0
\$\begingroup\$

I had been using Panasonic's MFT Lumix cameras for a while: DMC-GH1, DMC-GH3, now a DC-G9

The camera have "multi-format feature": You can shot in 4:3 (native), 3:2, 16:9, and 1:1 aspect ratios.

Sometimes I use raw (.rw2) images. The cameras before the G9 produced raw images in the specified aspect ratios (pixel sizes), or at least I never noticed otherwise.

However the raw files written by the G9 seem to contain always the full sensor data, even though a different aspect ratio was set for shooting. In Lightroom I see the default crop as intended, but I realized that I can "extend the image" outside of the default crop.

I wonder, what might be the reason the sensor data is not cropped when saving as RW2?

Specifically: As you don't actually see what you shot (the viewfinder and display only show the selected aspect ratio), uncropping the raw files can be surprising sometimes.

In the past I had used different aspect ratios also to reduce the size of the raw files, but that does not work any more.

Example

exiftool shows (among many other) these data for a 16:9 shot of the G9:

...
Panasonic Raw Version           : 0390
Sensor Width                    : 5264
Sensor Height                   : 3904
Sensor Top Border               : 8
Sensor Left Border              : 12
Sensor Bottom Border            : 3896
Sensor Right Border             : 5196
...
Panasonic Image Width           : 5184
Panasonic Image Height          : 3888
...
Crop Top                        : 492
Crop Left                       : 12
Crop Bottom                     : 3412
Crop Right                      : 5196
...
Image Height                    : 3888
Image Width                     : 5184
Image Size                      : 5184x3888

In contrast, here is an 1:1 shot from the GH-3:

...
Panasonic Raw Version           : 0320
Sensor Width                    : 3472
Sensor Height                   : 3472
Sensor Top Border               : 8
Sensor Left Border              : 8
Sensor Bottom Border            : 3464
Sensor Right Border             : 3464
...
Panasonic Image Width           : 3456
Panasonic Image Height          : 3456
...
Image Height                    : 3456
Image Width                     : 3456
Image Size                      : 3456x3456

\$\endgroup\$
2
  • \$\begingroup\$ If you want to compensate for distortion or color aberration you may need pixels that are beyond the final cropped area... \$\endgroup\$
    – xenoid
    Apr 28 at 22:16
  • \$\begingroup\$ "The cameras before the G9 produced raw images in the specified aspect ratios (pixel sizes), or at least I never noticed otherwise." \$\endgroup\$
    – Michael C
    Apr 29 at 7:13

2 Answers 2

2
\$\begingroup\$

That's kind of the definition of "raw", it's ideally a straight dump of what comes from the sensor. The sensor didn't get cropped just because you decided your picture should be. If you later decide your crop choice was wrong it can be a life saver.

\$\endgroup\$
2
  • \$\begingroup\$ You said "what comes out from the sensor". I might say "what is read from the sensor". There is a small, but significant difference: Why read more data from the sensor than needed? Also, despite of the different aspect ratios, there are also three sets of resolutions (L/M/H). If the lower resolutions just read every n-th pixel (CFA element), you could even have corresponding RAW files. For me RAW meant leaving the color interpretation to the post-processing. As it seems it also means leaving the aspect ratio and the resolution to the post-processing. \$\endgroup\$
    – U. Windl
    Apr 29 at 20:13
  • \$\begingroup\$ @U.Windl generating different resolutions is usually more complicated than just taking every nth pixel, and probably still requires a full dump. Much of the value of a RAW file comes from being able to change your mind about choices made when taking the picture, and that's only possible if all the information is retained. \$\endgroup\$ May 3 at 21:01
1
\$\begingroup\$

As far as I know all cameras do this, including your previous Panasonic Lumix cameras. The camera produced JPEG preview that is attached to the raw file will be in the instructed aspect ratio, but the actual raw image data will contain all of the information collected by the entire sensor.

The following is an excerpt from my answer to Why can software correct white balance more accurately for RAW files than it can with JPEGs? which you may find helpful:

Anytime you open a raw file and look at it on your screen, you are not viewing "THE raw file." You are viewing one among a near-countless number of possible interpretations of the data in the raw file. The raw data itself contains a single (monochrome) brightness value measure by each pixel well. With Bayer masked camera sensors (the vast majority of color digital cameras use Bayer filters) each pixel well has a color filter in front of it that is either 'red', 'green', or 'blue' (the actual 'colors' of the filters in most Bayer Masks are anywhere from a slightly yellowish-green to an orange-yellow for 'red", a slightly yellow green for 'green' and a bluish-violet for 'blue' - these colors more or less correspond to the center of sensitivity for the three types of cones in our retinas). For a more complete discussion of how we get color information out of the single brightness values measured at each pixel well, please see RAW files store 3 colors per pixel, or only one?

Here's a similar example where the shooter chose B&W but when they opened the raw files on a computer they were being rendered in color. When the raw data is saved with the camera set to 'Monochrome' or 'B&W' the preview image will be a B&W image, but the raw data will still contain all of the information collected by the sensor which can be used later to render the images in color. When the raw data is opened in different viewers or conversion applications, some will use the JPEG preview and display the B&W image generated in the camera, some will read the 'B&W' instruction in the metadata to create their own B&W rendering, and others will totally ignore the B&W instruction and apply their own default processing instructions that will display a full color image.

When you view "raw" images, you're looking at a very processed interpretation of the information collected by the camera and stored in the raw image data.

  • Some raw image viewers or raw processing apps will display the attached JPEG preview, in which case you will see the selected aspect ratio when you took the photo.
  • Other raw image viewers and conversion apps will do their own conversion of the raw data to produce one among a near infinite possible interpretations of the data contained in the raw file. How close this interpretation will be to the preview JPEG produced by the camera (or a JPEG produced by the camera if you saved Raw + JPEG) depends on how close the processing instructions used by the rendering viewer or app are to the processing instructions used by the camera to produce the preview image.

The instruction to use a certain aspect ratio is stored in the raw image file's EXIF Info that is part of the metadata included along with the actual data collected by the image sensor.

  • If the viewer or conversion app reads and applies this instruction then they will also display their interpretation of the image in the same aspect ratio.
  • If the viewer or app ignores this instruction, then they'll probably display a conversion of the full 4:3 image.

Sometimes the aspect ratio instruction may only be stored in a section of the EXIF Info known as the "Maker Notes" section. This section allows camera makers to store whatever they want in whatever form they want without having to comply with the standards that govern how the standardized fields in EXIF Info are defined.

For example, there's only one correct way to record "f/5.6" in the standard 'Aperture' field of the EXIF Info. Every manufacturer uses the same hex code to mean "f/5.6". The same is true for all other f-number values, all exposure time values (shutter speeds), ISO settings, etc. There is a standard way to record those values that is set out in the EXIF standard.

Information in the 'Maker Notes' section is not standardized. A hex code used by one manufacturer for a field they created to record something like "Lens Maximum Focal Length" may mean "500mm". The same hex code used by another manufacturer in a different field they created to describe a different thing, such as "Flash Exposure Compensation" may mean "-3EV".

Many raw image viewers and raw conversion apps do not have the needed information to properly interpret the 'Maker Note' sections of the EXIF Info from many cameras, so they ignore it. Adobe, for instance, has adamantly resisted using any 'Maker Note' information that they don't absolutely have to. When you export an image from any Adobe application, the 'Maker Note' section is stripped from the EXIF Info and not included in the exported image.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.