Page 1 of 1
4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 12:16 pm
by Allen Rice
I am trying to decide between the 4k and 4.6k camera and realize I don't really know the difference. I can't find example videos on youtube.
maybe this is a dumb question, but can people even play back 4.6k, since even expensive TV's only go up to 4k. Does anyone know of examples I can check out to see for myself or have any opinions?
thanks a lot.
4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 12:25 pm
by adamroberts
The usefulness of 4.6K is not as a delivery format but rather as an acquisition format.
If you shoot 4.6k and then downscale that in post to your delivery standard, say UHD, you are then oversampling and the result is a sharper image with more detail and more colour info.
You also have more resolution to work with if you need to do stabilisation in post or need to do some re-framing to better suit your edit.
Most people don't "need" 4K or even deliver 4K but they shoot 4K for the same reasons above.
Re: 4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 12:45 pm
by Tom
There are other differences in the two sensors which are worth considering also,
The 4.6k is said to be:
- Better in low light
- Have a higher dynamic range
- Switchable between rolling and global shutter mode
- slightly larger (not really a massive amount but it matters to some I guess)
The larger resolution means that you can either oversample for better resolving power or have the ability to post crop and stabilize without having to drop below a 4k or UHD resolution.
Re: 4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 1:24 pm
by Allen Rice
That makes perfect sense.
Do you guys have an opinion on how much of a difference it makes? I know that is an abstract question, but I have no examples to go on and I am curious what I can base my decision on.
Re: 4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 1:34 pm
by Jules Bushell
Allen Rice wrote:That makes perfect sense.
Do you guys have an opinion on how much of a difference it makes? I know that is an abstract question, but I have no examples to go on and I am curious what I can base my decision on.
You have to also remember it's not really 4.6k or 4k, its two greens for every red and blue. And you don't get any more info from interpolating. So at least with 4.6k you're getting closer to a true 4k more like 3k though.
Jules
Re: 4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 1:45 pm
by Tom
Jules Bushell wrote:Allen Rice wrote:That makes perfect sense.
Do you guys have an opinion on how much of a difference it makes? I know that is an abstract question, but I have no examples to go on and I am curious what I can base my decision on.
You have to also remember it's not really 4.6k or 4k, its two greens for every red and blue. And you don't get any more info from interpolating. So at least with 4.6k you're getting closer to a true 4k more like 3k though.
Jules
Let's not confuse resolution with resolving power.
With a 4096 x 2160 resolution sensor you get 4096 x 2160 photosites worth of bayer data which after processing results in a 4096 x 2160 RGB image.
You DO get more info from interpolating, by its very nature it is generating info by approximation. It's just that you get less accurate info than if you over sampled so that each final 4096 x 2160 pixel was derived from a greater sample set so that each final pixel got its own Red, Green and Blue photosite.
I know what you meant by your post and you are of course correct in principal - but if we are going to get technical - it is important not to confuse matters.
Will a 4096 x 2160 sensor resolve 2160 lines? no. Not in colour, not in black and white, bayer or not. This is a separate issue to bayer interpolation as a means of colour capture.
Allen - in order to "resolve" a given resolution, ie: capture detail of a certain size - you must Over sample the image. So a 4k camera doesn't resolve 4K worth of detail. This is going into quite a deep and complicated topic of which there is plenty of documentation online if you are interested.
This need to over sample applies even more for colour detail because of the way a bayer pattern sensor works. Each individual photosite of the 4k sensor only ever captures either Red, Green or blue data for that section. The remaining 2 colours are calculated based in the photosites around this particular photosite. There are not an equal number of red, green and blue sites, there are x2 green vs x1 red and x1 blue (we are more sensitive to green shades than any other so it makes sense to give this colour a bias). So if you wanted to ensure that you captured a colour image which did not use approximation or interpolation -you would need to make sure that the capture resolution was high enough vs the finished resolution so that there were all 3 photosites providing data to the finished pixel....... I am sure this will be worded better online somewhere. Look up debayering and over sampling.
Re: 4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 1:50 pm
by brent k
The 4.6k sensor is from Fairchild, and has 15 stops of dynamic range. The 4k sensor is from CMOSIS, and has 12 stops of dynamic range. Just on that alone, the 4.6k camera is far, and away superior to the 4k. The difference from 12 stops to 15 stops is massive.

- dynamic range.jpg (21.02 KiB) Viewed 10139 times
Re: 4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 1:59 pm
by Tom
brent k wrote:The 4.6k sensor is from Fairchild, and has 15 stops of dynamic range. The 4k sensor is from CMOSIS, and has 12 stops of dynamic range. Just on that alone, the 4.6k camera is far, and away superior to the 4k. The difference from 12 stops to 15 stops is massive.
dynamic range.jpg
Is that image from the Dolby Vision film? Saw it at IBC last year - it was way more impressive than the crazy 8K OLED curved displays
Re: 4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 2:24 pm
by Allen Rice
that's so much, this was all incredibly helpful.
Re: 4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 5:25 pm
by brent k
Tom wrote:brent k wrote:The 4.6k sensor is from Fairchild, and has 15 stops of dynamic range. The 4k sensor is from CMOSIS, and has 12 stops of dynamic range. Just on that alone, the 4.6k camera is far, and away superior to the 4k. The difference from 12 stops to 15 stops is massive.
dynamic range.jpg
Is that image from the Dolby Vision film? Saw it at IBC last year - it was way more impressive than the crazy 8K OLED curved displays
Yeah, it is much better looking live, but on the screen, that image should show the difference when you start getting to 15 stops.
I have the 4.6k ursa mini on preorder, 15 stops at BMD color science is worth $5k, who cares about the other stuff.
Re: 4k vs. 4.6k what's the real difference?

