A lot of people have been asking me for comparisons between the integrated camera in the Phantom 2 Vision and a GoPro carried by an original Phantom. The images in this comparison use a GoPro HERO3+: Silver Edition.

The Phantom 2 Vision’s camera is a 14-megapixel camera that shoots at 140° at full wide. The GoPro HERO3+ Silver shoots at 10 megapixels at around 170°. Unfortunately, in this test, the Phantom carrying the GoPro was at a higher altitude than was the Vision, so the images can’t easily be directly compared. I’ll try to do a better, next time.

From these images, you can see differences in color rendition, and you can get an idea of the kind of image that comes out of each camera. In these comparisons, I’ve scaled the images to be the same size.

Both cameras do a pretty good job of exposing, even though the sun and sky are really bright at the top of the frame. Note, however, that the Vision allows for exposure control during flight, so if I had wanted to expose for the sky instead of the ground, I could have dialed in underexposure. Most of my aerial photography has been done with a GoPro HERO3 Black Edition, and there have been many instances in which I’ve wanted to dial in underexposure, especially when taking aerial footage of the ocean. Over the ocean, anything that is light-colored on the surface typically gets blown out during auto-exposure.

The GoPro shoots much warmer than does the Vision, and I frequently see cyan skies. I spend a lot of time in post trying to get the yellow/green cast out of the images and getting the sky away from cyan, but it often isn’t easy to do. Doc Brown has a great video that shows how to successfully color correct aerial GoPro images using classic controls in Photoshop, and the old-school technique works much better than does tweaking sliders in Lightroom or Aperture.

In general, the pictures from the GoPro HERO3+ Silver look sharper, even though the camera is shooting at 10 megapixels instead of 14. Taking a closer look, neither camera produces images that look great at 100%, but neither camera is really designed to produce images that hold up for pixel-peeping. Given simple subject matter, both cameras will produce images that look excellent, even in when printed in, say, a magazine spread. But the subject matter captured here is a  torture test for small-sensor cameras, and I’d go so far as to say that most aerial shots are difficult for small-sensor cameras to deal with because of the amount of detail in aerial perspectives.

Trees and foliage have a lot of detail, and you can see in these crops that both cameras obliterate detail in favor of sharpening big edges. In the tight crop above, the GoPro image (on the left) is smoother and almost looks like a painting. The Vision’s image has more noise and looks like it is sharpening using a much smaller smaller radius, which produces more noticeable artifacts.

Note that none of this is really that noticeable when you look at one of these images on a typical monitor. Sure, they don’t have the detail, pop, and color that a larger-sensor camera might be able to capture, but getting a big camera up in the air requires much more commitment, and these are impressively-good for the amount of effort spent in getting them airborne. Look at them embedded (small) on the web, and you’re always hard-pressed to tell what camera a particular image came from.

Download the original JPGs from both cameras and check out the images, yourself.

I’ll write more about actually using the Vision in a future article. The first production model I received seems to have something wrong with its radio, and I am losing connection to the aircraft numerous times in a typical flight. DJI has sent me a replacement radio, which should be here in a couple of days. Once the Vision works as it should, I’ll spend more time testing it.

The good thing about the radio problem is that I got to thoroughly test the Failsafe feature, which brings the quad home when it loses connection with the radio. Each time I lost contact with the Vision, the bright LEDs on the bottom of the aircraft turned amber, and an overlay on the FPV view in the iOS app notified me that I had lost contact. After a short period of time, Failsafe would kick in, and the Vision would fly back to its launch coordinates, hovering above me and waiting for the radio to reconnect (which it would do every once in awhile) In any case, I’m looking forward to receiving the replacement radio, and will write more about the Vision in a future article.

Special thanks to Mike Knott for piloting one of the Phantoms for the test.