Blackmagic Forum

What is the advantage of using a 6K or 8K camera over a 4K camera?

Signage perhaps, VFX and the ability to reframe for 4K/HD. Although for over a 100 years the great cinematographers were able to frame beautifully, properly, on the shoot. Pixels are ultimately marketing IMO, since the average domestic viewer has difficulty distinguishing often between SD and HD, in the average front room, let alone 4K. 100% of the UK broadcast I work in is HD still, and even though the defacto documentary camera that they use is the 4K FS7, they almost always shoot in HD too. TV consumer manufacturers need you to buy a new telly every few years, hopefully every year and the race for pixels (or HDR, that we never missed), even though there is precious little content, if any, at 8K, is the current best way.

When I had £10k SD lenses from Canon and HD came in, I was categorically told that they would never resolve HD, even though they were just glass and now existing lenses are meant to work fine with BMD's 12K Ursa. Off the record a Canon engineer told me that was BS and I could see it was. Marketing needs you to believe that what you have today was only good for yesterday and you need to spend for tomorrow now - or be left behind to starve.

All just my opinion, of course.

For me, using such a camera would actually be a disadvantage.

I'd have to buy expensive new lenses and media.

The P4K works with what I already owned.

Mike1938 wrote:What is the advantage of using a 6K or 8K camera over a 4K camera?

With a more or less standard de-bayer algorithm.

A 4k bayer-patern sensor will give you a 2.8k output (what you see on the screen).
A 6k ~ 4k
An 8k ~ 5.5k
and the 12k ~ 8k

Mike1938 wrote:What is the advantage of using a 6K or 8K camera over a 4K camera?

IMO, not much - more of a marketing item. A better advantage is a camera having high dynamic range (13 stops or higher - the higher the better), global shutter, and support for RAW codecs. Also, 90% of big screen and broadcast movies are still delivered in 1080p or 2K due to investment in projectors and lower streaming bandwidth respectively.

To me “K” is a completely meaningless metric when discussing camera footage.

I regularly have to work with drone footage that was shot in “4K”, but it might as well be SD footage as far as I’m concerned, since the image is so terrible because of the quality of the glass combined with Long-GOP compression.

That said, when using a “real” digital cinema camera, shooting at high resolutions is great for reframing (assuming you’re shooting through decent lens), and also for downscaling to standard delivery resolutions like 1080p or UHD.

MishaEngel wrote:

Mike1938 wrote:What is the advantage of using a 6K or 8K camera over a 4K camera?

With a more or less standard de-bayer algorithm.

A 4k bayer-patern sensor will give you a 2.8k output (what you see on the screen).
A 6k ~ 4k
An 8k ~ 5.5k
and the 12k ~ 8k

this pretty much, given that all the other variables are bout the same.

And have a look at what Steve Yedlin has to say about resolution.

Uli Plank wrote:And have a look at what Steve Yedlin has to say about resolution.

"A Clear Look at the Issue of Resolution":
https://ascmag.com/articles/a-clear-loo ... resolution

ResDemo:
http://yedlin.net/ResDemo/

"Steve Yedlin Blows the Lid off of Resolution Myths":
https://nofilmschool.com/2017/08/yedlin ... tion-myths

The TL;dr version: "it's very hard to tell the difference between 2K/4K/6K, even on a 20' projected screen."

If it were up to me, I would tell the client (all things being equal) to shoot in 5K or 6K, then crop to 4K for final release. That way, we have a 4K frame with room on all four sides to reframe if needed in post for the final deliverables. Noted director David Fincher has done this on a lot of his feature & TV projects, and I think it makes a lot of sense:

Mike1938 wrote:What is the advantage of using a 6K or 8K camera over a 4K camera?

You can shoot hand held, crop, and use the extra image area surrounding the crop for stabilization.

This has been done very effectively with the BMD 12K camera - shoot 12K for 8K.

Don't expect too much from stabilization in post. Short bumps will become visible as a sudden blur.
Better fight the problem at the source by mechanical means.

For me those Steve Yedlin videos are pretty conclusive and kudos to him for taking the effort to demonstrate the marketing con. 8K makes no sense whatsoever as a target shooting or mastering format for theatrical films and television, and even less sense for a domestic TV. The limits we have today, with the tools at hand, are not technical but often self-imposed stylistic ones.

