My interest in testing these cameras originates from a mix of personal curiosity and questions from inquiring workshop participants — which camera creates the best image taken in low light conditions, primarily night time?
There are many cameras that tout incredible high ISO results, so I decided to choose a few that were interesting to me. I did not include any Canon cameras because in previous testing I have already determined that the Nikon D800 and D810 are both superior in low light to the 5DM3 and 6D. This time I was out to see just how close the least expensive Fuji XT-1 would stack up against the much more expensive Pentax 645Z. In addition, I wanted to see just how clean the file was from the Sony a7s vs. my Nikon D810.
- Fuji XT-1, Nikon 14 - 24mm F2.8 lens with Metabones Nikon G Lens to Fujifilm X-Mount Camera Speed Booster adapter, shot at shortest focal length (14mm). Sensor size: APS-C 23.6 x 15.6mm
- Sony a7s, Nikon 14 - 24mm F2.8 lens with Metabones Nikon F Lens to Sony NEX and E mount adapter, shot at 20mm ( to match Pentax ). Sensor size: Full frame 35.9 x 24mm
- Nikon D810, Nikon 14 - 24mm F2.8 lens, shot at 20mm ( to match Pentax ). Sensor size: 35.9 x 24mm
- Pentax 645Z, Pentax 25mm f/4 lens. Sensor size: 43.8 x 32.8mm
My method for comparing these sensors was very simple. I set the cameras up side by side and matched the composition in an environment and at a time where the light did not change. I used the same lenses and camera settings (ISO, Shutter Speed and Aperture) whenever possible. I also matched preferences such as long exposure noise reduction, high ISO noise reduction and file type.
Refugio State Beach park offered a decent view of the stars and a composition that included some foreground elements to offer additional detail to review. I arrived at the location, a couple hours before sunrise, where I set up the 2 tripods and the first two cameras to test.
The only variation in light occurred when a motion sensor light was activated by some raccoons during one of the exposures. This did not effect the stars nor did it effect the light on the distant mesa. You will notice the very first image file taken with the Fuji has a brighter tree trunk in the foreground.
In order to compensate for the different resolutions in the cameras, I have made similarly sized TIFF files for comparison. First, I want to show the files I chose to compare. I have not changed anything during import, not even white balance.
These next set of images are a 100% crop of the same region from each file. These files are all totally RAW, with no alterations from Adobe Lightroom. Click on an image to open it in Lightbox.
*Notice how useless the first Pentax file is, shot at ISO 3200. This is due to the F4 lens which is the brightest wide lens Pentax makes. I understand a 25mm F2.8 would be very large, but this becomes a considerable factor when comparing these cameras.
These next set of images are 100% crops of all the files after they have been processed and sized to that similar to the lowest resolution camera, the Sony a7s, ( 4240 x 2832 ).
I chose the ISO that made sense. This means that due to the aperture variable, I processed the files that yielded the luminosity I wanted without compromising additional noise.
Click on the images below to see the final, processed frames.
After normalizing the files it became quite clear that the Sony was the winner. This is based on two main factors. First, the file held more color information in the shadows, probably due to an increase in dynamic range because of the larger cell sites on the sensor. Second, the noise cleaned up better than the higher resolution files of the Nikon D810 and Pentax 645Z.
Even though the resolution of the Sony was much less than all the other cameras, it still yielded the very best low light files, making it a great tool for not only my still photography but for time-lapse work as well.
Ergonomics, Menus and Lens Selection
Menus and buttons don’t affect the quality of the final image file, but I always consider it a major issue if a particular camera is so poorly designed that it becomes a hurdle in my workflow. I am also quite critical about lens choice, and if the camera manufacturer does not offer lenses I am fond of, then I must weigh the compromise of file quality with creative vision.
Because there are so many features to compare, I wanted to focus on only the real important ones in regards to night photography.
Manual focus functionality
Fast wide lenses
Access to buttons in the dark and cold
The Fuji XT-1 offers a well designed menu interface and as a bonus, camera user profiles! Fuji does offer a wide variety of lenses, making it a great prospect for the one and only system for traveling light without compromising image file quality. For night photography, however, their 20mm F2.8 was not as sharp in the corners as the Nikon 14-24mm with metabones adapter. (I did not get the chance to test the Touit 12mm F 2.8!) Considering the lens size and versatility of this system, I would consider this the very best alternate camera system to the full size sensor DSLRs and mirrorless, including the Sony. My goal however, was not to minimize size and weight as it was to maximize quality of the image in low light.
The Sony also has a good menu system, although the camera user profiles are not very intuitive, they do work. The LCD is wonderful and offers the most light in the night, making manual focus easier. This is the most important aspect of night photography, as I can’t tell you how often I see focus misses ruin an entire evening of shooting. The Sony is a smaller camera body, but if you are considering it to replace your DSLR for weight, DONT! The lenses made by Sony are large and only F4! My plan is to use the new Nikon 20mm F1.8 lens for all my night photography. The lens just arrived on the market as I write this post.
I am amazed that Nikon has not updated it’s very antiquated “shooting menu bank” with a custom user profile similar to what they offer in the D610 and now D750. Since the Nikon D750 was not available at the time of this test it has not been included. Although, it appears to be as good as the D810 and maybe even have a slight advantage to the D810 in low light situations, most likely due to its slightly larger cell sites. It even offers a fold out monitor!
Pentax 645Z This camera was a bit larger than the normal full sensor DSLR, but very well designed with a great LCD. The real issue with this system is the limited lens choice and their slow apertures. The files do have more dynamic range than the Nikon D810 and handled better when radical post processing edits were applied, which is a practical sign of a greater dynamic range. I do not think the extra resolution and slightly greater dynamic range is worth the weight and price! Considering low light, I don't think the files handled any better than the D810 even when the resolution was decreased to match the smaller sizes of both the Nikon and the Sony.
Interesting To Note
The Metabones Speed Booster adaptor did not degrade sharpness. It reduces the focal length and increases the aperture, or makes your lens wider and brighter. This is then altered by the smaller size sensor to bring the focal length equal to, or very close to its original. In addition, the depth of field is increased due to the smaller sensor size. Nothing about the smaller sensor reduces the light, therefore, the end result is a brighter lens, which is magic, making it one of the coolest inventions in photography in my lifetime!
I also compared daylight scenes between the Nikon D810 and Pentax 645Z. I found that there was more dynamic range in the Pentax files, yielding better latitude for post processing. Prior to post processing the files were almost indistinguishable. This was a good advantage for the Pentax, however, the size, weight and poor lens choice made it a non-contender for the diversity of work that I do.
Life is short, make pictures!