ISO Sensitivity / Noise levels

Standard Test

ISO equivalence on a digital camera is the ability to increase the sensitivity of the sensor to enable faster shutter speeds and/or better performance in low light. The way this works in a digital camera is by "turning up the volume" on the CCD's signal amplifiers. Nothing is without its price however and doing so also amplifies any noise and can also affect colour saturation.

The DCS-14n provides eight ISO sensitivities (third stop steps) between ISO 80 and ISO 400 at full resolution and ISO 800 at lower resolutions.

Our noise comparison test involves shooting a GretagMacBeth ColorChecker at a selection of ISO sensitivities and then measuring luminance and RGB noise at a 'mid' grey patch (patch 22). We would really have liked to test the 14n without noise reduction, unfortunately Kodak don't offer this option either for in-camera JPEG or post-procesed RAW files.

Kodak DCS-14n vs. Canon EOS-1Ds

Camera settings / test notes

• Kodak DCS-14n: Normal NR, Low SHP, Product Look, Large JPEG III (4500 x 3000),
  Manual WB
   
• Canon EOS-1Ds: Parameters: Standard, Large/Fine JPEG (4064 x 2704),
  Manual WB
   
• Measurements taken at approximately 21°C (~70°F)
   
• Lighting was daylight

	ISO 80	ISO 50
	Kodak DCS-14n 1/20 sec, F8.0	Canon EOS-1Ds 1/15 sec, F8.0
Original crop
Red channel
Green channel
Blue channel
	ISO 100
	Kodak DCS-14n 1/30 sec, F8.0	Canon EOS-1Ds 1/30 sec, F8.0
Original crop
Red channel
Green channel
Blue channel
	ISO 200
	Kodak DCS-14n 1/60 sec, F8.0	Canon EOS-1Ds 1/60 sec, F8.0
Original crop
Red channel
Green channel
Blue channel
	ISO 400
	Kodak DCS-14n 1/125 sec, F8.0	Canon EOS-1Ds 1/125 sec, F8.0
Original crop
Red channel
Green channel
Blue channel

It's worth remembering that the EOS-1Ds provides sensitivities up to ISO 1250 at full resolution, not shown here. A quick glance at the luminance noise graph shown below tells you that (at these exposures) the DCS-14n exhibits less noise than the EOS-1Ds. Indeed it appears as though Kodak's noise reduction system does manage to keep noise down in flat areas such as the gray patch used for our measurements. Remember that we don't know what if any noise reduction is being carried out by the EOS-1Ds.

This however not the whole story, we also have to take a look at what happens to areas of detail and how the camera performs at slower shutter speeds (see below graphs).

Luminance noise graph

ISO sensitivity is on the horizontal axis of this graph, standard deviation (average) is on the vertical axis.

RGB noise graph

ISO sensitivity is on the vertical axis of this graph, standard deviation (average) is on the horizontal axis.

Noise levels in low light

As I hinted above the relatively low noise exhibited by the DCS-14n in our test wasn't the whole story, experience of shooting the DSC-14n at higher sensitivities indicated that it performs less well at slower shutter speeds and when the noise reduction system has to deal with areas of detail.

Kodak DCS-14n vs. Canon EOS-1Ds, ISO 100, low light

Kodak DCS-14n	Canon EOS-1Ds
1 sec, F11, 2,007 KB JPEG	1 sec, F11, 2,840 KB JPEG
Std. dev. 1.86	Std. dev. 2.86

At ISO 100 there appears to be little visible difference between the two cameras, the DCS-14n's sharpening algorithm appears to be slightly more sophisticated (although moiré is visible) and there are some subtle indications of noise reduction (loss of definition of the power button on the light meter) but otherwise an acceptable result.

Kodak DCS-14n vs. Canon EOS-1Ds, ISO 400, low light

Kodak DCS-14n	Canon EOS-1Ds
1/4 sec, F11, 4,483 KB JPEG	1/6 sec, F11, 3,542 KB JPEG
Std. dev. 5.24	Std. dev. 3.86

Here's where things start to turn nasty for the 14n, it's noise reduction system struggles to keep noise down and in doing so produces a 'watercolor like' appearance to the image, jagged edges and lumps of color, this side-effect of the Kodak's noise reduction system can be seen in a percentage of images, even at lower sensitivities.