Previous page Next page

ISO / Sensitivity accuracy

In a new addition to our reviews we are now measuring the actual sensitivity of each indicated ISO sensitivity. This is achieved using the same shots as are used to measure ISO noise levels, we simply compare the exposure for each shot to the metered light level (using Sekonic L-358), middle gray matched. We estimate the accuracy of these results to be +/- 1/6 EV.

Like many recent digital SLRs all four in this comparison proved to be right on the spot with their indicated sensitivity, that is to say an indicated sensitivity of say ISO 200 was exactly that (unlike some older digital SLRs which were slightly more or less sensitive than indicated).

Indicated
sensitivity
Olympus E-3
(actual sensitivity)
Canon EOS 40D
(actual sensitivity)
Nikon D300
(actual sensitivity)
Sony DSLR-A700
(actual sensitivity)
ISO 100 ISO 100 ISO 100 ISO 100 ISO 100
ISO 200 ISO 200 ISO 200 ISO 200 ISO 200
ISO 400 ISO 400 ISO 400 ISO 400 ISO 400
ISO 800 ISO 800 ISO 800 ISO 800 ISO 800
ISO 1600 ISO 1600 ISO 1600 ISO 1600 ISO 1600
ISO 3200 ISO 3200 ISO 3200 ISO 3200 ISO 3200
ISO 6400 n/a n/a ISO 6400 ISO 6400

ISO Sensitivity / Noise levels

ISO equivalence on a digital camera is the ability to increase the sensitivity of the sensor. The 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 1.4 in this review). Click here for more information. (Note that noise values indicated on the graphs here can not be compared to those in other reviews). Room temperature is approximately 22°C (~72°F), simulated daylight lighting.

Olympus E-3 vs. Sony DSLR-A700 vs. Canon EOS 40D vs. Nikon D300

  • Olympus E-3: Olympus 50 mm F2.0 Macro lens, Manual exposure, Manual WB,
    Default Parameters (Natural), JPEG Large / Fine

  • Sony DSLR-A700: Minolta 50 mm F1.4 lens, Manual exposure, Manual WB,
    Default Parameters (Standard), High ISO NR (Normal - default), JPEG Large / Fine
     
  • Canon EOS 40D: Canon 50 mm F1.4 lens, Manual exposure, Manual WB,
    Default Parameters (Standard), High ISO NR (Default; Off), JPEG Large / Fine
     
  • Nikon D300: Nikkor 50 mm F1.4 lens, Manual exposure, Manual WB,
    Default Parameters (Normal), High ISO NR (Normal), JPEG Large / Fine
Olympus E-3
ISO 100
Canon EOS 40D
ISO 100
Nikon D300
ISO 100
Sony DSLR-A700
ISO 100
Olympus E-3
ISO 200
Canon EOS 40D
ISO 200
Nikon D300
ISO 200
Sony DSLR-A700
ISO 200
Olympus E-3
ISO 400
Canon EOS 40D
ISO 400
Nikon D300
ISO 400
Sony DSLR-A700
ISO 400
Olympus E-3
ISO 800
Canon EOS 40D
ISO 800
Nikon D300
ISO 800
Sony DSLR-A700
ISO 800
Olympus E-3
ISO 1600
Canon EOS 40D
ISO 1600
Nikon D300
ISO 1600
Sony DSLR-A700
ISO 1600
Olympus E-3
ISO 3200
Canon EOS 40D
ISO 3200
Nikon D300
ISO 3200
Sony DSLR-A700
ISO 3200
Olympus E-3
n/a
Canon EOS 40D
n/a
Nikon D300
ISO 6400
Sony DSLR-A700
ISO 6400

As noted when we reviewed the A700, from ISO 100 to 400 there is virtually no difference from a noise point of view between any of these cameras, none exhibit any noticeable noise. At ISO 800 the E3 shows a little noise (and you can see that the noise reduction has kicked in because the image is softening). The Canon and Nikon do a better job, but Sony's strong noise reduction is causing problems at ISO 800.

At ISO 1600 the E-3 is certainly the noisiest image here (and the effects of noise reduction are really starting to take their toll), though only the Nikon D300 really shines; the Canon 40D and Sony A700 are retaining slightly more detail with noticeably less noise than the E-3, but it's not a vast difference.

At ISO 3200 the E-3's sensor is obviously struggling and the combined effects of noise and noise reduction have moved the output firmly into the 'emergency use only' camp. The Sony A700 is only slightly better, but both Canon and (especially) Nikon are noticeably less problematic. Olympus wisely chose to draw the line at ISO 3200; the A700 is a blurry, noisy mess (though the D300's output is surprisingly good, and isn't considerably worse than the E-3 was at ISO 1600).

Along with dynamic range (covered elsewhere), noise and high ISO performance has been seen as the biggest drawback to the Four-Thirds system; smaller sensors mean smaller pixels, so the refrain goes, meaning in theory a Four-Thirds camera can never quite keep up with an APS sensor camera with the same pixel count. With the E-3 Olympus did some work on the sensor and microlenses to increase the light gathering ability of each pixel, and the result has obviously borne some fruit, as the gap between the E-3's output and that of its competitors has narrowed considerably. But a gap there is, and it's obvious that once you get over ISO 400 the E-3 starts to struggle to keep up with the best cameras in this class. Even the A700, with its excessive noise reduction that smears away detail at higher ISOs, is outperforming the E-3 at ISO 1600

* It's worth noting that these results are with the cameras in their default modes, the EOS 40D for example does have an optional stronger chroma noise reduction option which delivers images with almost no chroma noise.

Noise graphs

Note that we normally show both gray and black results on the same graph, comparing five cameras this became too difficult to read hence we have two separate graphs, one for the gray patch (middle gray) and one for the black patch (shadows). For this comparison we've also thrown in the Panasonic DMC-L10 which we believe to have the same sensor (though as discussed elsewhere, there is a difference in the anti alias filer).

Luminance noise graph (gray patch)

As you can see the E-3 noise starts low but once you get above ISO 800 is a little higher than the competition (except the Panasonic L10, which is in a class of its own). As the crops above show the price of keeping noise down to the levels expected in this class is that detail is blurred away at higher sensitivities. The difference between the E-3 and L10 graphs clearly shows the different levels of noise reduction being applied by the two cameras.

Indicated ISO sensitivity is on the horizontal axis of this graph, standard deviation of luminosity on the vertical axis.

Luminance noise graph (black patch)

Once again the E-3's noise levels are higher than the competition (in this case at most ISO settings), with the gap at ISO 1600 and above showing quite a gap. Once you get above ISO 800 it's clear that the Canon D40 and (especially) the Nikon D300 give you a good two stop advantage when it comes to shadow noise.

Indicated ISO sensitivity is on the horizontal axis of this graph, standard deviation of luminosity on the vertical axis.

Chroma (color) noise graph

As the crops show, the E-3 controls chroma noise well (the noise reduction system seems to concentrate on color noise, leaving luminance noise, which looks a lot less unpleasant).

Indicated ISO sensitivity is on the horizontal axis of this graph, standard deviation of color on the vertical axis.

Previous page Next page
187
I own it
3
I want it
104
I had it
Discuss in the forums

Comments