Image Quality

Overall the a7's image quality is pretty much what you'd expect from a full-frame camera - making it easier to get shallow depth-of-field and offering excellent low light capability, simply as a result of it having such a huge sensor to gather light. However, despite this potential, we weren't always delighted with the images we got - particularly when we were shooting in JPEG mode.

Posterization

An odd artifact that we noticed several times in the JPEGs is the tendency toward posterization in subtle gradations, again, often in out-of-focus areas. It's certainly not a problem that appears in every image but it's something we noticed often enough to be concerned about it.

Note: We're mindful that the length of this dedicated section risks making this appear a more significant problem, but it's a nuanced enough issue that it requires several examples to show. The fact is, it won't be an issue for those users who shoot Raw. For those who will shoot only JPEG (and there are still plenty of folks out there) we think it's important to demonstrate.

JPEG 1/15 sec, F5.6, ISO 6400, using 55mm F1.8
RAW via ACR
100% crops (All channels)
Red channel (100% crop)

Further investigation revealed two issues - limited bit-depth in Raw files and clumsy/excessive JPEG processing. The bigger of the two problems is the JPEG processing. We suspect this stems from the compressed Raw files and the difficulty they have in describing smooth transitions, but the real problem is possibly being caused by the overly aggressive noise reduction in the JPEGs.

Turning the cameras' 'High ISO NR' to 'Off' improves matters (at all ISOs), because the standard processing appears to be supressing noise to such a degree that you lose the 'dithering' effect that noise usually brings to tonal transitions. The downside is that you have to put up pretty noisy high ISO images (noise reduction, in both JPEG and Raw is still being applied in the 'Off' setting).

JPEG posterization

Here we've set up a test scene that shows the stepped (posterized) effect that can occur in what are supposed to be smooth transitions. We've compared it to the Nikon D610 (which has a very similar sensor), to ensure we haven't created a scene that all cameras would stuggle with.

Although this is a simplistic simulation, designed to illustrate the point, any smooth transition in the camera's images, whether it be a blue sky or a diffuse background, risks exhibiting posterization and inconsistent tonality across those transitions.

Sony Alpha 7
ISO 100
F7.1, 5 sec
High ISO NR Normal
7.0MB
Sony Alpha 7
ISO 100
F7.1, 5 sec
High ISO NR Off
9.2MB
Nikon D610
ISO 100
F7.1, 5 sec
High ISO NR Normal
5.9MB
Sony Alpha 7 (X-Fine JPEG)
100% crop - All channels
Sony Alpha 7 (X-Fine JPEG)
100% crop - All channels
Nikon D610
100% crop - All channels
Sony Alpha 7 (X-Fine JPEG)
100% crop - Red channel
Sony Alpha 7 (X-Fine JPEG)
100% crop - Red channel
Nikon D610
100% crop - Red channel

As you can see, the problem isn't image-destroyingly bad - indeed, whether you can even see what we're talking about will depend to a large degree on the device you're viewing the site with. But, once you've looked at the image's red channel, it should be much more obvious what we're talking about. Look closely again at what should be a smooth background of the portrait on the Experience page, and you'll see it in a real-world setting.

Any users hoping to post-process their JPEGs will be particularly disappointed, since these hard steps between tones become exaggerated as soon as you start adjusting the image brightness. Obviously we'd recommend shooting Raw if you plan to do any significant post-processing (with any camera), but potential a7 owners need to be aware that its JPEGs offer even less processing latitude than usual.

A comparison between an X-Fine JPEG from the Sony (7.8MB) and a Nikon shot at the same exposure settings (6.9MB). Examine the dark regions closely and you'll find very dark green stripes across the lower sky on the Sony image, where the posterized steps between the red and blue channels are occuring out of sync with one another.

Push the two files (which we've done using the same tone curve on both images), and those color errors become very obvious, very quickly. The Nikon, one which noise dithers any stepping in its different channels, doesn't show the same errors.

Sony Alpha 7
ISO 100
F5.6, 4 sec
Nikon D610
ISO 100
F5.6, 4 sec
Sony Alpha 7 (X-Fine JPEG)
100% crop
Nikon D610
100% crop
Sony Alpha 7 (X-Fine JPEG)
With tone curve applied
Nikon D610
With tone curve applied

Raw posterization

From the file sizes alone, (they're fairly uniform in size, around 24MB per image, rather than the 36MB per file required by uncompressed 14-bit data) it's obvious that Sony is compressing its Raw files. Iliah Borg and Alex Tutubalin, who run the LibRaw Raw decoder project, have been analyzing Sony Raw files in considerable detail, and have shown that the 14-bit files contain essentially 11 bits of data.

They conclude that, so long as you're exposing files as the camera wants to (rather than ignoring the camera's metering in an attempt to 'expose to the right' or significantly underexposing to protect highlights), there's no significant photographic difference, though there is a reduction in processing latitude. Certainly it wasn't significant enough to prevent them choosing to buy Sony cameras.

When we looked into the issue, we found the camera would have certain steps missing even in the shadow regions of the file (where it's most important to be capturing as much information as possible). So, to see what impact this had, here we've underexposed the scene used above by 3EV, then push-processed both the a7's and D610's file back up to the correct brightness.

Sony Alpha 7
ACR conversion with +3.00EV exposure correction
Nikon D610
ACR conversion with +3.00EV exposure correction
Sony Alpha 7
Red channel from above conversion
Nikon D610
Red channel from above conversion

There's slightly more coarse noise in the Sony's file (the verical ), though it's only really visible if you start investigating individual channels, but there's little visual sign of the posterization that can occur in the JPEGs. If you wish to experiment for yourself, click here to download the Raw files, below.

Underexposed Raw files: