Description image

More Details on the Nikon D4's Video Features, and the Clean HDMI Feed in Action

01.11.12 @ 2:35PM Tags : , , ,

Since the official announcement, more details of the Nikon D4′s video features have emerged. Given Nikon is deeming it a “multimedia DSLR,” how will its features compare to its chief competitor, the Canon 1DX (both camera are high-end DSLRs, priced at $6-7k)? First up is a video by DSLR News Shooter’s Dan Chung, who got a hands-on look from Nikon’s James Banfied:

<embed src="" type="application/x-shockwave-flash" width="400" height="300"></embed>

Nikon also answered questions as part of a Q&A at Imaging Resource, which has been summarized in video blog format by some dude sponsored by Pepsi (just kidding, I think):

That’s Nathan Pawluck from LV Photo, to be precise. In response to the question as to whether the D4 is 8-bit video, the Nikon rep had this to say:

8-bit is correct, 4:2:2 is a standard digital video format. The numbers refer to the sub-sampling of the various channels (Luminance, and two chroma channels). 4:2:2 subsamples the chroma channels 2:1 along each line vs the luminance channel. The subsampling is only along each horizontal video line; there’s no sub-sampling between video lines.

Does this mean we’re going to run into aliasing problems — “no sub-sampling between video lines” — or am I misreading that statement? While I didn’t notice any aliasing on the first D4 footage, the Canon 1DX will supposedly eliminate or at least drastically reduce the pesky aliasing problem. If the Nikon exhibits significant aliasing or moire, then they’ll be back where they started — lagging Canon in the video department. Time will tell! More details in the Imaging Resource Q&A below.

Link: Nikon D4 Hands-on Preview & Q&A – Imaging Resource

[via CheesyCam, EOSHD]


We’re all here for the same reason: to better ourselves as writers, directors, cinematographers, producers, photographers... whatever our creative pursuit. Criticism is valuable as long as it is constructive, but personal attacks are grounds for deletion; you don't have to agree with us to learn something. We’re all here to help each other, so thank you for adding to the conversation!

Description image 26 COMMENTS

  • I just noticed this: “Note that when streaming to the HDMI port, the D4 doesn’t write to the CF or XQD card slots, however.” Wonder if that means no feed to external monitoring while recording in camera, because that would be a serious hindrance.

    • Oh, I noted that before — yes. It’s an either/or situation, I think — at least with regards to external recording (I assume monitoring as well?).

      • Take the external hmdi out into a pix 240 or 220, then use the sdi or hdmi out on the recorder for your monitor. Plus there are monitors on those recorders, the 240 at least.

    • I believe that it’s only for the 4:2:2 quality, it can record to cards and monitor out from what I gathered, I think it drops to standard or 720 or something. I don’t think nikon would elminiate monitoring out while recording.

  • I’m pretty sure the sub-sampling is not referring to resolution – it’s only referring to color sampling. I honestly can’t wrap my head around what exactly they are doing with the sensor unless I had a picture of it.

    As far as resolution, the original video – downloaded from youtube in 1080p, looks very, very soft to me. It looked even softer than 5D footage. This makes me believe Nikon is doing something stupid that’s really going to limit the camera resolution-wise.

    There was talk somewhere about the camera having a modified low-pass filter for video mode – I’m not sure if this materialized or was just a rumor. Either way, I’d like to see some close-ups and wide shots uncompressed recorded from the HDMI – that will really tell us if there is real resolution situation will be like

    • Your not seriously judging the quality based on a very compressed YouTube video are you?


      • That would be seriously dumb. You need to download original files (straight from the camera) to make any kind of decent judgement.

        • I was not passing judgement completely on the camera yet – like I said I want to see some uncompressed HDMI straight out of the camera.

          Yes – you can judge resolution based on compressed web videos. Does that mean it’s a correct assessment? Maybe not, but I’ve seen enough web video to know what I’m looking for.

          But let’s not kid ourselves – take a look at Koo’s new short or much of the GH2 web footage out there and tell me even in compressed form on the web that you can’t see resolution. Just because it’s a web video does not mean I have to give it a pass – I love Nikon as a company to death – I only use Nikon manual focus AIS lenses – but if they are spending millions of dollars on advertising I would expect them to have the best shooters and encoders working on these videos, making sure they look as good as they can.

          If youtube is decreasing quality that much, why are they using it?

          Honestly I don’t believe this has MUCH to do with web compression – but if it does – shame on Nikon for not giving us proper footage that we can judge. I’ve seen plenty of RED, GH2, and even 5D footage on the web that looks far sharper.

      • Yes, you can seriously judge resolution from a compressed vimeo video, especially when there’s an obvious difference. As Joe says, look at Koo’s most recent video on vimeo or a GH2 video on vimeo then watch the Why one.

