Sony A7ii vs FujiFilm X-T2 – High ISO Noise Test

High ISO Noise Test

Used prices on the X-T2 and A7ii are in the same ballpark at this point, and both cameras’ fans tout the great dynamic range and high ISO capabilities. But can the X-T2’s smaller sensor really keep up with the full-frame sensor in the A7ii?

High ISO Noise Test

There are several very well done comparisons of the FujiFilm X-T2 and Sony A7ii out there in both written and video format. Instead of comparing both cameras from an overall perspective, I’m going to focus on just one thing: a high ISO noise test to see how the two cameras stack up. I will point out, however, that since they are the same resolution (24MP), the A7ii *should* outperform the X-T2 in noise because the sensor pixels are less densely spaced. The same number of pixels spread over a larger area, all other things being equal, will produce less image noise. 

Most cameras these days do quite well with noise suppression up through at least ISO 1600, with many being fine up to ISO 6400. So I decided to test the upper limits of both cameras’ native ranges, and sample only ISO 6400 and ISO 12800.


Pixel Peeper’s Delight

I’m a confirmed pixel peeper. I like to see what images look like at 100% view to see how fine detail is rendered, noise characteristics, etc. With today’s high-resolution sensors, it seems like we all ought to care what’s actually happening at full resolution. If you primarily view images at 10% – 50% of original size, you probably don’t care about this kind of nit picking. But for all of my fellow pixel peepers out there, this high ISO noise test is for you!

 

Test Notes

  • For the X-T2, I used a Fujinon 23mm f/2. On the A7ii, I used a Canon 35mm f/2 IS (adapted via the Viltrox EF-NEX iv). The idea being that the effective focal lengths are more or less equivalent give the difference in sensor size. As you can see from the images below, they are not perfectly equivalent. 
  • I used 100% crops of the center of the image (the focus point). I used the exact same pixel dimensions for the crop on all images. 
  • Because of the difference in sensor size, the X-T2 images will have a deeper depth of field than the A7ii, which will make it look a bit sharper everywhere but the focal plane of the point of focus. Don’t be distracted by this. We’re looking at noise here, not overall sharpness. 
  • I did no editing to the images other than allowing Photoshop to automatically adjust the exposure. This removes my editing bias, but also means that the exposure adjustments between cameras is not perfectly consistent (X-T2 looks brighter). 
  • I left the default values on the Develop tab (click here to see the values), so there is ZERO luminance noise reduction in any of the below images. I did this to give an idea of RAW noise, not super-processed JPEG noise. I have also included the same images with the chrominance (color) noise reduction set to 0 for comparison. 
  • This is not a controlled test in lab conditions. It’s far from perfect. It’s just one guy’s attempt to show the differences between two cameras in a somewhat realistic shooting situation (unfavorable lighting). 

Comparison Images: ISO 6400

For both cameras, the exposure settings are f/8 at 1/1000 of a second. Click the images for full size. 

This comparison image is the fully default Adobe Camera RAW settings, other than having selected “Auto” for the exposure. 

Left = XT-2. Right = A7ii.

High ISO Noise Test

This comparison image is default except for changing the Color slider (chrominance noise reduction) from the default value of 25 down to 0). 

high ISO noise test

Comparison Images: ISO 12800

For both cameras, the exposure settings are f/11 at 1/1000 of a second. Click the images for full size. 

This comparison image is the fully default Adobe Camera RAW settings, other than having selected “Auto” for the exposure. 

Left = XT-2. Right = A7ii.

high iso noise test

This comparison image is default except for changing the Color slider (chrominance noise reduction) from the default value of 25 down to 0). 

high iso noise test

Comparison Images: ISO 25600

For both cameras, the exposure settings are f/11 at 1/2000 of a second. Click the images for full size. 

This comparison image is the fully default Adobe Camera RAW settings, other than having selected “Auto” for the exposure.

Left = XT-2. Right = A7ii.

high iso noise test

This comparison image is default except for changing the Color slider (chrominance noise reduction) from the default value of 25 down to 0).

High ISO Noise Test

 

Conclusions and Thoughts

To my eyes, and with these images, a few things seem clear to me:

  • The A7ii overall has less luminance noise than the X-T2 at all ISOs tested. But the difference is less than I had expected. Looks like half a stop, but that’s just my gut reaction.
  • A7ii noise seems to be more random, while X-T2 seems to be more uniform. I prefer the X-T2 noise from a qualitative perspective, but the lower amount of noise of the A7ii is also nice. 
  • The X-T2 has a significant edge when it comes to controlling chrominance noise (color). For any ISO below 6400, you can essentially disable color noise reduction on the X-T2. Less noise reduction equals more detail in the final image. However, the image quality impact of color noise reduction is fairly minimal, so it’s not as much of an advantage as it might seem. 
  • The X-T2 has a more “film” look at high ISO than the A7ii. This is consistent with Fuji’s focus on vintage film processing profiles. This is highly subjective, and you may or may not find this desirable.