Better K's instead of 8k.

The place for questions about shooting with Blackmagic Cameras.
  • Author
  • Message
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Better K's instead of 8k.

PostSat Aug 31, 2019 4:01 pm

I've had some bemusing conversation with some lost Ronin around here lately. The idea of better K's rather than more K's came up. Which seems like the start of a new mantra or something. But the guy is onto something, we don't need all these K's.

Instead of 8k we could have 1k, or 1.28k, which is close enough to a film standard. We would have bigger pixels, for lower noise for even lower low light. We could record uncompressed for maximum quality. We could record 25mbit/s or less to save precious storage space With Sony, maybe we would have the same dynamic range, but it's all about the lower K's isn't it?

The pro cameras could be lighter and have longer battery life. The mini could be as thin as a tablet, with a lens mount out the front. Once you add fold out display, its still going be half a centimeter or more, with all day battery and storage. Why, you could even then use it as a tablet, and editing computer! Stripping the K's off really pays dividends.

But why stop there we, don't need k's, we could go back to c's, and use standard definition. Even 128-160 pixels or so across in 16 shades of grey conveys a picture, like they used to film in that audio cassette recording camera in the 1980's. Wow, the camera might even fit in a price of wire. Just moving it around might charge it's battery. You could have a little Handel to turn, like they used to have on real cinema cameras. In a credit card sized camera stuck on the back of a full frame lens, you could film maybe all week, and cost a $1 to make. What are people saying their development their development dollars on these 8k monsters.

Or, you could just get enough low light, low noise, and 17+ stops dynamic range on an 8k camera today, which is going debayer and downres into a superior color image, for less than $5k soon enough, and record to cheap copious storage at 100MB/s+, rather than a $1 1.28C credit card camera.
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline

Roberto de la Torre

  • Posts: 107
  • Joined: Mon May 27, 2013 1:01 pm

Re: Better K's instead of 8k.

PostSat Aug 31, 2019 10:59 pm

:lol: Go out and shot something with whichever camera you've.
Roberto de la Torre

Comic book,storyboard,cinematography........and more
Offline
User avatar

Ulysses Paiva

  • Posts: 804
  • Joined: Sun Sep 01, 2013 8:32 pm
  • Location: Pernambuco, Brasil

Re: Better K's instead of 8k.

PostSun Sep 01, 2019 8:18 pm

Roberto de la Torre wrote::lol: Go out and shot something with whichever camera you've.

+1
Ulysses Paiva
UPMD Filmes
http://www.upmdfilmes.com.br
Offline

colink_mike

  • Posts: 7
  • Joined: Tue Aug 27, 2019 12:44 pm
  • Real Name: Mike Losch

Re: Better K's instead of 8k.

PostThu Sep 05, 2019 5:08 pm

Oh Wayne, such a romantic ))))
Offline
User avatar

Kim Janson

  • Posts: 907
  • Joined: Wed Jul 15, 2015 6:54 pm
  • Location: Finland

Re: Better K's instead of 8k.

PostThu Sep 05, 2019 6:19 pm

160x120 pixels, global shutter, 14 bit depth :)
Attachments
Screenshot 2019-09-05 at 21.17.40.png
Screenshot 2019-09-05 at 21.17.40.png (806.84 KiB) Viewed 1013 times
LeViteZer Smooths the movement, www.levitezer.com
Offline

Que Thompson

  • Posts: 303
  • Joined: Thu Jan 24, 2013 10:19 pm

Re: Better K's instead of 8k.

PostThu Sep 05, 2019 7:01 pm

Exactly. Better K's. Watch this in 720p (Alexa LF 4K):



Now go watch any Blackmagic 4k video uploaded to YouTube, in 720p. It doesn't look bad, but it doesn't look like that.

Better K's for the future!
Offline

Denny Smith

  • Posts: 11116
  • Joined: Thu Aug 01, 2013 4:19 pm
  • Location: USA, Northern Calif.

Re: Better K's instead of 8k.

