Review: Atomos Ninja2

For my short film Over Fjellet I had the pleasure of testing out Atomos Ninja2. If you’re not familiar with it; the Atomos Ninja is a small HDMI video recorder. It records in Apple ProRes files onto a harddrive or SSD.

By sending an uncompressed* HDMI signal out of the camera, in my case Nikon D800*, I could record it as Apple ProRes 4:2:2 files on the Ninja. This is much better quality than the built-in recorder gives me (h.264, 25mbit/s). For this project the main advantage of the higher bitrate and color sampling comes in post production when it comes to color correction and grading. For others it will be important when it comes to incorporating CGI and special effects.

All you need to get started, beside the Ninja-kit, is a HDMI-cable and a 2,5″ storage drive. The Ninja can use both regular spinning hard drives (HDD) and solid state drives (SSD) I used a regular spinning hard drive and this worked perfectly. Though I did not have any intense camera-movement or bring the camera on a bumpy car ride or anything. For that you would probably be better of using a SSD. As for the HDMI-cable, I picked up a cheap one at a local electronics store. Only requirement is that it is a mini-HDMI (Type-C) to regular (Type-A) cable as the camera has a mini-HDMI connector and the Ninja uses a regular.

The Ninja is really a joy to use! It’s just plug and play and it’s so intuitive and simple. It took me no more than 5 minutes to get comfortable with it, and know every function there is. This simplicity is truly one of it’s strengths. It doesn’t have mile-long menus and a lot of extra functions you never use, and you don’t miss it. You control everything through the touch-screen. I’m no fan of touch screens actually, but for the Ninja it really works. They’ve built a user interface from scratch and made it specifically for touch screen and not just converted a regular one like most cameras do. Because of this it’s really fun to use the Ninja and it’s quick to navigate.

In addition to being a recorder, the Ninja also functions as a monitor. It’s not really a high-quality monitor, but it works just fine. The main advantage is that it has some important functions most DSLRs lack, like zebra-stripes and focus peaking. You can also switch between black and white and color. It can also double as a monitor for the camera assistant when pulling focus.

I didn’t really miss anything with the Ninja. It has a good system of naming clips where you set scene and shot and it automatically adds number of take. I don’t think you can give the clips names of your own though. While there were sufficient with accessories in the kit I did miss some sort of mounting-device for attaching the Ninja to the camera. A simple cold-shoe mount would’ve been enough. Also I wouldn’t mind a better build quality. It’s plastic and got a few bumps after only two days of regular use. On the plus, side the battery capacity is really impressive. Handled an 8 hour day with intense use  with no problems at all!

All in all the Atomos Ninja2 really impressed me! It’s powerful, and really useful when shooting on a DSLR, yet it’s just so simple, playful and really fun to use.

*Well it’s a 4:2:2 signal and not a 4:4:4 raw RGB signal. It’s labeled uncompressed because in the TV-world 4:2:2 color sampling is considered uncompressed as it has no visual quality-loss.

*One note on setting up Nikon D800: During set-up I found it a bit illogical and difficult to set up the D800 to output a clean, high-quality HDMI-feed. However, if you just follow the steps described on the Atomos website it should work just fine!

First DSLR video shoot!

Believe it or not, but today was my first ever DSLR video shoot! That is, I’ve used it briefly as part of a project mainly shot on regular video camera a couple of years ago, and I’ve worked on productions shot on DSLRs, but I’ve never shot a production entirely on a DSLR as a DP before.

That being said, this was a school assignment and not an entire short film. It was a single scene we wrote and shot, as a collaborative project between our TV-production class and the drama class. It was a simple and quick production with less focus on technical qualities and more on storytelling and acting which is why I thought it would be a perfect production to test out DSLR-shooting and put the video mode of my D800 through some practical testing.

While I’ve recognized the potential of DSLR video, I’ve remained skeptical. There are several reasons, mainly quality and practical use. The bit rate is usually low, image quality questionable (image quality is not the same as depth of field…) It’s hard to focus, lack of dedicated buttons and lack of proper audio functions for some. I’ve only seen the material from today’s shoot on the camera LCD and a laptop, but not a proper monitor yet, so can’t really say much about the quality in this post. However, I can share with you some of my practical experiences.

