Pentax 645Z Review

Full-Resolution Studio Tests

We performed a series of studio image tests to gauge the 645Z's color and noise performance in a controlled environment.  While doing so, we also compared the 645Z to the Nikon D810 and the Pentax K-3: two high-resolution cameras using smaller sensor formats that you may see as alternatives to the 645Z.

Test Conditions

Our test photos were all taken on a tripod with mirror lock-up/self-timer.  We used prime lenses that maintained roughly the same angle of view across all 3 cameras, and adjusted the aperture settings accordingly to maintain a comparable depth of field.  We then ran through the entire ISO range of each camera, at 1-stop increments, adjusting only the shutter speed to maintain proper exposure.

The file format was set to RAW+ with default JPEG color rendering and normal (Nikon) or auto (Pentax) noise reduction for the JPEG files.

The comparative photos were illuminated with dual 500W studio floodlights, resulting in an EV value around 13EV.  We also took a series of low-light photos with the 645Z where the only light source was a single 60W CFL bulb, or approximately 4EV.

Full-Size Downloads

Before we move on to the test results, be advised that RAW files from each camera, at all ISOs from minimum to maximum, are available for download below.  JPEG files with normal in-camera noise reduction are also available:

The files are ordered by ISO.  Note that below, we will only be looking at a handful of ISOs to keep things simple.

The photo below shows our test scene.  The four areas for which we compare full-size crops are highlighted in red.

Studio Test SceneTest Scene
Area 1: Left, Area 2: Right, Area 3: Center, Area 4: Top

Without further ado, let's look at the results.  As a reminder, all images in the frames below are unscaled and unedited (full-size) crops.  They were developed in Adobe Camera RAW using baseline settings without any noise reduction.  The out-of-camera JPEG files are not shown below and only available via the links above.

645Z vs. D810 vs. K-3 at ISO 100, 400, 1600, and 6400

All four sets of photos below were shot will all three cameras.  If you have trouble using the dropdowns, please try a supported browser such as Firefox.

Area 4 - 100% Crops

Area 3 - 100% Crops

Area 2 - 100% Crops

Area 1 - 100% Crops

645Z In Low Light

Finally, let's take a look at how the impact of noise changes in low light.

All Areas - 645Z Only - Low Light

Commentary

The 645Z offers an impressive level of detail in its images.  If files from the 645D left a positive impression on you, then you'll simply be spolied by the 645Z. This is not just due to the extra 11 megapixels, but also thanks to its greatly improved high-ISO performance from its CMOS sensor.

These results should speak for themselves.  Files up to ISO 400 are remarkably clean from the 645Z.  Even in very low light, the increase in noise and decrease in color accuracy is so insignificant that will most likely not necessitate software corrections.  Thus, there is no significant practical effect on image quality.  This is good, since for those who wish to hand-hold the 645Z, shooting a couple of stops above the base ISO is advised.

At ISO 1600, the 645Z still performs exceptionally well.  If we compare it to the 645D, which lost plenty of shadow detail at maximum ISO, pictures from the 645Z really hold their ground.  There is very little noise in solid colors that are well-lit, and even fine details remain visible. There is but a slight loss of shadow detail, and noise levels are easily manageable overall.  So, we'd like to conclude that you can still expect very high-quality prints at ISO 1600.

At ISO 6400 noise starts to become noticeable at smaller image sizes, but image quality overall is still good despite being a noticeable drop compared to ISO 1600 or even 3200.  In poorly-lit areas, noise can eat away at details, however, as we can observe in area 1 (low light) and area 3 (studio light).  At this ISO we can start to appreciate the importance of pixel pitch.  The K-3 struggles in all test areas, while the D810's noise levels are similar to those of the 645Z (both cameras have comparable pixel pitch).  Even though the 645Z generally holds an advantage over the D810 in terms of detail and exhibits a little bit less noise overall, the Nikon is neck-and-neck with the Pentax in the shadows (see area 3), despite the lower resolution.

What's the highest ISO we'd recommend using?  We feel that ISO 12,800 is about the limit for getting quality prints from the 645Z, with ISO 25,600 being the worst-case scenario.  We say this because at ISO 12,800 you can still make out the individual grooves between the staples atop the Pentax lens box, whereas, at ISO 25,600, it becomes tough amid the increasing noise.  At ISO 51,200, web-sized files of well-lit scenes will still be moderately clean, but you'll have trouble above that. 

We recommend that you also take a look at the full-size RAW files linked at the top of this page to see for yourself how the three cameras fare.  Experiment with noise reduction and scaling to see when each camera meets your expectations.  Other considerations aside, once you start seeing results you're satisfied with, you'll know which format is right for you.

Verdict

The 645Z is a remarkable improvement over the 645D in the studio, which is a testament to the evolution of sensor technology and thus image quality.  Its RAW files passed our tests nearly with flying colors.

But there's one thing we're dissatisfied with, and that's JPEG performance.  This may not be the biggest consideration for most users, but we feel that it is important to make note of.  If you took a look at the full-size JPEG files alongside those of the K-3, you would have noticed that the colors appear roughly the same.  Indeed, the two cameras use the same image processing engine (PRIME III) for JPEG conversions and noise reduction.  This means that little to no tweaks have been made to improve the JPEG performance of the 645Z compared to its smaller cousin.  When comparing its JPEGs to those of the D810, not only is the 645Z's noise reduction more lossy, but the colors it produces are also more saturated than they out to be in the default Bright custom image mode. While the 645Z's white balance (like that of the K-3) is very accurate, its reds simply appear too red in JPEG mode.  The box of Horizon Classic Mac from our studio setup has a reddish orange color close to what we see in the D810's JPEGs, rather than the bright red depicted in the two Pentax files.  The reds are more accurate in Natural mode, but still not perfect.  We therefore recommend that you make adjustments to your JPEG preset (see below) and apply post-processing as needed for the best results.

JPEG Presets

Below is a comparison of the 645Z's different JPEG presets.  Click on any thumbnail to enlarge and browse.

Bright

Natural

Landscape

Vibrant

Portrait

Radiant

Bleach Bypass

Monochrome

Reversal Film

Muted

D810 JPEG

645Z RAW w/ Adobe Camera Raw Defaults

Each mode includes a context-sensitive adjustment screen that lets you tweak numerous parameters, such as saturation, hue, contrast, sharpness, and brightness.  These parameters can also be adjusted on RAW files opened in the included Pentax Digital Camera Utility software, thus allowing you to determine the best JPEG preset based on a shot you've already captured.  Once you find the perfect mix of settings, we recommend that you also apply them in-camera to improve your future JPEGs.


facebook.com/PentaxForums PentaxForums @PentaxForums News | Reviews | Forum

Support Pentax Forums Donate to Pentax Forums Support Pentax Forums