Saturday, October 24, 2015

Nikon D3300 vs. Sony RX100 II - An Image Comparison

The Nikon D3300 DSLR vs. the Sony RX100 II. Which has better image quality? Can Sony's small 1" sensor really stand up to the APS-C sensor found in the Nikon?

I don't usually do these kind of side-by-side comparisons. I prefer real-world-use impressions. After all, outside of blogs and camera magazines, nobody's photographing bookshelves or playing cards or paper graphs. These controlled tests are often silly exercises.

And no matter how much care you take to ensure that everything is fair and accurate, someone will nitpick your methods. You should have done this instead of that. Whatever. I don't care. If you think you could do it better, go ahead and do so.

This may seem like an apples-to-oranges comparison. One's a DSLR, the other a pocket fixed-lens camera. I'm downsizing and simplifying my gear, and I'm replacing my D3300 with a Sony RX100 II. So I'm personally interested in the similarities and differences in image quality between these two cameras. I still have them both, so why not see how they stack up to each other?

The lens I used for the D3300 is a Nikkor 40mm AF-S DX f/2.8G Micro. This is one of the sharpest DX prime lenses that Nikon makes. I've used it to create many of my favorite photographs.

I found a scene to photograph: a blue metal R and some carnations in a mason jar sitting on the mantel above my fireplace. It's a simple scene, and not something that I'd likely photograph outside of this exercise.

I set up my tripod and placed the D3300 on it, capturing one image at ISO 100 and another at ISO 1600 (while my 20-month-old was playing underneath it). I then replaced the D3300 with the RX100 II and discovered my first issue: the Sony camera is much smaller than the Nikon. The RX100 II needed to be moved closer to the subject so that the image would be the same. Just how much closer was a guess.

And then I quickly discovered the second issue: the Sony camera, as far as I can tell, doesn't tell you the focal length in millimeters. It tells you how much you've zoomed (e.g. 1.8x), which isn't helpful to this test. The Nikkor lens has a 60mm equivalent focal length in full-frame terms. I guessed on the Sony and ended up with a 65mm equivalent focal length (which I figured out after the fact by looking at the EXIF data).

The cameras were both set in aperture-priority mode with the f-stop at f/8. I captured one image at ISO 100 and another at ISO 1600 with each camera. I let the cameras figure everything else out. The images were shot in RAW format.

Let's take a look at the originals:
Nikon D3300 at ISO 100
Sony RX100 II at ISO 100
There's some small differences in exposure and white balance, but overall the two images are very similar. This is basically an overview and we'll have to look a little closer at the photographs to really learn anything useful.

The images were captured in RAW format, which is like undeveloped film. It's the raw data from the sensor, and it needs to be processed in order to make an actual viewable photograph. I used Alien Skin's Exposure 7 software for the D3300 and was careful to be light-handed on the editing. Exposure 7 can't read the RX100 II's RAW files, so I converted to TIFF format using Phase One's Capture One Express software, and then used Exposure 7 to apply the same exact post-processing treatment that I gave the Nikon image. 

Here are 50% crops from the above images:
Nikon D3300 at ISO 100 - 50% Crop
Sony RX100 II at ISO 100 - 50% Crop
If the uncropped images were printed and you were looking at this section this big, you'd likely be viewing an 8" x 12" print (give or take, depending no the size of your monitor). The print would actually look better because your monitor, no matter how good it might be, has to down-sample the file in order to display it. In other words, you're not viewing the 50% crop at full-resolution because your monitor doesn't have the ability to display it at full resolution.

What can we gather from looking at these two crops? The Sony RX100 II's lens is sharp. Really sharp. I read prior to purchasing the camera that the lens was better than most zoom lenses, but not prime lens tack sharp. Nonsense! It's sharper than the Nikkor prime lens. The Nikon has a 4 megapixel resolution advantage and doesn't have a sharpness-stealing anti-aliasing filter, and it still can't match the sharpness of the Zeiss lens on the Sony. Amazing!

You might notice that the D3300 seems to have a small dynamic range advantage in the shadows. I think part of this may be attributed to the post-processing, and doing more rigorous editing could lighten those dark shadows up a bit.

