Example of Manufacturer-Specific Exif Data | Manufacturer-Specific Exif Data in GraphicConverter | Putting Things Together... | Links to ExifTool
Archive
Considering my issues with the Ricoh GXR's AF behavior, I believed that it would be of some help to know which focus mode was set and which distance the camera or the user selected. I therefore inspected the manufacturer-specific fields in the Exif data with ExifTool. At the moment, I can give away only little information, though.
To inspect Exif data, you need a tool that displays them. Many photo applications, for example, do so, but most applications and tools display only a subset of the Exif data. Moreover, camera manufacturers hide manufacturer-specific data in Exif fields that are not documented publicly. Therefore, it is difficult to decipher this data, if if may be of primary importance for judging what happened to certain photos.
The probably largest set of Exif data is displayed by the ExifTool application provided by Phil Harvey. This is a command-line tool that can be installed on Unix, Windows, and Apple Macintosh computers. For Windows, there is also a GUI shell available. For copyright reasons, I do not duplicate the information that Phil Harvey provides on his site and provide just a number of links to specific content and topics.
Phil Harvey also provides a page on which he lists Ricoh-specific Exif tags. This page is quite useful for understand the output of the Exiftool application for images taken with Ricoh cameras like the Ricoh GXR. However, I found that Phil Harvey still does not include the focus-specific tags that Pavel Kudrys and I have revealed. Perhaps, I should send him an e-mail...
This is an example of how ExifTool 9.46 lists the manufacturer-specific Exif data for the Ricoh GXR:
---- Ricoh ----
Maker Note Type : Rdc
Firmware Version : 1.51
Ricoh 0x0003 : 1006
Serial Number : (20140000)04100058
Ricoh 0x0006 :
Ricoh Image Width : 4928
Ricoh Image Height : 3264
Ricoh Date : 2013:02:17 13:45:07
Preview Image Start : 12
Preview Image Length : 0
Flash Mode : Off
Macro : Off
Sharpness : Normal
White Balance : Multi-pattern Auto
ISO Setting : 400
Saturation : Normal
Ricoh 0x1002 : 1244
Sharpness : Normal
Ricoh Subdir 0x0001 : 165 216
Ricoh Subdir 0x0002 : 50331860
Manufacture Date 1 : 2009:11:22
Manufacture Date 2 : 2002:05:11
Ricoh Subdir 0x0006 : 263 32768
Ricoh Subdir 0x0007 : 0 0
Ricoh Subdir 0x0008 : 21954608
Ricoh Subdir 0x0009 : 1792
Ricoh Subdir 0x000a : 0.1152238806
Ricoh Subdir 0x000b : 0.1152238806
Ricoh Subdir 0x000c : 6614 51
Ricoh Subdir 0x000d : 1024 65462
Ricoh Subdir 0x000e : 0
Ricoh Subdir 0x000f : 138 93
Ricoh Subdir 0x0010 : @@. .[...]
Ricoh Subdir 0x0011 : 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0[...]
Ricoh Subdir 0x0012 : 65 0 0 0
Ricoh Subdir 0x0013 : 0 0 0 0
Ricoh Subdir 0x0014 : 0 0 2 234 2 76 0 233 0 160 0 220 0 179 0 222 0 171 90
0[...]
Ricoh Subdir 0x0015 : 0 32
Ricoh Subdir 0x0016 : 57 2 6 10 0 0 254 0 0 0 0 0 0 4 0 0
Ricoh Subdir 0x0017 : 144 0 15 15
Ricoh Subdir 0x0018 : 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0[...]
Ricoh Subdir 0x0019 : 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0[...]
Faces Detected : 0
Face Detect Frame Size : 0 0
Ricoh Subdir 0x001c : 255 209 3 244 0 129 0 0 0 0 0 0 0 5 125 186 255 255 252[...]
Ricoh Subdir 0x001d :
Ricoh Subdir 0x001e :
Ricoh Subdir 0x001f : 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0[...]
Ricoh Subdir 0x0020 :
Ricoh Subdir 0x0021 : 56 80 93 99 129 158 122 164 184 191 180 145 112 117 121[...]
Ricoh Subdir 0x0023 : 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0[...]
Ricoh Subdir 0x0025 : 16791297
Ricoh Subdir 0x0026 : 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0[...]
Ricoh Subdir 0x0027 : 0
Ricoh Subdir 0x0028 : 0
Firmware Revision : Rev01510000
Firmware Revision 2 : Rev01510000
Noise Reduction : Off
Ricoh Subdir 0x002b : 0 0 0 0 0 0 0 0 1 0 0 64 0 54 13 36 0 0 0 0 0 0 0 0 0
0[...]
Body Firmware : RS1 :V01510000
Body Serial Number : SID:14100923
Lens Firmware : RL6 :V01510000
Lens Serial Number : LID:04100058
Ricoh Subdir 0x002d : 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Ricoh Subdir 0x002e : 0 0 63 167 0 0 64 89 0 0 64 90 0 0 64 102 0 0 64 90 0
0[...]