PostFri Sep 06, 2019 4:41 pm

For a fair test, you would real,y need to have **** BMD camera in same situation, lighting, actors, etc.
A camera’s pixels only capture the image data, the mood is established by lighting, everything else is data processing, getting the most out of the pixel data. A 24MP (6/8K) Sensor, is really, only a 12/6/6MP Sensor, with chroma data only having 6/6MP, luminance has the other 12MP. It is how a camera system/you use those pixels and the data they generate that makes the difference. ;)
Denny Smith
SHA Productions
Offline

Que Thompson

  • Posts: 303
  • Joined: Thu Jan 24, 2013 10:19 pm

Re: Better K's instead of 8k.

PostFri Sep 06, 2019 7:50 pm

Denny Smith wrote:For a fair test, you would real,y need to have **** BMD camera in same situation, lighting, actors, etc.
A camera’s pixels only capture the image data, the mood is established by lighting, everything else is data processing, getting the most out of the pixel data. A 24MP (6/8K) Sensor, is really, only a 12/6/6MP Sensor, with chroma data only having 6/6MP, luminance has the other 12MP. It is how a camera system/you use those pixels and the data they generate that makes the difference. ;)


I knew this argument would be made... I tried to find a similar BMD test scenario but couldn't. However, we both know the Arri has "better" K's. That was the point. Better K's for the future rather than more K's.
Offline
User avatar

Ulysses Paiva

  • Posts: 804
  • Joined: Sun Sep 01, 2013 8:32 pm
  • Location: Pernambuco, Brasil

Re: Better K's instead of 8k.

PostFri Sep 06, 2019 8:02 pm

Guys, you really think what you already have is not enough?

Then I think the problem is not the camera...
Ulysses Paiva
UPMD Filmes
http://www.upmdfilmes.com.br
Offline

lee4ever

  • Posts: 277
  • Joined: Tue Jan 08, 2019 9:59 pm
  • Real Name: Aki Lee

Re: Better K's instead of 8k.

PostFri Sep 06, 2019 8:16 pm

Full agree! Better K's for the future than more K's.
Offline

Que Thompson

  • Posts: 303
  • Joined: Thu Jan 24, 2013 10:19 pm

Re: Better K's instead of 8k.

PostFri Sep 06, 2019 8:20 pm

Ulysses Paiva wrote:Guys, you really think what you already have is not enough?

Then I think the problem is not the camera...


Hilarious. The irony.

This thread was started (I'm sure) to say "guys, you really think what you already have is not enough?" and to explain why 8k is not needed. You take that explanation and say the same thing to it.
Offline
User avatar

rick.lang

  • Posts: 11450
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: Better K's instead of 8k.

PostFri Sep 06, 2019 9:16 pm

For the work we do, we would appreciate better Ks once you hit the BMPCC6K or even BMPCC4K. Admittedly it is very useful to have a 6144x3456 sensor for 4K deliverables. Given I’m only delivering in 2K/HD, I’m good too with 4096x2160.

To me, it is amazing how decent the phone cameras are with their tiny sensors. Lots of other tech coming in to play to make them look decent (if they’re not pushed generally.

Still I lament the continual shrinking of the pitch of the photosites that began at 6.5 microns. The actual circuitry and design of the photosite has improved and a smaller pitch has not meant a linear loss of the electron well or increased noise. I do wonder if the same recent technology was applied to a K count like the ARRI Alexa’s sensor, if a BMD camera would be virtually identical in quality at a magnitude less cost.

Part of the race to add Ks is the pursuit of the ‘full frame’ 135 film sensor or medium format camera as if that will be some ultimate Nirvana for narrative film storytelling. Sure that’s great for viewing stories featuring landscapes in an IMAX theatre. But I don’t even see the value for 90% of narrative storytelling to emulate these very large 70mm film gates.

As anyone who has shot 16mm or Super16, there’s something interesting about the constraints of that frame. It ends up being an aid, not a hindrance, to those shooters. I mean beyond the increase depth of field. 35mm or Super35 achieves a similar affect after several decades of dominating what we view.

