Description of Exif file format

Currently, most of new digicams 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 which can get from internet.

I believe this document is basically based on Exif2.1 specification, but actually I have not read official specification yet, there are some 'unknown' items and it may have some of errata.

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.

rev. 1.1 Dec.19,1999
rev. 1.0 May.28,1999

• What is Exif format?
• JPEG format and Marker
• Marke used by Exif
• Exif data structure
  • TIFF Header
  • IFD : Image file directory
  • Data format
  • IFD data structure
  • href="Exif file format.html#ExifThumbs" Thumbnail image
    • JPEG format thumnail
    • TIFF format thumbnail
• Tag number used by Exif/TIFF
  • IFD0 (IFD of main image) section
  • Exif SubIFD section
  • IFD1 (IFD of thumbnail image) section
  • Misc Tags
• Appendix 1: MakerNote of Olympus digicams
• History
• Acknowledgement

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.

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.

Marker used by Exif

The marker 0xFFE0~0xFFEF i named "Application Marker", not necessary for decoding JPEG image. They are used by user application. For example, older olympus/canon/casio/agfa digicams 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;

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 APP 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 align and it contains JPEG format thumbnail. As described above, Exif data is starts from ASCII character "Exif" and 2bytes of 0x00, then Exif data follows. Exif uses TIFF format to store data.

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 Motrola 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 digicams uses Intel align. Ricoh uses Motorola align. Sony uses Intel align except D700. Kodak DC200/210/240 use Motorola align, but DC220/260 use Intel align 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 alignment. 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 "0x2a00". If it uses Motorola, they are "0x002a". 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'.

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.

'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 value, and the first represents the numerator, the second, the denominator.

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 i 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.

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;

• The first 2bytes are "I I", byte align is 'Intel'.
  
• Address 0x0004~0x0007 is 0x08000000, IFD0 starts from address '0x0008'
  
• Address 0x0008~0x0009 is 0x0200, number of directory entry of IFD0 '2'.
  
• Address 0x000a~0x000b is 0x1A01, it means this is a XResolution(0x Tag, it contains a horizontal resolution of image.
  
• Address 0x000c~0x000d is 0x0500, format of this value is unsigned rational(0x0005).
  
• Address 0x000e~0x0011 is 0x01000000, number of components is '1'. Unsigned rational's data size is 8bytes/components, so total data length is 1x8=8bytes.
  
• Total data length is larger than 4bytes, so next 4bytes contains an offset to data.
  
• Address 0x0012~0x0015 is 0x26000000, XResolution data is stored to address 0x0026
  
• Address 0x0026~0x0029 is 0x48000000, numerator is 72, address 0x002a~0x002d is 0x0100000000, denominator is '1'. So the value of XResoultion is 72/1.
  
• Address0x0016~0x0017 is 0x6987, next Tag is ExifOffset(0x8769). Its value is an offset to Exif SubIFD
  
• Data format is 0x0004, unsigned long integer.
  
• This Tag has one component. Unsigned long integer's data size is 4bytes/components, so total data size is 4bytes.
• Total data size is equal to 4bytes, next 4bytes contains the value of Exif SubIFD offset.
  
• Address 0x001e~0x0021 is 0x11020000, Exif SubIFD starts from addre '0x0211'.
  
• This is the last directory entry, next 4bytes shows an offset to n IFD.
  
• Address 0x0022~0x0025 is 0x40000000, next IFD starts from address '0x0040'

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.

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 by JpegIFOffset(0x0201) Tag in IFD1, size of thumbnail by JpegIFByteCount(0x0202) Tag. Data format is ordinary JPEG format, starts from 0xFFD8 and ends by 0xFFD9. It seems that JPEG format and 160x120pixels of size are recommended thumbnail format for Exif2.1 or later.

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 an 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 use this format.

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.

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

The numerics in parenthesis are 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.

: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  ................
        

History

rev. 1.1

• Added byte align explanation to TAG Mark of TIFF header
• Corrected data format of some TAGs to "unsigned short/long"
• Corrected value of FocalPlaneResolutionUnit
• Added Appendix 1: MakerNote of Olympus

rev. 1.0

• First release