ISO Sensitivity / Noise levelsISO equivalence on a digital camera is the ability to increase the sensitivity of the sensor. This works by turning up the "volume" (gain) on the sensor's signal amplifiers (remember the sensor is an analogue device). By amplifying the signal you also amplify the noise which becomes more visible at higher ISO's. Many modern cameras also employ noise reduction and / or sharpness reduction at higher sensitivities.To measure noise levels we take a sequence of images of a GretagMacBeth ColorChecker chart (controlled artificial daylight lighting). The exposure is matched to the ISO (ie. ISO 200, 1/200 sec for consistency of exposure between cameras). The image sequence is run through our own proprietary noise measurement tool (version 1.5 in this review). Click here for more information. Room temperature is approximately 22°C (~72°F), simulated daylight lighting. Olympus E-P1 vs. Nikon D5000 vs. Canon EOS 500D vs Panasonic DMC-G1
Although measured noise is slightly higher on the E-P1 than the Canon or Nikon (and you can see that - even with the fairly strong default noise reduction, the E-P1 is a little noisier throughout the range), the difference is hardly visible and won't impact on images at normal magnifications until you hit ISO 3200. Interestingly the E-P1 does a better job than the E-620 on JPEGs (with a finer grain and less color noise), and - of course - you can tweak the noise filter settings to your own taste (we'd use a lower setting at anything under ISO 800). Overall there's very little to complain about. Noise graphs
Indicated ISO sensitivity is on the horizontal axis of this graph, standard deviation of luminosity on the vertical axis. RAW noiseAt the moment the E-P1 isn't supported by our benchmark raw converter (Adobe Camera Raw), so our normal comparison between the E-P1 and its competitors has to be viewed with a couple of qualifications (see below). Let's start by looking at the output from the various raw converters that either support the E-P1 (Olympus Studio and DCRaw/Graphic Converter) or those we can trick into opening the files by editing their metadata to make the programs think they were taken on an E-30 (Capture One and Adobe Camera Raw).
In all cases all noise reduction was turned off (or set to minimum), all other parameters (including sharpening) left at default settings. * Important: the ACR and Capture One files were produced using the Olympus E-30 parameters and therefore are unlikely to represent fully optimized output (most specifically, the color mapping and sharpening will be tweaked when 'true' support for the E-P1 is offered in future updates). This review will be updated at that point if the output changes significantly.
Looking at the crops above it's obvious that even with the noise filter set to 'off' Olympus Studio is using some (color) noise reduction, as a quick glance at the output from the other converters clearly shows that at the highest ISO settings noise (both luminance and chroma) is pretty intrusive. Putting aside the sharpening differences (which in the Capture One crops exaggerates the noise), all three third-party converters produce similar levels of noise, with DCRaw giving perhaps the most 'honest' output, with apparently no attempt made to reduce the noise. There's no doubt that - even as low as ISO 200 - the E-P1 is showing more noise than we'd like to see in an ideal world, but it's unlikely to be an issue for 99% of users. Raw noise compared (ACR)As mentioned in several places in this review, our benchmark raw converter, Adobe Camera Raw, doesn't currently support the E-P1, making our normal side by side comparisons difficult. As before we've tricked ACR into opening the E-P1 raw files by making it think they're taken on an Olympus E-30. The results from ACR (which doesn't seem very well optimized for Olympus SLRs) shown here are probably pretty close to those we'll get when Adobe finally adds E-P1 support officially, but until then (at which point we'll update this review if it does change) you should be able to get an idea of the relative noise in these four cameras. As we saw when we recently reviewed the E-30 and E-620 (which we presume share the E-P1's sensor - albeit with a different low pass filter), any inherent noise advantage offered by the slightly larger APS-C sensors used in the Canon and Nikon models has been minimized in the latest generation of cameras; possibly due in part to Olympus's decision to stick at 12 megapixels.
Raw noise graphs
Indicated ISO sensitivity is on the horizontal axis of this graph, standard deviation of luminosity on the vertical axis. Note: the vertical scale on this graph has been doubled compared to the other graphs in this section. Four Thirds cameras compared (JPEG, default)Before we move on let's take a quick look at how the E-P1 compares with the other two Micro Four Thirds cameras (the Panasonic G1 and GH1) - and the nearest Four Thirds camera, spec-wise (the Olympus E-620). Three of these cameras are presumed to share the same 12 megapixel LiveMOS sensor, with the GH1 sporting a slightly higher-spec multi aspect ratio version. Unsurprisingly there's not a lot between them up to ISO 800. At higher settings it's all about the amount and type of in-camera noise reduction being applied (remembering these are all at the default setting), with the Panasonic GH1 probably the best of the bunch by a whisker.
|