First of all let’s talk equipment. I had my D800, a range of Nikon lenses: 17-35mm 2.8, 50mm 1.8 D, 85mm 1.8 D, 105mm 2.8 D micro. A Zacuto 3x viewfinder, Sevenoak shoulder rig and of course a regular video tripod and a standard light kit of a couple of 800w oranges and some LitePanels. In addition we were planning on using an Aja Ki-Pro recorder to record a 4:2:2 output and then use the SDI-out from the Ki-Pro to get a proper field monitor. unfortunately the Ki-Pro was in use on an other production and we had to do heavily compressed in-camera recording in steed.

I actually used all of the lenses. My favorite was by far the 85mm. Not only the sharpest of the bunch, but also the easiest one to pull focus with. However, none of them are particularly good for video. Then again, they’re stills photo lenses and not film or video lenses. Luckily it was a fairly simple set up. Most of the shots were tripod mounted and static with some simple focus work. On a video camera it would’ve been easy, but it was much harder on the DSLR. To make it easier for my self I worked on a 4-5.6 aperture.

I have some mixed feelings about the Zacuto viewfinder, though. It made it a lot easier to see focus on the LCD screen and actually made it possible to somewhat finetune focus, but I found it clumsy and bulky to work with. However, the worst part was that it was hard to see the entire frame. I could only focus on one part at the time and did not get a good overview of the entire frame. This made it actually quite hard to work with and I felt I lost some control. I used the 3x magnification, perhaps 2,5x would’ve been enough.

I must admit it was fun to work with the D800. I ran the camera on full manual mode. This gave me control of shutter, aperture, ISO, and the works. In addition I put the white balance on manual and then I felt I had control over the camera in much the same was as a video camera. Of course I missed the zebra or a waveform monitor in the camera, but I had the histogram which gave me some control over the exposure. I did not fool around with any of the audio options on todays shoot, though. So, I have no practical info about that. We recorded everything with an external sound recorder and will be syncing it later. I’m also a bit disappointed I did not get to try the 4:2:2 output via HDMI. Hence, I look at this as more of a testing of the practical issues related to DSLR video shooting as quality probably would’ve been much better with a high quality recording.

I’ve not yet converted to the DSLR video religion and don’t think I will either. Don’t get me wrong, I really liked shooting DSLR for this project and hope to do more of it in the future. That being said I will not shoot DSLR on every single project just because “oh my God, it’s so awesome, it looks like film!”. It requires more than just the camera to make the video look good and with a DSLR perhaps even more so, because of the limitations when it comes to focus and bit rate etc. However, on certain projects that require that particular look and where I have the time and resources required, I will definitely consider DSLR as a real alternative!

EBU test of Nikon D800 and D4

Just accidentally stumbled upon this. It’s an assessment by the European Broadcasting Union, the EBU, of the video capabilities of Nikon’s D800 and D4 cameras.

Check the links below for links for the PDF-files:

EBU test of Nikon D800
EBU test of Nikon D4

To sum up the tests; The D4 comes out a little better than the D800. The D4 is said to give 13 stops exposure range as to 12 on the D800. Both test best at 1080p recording and are deamed “not good” at 720p. They conclude that both can be used on a ISO value up to 6400 with reasonable results. Rolling shutter is at a minimum on both cameras and both cameras have good color performance.

However, the D4 performs better than the D800 as the conclusion for the D800 reads:”This camera cannot be recommended for serious programme-making.”

The EBU also tried to do a test on a Canon 5D. Link below for PDF:

EBU test of Canon 5D
However, this was a pre-production model which obviously lacked a low-pass filter. It failed the first test horribly and was not tested further.

Nikon D800 or Canon 5D Mark III?

That’s a question a lot of people is asking these days. Finally both cameras are out and people are testing them. I have tested the D800 a little bit. I’ve even done some quick and rough testing of the video function. The 5D however I have not yet tested much.

Never the less, I’ve been asking my self that very question and have plenty of thoughts on the subject. I’ve checked the pros and cons and decided what camera I’m going to buy. With emphasis on the I. While I’ve figured out what camera is right for me, this does not mean it’s the better camera of the two. It’s simply the best for me. Anyway,I thought I’d share my thoughts on the matter. I’ll present some facts and specifications and other reasons I’ve taken into account. If you want to just read my conclusion, scroll to the end.

