[Rondec]

There must come a point where a single frame from a video stream is a perfectly decent still image, which will revolutionise portraiture and action photography.

Maybe we're there already? Naturally a lot hangs on what "perfectly decent" means for you.

Problem with yanking a still image from video stream is that the dynamic range isn't the best, shutter speed, if optimal for still capture, won't necesarily produce good video results, and composition may not be ideal either. Beyond which, I can't imagine searching through 30 or 60 frames a second to try to find the ideal still image. I'd rather shoot two or three and delete the two that don't work.

[Nicolas06]

I repeat the most important part of my previous post:

Most likely Pentax would have to completely redo all of their programming if they moved to another processor family; I don't know how much programming is involved, so I don't know how this would affect near-term development efforts {at one time there was a rule-of-thumb saying that a given developer would produce roughly 100 lines of debugged code per day}

I agree programming and maybe hardware chip design. This again a problem of fixed cost.

Maybe when you change hardware on the next gen phone as a manufacturer, you may have to redo lot of things. And maybe that mean you need say 200 persons working for 6 months doing that. So that 100 man year worth of work. The cost being about 10-20 millions dollard depending of your structure cost.

Now you are samsung and you sell 50 million unit a year. The cost is 20 to 40 cent per phone. Negligible.

You do the same, you are Pentax. Maybe you would not need 100 man years but 200 , because you don't have the same experience. Maybe samsung do it every year. That's business as usual. So the cost maybe 20-40 millons dollars to you.

If you do that for K1, you may sell say 300K unit over the life of the camera. The cost per unit is in the range of $60-120 instead of less than one dollar.

What more there only 30 persons in the company right now that could possibility do that. Maybe you could accept to hire 5 new persons and dedicate 3 person of the current team to help them. As the other have to implement actual new feature that would give sales than just doing a technical migration...

Remember a few year ago already apple explained they have a dedicated unit of about 1000 persons working only on their embedded camera for their smartphone. Not the factory. The persons working at apple. I am not sure whole Pentax is 1000 people.

Scale is really important. When you sell 100, 1000 time more, you can afford to spend 10,100 time more for R&D and find it cheap.

[surfar]

I have a 4K UHD set myself and it's great. But image quality isn't a dramatic leap over 1080p -

Its funny, I don't have a 4K set but when I play4K footage on a 1080p set I immediately notice the difference....Then when I visit a close friends house and play it on his 4k set I notice a bigger difference....THEN when we project it wall size in his home theatre room....theres a WOPing big difference.AND with media rooms built into most new houses in OZtralia, projected images will be viewd more and more.

[Tatouzou]

I agree programming and maybe hardware chip design. This again a problem of fixed cost.

Maybe when you change hardware on the next gen phone as a manufacturer, you may have to redo lot of things. And maybe that mean you need say 200 persons working for 6 months doing that. So that 100 man year worth of work. The cost being about 10-20 millions dollard depending of your structure cost.

Now you are samsung and you sell 50 million unit a year. The cost is 20 to 40 cent per phone. Negligible.

You do the same, you are Pentax. Maybe you would not need 100 man years but 200 , because you don't have the same experience. Maybe samsung do it every year. That's business as usual. So the cost maybe 20-40 millons dollars to you.

If you do that for K1, you may sell say 300K unit over the life of the camera. The cost per unit is in the range of $60-120 instead of less than one dollar.

What more there only 30 persons in the company right now that could possibility do that. Maybe you could accept to hire 5 new persons and dedicate 3 person of the current team to help them. As the other have to implement actual new feature that would give sales than just doing a technical migration...

Remember a few year ago already apple explained they have a dedicated unit of about 1000 persons working only on their embedded camera for their smartphone. Not the factory. The persons working at apple. I am not sure whole Pentax is 1000 people.

Scale is really important. When you sell 100, 1000 time more, you can afford to spend 10,100 time more for R&D and find it cheap.

I agree: software development is very expensive, and massive sales are the key factor in software industry.
Which means the winner takes it all, and the followers have a very hard time.