By the time you get to shooting ‘full frame’ or higher, I don’t know if it helps anything and might make it very difficult for indies to concentrate on a story about characters when the audience is counting pimples or noticing flaws in the set. And the costs are likely exponentially increased as the film gate increases. That’s why your phone costs less than a good lens.

So when I say better Ks, I’m not thinking of ever larger photosites forcing ever larger sensors, I’m thinking better to apply better technology in making better use of the space you’re using today. Better Ks not more Ks, but few of us may need or benefit from bigger arenas.
Rick Lang
Offline

Denny Smith

  • Posts: 11116
  • Joined: Thu Aug 01, 2013 4:19 pm
  • Location: USA, Northern Calif.

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 12:20 am

Well put a Rick, I made my first film with 16mm, in college. It forces you to think inside the box so to speak, keeping your perspective in check. I was fortunate enlightened to have a reflex Angenieux Zoom lens on that school camera, which made my shots easier to blockout. I shot mostly outdoors, no inside sets, everything shot on location, lots of flags and reflectors.
Cheers
Last edited by Denny Smith on Sat Sep 07, 2019 2:08 am, edited 1 time in total.
Denny Smith
SHA Productions
Offline
User avatar

Kim Janson

  • Posts: 907
  • Joined: Wed Jul 15, 2015 6:54 pm
  • Location: Finland

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 1:07 am

I have often wondered why the phone technology is not used on computers and cameras. They have so powerful, tiny and energy efficient processors and aplications run with much less memory than in computers.

Not to speak of the tiny camera and lens, and what they can do with that, to make the image and video pleasing and how easy it is to use, anyone can take pleasing looking video and images with phones. It is much thanks to the enormous calculation power and some smart software that the huge volumes the phones are sold have made possible.

However I hope the 'profesional' cameras would stay true to the original image and improve capturing the photons as they hit the sensor, instead of using the 'phone technology' to improve porely captured image to make it pleasing.

I would take better pixels and glopal shutter any day over 8k, even 4k. Not only it looks more natural, but makes virtual studio kind of stuff much easier and is easy and fast to edit.

Edit: just noted this at the forum, a good examble of what I mean viewtopic.php?f=2&t=98282
(This could be also post processing or in camera proicessing error, but we all have seen it also on correctly processed images)
LeViteZer Smooths the movement, www.levitezer.com
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 3:44 am

Ulysses Paiva wrote:Guys, you really think what you already have is not enough?

Then I think the problem is not the camera...


It's not them. It's professionally being able to match the human visual system with the least effort.
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 3:54 am

Que Thompson wrote:
Ulysses Paiva wrote:Guys, you really think what you already have is not enough?

Then I think the problem is not the camera...


Hilarious. The irony.

This thread was started (I'm sure) to say "guys, you really think what you already have is not enough?" and to explain why 8k is not needed. You take that explanation and say the same thing to it.


Probably the reverse actually, satire humor I think. The point is we haven't hit the limit, and people act like more K's mean poor K's, but we can have both better K's, and more K's at the limit. The more K's at the limit are also good for making better 4k and 2k delivery. That, satirically, you can deliver an feature film image at 160x120 or less with quality pixels, but, isn't 8 good k better then that. So, better K, IS not the full picture, appropriately more K is part of better K! Isn't it simple?
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline
User avatar

Kim Janson

  • Posts: 907
  • Joined: Wed Jul 15, 2015 6:54 pm
  • Location: Finland

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 6:05 am

Sure 160x120 pixels is not enough for normal use, but perfect for what we use it. Sure I would take 32k if:

1. the image would be othervice perfect,

2. an hour of storage would cost the same as for FHD today and to transfere the files would take the same time,

3. camera battery life heat generation, size and weight would be ok,

4. editing would be smooth with a system I have or will next buy at reasonable price.