Why I’m upgrading

To start I’ll explain my reason for upgrading to a new camera. I have a Nikon D90 today. A good camera, that’s served me well. However, it’s getting older and I’m ready for an upgrade. I feel that when first upgrading I don’t want a newer camera in the same range (say D300s, D7000 or 7D) I want a full frame sensor camera. An important reason being less noise on higher ISO. I like to do late night photography, and while I usually use a tripod some times I would still like to go a little further on the ISO. I also generally do a lot of hand-held and being able to go higher on the ISO will let me shoot hand-held in harder conditions. Another reason for going pro full frame is the weather seal. I’ve used my D90 on windy beaches and in rainstorms. Quite frankly I’m impressed it still works. That being said, I’m upgrading my photo camera! I’m buying a camera to take stills primarily. I’ve used the video function on the D90 once. It’s crap. End of story. But new cameras have come a long way since then, so when I’m upgrading I do want a camera with good video capabilities.

Why not other brands?

Some may ask the question why Nikon and Canon? Why not Panasonic, Sony, Pentax or others? Well first off because of the marked situation in Norway. Nikon and Canon control the DSLR marked. While Sony is gaining some percent, Nikon and Canon are the two major brands. Second; they are the leading brands internationally. They’ve been doing this for a long time with lots of experience. Third; the lenses. They produce quality lenses and have a large lens collection available. In addition there are a large number of third-party lenses available. And we all know how important the lenses are!

Video

Ok, so that’s my background. What to choose? We’ll start off with the video. On paper Canon’s all-I codec at almost 100Mbit/s is much better than Nikon’s 24 Mbit/s P-frame based codec. However, I haven’t seen any tests yet. That said, Nikon’s H.264 24Mbit/s codec should match that of AVCHD, maybe even better since it’s based more on P-frames and less B-frames. Meaning it should be a decent in camera codec. Not to mention Nikon offers an 8 bit, 4:2:2 1080i (or 720p) output via the HDMI, so for those heavy projects you could get an external recorder and blow Canon’s all-I out of the park. But, it’s not just the codecs and outputs that’s important. Down scaling is another matter. I have not seen any comparison between Canon and Nikon here, so it’s really hard to say anything, but poor down scaling could give a less sharp image. This is a bigger issue for the D800 than the Mark III due to the higher resolution. Another common problem on DSLR is rolling shutter etc. I did a quick test on the D800; in a dark room I fired a flash. ideal this should light the whole frame. I shot in 720p 50 mode with a shutter of 1/50 so standard video. The flash fired at high speed and it lit the whole frame! This should mean rolling shutter is drastically improved and not a big problem in regular conditions! (with a faster shutter it may become a problem, I don’t know) I can only assume it’s the same or better for Canon. On one note however, D800 do have full autofocus on their video, the 5D does only have semi autofocus. If any of them are any good however I don’t know.

So, both cameras have good video. I would assume (because of the history) that the Mark III may be better. SMPTE time code and all-I codec is ideal if you do documentary. You can sync a sound-recorder and get good quality video straight out of the camera. However D800’s uncompressed output (What? It’s not 10-12 bit 4:4:4 like RED so why do you call it uncompressed?! Because professionally (8)10 bit 4:2:2 is the SDI and HDSDI standard and considered uncompressed in studio and TV production settings) is a great if you want the best quality and do mostly controlled productions like films and commercials and can use an external recorder. It’s also great if you want to use DSLRs in a multi-camera production.

Stills photography

Now, over to the still photography abilities of these cameras. After all they are photo cameras and they’re made to capture still images and not video. I’m not going to go into every detail of the specifications here. Both of these are good cameras, and the specifications for both are good! However, most of the people I’ve talked to have not been impressed with the 5D Mark III. There are no real big upgrades like on the D800. It’s pretty much just what the Mark II should have been. An updated autofocus-system seems to be what everyone is talking about. It also has new ergonomics and improved building quality and of course it has improved ISO-range and less noise so it’s closing in or passing the D700 there. However, according to most tests is does not beat the D800. According to DxOMark the D800 is by far the better camera. It has a dynamic range superior to just about any camera, and if you downscale the ISO performance is better than that of the Mark III. That said the Mark III can be pushed further than the D800, right up to 25600 (expandable to 102400) while the D800 stops at 6400 (expandable up to 25600).

