Testing the ATOMOS NINJA-2 on the Panasonic GH3 May 6, 2013
As you may know, there are many options for recording on your DSLR or other similar camera using the HDMI-out (or SDI, if equipped) to an EXTERNAL recorder. Often, this is beneficial for several reasons. Probably the most important one is that you get a substantial bump-up in quality and have a larger array of more professional codecs to record to. But also you can increase your shooting time since you’re generally recording to SSD drives that can hold much more data than what you can currently buy in smaller-capacity Flash memory cards.
I recently purchased a Panasonic GH3 “mirrorless” micro four-thirds camera system to replace my aging Canon T3i. One big selling point for me was that it shoots video at an amazingly-high (for it’s class) 72mbps (the T3i shoots at about 42mbps). It can also record indefinitely, whereas other DSLR cameras tend to have a max recording time of 20 or 30 minutes.
At work, we have a Nikon 4D. This DSLR is a fantastic still camera, but since it’s video is capped at 24mbps and has a max recording time of 20 minutes (in 1080 mode), we use an ATOMOS NINJA-2 that records the 4D’s HDMI output with the professional ProRes 422 HQ codec at 220mbps. Which leads me to ask…
How would the GH3’s 72mbps recording do against the NINJA-2’s 220mbps?
Since the GH3 permits simultaneous recording full HD at 1080 resolution to the internal SD card while outputting video via it’s HDMI port (the Nikon 4D does not), I thought it would be interesting to do a side-by-side test. The following is the result.
Granted, you’re looking at this in a compressed form from Vimeo (and I apologize for the poor camera work), but when I looked at the raw on my system, I can tell you, there was very little difference. The blacks seemed richer on the NINJA-2, which is neither here-nor-there for me since it’s a personal preference that can be achieved with some minor color correcting in post on the GH3’s side. The colors seemed more vivid on the GH3, but again, that’s easily achievable on the NINJA-2’s footage in post, too. (But, I did have the GH3 set to the “vivid” profile, which in hindsight, probably was a mistake for this testing purpose. But then again, this is not a scientific test.) There was also slightly less “noise” or artifacting on the NINJA-2’s video, especially in the bricks on the building. But this wasn’t enough to make me concerned about my GH3 purchase or make me want to run out and get a NINJA-2 for myself.
My conclusion? Personally, after looking at this side-by-side test, I don’t think the added bulk, weight, and complexity to the camera is enough to persuade me to record on anything but the GH3’s internal SD card. Plus, as you can see at the 1:10 mark, I must have bumped the HDMI cable which resulted in a few frames of black that would have been disastrous on a “real” shoot if it had happened during an important take.
So, with our Nikon 4D at work where we generally shoot with a crew, this is a usable and practical configuration. But for me and my one-man shop with the GH3, I think I’ll skip it. And since the HDMI out on the GH3 is still only 8-bit 4:2:0, I see no substantial benefit to use the NINJA-2.
For those who are curious, the lens I used in this test (as pictured above) is the Lumix 12-35mm f2.8. (Yes, this is an expensive hobby.)0