5. The camera would be in same price range as BMD cameras today.

As it is today, I just think there is far more important areas to develope, than to get beyon 4k, except for 360 video.
LeViteZer Smooths the movement, www.levitezer.com
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 8:34 am

Yep, we are years away from your spec being cheap Kim. I've got my own magnetic processing technology design I want to develop, but without investment to develop, manufacture, distribute and market, it's hopeless, but otherwise that would be a way to do it cheap. I'm estimating mobile phone like for your spec super cheap within 3 or so years on my system. But normally you are talking 10+ years, to wait for development to complete and for it to be sorted out which is going be the most profitable to sell, which can slow development expenditure. Then when they decide on possible winners, it is a premium product, and likely with multi billion dollar manufacturing lines (once investment in developing that equipment pans out). So, they charge against existing processing costs, so the price might be half the equivalent sever farm, but if you need say 100x the processing power in your pocket, guess how much they are going charge for such a chip at first, and keep dropping the price as they need over the years, or decades to increase sales and react to competition. If however, you can develop such a thing cost effectively, you have the liberty to include it in your product at cost price. We are talking about figures of a millionth the power consumption, up to a couple of terahertz, and FPGA varieties I expect are not designed for that speed, but 500mhz+ in parallel, meaning the next data follows physically sequentially behind the previous data each clock step. Talk about super pipelining. You can have millions of sets of data loaded up on a chip, being processed in parallel as well as sequential parallel (my term). So, that sort of fpga might be a very good solution in coming years, at reasonable prices. Years ago, when I was trying to determine how to do a home chip foundry). I determined the most convenient solution was to do FPGA chips. Speeds and density past current FPGA should be possible at low energy. 96k+ holographic pro cameras should eventually be possible (Sony has been doing research on this with a 16k chip years ago to yield 720p). People accuse me of wanting perfect this or, that, but the stuff I talk about is just practical quality nowhere near this holographic stuff, which actually will be costly to do for now. Your real issue is, they don't want to release consumer stuff that threatens there over priced pro sales, even though they could at current pricing. They got to pretend many advancements don't apply, somehow they cost many times more, ignore mores law, ignore ignore cost savings to jack up the price of flash and other things. All the same BS, as years ago, which gave us the chance to attract SI and Red, BM, zcam, and many others to the low cost pro camera market.
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline
User avatar

Kim Janson

  • Posts: 907
  • Joined: Wed Jul 15, 2015 6:54 pm
  • Location: Finland

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 1:45 pm

If you have an patentable idea, why not apply for patent to get priority date and start marketing the idea. Everything else is over thinking.

Applying patent is not expensive, and with that you have priority date for many years.

If the idea is good, you will find buyer for it, or investor if you so prefer and eventually it will find its way to markets.
LeViteZer Smooths the movement, www.levitezer.com
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 5:30 pm

Kim Janson wrote:If you have an patentable idea, why not apply for patent to get priority date and start marketing the idea. Everything else is over thinking.

Applying patent is not expensive, and with that you have priority date for many years.

If the idea is good, you will find buyer for it, or investor if you so prefer and eventually it will find its way to markets.


That's everything wrong with the patent system, non of it is really that simple if you don't want minimal go away money. Applying for a patent covers a market, or markets, but I think you still get hit for fees for each if the markets, and fees to keep it alive. Try to do that for every country and see his much it cost. But buying a patent is like buying a certificate, you need really big money to actually defend it. Otherwise there are companies and law firms looking for private patents to bust and use. Any good system, is likely to require heaps of patents on top on this. That's my problem, I work on systems and discover to many related things for the product. Forget ideas debate, good inventions are concrete solutions. Because of hassles big companies have been avoiding investment in outside ideas. Your investors start at 80-90$ of your profits, and control of your company while they wait to axe you in the neck. That's why I'm looking at trying to sell simpler things still valid, that aren't a long term strategy for me. Ideas are mere notions of the shape of the desire of something, solutions systems are the substance.
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline
User avatar

Kim Janson

  • Posts: 907
  • Joined: Wed Jul 15, 2015 6:54 pm
  • Location: Finland

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 5:40 pm

It is enough that you get a priority date in one country. For me that cost 450 euros. With the priority date you can claim patent anywhere within 3 years. (If I remember correctly)

Sure word wild patent will set you back some 50 to 100 keuro, but you have 3 years to deside if that is needed.

For something like a new type sensor that has mass markets, you just need to cover couple of main markets anyway. No company is going to make a phone or camera they can not sell in US.