  • Wow!
    If this camera does what it says it does; Canon’s ($14,000) C300 neck better tuck in because Nikons ($6000) boot is on the prowl. Yikes!

  • @ Joe,

    Softer than the 5D? If you are referring to the ‘Why’ video, then that’s an absolutley assinine comment.

    • Well if you look at how cameras like the 5D and 7D actually produce their images, they gain an incredible amount of fake sharpness from aliasing. The downside is obviously on wide shots or fine patterned material.

      So if Nikon is getting rid of aliasing in some way, you’re removing all of that fake sharpness – and if your downsampling method in camera is sub-par – which it very well could be – you’re not going to get much higher than around 720p actual resolved lines.

      For a next generation camera – it should be using a smarter algorithm to get down to 1080p. There’s no reason why in resolved detail it’s not somewhere around 850-900 lines, if not 1000.

      When I look at that video, at full 1080 on a large 1080p monitor, it’s not very sharp – as compared to web videos of the GH2 – or RED for that matter.

      Is this a codec issue? It could be, that’s why I’d really like to see some video shot uncompressed before I make a final judgement.

    • If you’re going to use a word like asinine you should probably use it correctly and spell it right. Joe is not being foolish, the Why video, in 1080P, downloaded from the original on Vimeo looks as soft as 5D footage downloaded from Vimeo, try it for yourself. It’s definitely softer then GH2 footage. I imagine with the production value that went into Why, that they didn’t hand the final 1080P footage to a chimpanzee to throw it in a blender before uploading, so it seems that the Nikon in camera downres isn’t a progression from the three year old 5D and is in fact a regression from the GH2, at least resolution wise. No amount of lovely uncompressed 4:2:2 options will fix a poorly resolved image. The 1DX doesn’t line skip and though neither have really been tested, judging by that alone and Nikon’s best D4 video, it looks like the 1DX will stomp on it for image quality. Which is a huge shame really, Nikon almost have a real winner here and I bet it’s a beast in low light but I want something with better than 720P resolution for this price.

  • I thought the ‘Why’ project was shot on CF cards not HDMI out

  • For the price, I’d rather pick up an AF100 and a Atomos Ninja.

  • they’re definitely not talking about line-skipping, only explaining what 422 does as opposed to 420 (“The subsampling is only along each horizontal video line; there’s no sub-sampling between video lines”)

  • ASSinine (end/sacrcasm)

    Anyway – the D4, is primarily a stills camera – with heavily marketed new video features that no hdslr has yet – are there shortcomings? You bet. Everything is up for criticism, but please keep away from the blatant Canon fanboyism. How can you say the 1DX will blow it away without having seen any completely produced 1DX footage whatsoever? It’s bad enough that posers like LaForet and Bloom are totally mum about the D4 – but heck they are paid by Canon, but they are also supposedly hdslr ‘convergence’ gurus that tweet about any crap gear that comes out.

    • @agitprop. Dude, relax with the personal insults at genuinely helpful people.

      The D4 looks great on paper and I mentioned some of it’s great features but the well shot footage looks a lot like 5D resolution and not enough like a GH2, we should be progressing here, not regressing. The 1DX at least looks like it will be the best of both of those as it doesn’t line skip. Which will be better? Nobody knows, I’m guessing based on a dissapointing video and an interesting spec list as we all are.

  • “Does this mean we’re going to run into aliasing problems” [re: 4:2:2]

    What the hell dude?

    The current gen EOS HDSLRs (7D and 5D etc) do colour in 4:2:0. Most sub-$5K dedicated film cameras do 4:1:1. Each of which (4:2:0 and 4:1:1) mean 1 quarter of pixels is assigned to colour for 2D resolution (more rods than cones in your retina). Saves space. Better cameras offer 4:2:2 (half pixels for colour sampling) – but none are going to deliver a quarter and half streams simultaneously. So yeah – “either/or”. Flash cards are generally not fast enough to record the extra data in 4:2:2 (unless horrid compression is involved). The canon XF1xx series and C300 gets it due to dual CF slots. Generally, with cameras like the AF100, you need to get your 4:2:2 out of HD-SDI to a device that is capable of handling the bitrate (Canon uses 50MBps for 4:2:2).

    If the D4 is brining 4:2:2, then rejoice and be very very glad indeed and go get an Atomos Ninja to monitor/record. That’s a canon crushing leap forward for the humble HDSLR. Cameras like the 5D and 7D export a giant red dot with their image so that recording via HDMI is not possible … ARGGGHHH… Canon suck dogs balls.

    • Flash Cards are not fast enough? Most modern cards do 40-50 MB/s, some are even faster. That is MegaBYTES, not Mbits. I think with a Datarate of 160 Mbit/s (= 20 MB/s) you should be able to store 4:2:2 information nicely without horrid compression artefacts…

  • Could be some usefull info:

Description image1 pingbacks