripped, this value is single and means whole data size of image.
0x011a XResolution unsigned rational 1 Display/Print resolution of image. Large number of digicam uses 1/72inch, but it has no mean because personal computer doesn't use this value to display/print out.
0x011b YResolution unsigned rational 1
0x011c PlanarConfiguration unsigned short 1 When image format is no compression YCbCr, this value shows byte aligns of YCbCr data. If value is '1', Y/Cb/Cr value is chunky format, contiguous for each subsampling pixel. If value is '2', Y/Cb/Cr value is separated and stored to Y plane/Cb plane/Cr plane format.
0x0128 ResolutionUnit unsigned short 1 Unit of XResolution(0x011a)/YResolution(0x011b). '1' means inch, '2' means centimeter.
0x0201 JpegIFOffset unsigned long 1 When image format is JPEG, this value show offset to JPEG data stored.
0x0202 JpegIFByteCount unsigned long 1 When image format is JPEG, this value shows data size of JPEG image.
0x0211 YCbCrCoefficients unsigned rational 3 When image format is YCbCr, this value shows constants to translate it to RGB format. In usual, '0.299/0.587/0.114' are used.
0x0212 YCbCrSubSampling unsigned short 2 When image format is YCbCr and uses subsampling(cropping of chroma data, all the digicam do that), this value shows how many chroma data subsampled. First value shows horizontal, next value shows vertical subsample rate.
0x0213 YCbCrPositioning unsigned short 1 When image format is YCbCr and uses 'Subsampling'(cropping of chroma data, all the digicam do that), this value defines the chroma sample point of subsampled pixel array. '1' means the center of pixel array, '2' means the datum point(0,0).
0x0214 ReferenceBlackWhite unsigned rational 6 Shows reference value of black point/white point. In case of YCbCr format, first 2 show black/white of Y, next 2 are Cb, last 2 are Cr. In case of RGB format, first 2 show black/white of R, next 2 are G, last 2 are B.
Misc Tags
Tag No. Tag Name Format CompoNo Desc.
0x00fe NewSubfileType unsigned long 1
0x00ff SubfileType unsigned short 1
0x012d TransferFunction unsigned short 3
0x013b Artist ascii string

0x013d Predictor unsigned short 1
0x0142 TileWidth unsigned short 1
0x0143 TileLength unsigned short 1
0x0144 TileOffsets unsigned long

0x0145 TileByteCounts unsigned short

0x014a SubIFDs unsigned long

0x015b JPEGTables undefined

0x828d CFARepeatPatternDim unsigned short 2
0x828e CFAPattern unsigned byte

0x828f BatteryLevel unsigned rational 1
0x83bb IPTC/NAA unsigned long

0x8773 InterColorProfile undefined

0x8824 SpectralSensitivity ascii string

0x8825 GPSInfo unsigned long 1
0x8828 OECF undefined

0x8829 Interlace unsigned short 1
0x882a TimeZoneOffset signed short 1
0x882b SelfTimerMode unsigned short 1
0x920b FlashEnergy unsigned rational 1
0x920c SpatialFrequencyResponse undefined

0x920d Noise undefined

0x9211 ImageNumber unsigned long 1
0x9212 SecurityClassification ascii string 1
0x9213 ImageHistory ascii string

0x9214 SubjectLocation unsigned short 4
0x9215 ExposureIndex unsigned rational 1
0x9216 TIFF/EPStandardID unsigned byte 4
0xa20b FlashEnergy unsigned rational 1
0xa20c SpatialFrequencyResponse unsigned short 1
0xa214 SubjectLocation unsigned short 1


Appendix 1: MakerNote of Olympus Digicams

The data below is analyzed at Olympus D450Z(C-920Z) by Peter Esherick.

MakerNote of Olympus Digicam starts from ASCII string "OLYMP". Data format is the same as IFD, but it starts from offset 0x07. Example of actual data structure is shown below. 
:0000: 4F 4C 59 4D 50 00 01 00-0B 00 00 02 04 00 03 00  OLYMP...........
:0010: 00 00 0E 04 00 00 01 02-03 00 01 00 00 00 03 00  ................