Posted:
Fri Jul 03, 2015 6:21 pm
by Allen Rice
the audio option is nothing to shake a stick at either.
Re: 4k vs. 4.6k what's the real difference?

Posted:
Mon Jul 06, 2015 7:23 pm
by shanepeters@bellsouth.net
I have asked myself that same question. While I understand and agree with the technical reasons, for me the question is....should I spend a few thousand extra for a camera that might be outdated, by it's own manufacturer, in a year or two?
I am a relatively happy owner of several BM camera products, and as an early adopter, learned quickly that BM will continually knock new products out in different form factors and configs for the next several years. Perhaps eventually they will evolve into something like Red has...a single, upgradeable camera body that can keep up with the tech improvements (Ursa is getting closer to this), but for now, I have to expect that as soon as I buy the mini-ursa, by next NAB, there will be a new contender that trumps the abilities of what I just purchased. BM is growing, and adjusting as they develop a camera product line.
Our industry is moving faster than I have ever seen it with regards to the tech. I think buying the 4.6K is a good move if you plan to pay it off quickly....but I might be hard-pressed to spend that extra money on something that is not a huge noticeable difference for the bulk of my work... that is a nice chunk of change for other gear acquisition.
Re: 4k vs. 4.6k what's the real difference?

Posted:
Mon Jul 06, 2015 8:43 pm
by Soeren Mueller
shanepeters@bellsouth.net wrote:I am a relatively happy owner of several BM camera products
Wait what?! Oh gosh I think I have stepped into some alternative reality thing... *cue twilight zone theme*
4k vs. 4.6k what's the real difference?

Posted:
Mon Jul 06, 2015 9:00 pm
by rick.lang
All true, Shane. RED has made obsolescence obsolete, but at quite a cost regardless. For instance, at NAB 2015 I believe, they announced the Dragon 8K sensor and the Weapon camera that will come in three flavours, Forged Carbon will be first out of the gate shortly, followed this year by Weapon Carbon Fibre, and lastly Weapon Magnesium. These will all take the Dragon 8K eventually but if you want a camera now, you must buy the Forged Carbon Weapon with the Dragon 6K, with no date given for when you'll then be allowed to upgrade the sensor to the Dragon 8K announced at NAB. There is no direct option to simply buy a new camera with the 8K Dragon, you must get the 6K first and then upgrade later.
You must think I'm crazy posting such drivel and illogicality, but it's true unfortunately. Not everyone thrilled about this extra step. It's not quite like buying the URSA 4K today and then upgrading to the 4.6K sensor, because once the new sensor is released, you'll be able to buy the camera new with the new sensor. You aren't being forced to jump through the 4K hoop to get to 4.6K if you can wait a bit.
So that does make BMD seem like they have our best interest at heart at least, but they're not going to stop innovating as you mentioned. It comes down to accessing if the new 4.6K sensor will satisfy your needs now and for whatever period of time you can wait before you decide you want a new turret on the URSA or a newer version of the URSA Mini. Since the 4.6K sensor was a couple of years in development, it might be a couple of years before BMD develops something else that's irresistible. Maybe the sensor will last more than two years, but the camera will be updated with something irresistible. The cost of change is inevitably depreciation of current assets.
Given all the requests users are suggesting here, no wonder BMD may continue to burn the midnight oil working on next year's NAB announcements. Looking on the bright side, going with the 4.6K sensor in an URSA or URSA Mini, may give you a competitive edge in the quality of your work that justifies and compensates for annual or biannual upgrades.
Sent from my iPad using Tapatalk
Re: 4k vs. 4.6k what's the real difference?

Posted:
Mon Jul 06, 2015 9:10 pm
by Rakesh Malik
Red also announced the end of upgrades for some models, which means that even their high-priced modularity has its limits.

Even a 4K Ursa Mini is going to be a great camera. I don't think you have much to worry about unless you're primarily interested in resale value.
Re: 4k vs. 4.6k what's the real difference?

Posted:
Mon Jul 06, 2015 11:07 pm
by PaulDelVecchio
Rakesh Malik wrote:Red also announced the end of upgrades for some models, which means that even their high-priced modularity has its limits.

Even a 4K Ursa Mini is going to be a great camera. I don't think you have much to worry about unless you're primarily interested in resale value.
Not taking anything away from RED because their cameras are spectacular, but if we're comparing prices, BMD basically has "obsolescence obsolete" as well. I mean, $3-5K for a camera... And 5K for a camera that does 4.6K 15 stops of DR and switchable rolling/global shutter. The only difference is instead of replacing a "brain," you get a whole new camera. Between the features and the price, to me, that's truly "obsolescence obsolete."
Re: 4k vs. 4.6k what's the real difference?

Posted:
Tue Jul 07, 2015 1:21 am
by rick.lang
Paul, I agree with your reasoning. BMD sells cinema cameras for less than the cost of an ARRI viewfinder. If you're happy buying an URSA, the turret upgrade costs considerably less than a new camera, but the cost of adding an updated version of the URSA Mini after a couple of years still seems quite reasonable. It's not like you have your life savings invested in a $60,000 camera with an amazing sensor when you can buy a Mini camera with an amazing sensor for a tenth that cost.
Sent from my iPad using Tapatalk