Review or Mini-Review DJI Phantom 3 vs. GoPro Hero4 Still Picture shootout!

And the winner is.....well, you decide for yourself.
To my eyes the Phantom 3 pictures are superior.
I hooked the GoPro4 Silver to the base of the Phantom and set it on 12MP and interval (2 sec) shooting. I took 8 pictures from each of the P3 and the GP cameras which represent various views and lighting.
No post processing was done. No special settings were used other than a few small bumps in exposure on the DJI as I was taking the shots.

NOTE: The Phantom 3 camera is on a stable gimbal while the GoPro is hanging off the bottom of the quadcopter. This could make a small difference in the overall stability (and therefore quality) of the pictures.


Phantom with GoPro4 on bottom

Here are the photos direct from each camera - and/or a smaller (2000 pixels across) rendition of each.
The entire file of full sized is 50 MB and is at this link (photos are named correctly so you can tell which cam they came from).
https://www.droneflyers.com/images/shootout.zip
The entire zip file of downsized pics is 20 MB - photos are named the same as the full sized, so make sure you keep them in a different directory!
https://www.droneflyers.com/images/shootout2000.zip

Here are some examples of enlargements - click to enlarge

This is from the DJI (above)


This is from the GP4 Silver



This is from the DJI P3


This is a similar view with the GP4 Silver

Conclusion - make your own! To my eyes the GoPro still suffers from some "watercolor" processing which makes a blow-up of any photo look like water was spilled on a photography.
The colors of the P3 seem to pop more and are more accurate - less faded and blurry.

If we used a scale of 1 to 10 for for sport and aerial cameras with a Mobius (popular sports cam) at about a 4, the GoPro4 Silver would rate about a 6.5 (a 6 in this test, but adding a 1/2 point because it was not on a gimbal) while the Phantom 3 would make it to 8. There is always room for improvement - however, in terms of a lightweight aerial camera the Phantom 3 delivers quality which is good enough for more hobby and light commercial applications. To my eyes the GoPro4 does not seem as pleasing, but these types of opinions are best left to the eye of the person viewing.

Links:
Phantom 3 Advanced - http://click.dji.com/1yk3KUe9-_sqNSln
Phantom 3 Pro - http://click.dji.com/1yk3KUe9-_sqNSln
GoPro 4 Silver - http://amzn.to/1bB1c6z

 

One reader noted that perhaps the GoPro samples were affected by vibrations - my answer and samples below:

There is certainly a difference in the stabilization - but most of the photos are flying very lows and in no wind. Also, the shutter speed of the GoPro is very fast and I took photos every 2 seconds and selected the best of them.
Also, the GP in this test is very isolated from the main vibrations since it is hung off a plastic bottom protector (corrugated) and then off a rubber GoPro brand sticky mount, etc.
Just to confirm that part I will take some of those same angles from the ground with the GoPro and test them.

So here are the photos taken with the GoPro from the ground - fully stable!
Note - a couple hours have gone by so the light is a little different.

Also, it's now a little unfair to the Phantom 3 - because we are comparing it's flying camera to a ground based model.

I see a slight improvement in the GoPro detail from the same pic flying - however, the picture seems quite dull compared to the same Phantom screenshot.

Another reader stated that the GoPro exposure could have been set differently - however, this test used both cameras in their stock and default modes...with the the GoPro it is not easy in terms of exposure and other settings changes, especially when flying on a quadcopter.

Another point which should be mentioned is that the GoPro is taking a much wider shot - so having more real estate in the shot makes for fewer pixels in a similar sized object like the houses we are using for the example. Those who want a very wide field of view will prefer the GoPro - or will have to stitch together a panorama in the Phantom 3.

Here is a comparison on one pic...click to enlarge.


Here is a comparison of a bush and some shade and sun.

 

Here is a video comparison - and also a still image.
http://www.provideocoalition.com/dji-cams

As noted, the GP4 Black is probably equal in overall video quality to the P3 and Inspire. It becomes more a matter of whether one needs the wide angles and fish-eye (or wants to remove them in post production).

For still photography I prefer the P3 Inspire due to the lens and the ease of control. The P3 camera allows easy change of the focusing and metering which is very difficult with the GoPro. Perhaps this will change when GoPro releases their new interface that allows external control (with the Solo, etc.).

 

Yep, you can definitely crop and straighten, etc. and for most use (online, etc.) it hardly makes any difference.
Vimeo, for example, defaults to 720p - so you could shoot in 1080 and crop, straighten, etc. plenty and still have good output.

People don't watch online video (in general) in way where little burps in quality are noticed.

Everyone has their prefs - some folks say to shoot in 4K and then when you compress it down to 1080 it will be better.

My current pref on the quads is to shoot in 60fps due to the motion - so I am using 1080 60. It's more than good enough for anything I have to output.

Video uses many "magic tricks" anyway - the compression is done using lots of fancy mathematics, etc.
For example, a compression program may determine "hey, those 7 frames only differed by 10% in what was showing, so I'll take the other 90% as one frame (x7) and only change the 10% that moved".

GoPro studio doesn't have any idea what your final output screen is or what you are even editing on. If you worked at NFL films you'd have all these fancy monitors around you so that you would know if you've gone too far......

Actually, I toured NFL Films a couple times and their famous editors worked on footage which was relatively low res. After they finished the main processing computers would then use all the decisions (timelines) made by the editor but grab the original hi-res from the central servers and put together the Final Cut.

We peons don't have to worry about such things - since our output is all going on relatively low-res screens.