Tag No. Tag Name Format CompoNo Value
0x0200 SpecialMode Unsigned Long 3 Shows picture taking mode. First value means 0=normal, 1=unknown, 2=fast, 3=panorama. Second value means sequence number, third value means panorama direction, 1=left to right, 2=right to left, 3=bottom to top, 4=top to bottom.
0x0201 JpegQual Unsigned Short 1 Shows JPEG quality. 1=SQ,2=HQ,3=SHQ.
0x0202 Macro Unsigned Short 1 Shows Macro mode or not. 0=normal, 1=macro.
0x0203 Unknown Unsigned Short 1 Unknown
0x0204 DigiZoom Unsigned Rational 1 Shows Digital Zoom ratio. 0=normal, 2=digital 2x zoom.
0x0205 Unknown Unsigned Rational 1 Unknown
0x0206 Unknown Signed Short 6 Unknown
0x0207 SoftwareRelease Ascii string 5 Shows Firmware version.
0x0208 PictInfo Ascii string 52 Contains ASCII format data such as [PictureInfo]. This is the same data format of older Olympus digicam that not used Exif data format (C1400/C820/D620/D340 etc).
0x0209 CameraID Undefined 32 Contains CameraID data, which is user changeable by some utilities
0x0f00 DataDump Unsigned Long 30 Unknown

Appendix 2: MakerNote of Nikon

There are 2 formats of Nikon's MakerNote. MakerNote of E700/E800/E900/E900S/E910/E950 starts from ASCII string "Nikon". Data format is the same as IFD, but it starts from offset 0x08. This is the same as Olympus except start string. Example of actual data structure is shown below. 
:0000: 4E 69 6B 6F 6E 00 01 00-05 00 02 00 02 00 06 00 Nikon...........
:0010: 00 00 EC 02 00 00 03 00-03 00 01 00 00 00 06 00 ................

Tag No. Tag Name Format CompoNo Value
0x0002 Unknown Ascii string 6 Unknown. E900/E900S/E910:"09.41", others:"08.00".
0x0003 Quality Unsigned short 1 Shows quality setting. At E900, 1:VGA Basic, 2:VGA Normal, 3:VGA Fine, 4:SXGA Basic, 5:SXGA Normal, 6:SXGA Fine
0x0004 Color Mode Unsigned short 1 1:Color, 2:Monochrome.
0x0005 Image Adjustment Unsigned short 1 0:Normal, 1:Bright+, 2:Bright-, 3:Contrast+, 4:Contrast-.
0x0006 CCD Sensitivity Unsigned short 1 0:ISO80, 2:ISO160, 4:ISO320, 5:ISO100
0x0007 White balance Unsigned short 1 0:Auto, 1:Preset, 2:Daylight, 3:Incandescense, 4:Fluorescence, 5:Cloudy, 6:SpeedLight
0x0008 Focus Unsigned rational 1 If infinite focus, value is '1/0'.
0x0009 Unknown Ascii string 20 Unknown
0x000a Digital Zoom Unsigned rational 1 '160/100' means 1.6x digital zoom, '0/100' means no digital zoom (optical zoom only).
0x000b Converter Unsigned short 1 If Fisheye Converter is used, value is '1'.
0x0f00 Unknown Unsigned long 25~30 Unknown


At E990, there is no Ascii string. As like the usual IFD (e.g. IFD0, SubIFD), IFD starts from the first byte of data. Nikon D1 also uses this format. Example of actual data structure is shown below. 
:0000: 10 00 01 00 07 00 04 00-00 00 00 01 00 00 02 00  ................
:0010: 03 00 02 00 00 00 00 00-64 00 03 00 02 00 06 00  ........d.......

The data below is analyzed by Max Lyons .

Tag No. Tag Name Format CompoNo Value
0x0001 Unknown Undefined 4 Unknown. Always "0100". version?
0x0002 ISO Setting Unsigned short 2 0,100=ISO 100
0,200=ISO200
0,400=ISO400
etc.
0x0003 Color Mode Ascii string varies "COLOR", "B&W"
0x0004 Quality Ascii string varies "NORMAL", "FINE", "BASIC"
0x0005 Whitebalance Ascii string varies "AUTO", "WHITE PRESET" etc.
0x0006 Image Sharpening Ascii string varies "AUTO", "HIGH" etc.
0x0007 Focus Mode Ascii string varies "AF-S" means Single AF, "AF-C" means Continuous AF.
0x0008 Flash Setting Ascii string varies "NORMAL", "RED-EYE" etc.
0x000a Unknown Unsigned rational 1 Unknown, Always '8832/1000'?
0x000f ISO Selection Ascii string varies "MANUAL":User selected, "AUTO":Automatically selected.
0x0080 Image Adjustment Ascii string varies "AUTO", "NORMAL", "CONTRAST(+)" etc.
0x0082 Adapter Ascii string varies "OFF", "FISHEYE 2", "WIDE ADAPTER" etc.
0x0085 Manual Focus Distance Unsigned rational 1 Distance in Meters if focus was manually selected, otherwise 0
0x0086 Digital Zoom Unsigned rational 1 '100/100' means no digital zoom (optical zoom only), '140/100' means 1.4x digital zoom.
0x0088 AF Focus Position Undefined 4 '0,0,0,0':Center, '0,1,0,0':Top, '0,2,0,0':Bottom, '0,3,0,0':Left, '0,4,0,0':right
0x0010 Data Dump Undefined 174 Unknown.