You are over thinking it. What alternatives you have than get some protection and start fundraising and then you can better protect it.

A big company that want to start using it, absoluty wants to be clear on the IPR. They are happy to pay you to be clear on it. I am pretty sure of that.
LeViteZer Smooths the movement, www.levitezer.com
Offline
User avatar

Rakesh Malik

  • Posts: 2594
  • Joined: Fri Dec 07, 2012 1:01 am
  • Location: Lynnwood, WA

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 9:47 pm

rick.lang wrote:So when I say better Ks, I’m not thinking of ever larger photosites forcing ever larger sensors, I’m thinking better to apply better technology in making better use of the space you’re using today. Better Ks not more Ks, but few of us may need or benefit from bigger arenas.


I agree, but I also think that Sony + BMD have done a really good job of this, because they're taking advantage of BSI sensors to improve the sensitivity of their sensors and cameras, and BMD's image processing is continually improving.

On top of that BMD is now offering the sorts of workflow options that used to be out of reach -- the metadata sidecar file that comes with braw clips and is passed on throughout the pipeline is an understated in for Black Magic users.
Rakesh Malik
Cinematographer, photographer, adventurer, martial artist
http://WinterLight.studio
System:
Alienware M15 Hexacore i7/32GB/2070 Max-Q
Offline

Roberto de la Torre

  • Posts: 107
  • Joined: Mon May 27, 2013 1:01 pm

Re: Better K's instead of 8k.

PostSat Sep 07, 2019 11:35 pm

Kim Janson wrote:I have often wondered why the phone technology is not used on computers and cameras. They have so powerful, tiny and energy efficient processors and aplications run with much less memory than in computers.

)


This make my think many times why digital cinema cameras haven't rangefinder mount. Same as cinema lenses.
Rangefinder is small, full frame and there's some ones with a pretty well performance.
I had shot much film with Leica M and the lenses fit on the palm of the hand.
Even Xpan lenses which are capable of 65mm film still pretty small.
Roberto de la Torre

Comic book,storyboard,cinematography........and more
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostSun Sep 08, 2019 7:45 am

Kim Janson wrote:It is enough that you get a priority date in one country. For me that cost 450 euros. With the priority date you can claim patent anywhere within 3 years. (If I remember correctly)

Sure word wild patent will set you back some 50 to 100 keuro, but you have 3 years to deside if that is needed.

For something like a new type sensor that has mass markets, you just need to cover couple of main markets anyway. No company is going to make a phone or camera they can not sell in US.

You are over thinking it. What alternatives you have than get some protection and start fundraising and then you can better protect it.

A big company that want to start using it, absoluty wants to be clear on the IPR. They are happy to pay you to be clear on it. I am pretty sure of that.


Not unless its changed in the last 20 years. You have time outs, to lodge subsequent patents, usually running out in 12 months. Your 450 euros covers hardly anything, and definitely not world wide, just that market. You are eight that one of the strategies is to limit the market based on the best markets, but three are not enough. What are we up to now, the g20? Plus, stuff I work on is sellable in most all countries, to poor people. You might not realise, that we now have a online grey import problem. So, any market not covered can produce their own version, which can be online shopped through any other county. A expensive nightmare to police, and trying to prosecute individuals buying is an expensive issue. However, patent gives you effective control of sales, use, and distribution (most of which I don't need). It's cheaper to have a patent in every market to close down manufacture, than trying to stop leakage from these markets. What I'm working on is often not what you want to let go, as it globally applies. You can miss out on over 50% of sales by going three markets. Plus, 450 euro, is that just to apply, and not including the fees until it guest to grant, and fees to keep it alive after the grant? Also, what's over thinking about these realistic figures, when you work through a complex solution and can easily get ten to a hundred patents, where just going for one in three markets, exposes the technology to other people applying patents in what they see blocking your other patents, so you land up having to get a license from them to use whatever else you have left to patent. When you do a good solution, it can often be out there new different, the while design them becomes interconnected along this different way of doing it, naming a single patent reveal what the network is about, and all it requires is for one person to say, aha, and build on your work with note patents before you are ready. Remember, you can't afford the millions to defend even the single patent. So, no, I think me selling off simple dead end things to produce enough money to get proper patent coverage and development is a much better idea.