Ricoh Subdir 0x002f : 381146496
Ricoh Subdir 0x0030 : 3407924
Ricoh Subdir 0x0033 : 0
Ricoh Subdir 0x0034 : 0
Ricoh Subdir 0x0035 : 0
Ricoh Subdir 0x0036 : 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0[...]
GraphicConverter lists some of the manufacturer-specific fields, but does not mention their HEX numbers (labels are in German):
---- Ricoh ----
Maker Note Type: Rdc
Firmware-Version: 1.51
Seriennummer: (20140000)04100058
Ricoh-Bildbreite: 4928
Ricoh-Bildhöhe: 3264
Ricoh Datum: 2013:02:17 13:45:07
Vorschaubild-Datenposition: 12
Vorschaubild-Datenlänge: 0
Blitz-Modus: Aus
Makro: Aus
Weißabgleich: Multi-pattern Auto
ISO-Einstellung: 400
Farbsättigung: Standard
Schärfe: Standard
Manufacture Date 1: 2009:11:22
Manufacture Date 2: 2002:05:11
Gesichter erkannt: 0
Gesichtserkennung Bereichsgröße: 0 0
Firmware-Revision: Rev01510000
Firmware-Revision 2: Rev01510000
Rauschreduktion: Aus
Body Firmware: RS1 :V01510000
Body Serial Number: SID:14100923
Lens Firmware: RL6 :V01510000
Objektiv-Seriennummer: LID:04100058
Elsewhere on my pages, I describe the issues that I had with my S10 module. As I was often not able to memorize at which distances I took my test shots, I would have been happy to find the selected focus distance, either set manually or by the autofocus, in the Exif data. Some companies provide such information, but regrettably, Ricoh did not seem to provide it. In July 2010, I had an e-mail exchange with Pavel Kudrys from Ricohforum.com regarding distance information in the Exif data. Pavel told me that this information is indeed provided, however in some manufacturer-specific fields that are typically not displayed by Exif viewers. Pavel wrote:
Good luck meant, that Pavel had no idea of how to decipher the information provided in the 0x0008 field. I took a lot of test shots with different camera units and different focus modes at different distances and collected the values of the tag8 field in an Excel sheet. The values decreased from 1m to infinity and were between 48 and 50 at infinity. Being a physicist, I have some training with numbers, and the natural thing was to normalize the numbers, that is, to divide all values by he value for 1m. That was already the solution: Simply divide the tag8 value by 67108914, the value for 1m, and you get the distance in meters.
Here is Pavel's explanation from a recent thread at dpreview (with a correction by me...):
I would like to add that Multi AF seems to provide a value of 50 for infinity, whereas manual focus seems to provide a value of 49.
Pavel also deciphered the tag for focus mode (Ricoh Image Info 0x0023). Here are the values for the different modes that this tag displays:
However, first I was not able to find this tag in the Exiftool output, because it simply was not listed there, and I wrote an e-mail to Pavel asking for help. But after tinkering around with Exiftool's verbose output mode, I was able to spot the tag without Pavel's help. You have to set the verbose mode to 3 at least (-v3). Here is a sample command line and section of the output indicating "spot AF":
Command: exiftool -a -u -g1 -v3 /usr/R0065824.JPG
| | | 5) ImageInfo (SubDirectory) -->
| | | - Tag 0x1001 (224 bytes, undef[224]):
| | | 04e8: 13 40 0c c0 01 18 20 13 06 10 16 57 51 00 01 00 [.@.... ....WQ...]
| | | 04f8: fa 02 30 00 64 74 fc 00 00 00 02 00 00 00 00 00 [..0.dt..........]
| | | 0508: 00 00 01 06 02 0d 09 08 01 00 00 00
00 00 00 00 [................]
| | | 0518: 00 03 04 03 09 09 00 00 02 00 00 18 00 00 00 00 [................]
| | | 0528: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [................]
| | | [snip 144 bytes]
Plase note that 0x0023 is a hexadecimal number, meaning 35 in decimal (there are 16 bytes in each row).
I was less successful with deciphering field Ricoh Subdir 0x0009, which seems to be related to Multi AF. This field contains numbers that look like powers of 2 or sums thereof (512, 1536, 2048, 2056, 3072, 3585, and many more). Being trained as a physicist, I thought that the 9 AF fields were assigned to certain powers of two and that the sum would indicate, which AF indicators "fired." That would indeed be useful information. However, when I tested my hypothesis, I could not find any consistent pattern in the data. While in some cases, it seemed to work, in other cases, different AF indicator patterns led to the same number and vice versa. Perhaps, someone else will have the right idea here...
For more links around Exif data see page Using ExifTool on the Apple Macintosh, ExifTool Links.
19.11.2020 |