Just about 95% of all films today seem to be shot low key and with very shallow depth of field - 70% of the image is out of focus and dark, yet people still obsess about resolution and HDR. A Technicolor VistaVision film like Vertigo, can still look just as vibrant, saturated and sharp on a modern TV, as it did in the cinema (probably better) so it's not a technical thing, in a purely perceptual way.

The demonstration also gives clearly, I think, the lie to all those who held and still do that digital cinema would never match 35mm film for resolution.

Yedlin shows what is possible today but hopefully AI and ML, in the future, will improve scaling algorithms beyond current recognition. We just need the very best of pixels not more of them.

Uli Plank wrote:Don't expect too much from stabilization in post.

Some might disagree.

Mike1938 wrote:What is the advantage of using a 6K or 8K camera over a 4K camera?

If you shoot, say, 6K, confine your subject to the center 4K, and crop to 4K in the edit, you then have a border that could be helpful for stabilizing. And you can recompose if, for example, you see that the center 4K crops something awkwardly and you want to recompose, or even pan-n-scan. There are also things like viewer-focus matching -- I'm not sure what the official name is, it's not eyeline-matching -- where you cut from one shot to another and want the viewer's main focus in the second shot to match where they are focused in the first shot. That would generally involve repositioning the second shot, and if you have extra video room you might not need to scale the image to cover the repositioning.

(Depends on lots of circumstances as to how much scaling is acceptable to you. People scale images a bit to account for mild stabilization, or to adjust eye lines and it looks pretty good.)

If you shoot, say 6K, with the intent to use the full frame, you can still do things like the pseudo second camera ("Scale Up" feature to scale up 50%) to make a jump cut less jarring. And I think it could be helpful for compositing/keying purposes where in general more detail and more bits is better. (I'm pretty sure it's not as beneficial in keying as I might naively assume, but I think it does have some positive impact, unless you're recording fewer bits to try to have your 6K use only as much bandwidth as a higher-quality 4K, which would be bad.)

Then there's the possibility of "future proofing". If you think that 8K TV might become a standard and you think your video should live large in that glorious world, shooting in 8K would be nice, and even shooting in 6K means a 33% scale-up rather than a 100% scale-up. Of course, in that glorious (future) world, machine-learning algorithms may allow you to scale your 4K up to 8K with pretty much as much quality.

In addition to stabilization, you can also un-stabilize, i.e., add camera shake.

Noel Sterrett wrote:In addition to stabilization, you can also un-stabilize, i.e., add camera shake.

That's an expensive way just to be able to do stabilization in post.

Marc Wielage wrote:"it's very hard to tell the difference between 2K/4K/6K, even on a 20' projected screen."

In my home theater, the most prominent 'improvement' of 4K is HDR.

Ellory Yu wrote:That's an expensive way just to be able to do stabilization in post.

Let them eat Steadicams.

Ellory Yu wrote:

Noel Sterrett wrote:In addition to stabilization, you can also un-stabilize, i.e., add camera shake.

That's an expensive way just to be able to do stabilization in post.

very common for VFX in greenscreen where it is much easier to do the compositing and add the camera shake after.

Noel Sterrett wrote:

Uli Plank wrote:Don't expect too much from stabilization in post.

Some might disagree.

First of all, this kind of almost static shot should always be done on sticks. Then there's not much of a need for stabilization in post.

Ellory Yu wrote:First of all, this kind of almost static shot should always be done on sticks.

The handheld shot was done as a proof of concept. It did.

Noel Sterrett wrote:In addition to stabilization, you can also un-stabilize, i.e., add camera shake.

Yes, that's a big Fincher trick. He'll shoot a conversation between two people completely static, on 6K cameras, and then introduce all the shake and "hand-held" vibration as a specific reposition in post. The advantage to that is that the bobs & weaves are all carefully planned and happen specifically on word cues, actor reaction, or relative position. Trying to do that on set would be maddening and could easily go up into 50-60-70 takes before you got one that was absolutely perfect; if the cameras are locked down, you only have to worry about performance and focus. I think that's a very clever use of technology in a creative way.

I had a show a few weeks ago where they had done a pickup on the final scene, and we had shaky-shot1, shaky-shot2, static shot, shaky-shot3, and shaky-shot4. I turned to the director and said, "hey, how about we shake up that third shot to match the others." He was reluctant, but I did a quick find and he was really pleased, admitting the added motion helped disguise the fact that it was a pickup done weeks later (shot by a different DP and different crew).

same here, in a truck chase sequence, at one point one inbetween shot was a bit too static and we had enough side material to shake it and not zoom in... (this just a week ago or so...)