It might pay to talk to people with a lot of lean and mean experience in these things. The strategy is normally to license to a company that needs and can afford to develop such things. Under NDA, you negotiate their interest in it, once you have made a business case with research and they are willing to seriously want to sign, you lodge your application, getting you on the treadmill, and reveal it under NDA with non compete (usually 12 months, which they don't want to sign), you then have the patent application number to prove you are serious and they have to go through you. Something like that. Now, if you can get 1% if the retail price, you are doing OK, 3% is higher, I think you can get more, but if you can make water into gold, maybe you are better off doing it yourself than getting even 10%. If you are going to dominate the solution pile, you are better off trying to get a better deal.
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostSun Sep 08, 2019 8:37 am

Roberto de la Torre wrote:
Kim Janson wrote:I have often wondered why the phone technology is not used on computers and cameras. They have so powerful, tiny and energy efficient processors and aplications run with much less memory than in computers.

)


This make my think many times why digital cinema cameras haven't rangefinder mount. Same as cinema lenses.
Rangefinder is small, full frame and there's some ones with a pretty well performance.
I had shot much film with Leica M and the lenses fit on the palm of the hand.
Even Xpan lenses which are capable of 65mm film still pretty small.


It's costly but easier and less costly than asic.


It's a lot of semantics and business politics, plus asic like Ambarella just busts it, and is used in the camera industry. So, an ambarella based can be made cheaper better, than sold in a $3k+ prosumer, pro cameras at great price.

The intel juggernaught paid for a lot more performance development and locked people into supporting x86 code. This produced a lack of will to chsnee, but that is changing. You can but windows laptop with arm that run x86 32 bit now I believe. There has been arm server chips, but custom code or asci, can do a lot of operations. Server farms demand a lot of power, so I think change is inevitable. If they really wanted to save energy today, memory/storage node processs based asic with processing designs is the way to go. Using a non volatile memory system is vest, like Intel has, but the 1 transistor+ dram memories that retain the memory image without power as a capacitance build up in the substrate, as a good option. You basically just get networked memory chips with few if any, external processors. Cheap to make, high density, low power, high performance. You could be looking at 10x or more energy reductions. Sony uses some processing and memory on die on those mobile sensor chips.

But yes, you could do a mobile phone reference design based camera relatively cheaply, with 4+ sensors to give you multipoint 3D post computational image capability. Here you can emulate the best cameras and lenses. The real issue is low light using those tiny sensors. The imperfection of the small lens systems can be ironed out by post processing. Sure, this is going require ASIC or magnetic computing etc, to do quickly. So, the equipment is cheap, the post processing nay not be for sometime. But look at it this way, you take it home, you set the system off into an auto colour correction, an auto framing etc, and post processing. You get on with selecting footage which is prioritised by the system. Hopefully by the time you finish that, the auto colour correction is in to your look, you then further refine the footage you want to use as you adjust the auto correction suggestion, hopefully once you finished that the auto framing etc suggestion is in, prioritising the footage you are working on. You work on finishing that aspect. Then you go to bed, and hopefully by then the accurate calculation of the pixels etc is done. Overview and refine again as needed, then out. So, you work on draft a lot of the time. So, on the cheap, low end film makers can simply use extra time to process on cheaper hardware it while they while they sleep. But maybe they cab just real time process it in draft/proxy, and leave it render latter. The interesting thing here, is you can maybe extract 8k detail on everything, despite movement. You want to do that our of proxy, unless you want proxy resolution detail to be your target.
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline

John Griffin

  • Posts: 217
  • Joined: Sun Aug 23, 2015 4:18 pm

Re: Better K's instead of 8k.