A deal breaker for many is the fact that the D800 is a slower camera than the Mark III. The D800 only delivers around 4 fps while the Mark III can push 6 fps. This may be a deal breaker especially for sports and maybe wedding photographers. When it comes to battery capacity the Mark III (uses the same battery as the Mark II and 7D) is said to have improved battery capacity where as the D800 (same battery as the D7000. Why, Nikon?) is said to have less capacity than the D700 and less than the Mark III.

The elephant in the room is of course the D800’s 36Mpx resolution. Bordering on medium format resolution it’s great for studio use, but it gives you RAW-files at 70MB. It also requires more of your computer and more of you lenses. I’m not happy about it and would much rather have a 22Mpx sensor like the Mark III. I simply just don’t need 36Mpx and many others probably feel the same.

Nikon D800 is probably better in a studio than Canon Mark III due to its resolution. For landscape the D800 is the obvious choise due to the dynamic range. For action and news photography the Canon’s increased ISO range and quicker shooting rate may be the better choise.

Conclusion:

So to sum up. This is how I choose what camera I’ll be getting.

1: Video is good enough: Both have decent video. However, the 8 bit 4:2:2 is more appealing to me for larger projects than the all-I codec. The in camera 24Mbit/s codec is OK for smaller projects, and the quality is much better than that of 5D Mark II. From what I’ve seen the video is good on the D800. The Mark III may have better video, but the D800’s video is still good and good enough.

2: D800 is the better photo camera: As I said, I’m upgrading my photo camera. Nikon’s D800 is the better camera and few seems to be arguing against it. It has a sensor superior to just about any cameras today. I don’t do much sport or action photography so I don’t need the extra speed the 5D offers, nor do I need the increased ISO range. Yes, I’m upgrading to get more ISO range, but I don’t need the range of the Mark III. I’m however very interested in the dynamic range. I do a lot of nature photography and don’t use a lot of flash or lighting. I love to explore natural light and the dynamic range of this camera is a great for my use.

3: Pricing and economics: Today I am a Nikon user, hence I have Nikon accessories and lenses. While I could use Nikon lenses on a Canon body, it would require me getting an adapter. Nikon also offers the DX crop mode. Meaning I can use my old DX lenses and also my FX lenses becomes more versatile than the Canon counterpart. Not to mention that the Nikon D800 itself is cheaper than the Canon Mark III.

There is also the fact that I’m used to working with Nikon and like their cameras, though at one point I was sure I would get a Canon camera this time. Anyway, these are of course the things and reasons I’ve been focusing on when I’ve chosen what camera to get. Because these are the right reasons for me, they may not be the right reasons for you. Only you can answer what is the best camera for you and your use.

Happy Halloween!

We were carving out a pumpkin the other day and I found a lot of cool stuff I wanted to shoot, so I did and edited it into a little halloween video.

Now, I shot this using a Nikon D90. While it’s a great DSLR that takes excellent photos it’s video quality is absolutley crap! End of story. The rolling shutter is a big issue, there’s a total lack of manual control, and there are no sound control at all. As always you can work around it, but it’s frustrating.

The timelapse in the beginning was more of a test. I want to do some more timelapse shoots later. So this one was a quick set up, just to try out the camera and software. I’ll come back with more details when I do a proper timelapse shoot.

This is the second time I use it to shoot video. The first time to test it, this time because it was the only camera I had at hand. This not being a serious project I figured the quality wasn’t that important.

But I lacked more than a proper camera for this little pet project. I usually edit on Final Cut Pro, but I didn’t have access at the moment, so I ended up editing on iMovie. I’ve never really like the “simple” editing softwares. I actually think they’re much harder to use, believe it or not.

So a Happy Halloween to you all! So far we’ve had less than ten kids trick or treating at our door and we’ve started to eat the rest of the candy ourselves.