Now let's see the 100% crops:
Nikon D3300 at ISO 100 - 100% Crop
Sony RX100 II at ISO 100 - 100% Crop
If the uncropped images were printed and you were looking at this section this big, you'd likely be viewing a 16" x 24" print (again, give or take depending on the size of your monitor). And, even though these are 100% crops, your monitor still has to down-sample the images to display them.

The takeaway from looking at these 100% crops is that at ISO 100 the cameras have nearly identical and pretty much indistinguishable image quality. The 1" sensor in the Sony camera matches the image quality of the APS-C sensor in the Nikon camera at low ISO.

But what about high-ISO? Let's take a look:
Nikon D3300 at ISO 1600
Sony RX100 II at ISO 1600
So why ISO 1600? Why not 3200 or 6400? From the little that I've used the RX100 II (I've only had it for a couple of days), it seems that ISO 1600 is the practical limit for the Sony camera. Yes, you could use ISO 3200, but there's a significant increase in noise. In real life photography I would only use that high of ISO with this camera if I were converting the images to monochrome and wanted a grainy look. The D3300's practical limit for high-ISO is 3200, and ISO 6400 is usable only for grainy-looking black-and-white photographs. There's a known one-stop high-ISO advantage for the Nikon.

You can't tell much from the uncropped images captured at ISO 1600. They look similar to each other and you might even have a hard time differentiating them from the ISO 100 photographs.

Here are the ISO 1600 50% crops:
Nikon D3300 at ISO 1600 - 50% Crop
Sony RX100 II at ISO1600 - 50% Crop
You can see that both cameras are showing some noise. Some of this could be removed in post-processing by applying noise reduction. It's not terrible in either camera, although full-frame users are probably not impressed with the results.

What might be surprising is that the RX100 II holds up pretty well against the D3300. According to DxOMark, the D3300 is the second best APS-C camera at high-ISO (just barely behind the D5500). At ISO 1600 looking at 50% crops it's difficult to tell them apart.

Now let's look at the 100% crops:
Nikon D3300 at ISO 1600 - 100% Crop
Sony RX100 II at ISO 1600 - 100% Crop 
The noise at ISO 1600 is similar in both images, but what we can notice is that the look of the noise is different. To me, the noise in the RX100 II photograph looks a little larger but more film-grain-like. The noise in the D3300 photograph looks digital. So even though the noise is just a bit more pronounced in the Sony camera, I prefer the look of it over the noise in the Nikon camera. At least in this image.

Conclusions

What are some takeaways from this?

One conclusion is that it doesn't make much of a difference if you use a camera with a 1" sensor or one with an APS-C sensor. You have to study 100% crops side-by-side to even notice any dissimilarities, and even then the differences are incredibly tiny. This is a testament to how far digital cameras have come!

Another conclusion is that the Zeiss lens that Sony put on the front of the RX100 II is pretty fantastic. I was surprised at just how much sharper it is than the 40mm Nikon prime.

The only other conclusion I have is that, perhaps, I've wasted my time. All that I've really done is capture some extraordinary boring photographs of a blue metal R and some flowers. I should have used these cameras to make some real photographs instead. That would have been a much better use of my time.

3 comments:

  1. Sheesh! The sharpness on that little beast is amazing! I like the color reproduction of it too,
    The blue "R" looks more detailed and less flat than the Nikon.

    I want it so bad but am trying to hold on for a bit longer, maybe the next version or the A7S2.

    ReplyDelete
    Replies
    1. I was surprised at the sharpness from the RX100 II lens, it's nothing short of amazing! I had to be careful not to confuse depth of field with lens sharpness, but, as you correctly pointed out, the R is where to spot the difference.

      As far as the color, it's clearly superior on the Sony images, but a little more rigorous editing of the Nikon images could make them look the same, so I wouldn't read too much into that.

      Thank you for commenting, Nuyoka Co!

      Delete
  2. I lost I'm sure natcbinv mavic could be done, I love the process but bejbf able to get those colors straight out of camera is just too awesome!

    Was looking around in Flckr to see how it handled skin or night photography but all I find is cat pics! :)

    Looking forward to more of your captures with your new toy.

    ReplyDelete