Back when we were briefed by Canon about its update to one of the most popular line of digital SLRs in existence, Canon told us that dynamic range was one of the top concerns of current 5D-series owners.
We're happy to report that the 5D Mark IV represents a major leap forward in this regard, and it's thanks to the same move to on-chip analog-to-digital conversion (ADC) that the 1D X II and 80D saw. It's been a long time coming for Canon, as placing the ADCs on the imaging chip itself is a staple of modern image sensor design.
This design ensures less noise is added to the analog signal captured by the sensor by digitizing* as early as possible, resulting in lower read noise, which you can think of as background amplifier hum. The lower 'hum' means its easier to distinguish between captured information and background noise in areas of low signal. In turn, this means more malleable Raw files with more useful information available when you try to brighten shadows, allowing you in turn to more confidently expose high contrast scenes, like the one below, for the highlights.
This Nikon D810 shot was exposed for the highlights by keeping ISO near base. In M mode I dialed in a wide-open aperture and the shutter speed necessary to freeze the action (1/640s). In order to ensure I didn't blow any of the sky, I kept my ISO down at 140, despite a meter reading suggesting ISO 2200 for a 'proper' exposure. Pushing shadows 4 EV in post, while holding back highlights, I was able to tone-map the scene to my desired vision, without tones becoming too noisy. All thanks to a high dynamic range sensor. Photo: Rishi Sanyal
Below we look at just how malleable the 5D Mark IV's Raw files are, and on the next page we'll assess the 'ISO-invariance' of the camera, the property that allowed for the capture of the above image without noise penalty.
Exposure Latitude
One of the most easily understandable ramifications of increased base ISO dynamic range is increased processing latitude; that is, the ability to brighten Raw exposures without drastic noise penalty (there will always be some penalty due to shot noise). This can be particularly useful in dealing with high contrast scenes, which require conventional underexposure to prevent bright tones from clipping to white, with requisite shadow brightening - or tone-mapping - to make dark tones visible on our current dim display technologies.
So in this test we look to see how tolerant of pushing exposure the 5D Mark IV's Raw files are. We've done this by exposing our scene with increasingly lower exposures, then pushing 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. **
$(document). ready(function() { ImageComparisonWidget({"containerId":"reviewImageComparisonWidget-34339389","widgetId":397,"initialStateId":2773}) })
The 5D Mark IV shows significant improvements in exposure latitude thanks to its increased dynamic range. After a 5 EV push, it's well ahead of the 5DS which, despite its old sensor design with off-chip analog-to-digital conversion, was already 2/3 EV ahead of the 5D Mark III. That places the 5D IV well ahead of its predecessor, nearly catching up to the excellent Sony a7R II. Despite it's improvements, it's not at the level of the current industry leader, the Nikon D810, though. After a 6 EV push$(document). ready(function() { $("#imageComparisonLink2774"). click(function() { ImageComparisonWidgetLink(2774); }); }), the 5D Mark IV falls further behind the a7R II and D810, but the improvement over the 5DS is even more dramatic.
Digging a bit deeper: the 5D IV shows improvements over the 5DS with even more moderate 3-4 EV pushes$(document). ready(function() { $("#imageComparisonLink2768"). click(function() { ImageComparisonWidgetLink(2768); }); }), but especially so when files are pushed 5-6 EV$(document). ready(function() { $("#imageComparisonLink2770"). click(function() { ImageComparisonWidgetLink(2770); }); }). This means you'll see the advantages of the on-chip ADC in the form of less noise not just with drastic exposure adjustments, but even more moderate ones. Results are about on par with, if not slightly better than, the 1D X II$(document). ready(function() { $("#imageComparisonLink2769"). click(function() { ImageComparisonWidgetLink(2769); }); }). And while the 5D IV falls just slightly short of the Sony a7R II, the differences are really only visible after fairly extreme pushes$(document). ready(function() { $("#imageComparisonLink2771"). click(function() { ImageComparisonWidgetLink(2771); }); }).
Differences against the current dynamic range market-leader, the D810, start becoming apparent$(document). ready(function() { $("#imageComparisonLink2767"). click(function() { ImageComparisonWidgetLink(2767); }); }) after even a 3 EV push, and fairly significant after a 6 EV push. The D810 can perform so well because of even lower read noise, and increased sensor capacity for light at ISO 64 that gives its files a nearly medium format-esque quality.
On the next page, we'll look at how ISO-invariant the 5D Mark IV is, and also directly compare the Mark IV to its predecessor, the Mark III.
* Digital signals tend to be more 'protected' or immune to noise than analog signals, because they're binary. It's the same reason CDs (digital) don't exhibit the pops and crackle of records (analog).
** Differences in noise performance in our Exposure Latitude test are caused by both read noise and shot noise, the latter of which is mainly determined by the amount of light the camera has had access to. Therefore, one might argue the results are only directly comparable between cameras of the same sensor size. However, sensor size differences will also be relevant in real-world shooting if you're limited by what shutter speed you can keep steady, so this test also gives you an idea of the amount of processing latitude different formats give when light-limited.
. dpreview.com2016-9-1 21:16