Although I'm not a huge fan of DxOmark since it reduces photography down to raw numbers, I recently became concerned when I noticed that DxOmark is reporting that Olympus camera ISOs aren't as accurate as other cameras: This isn't a scientific test, but I have both the Nikon D3100 and Olympus OM-D E-M5 so I put the Olympus 45mm 1.8 on the Olympus and the 85mm 1.8G on the Nikon and set them both to Aperture priority at f/1.8 and ISO 1600 with AWB. I then spot metered the same subject in my office with room light only (both cameras suggested the same shutter speed of 1/80th sec.) and took the photos. I shot handheld so excuse the minor difference in composition. Nikon: Olympus: If the Olympus ISO was as low as what DxOmark claims the camera meter should have suggested a slower shutter speed or the Oly shot should have a darker exposure. I'd say both cameras delivered essentially the same exposure. I shot the Oly in JPEG mode with a custom curve to reduce contrast -- to prevent highlight/shadow clipping -- so that is why it looks less contrasty than the Nikon shot which was set to the default curve, but it's worth mentioning that the Olympus shot has sharper detail in the face of the monster model thanks to the IBIS and the Nikon shot has a hint of motion blur due to the lack of image stabilization.