Exposure Latitude

In this test we look to see how tolerant of pushing exposure the NX1's Raw files are. We've done this by exposing our scene with increasingly lower exposures, then pushed them back to the correct brightness using Adobe Camera Raw. Examining what happens in the shadows allows you to assess the exposure latitude (essentially the dynamic range) of the Raw files.

As we'd expect, as the exposure is decreased, less light is collected, which leads to more overall noise visible especially in shadows and midtones. What's interesting is what happens when we compare it to other cameras of the same sensor size.

Compare it with Nikon D5500, for instance, and the two cameras show similar amounts of noise when underexposed and pushed by as much as 6EV. The alternative to this is a camera whose noise floor starts to appear when you push the images too far: as was the case with previous Samsung models. For example, the Samsung NX3000, begins to look a fraction noisier than the NX1 from about a 4EV push onwards. Once you get to a 6EV push, it's really apparent that the camera is contributing more to the noise than the NX1 does, though still not as much as what we see with a Canon 70D.

Photographically, this means the NX1's Raw files have a lot of information in them - more than what's available in the camera JPEG, where the tone curve gives up dynamic range to provide a punchy, contrasty image. If you're shooting high contrast scenes like the sunset in our next section below, or if you just want a lot of processing latitude for those times you didn't quite nail your exposure, you'll want to shoot Raw in order to recover information in shadows you didn't even know were there.

Real world impact:

Here we look at the real-world impact of this performance, comparing an image shot at the same time as the Nikon D5500 image we used in that camera's review. Both images were shot so that the brightest tones in the image are represented by the highest-possible Raw value, without clipping (they are near-perfectly 'exposed to the right').

The first thing you'll notice, by clicking on the [i] button at the bottom right of each crop, is that we were able to give the Samsung less exposure, indicating that its sensor is slightly more efficient.

Despite this, if you look around the images, you'll see that the NX1 gives a very similar performance in deep shadow regions, showing that it's almost matching the class-leading Nikon D5500. That's impressive.

ISO Invariance

A camera with high (base ISO) dynamic range has a very low noise floor, which has an interesting implication: the low noise floor can reduce the need to amplify the sensor's signal in order to keep it above that noise floor. This can afford you benefits in situation conventionally demanding higher ISO settings.

Here we've done something that may seem counter-intuitive: we've used the same aperture and shutter speed at different ISO settings to see how much difference there is between shooting at a particular ISO setting (and using hardware amplification) or digitally correcting the brightness, later.

As you can see, there's essentially little to no difference between the camera's native ISO 3200 performance and the results you get from shooting at ISO 100 and pushing by 5EV. This is an example of near perfect ISO-invariance. Contrast this to an 'ISO-variant' camera like Canon's 7D Mark II and you'll see that can't achieve the same performance from underexposing and pushing as opposed to shooting at a higher ISO.

But why would you want to shoot in this manner? With cameras like the NX1, rather than using a high ISO, it's quite possible to use a lower ISO setting and push it up up to at least 5EV without paying an additional noise cost. The advantage of this approach is that, while the shadow noise is likely to be very similar, you should gain several extra stops of highlight detail, since you're using lower levels of hardware amplification and hence won't have amplified that highlight data to clipping.

The idea is that if the scene has highlights that are clipping at ISO 3200, you will be better off by keeping the same aperture value and shutter speed, but turning the camera back down to ISO 100 and brighten it later, from Raw. This way you can get essentially the same shadow noise performance but with an extra 5EV of highlight information. Ultimately, you can use this type of shooting to extend the effective dynamic range of your camera beyond what it would typically offer at higher ISOs.

What this tells us about the sensor's noise floor:

This test also makes clearer just how little electronic noise is being added by the camera's electronics: by using the same exposure for all the shots, we've made sure all the images have the same amount of shot noise (the visible effect of the random nature of light), which means any differences we see are purely the result of electronic noise. The fact that you still can't see any additional noise, even after a 5EV push is very impressive, and a big step forward for Samsung.

The previous generation of NX models, exemplified here by the 20.3MP NX 3000, started to show additional noise after a 2EV push of ISO 800 (bottom right tile), and this had increasing impact on the image quality if you pushed it any further, which you can see in the ISO 100 and 200 pushed crops above.