Appendix 3: MakerNote of Casio


Casio started to use Exif format from QV2000/QV8000. Casio's MakerNote format is the same as usual IFD (e.g. IFD0, SubIFD0). Example of actual data structure is shown below. 
:0000: 00 14 00 01 00 03 00 00-00 01 00 0A 00 00 00 02 ................
:0010: 00 03 00 00 00 01 00 03-00 00 00 03 00 03 00 00 ................
The data below is analyzed by Eckhard Henkel .

Tag No. Tag Name Format CompoNo Value
0x0001 RecordingMode Unsigned Short 1 1:Single Shutter, 2:Panorama, 3:Night Scene, 4:Portrait, 5:Landscape
0x0002 Quality Unsigned Short 1 1:Economy, 2:Normal, 3:Fine
0x0003 Focusing Mode Unsigned Short 1 2:Macro, 3:Auto Focus, 4:Manual Focus, 5:Infinity
0x0004 Flash Mode Unsigned Short 1 1:Auto, 2:On, 3:Off, 4:Red Eye Reduction
0x0005 Flash Intensity Unsigned Short 1 11:Weak, 13:Normal, 15:Strong
0x0006 Object distance Unsigned Long 1 Object distance in [mm]
0x0007 White Balance Unsigned Short 1 1:Auto, 2:Tungsten, 3:Daylight, 4:Fluorescent, 5:Shade, 129:Manual
0x0008 Unknown Unsigned short 1 Unknown
0x0009 Unknown Unsigned short 1 Unknown
0x000a Digital Zoom Unsigned Long 1 0x10000(65536):'Off', 0x10001(65537):'2X Digital Zoom'
0x000b Sharpness Unsigned Short 1 0:Normal, 1:Soft, 2:Hard
0x000c Contrast Unsigned Short 1 0:Normal, 1:Low, 2:High
0x000d Saturation Unsigned Short 1 0:Normal, 1:Low, 2:High
0x000e Unknown Unsigned short 1 Unknown
0x000f Unknown Unsigned short 1 Unknown
0x0010 Unknown Unsigned short 1 Unknown
0x0011 Unknown Unsigned long 1 Unknown
0x0012 Unknown Unsigned short 1 Unknown
0x0013 Unknown Unsigned short 1 Unknown
0x0014 CCD Sensitivity Unsigned short 1 QV3000:   64:Normal, 125:+1.0, 250:+2.0, 244:+3.0
QV8000/2000:   80:Normal, 100:High

Appendix 4: MakerNote of Fujifilm


Fujifilm's digicam added the MakerNote tag from the Year2000's model (e.g.Finepix1400,Finepix4700). It uses IFD format and start from ASCII character 'FUJIFILM', and next 4 bytes(value 0x000c) points the offset to first IFD entry. Example of actual data structure is shown below.
:0000: 46 55 4A 49 46 49 4C 4D-0C 00 00 00 0F 00 00 00 :0000: FUJIFILM........
:0010: 07 00 04 00 00 00 30 31-33 30 00 10 02 00 08 00 :0010: ......0130......

There are two big differences to the other manufacturers.
I think it's a BUG, but it can't be helped now... The data below is analyzed at Fujifilm FinePix4900Z.