PostMon Sep 09, 2019 5:50 am

Say we had a perfect sensor tomorrow with infinite dynamic range, resolution and perfect colour science - there would be forums full of complaints about how it looked too good or real and with various methods on how to degrade the image to make it look more /cinematic/filmic or whatever.
Perfection is the enemy of the good.
Dell XPS 9560 i7 7700HQ,GTX 1050m, 32gb 1Tb NVME C drive, 1TB SSD 2nd internal as Cache. GTX 1080ti eGPU, 4x 7200rpm 1.5tb HD's in RAID 0 on USB-C as media storage. Win10 x64, Resolve Studio (all latest versions and up to date drivers)
Offline

David Hutchinson

  • Posts: 172
  • Joined: Thu Jun 05, 2014 11:00 am
  • Location: East Yorkshire, UK

Re: Better K's instead of 8k.

PostMon Sep 09, 2019 8:56 am

Que Thompson wrote:Exactly. Better K's. Watch this in 720p (Alexa LF 4K):



Now go watch any Blackmagic 4k video uploaded to YouTube, in 720p. It doesn't look bad, but it doesn't look like that.

Better K's for the future!


Interesting..... Que remind me of the Alexa LF price again?
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostMon Sep 09, 2019 12:00 pm

John, you can just buy a Red Monstro camera today and possibly auto cut out the range you need, in the same sort of way.
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline

François Giroux

  • Posts: 87
  • Joined: Tue Sep 30, 2014 4:58 pm

Re: Better K's instead of 8k.

PostMon Sep 09, 2019 12:58 pm

I miss the global shutter of my production 4k..... I do not like jello... I got the pocket 4k and it is not that hi-end. Sometimes vibrations can mess the image in an ugly way. Has way more noise at 3200 iso when baking prores with lut than expected, side to side with my old 4k doesnt it look that great... at that price it is useful but I am not thrilled.
Offline

Frank Engel

  • Posts: 187
  • Joined: Thu May 17, 2018 11:09 pm
  • Real Name: Frank Engel

Re: Better K's instead of 8k.

PostMon Sep 09, 2019 1:22 pm

I have a calculator with 32K.

Too bad it can't take pictures...
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostMon Sep 09, 2019 1:45 pm

Yes, and we have special K too. :)

Image
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline
User avatar

rick.lang

  • Posts: 11450
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: Better K's instead of 8k.

PostMon Sep 09, 2019 3:44 pm

Finally something we can sink our teeth into.

Must remember that “perfection is the enemy of good” because the last thing we honestly want to deliver is perfection defined as reality. Colour Science to give us the correct hues to capture. But remember we deliver the stuff dreams are made of.


Sent from my iPhone using Tapatalk
Rick Lang
Offline

Que Thompson

  • Posts: 303
  • Joined: Thu Jan 24, 2013 10:19 pm

Re: Better K's instead of 8k.

PostMon Sep 09, 2019 4:15 pm

David Hutchinson wrote:Interesting..... Que remind me of the Alexa LF price again?


:lol: This is exactly my point. How much is the sensor? If we as customers pushed for better K's instead of more K's, BMD could put this sensor in a new camera next year instead of scrambling for some 8k sensor.

My TV was $7000 brand new, I did not purchase it new. I bought it for $2500 (refurbished) 1.5 years ago, it's now 3-4 years old, it still looks GREAT. 75" 4K 10 bit panel and the blacks are amazing.

I said that to say, how much is that sensor today, how much will it be next year, the year after that? Being on the cutting edge will cost you, but I think a high quality 4k sensor that is a few years old will deliver a superior image to a brand new, cheap to equally priced, 8k sensor.

That video was simply an example of "better" K's.

Does anyone know the difference in sensor prices? Super35 high end vs. the one in the Pocket 6k. Or anything else is fine too...
Last edited by Que Thompson on Mon Sep 09, 2019 5:33 pm, edited 1 time in total.
Offline

Roberto de la Torre

  • Posts: 107
  • Joined: Mon May 27, 2013 1:01 pm

Re: Better K's instead of 8k.

PostMon Sep 09, 2019 4:41 pm

Low DR is enough with the support of huge HMIs.
To my eye, mostly feature films have the key around 300 LUM very well lit it, and enthusiast still stretching out 0 to 896 looking for his key around 500-600.
Naturalistic cinematography is quite hard.
**** LOG.. :lol: let the camera doing his best in REC 709, 2020..or whatever final look.

