Image Size / File Quality Options
The E-1 provides five options for image size / quality; TIFF (8-bit), RAW (Olympus .ORF), SHQ (JPEG 1/2.7), HQ (JPEG 1/8) and SQ (customizable). The SQ option can be reprogrammed to any combination of four image sizes and two JPEG compression levels. The E-1's RAW option produces 10 MB .ORF files which contain a RAW 'dump' of the data directly from the sensor (12 bits per pixel). Most people consider the RAW format to be the 'digital negative' because it is lossless and has had no image processing applied to it (tone, white balance, sharpening etc.) all of which can be adjusted at a later date before outputting the final image (using the included Olympus Viewer or optional Olympus Studio).
One oddity about the RAW files was their size (just over 10 MB), assuming 2614 x 1966 active pixels (from the Kodak spec sheet) at 12 bits per pixel that should produce a 7.7 MB file (excluding header data). Instead it appears to me at least that Olympus are storing 16 bits for each pixel, assuming the analog to digital converter doesn't deliver more than 12 bits then this is a waste of around 3 MB of storage for each RAW file.
Standard Test Scene
To give an impression of what some of the combinations of image size and quality produce the table below is a cross reference of some of them:
- 2560 x 1920 RAW (to TIFF using Olympus Viewer 1.0)
- 2560 x 1920 TIFF
- 2560 x 1920 SHQ JPEG
- 2560 x 1920 HQ JPEG
- 1280 x 960 SQ JPEG (1/2.7 quality)
Crops below are of the same 240 x 100 area of each image nearest neighbour magnified 200%.
Settings: ISO 100, E 14-54 mm F2.8-F3.5 @ F5, Sharpness 0
|2560 x 1920|
10,412 KB .ORF (not available for download), VGA crop - 688 KB .TIF
15,696 KB .TIF (not available for download), VGA crop - 646 KB .TIF
|1280 x 960|
There is a clear difference between the quality of the RAW converted image and TIFF / JPEG straight from the camera. A little further analysis revealed that Olympus appear to have two different RAW development engines, Olympus Studio provides access to both of these. The faster 'High Speed' engine produces the same artifacts we see from in-camera TIFF / JPEG, the better quality 'High Function' engine delivering the results seen above (Olympus Viewer also uses the High Function engine).
JPEG artifacts were only visible in the 1/8 quality image (JPEG HQ) and these were mostly isolated to the 'mosquito' type effect around contrast detail.
The E-1 provides output color space options of sRGB and Adobe RGB. Switching to Adobe RGB color space changes the mask used for filenames, instead of being prefixed with a 'P' they are prefixed with an underscore '_'. Additionally the E-1 does embed the Adobe RGB color profile into images, this means that color space aware applications such as Adobe Photoshop automatically assign the correct profile upon opening.
Place your mouse cursor over the labels shown below the color chart to see the difference between shooting in sRGB mode and shooting in Adobe RGB and then converting the image to sRGB. Note that the Adobe RGB image has been converted to the sRGB color space for correct color on your monitor.
Settings: ISO 100, ED 50 mm F2.0 Macro, 1/50 sec, F5
|sRGB||Adobe RGB -> sRGB|
Below are two studio shots taken as both sRGB and Adobe RGB. Note that in these samples the Adobe RGB image has NOT been converted to sRGB and so to view it correctly you will have to load it into a color space aware photo application and assign the Adobe RGB color space. Below each sample is that image's CIE u'v' Color Distribution chart; larger gray triangle approximately represents the range of color which the human eye can resolve, the inner triangle the available gamut in each color space (sRGB or Adobe RGB).
Settings: ISO 100, E 14-54 mm F2.8-F3.5 @ F5, Sharpness 0, HQ JPEG