Tag No. Tag Name Format CompoNo Value
0x0000 Version Undefined 4 Version of MakerNote information. At present, value is "0130".
0x1000 Quality Ascii string 8 Quality setting. Ascii string "BASIC","NORMAL","FINE"
0x1001 Sharpness Unsigned Short 1 Sharpness setting. 1or2:soft, 3:normal, 4or5:hard.
0x1002 White Balance Unsigned Short 1 White balance setting. 0:Auto, 256:Daylight, 512:Cloudy, 768:DaylightColor-fluorescence, 769:DaywhiteColor-fluorescence, 770:White-fluorescence, 1024:Incandenscense, 3840:Custom white balance.
0x1003 Color Unsigned Short 1 Chroma saturation setting. 0:normal(STD), 256:High, 512:Low(ORG).
0x1004 Tone Unsigned Short 1 Contrast setting. 0:normal(STD), 256:High(HARD), 512:Low(ORG).
0x1010 Flash Mode Unsigned Short 1 Flash firing mode setting. 0:Auto, 1:On, 2:Off, 3:Red-eye reduction.
0x1011 Flash Strength Signed Rational 1 Flash firing strength compensation setting. Unit is APEX(EV) and value is -6/10, -3/10, 0/10, 3/10, 6/10 etc.
0x1020 Macro Unsigned Short 1 Macro mode setting. 0:Off, 1:On.
0x1021 Focus mode Unsigned short 1 Focusing mode setting. 0:Auto focus, 1:Manual focus.
0x1030 SlowSync. Unsigned short 1 Slow synchro mode setting. 0:Off, 1:On.
0x1031 Picture Mode Unsigned short 1 Picture mode setting. 0:Auto, 1:Portrait scene, 2:Landscape scene, 4:Sports scene, 5:Night scene, 6:Program AE, 256:Aperture prior AE, 512:Shutter prior AE, 768:Manual exposure.
0x1032 unknown Unsigned Short 1 Unknown
0x1100 ContTake/Bracket Unsigned Short 1 Continuous taking or auto bracketting mode setting. 0:off, 1:on.
0x1200 unknown Unsigned Short 1 Unknown
0x1300 Blur warning Unsigned Short 1 Blur warning status. 0:No blur warning, 1:Blur warning.
0x1301 Focus warning Unsigned short 1 Auto Focus warning status. 0:Auto Focus good, 1:Out of focus.
0x1302 AE warning Unsigned short 1 Auto Exposure warning status. 0:AE good, 1:Over exposure (>1/1000s,F11).

Appendix 5: MakerNote of Canon

The data below was primarily analysed by David Burren and the master version of this information is available at: http://www.burren.cx/david/canon.html. Please send any updates to David.
This document is based on his Rev.1.11(2001/01/30) of document.

Canon's MakerNote data is in IFD format, starting at offset 0.

Some of these tags and fields are only produced on cameras such as the EOS D30, but (to current observation) all this is valid for all Canon digicams (at least since the A50). If the tag is not found, or is shorter than shown here, it simply means that data is not supported by that camera.