Other than that are just K's which all depends of the industry's requirements. nobody gonna purchase a 1K film because haven't future to show.
Roberto de la Torre

Comic book,storyboard,cinematography........and more
Online
User avatar

Dmitry Shijan

  • Posts: 751
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: Better K's instead of 8k.

PostWed Sep 11, 2019 5:23 pm

Just found nice example of better K's:

Image
Image


Image
Image
All my custom made accessories for BMMCC/BMMSC now available here https://lavky.com/radioproektor/
Offline

John Brawley

  • Posts: 2115
  • Joined: Tue Aug 21, 2012 7:57 am
  • Location: London UK

Re: Better K's instead of 8k.

PostWed Sep 11, 2019 5:49 pm

Dmitry Shijan wrote:Just found nice example of better K's:



That horizontal noise seems to be a big problem.

JB
John Brawley
Cinematographer
London UK
Offline
User avatar

MScDre

  • Posts: 129
  • Joined: Mon Apr 30, 2018 8:31 pm
  • Real Name: Andrea Domenichini

Re: Better K's instead of 8k.

PostWed Sep 11, 2019 8:16 pm

I was researching something about medical imaging and found this video



and apparently these guys created a sensor that measures not just intensity of the photons but their wavelength.

Admittedly its for X-rays but imagine a visible light version that stores a 32bit value for light intensity at a site as well as a 32bit value of specific wavelength of light measured :shock:

might actually get that 20stops and perfect colour magic people joke about, better Ks indeed.
Offline
User avatar

rick.lang

  • Posts: 11450
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: Better K's instead of 8k.

PostWed Sep 11, 2019 8:42 pm

All in the X-ray end of the electromagnetic spectrum where colour is assigned by us. BMD will be bringing out Colour Science gen X for that camera!



Sent from my iPhone using Tapatalk
Rick Lang
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostThu Sep 12, 2019 2:15 am

That's the sort of thing I was aiming for. See my other thread on getting better colour from a colour sensor.

The bit where I refer to something I was working on was to achieve this. I can comment though, it's actually rather complex. In each pixel is a range of frequencies, which has to be sensed without touch time lag, to get the right mapping. However, in my thread I am going into techniques to refine better to how the eye sees, to better match it, but you don't need to go as far as the individual frequencies. 5-7 bands is sufficient. The best original Bayer pattern was actually a complementary colour pattern, but was too complex in calculation. It shouldn't he now. But what I'm saying is a 5+ color pattern would maybe help, but your colour resolution accuracy might suffer. Sony did a 4 color once. But 6 color might be better, and 9 color for scientific and security use.

viewtopic.php?f=2&t=98485
Often people deceive themselves so much they do not understand, even when the truth is explained to them.
Offline
User avatar

rick.lang

  • Posts: 11450
  • Joined: Wed Aug 22, 2012 5:41 pm
  • Location: Victoria BC Canada

Re: Better K's instead of 8k.

PostThu Sep 12, 2019 3:45 am

Wayne, was the “best original” Bayer using CMYK?


Sent from my iPad using Tapatalk
Rick Lang
Offline

Wayne Steven

  • Posts: 2117
  • Joined: Thu Aug 01, 2013 3:58 am
  • Location: Earth

Re: Better K's instead of 8k.

PostThu Sep 12, 2019 6:56 am

Could be. He actually talks about different luminances in the pattern in the patent I see, so I was spot on about the use of highlight recover. That pattern is more used for scanning print or film.

You know, the Kodak version with clear tile is a bit like that. You could maybe get a bit more latitude that way.


Can they bypass the Red patent by simply using a different color filter pattern? The random filter pattern was supposedly good?
Often people deceive themselves so much they do not understand, even when the truth is explained to them.

Return to Cinematography

Who is online

Users browsing this forum: DarrenMcPhee, Dmitry Shijan, pnguyen720, Rrrbbb, SkierEvans and 36 guests