Tag No. Tag Name Format CompoNo Value
0x0 Unknown Unsigned Short 6 Always 0
0x1 Unknown Unsigned Short varies
Offset within tag Meaning
0 Length of tag in bytes (i.e. twice the number of shorts)
1 Macro mode 1: macro
2: normal
2 If non-zero, length of self-timer in 10ths of a second
3 unknown
4 Flash mode 0: flash not fired
1: auto
2: on
3: red-eye reduction
4: slow synchro
5: auto + red-eye reduction
6: on + red-eye reduction
16: external flash (not set on D30)
5 Continuous drive mode 0: single or timer (see field 2)
1: continuous
6 unknown
7 Focus Mode 0: One-Shot
1: AI Servo
2: AI Focus
3: MF
4: Single (but check field 32)
5: Continuous
6: MF
8, 9 unknown
10 Image size 0: large
1: medium
2: small
11 "Easy shooting" mode 0: Full Auto
1: Manual
2: Landscape
3: Fast Shutter
4: Slow Shutter
5: Night
6: B&W
7: Sepia
8: Portrait
9: Sports
10: Macro / Close-Up
11: Pan Focus
12 unknown
13 Contrast 0xffff: low
0x0000: normal
0x0001: high
14 Saturation 0xffff: low
0x0000: normal
0x0001: high
15 Sharpness 0xffff: low
0x0000: normal
0x0001: high
16 ISO If zero, use ISOSpeedRatings EXIF tag instead
15: auto
16: 50
17: 100
18: 200
19: 400
17 Metering mode 3: Evaluative
4: Partial
5: Center-weighted
18 unknown
19 AF point selected 0x3000: none (MF)
0x3001: auto-selected
0x3002: right
0x3003: center
0x3004: left
20 Exposure mode 0: "Easy shooting" (use field 11)
1: Program
2: Tv-priority
3: Av-priority
4: Manual
5: A-DEP
21, 22 unknown
23 "long" focal length of lens (in "focal units")
24 "short" focal length of lens (in "focal units")
25 "focal units" per mm
26 - 28 unknown
29 Flash details Bits 15..0:
14: external E-TTL
13: internal flash
11: FP sync used
4: FP sync enabled
other bits unknown
30 - 31 unknown
32 Focus mode G1 seems to use this in preference to field 7
0: Single
1: Continuous
0x3 Unknown Unsigned Short 4 unknown
0x4 Unknown Unsigned Short varies
Offset within tag Meaning
0 Length of tag in bytes (i.e. twice the number of shorts)
1 - 6 unknown
7 White balance 0: auto
1: Sunny
2: Cloudy
3: Tungsten
4: Flourescent
5: Flash
6: Custom
8 unknown
9 Sequence number (if in a continuous burst)
10 - 13 unknown
14 AF point used Only set in One-Shot mode?
If none used, AF failed or manual focus was used (e.g. on a lens with full-time manual focus)
Bits 15..0:
15-12: number of available focus points
2: left
1: center
0: right
15 Flash bias 0xffc0: -2 EV
0xffcc: -1.67 EV
0xffd0: -1.50 EV
0xffd4: -1.33 EV
0xffe0: -1 EV
0xffec: -0.67 EV
0xfff0: -0.50 EV
0xfff4: -0.33 EV
0x0000: 0 EV
0x000c: 0.33 EV
0x0010: 0.50 EV
0x0014: 0.67 EV
0x0020: 1 EV
0x002c: 1.33 EV
0x0030: 1.50 EV
0x0034: 1.67 EV
0x0040: 2 EV
16 - 18 unknown
19 Subject Distance Units are either 0.01m or 0.001m (both have been observed). Still investigating.
In any case, the SubjectDistance EXIF tag is set by Canon cameras.
0x6 Image type Ascii string 32 e.g.: "IMG:EOS D30 JPEG"
Has trailing whitespace.
0x7 Firmware version Ascii string 24 May have trailing NULs and whitespace.
0x8 Image Number Unsigned Long 1 Normally reported as FFF-XXXX.
FFF is this value divided by 1000, XXXX is this value mod 1000.
0x9 Owner name Ascii string 32 May have trailing NULs and whitespace.
0xa Unknown Unsigned Short varies unknown
0xc Camera serial number Unsigned Long 1 High 16 bits are printed as a 4-digit hex number.
Low 16 bits are printed as a 5-digit decimal number.
These are concatenated to form the serial number. Example printf() format string would be "%04X%05d".
0xd Unknown Unsigned Short varies unknown
0xf Custom Functions Unsigned Short varies First short is the number of bytes in the tag (i.e. twice the number of shorts)
For each other value: the top 8 bits are the C.Fn number, and the lower 8 bits are the value.

EOS D30 Custom Functions

C.Fn Name Value
1 Long exposure noise reduction 0: Off
1: On
2 Shutter/AE-lock buttons 0: AF/AE lock
1: AE lock/AF
2: AF/AF lock
3: AE+release/AE+AF
3 Mirror lockup 0: Disable
1: Enable
4 Tv/Av and exposure level 0: 1/2 stop
1: 1/3 stop
5 AF-assist light 0: On (auto)
1: Off
6 Shutter speed in Av mode 0: Automatic
1: 1/200 (fixed)
7 AEB sequence/auto cancellation 0: 0, -, + / Enabled
1: 0, -, + / Disabled
2: -, 0, + / Enabled
3: -, 0, + / Disabled
8 Shutter curtain sync 0: 1st-curtain sync
1: 2nd-curtain sync
9 Lens AF stop button Fn. Switch 0: AF stop
1: Operate AF
2: Lock AE and start timer
10 Auto reduction of fill flash 0: Enable
1: Disable
11 Menu button return position 0: Top
1: Previous (volatile)
2: Previous
12 SET button func. when shooting 0: Not assigned
1: Change quality
2: Change ISO speed
3: Select parameters
13 Sensor cleaning 0: Disable
1: Enable


History

rev. 1.4

rev. 1.3

rev. 1.2

rev. 1.1

rev. 1.0

Acknowledgement

I would like to thank to;

Daniel Switkin: Byte align of TAG Mark, Format of ImageWidth/ImageLength
Peter Esherick: MakerNote of Olympus
Matthias Wandel: Value of FocalPlaneResolutionUnit
Max Lyons: Makernote of Nikon  ...His webpage
Eckhard Henkel: Makernote of Casio  ...His webpage
David Burren: FocalPlaneResolutionUnit of Fujifilm's / Makernote of Canon  ...His webpage