[biz-engineer]

I agree: software development is very expensive, and massive sales are the key factor in software industry. Which means the winner takes it all, and the followers have a very hard time.

That's true for standalone electronic devices because there is no cost to change from one brand to another. For the ILC market, that's a mixed bag of things. Lenses and especially good and expensive less make a barrier for the ones who would want to switch. In order to increase revenue and keep it growing, ILC companies have to acquire more new customers vs what losing existing customers. Canon had nearly perfect strategy, get people onboad with the cheapest ILC cameras in mass retail, and have a large set of lenses up to 600mm or 800mm, to increase customer retention. Basically, Canon allow they customers to track themselves in the Canon mount, with ensure repeated business over time by selling camera upgrades. No wonder why Canon have the largest market share, it's not thanks to their camera software, it's very much thanks to the wide range of glass they offer. Look at Pentax, the DA560 is the evidence that these guys in Hoya time, were driven by lab engineers and financial officer rather than giving any power to their marketing staff. I mean, once you get a $6K lens in K mount , you aren't going to leave Pentax that easily, even if another brand has the better software but not the lens. Due to lens mount no being compatible across brands, the levers of the ILC market is different from the mobile phone market.

[monochrome]

I feel so old. I'm streaming NFLX from a FIRE box on a 2006 42" 1080p LG sitting 14' away in a traditional family room, and I'm perfectly happy to keep doing that until the end of time or the monitor fries.

Even the wealthy and young people I know do the same. One television in the home. Turn it on intentionally to watch - maybe DVR'ed fare - turn it off when finished. I suspect most people are similar. It isn't broken. Keep it.

[GlassJunkie]

That's true for standalone electronic devices because there is no cost to change from one brand to another. For the ILC market, that's a mixed bag of things. Lenses and especially good and expensive less make a barrier for the ones who would want to switch. In order to increase revenue and keep it growing, ILC companies have to acquire more new customers vs what losing existing customers. Canon had nearly perfect strategy, get people onboad with the cheapest ILC cameras in mass retail, and have a large set of lenses up to 600mm or 800mm, to increase customer retention. Basically, Canon allow they customers to track themselves in the Canon mount, with ensure repeated business over time by selling camera upgrades. No wonder why Canon have the largest market share, it's not thanks to their camera software, it's very much thanks to the wide range of glass they offer. Look at Pentax, the DA560 is the evidence that these guys in Hoya time, were driven by lab engineers and financial officer rather than giving any power to their marketing staff. I mean, once you get a $6K lens in K mount , you aren't going to leave Pentax that easily, even if another brand has the better software but not the lens. Due to lens mount no being compatible across brands, the levers of the ILC market is different from the mobile phone market.

Absolutely. That's why I keep looking back at the 2 year ago CEO/ R&D interviews. Bigger/ wider glass, upgraded APS-c for wildlife (with appropriate glass upgrades) big glass for 645, HD/ better coatings, all make sense... They are going uptown, and if their 2Gen "follower strategy works (like 645z did) whoa... I don't see Leica bragging on video capability.... Makers me wonder... If Canon plays more and more on video, why should Pentax fight them on common ground? Asymmetric warfare is what Pentax has a history of, and I don't see that changing...

[timcatn]

I have a 4K UHD set myself and it's great. But image quality isn't a dramatic leap over 1080p - not in the way HD was a leap over the standard-definition TV picture we all grew up with.

I agree. But the higher dynamic range of the newer panels is quite noticable. Escpecially if you play HDR content.

[Nicolas06]

Its funny, I don't have a 4K set but when I play4K footage on a 1080p set I immediately notice the difference....Then when I visit a close friends house and play it on his 4k set I notice a bigger difference....THEN when we project it wall size in his home theatre room....theres a WOPing big difference.AND with media rooms built into most new houses in OZtralia, projected images will be viewd more and more.

For years, many movies used 2K master at cinemas... Now as the difference is obvious to you between 2K and 4K... Can you name the movies that use 2K from the one that use 4K ?

Example here: 4K or 2K Resolution for my DCP? ? Ask the DCP Master

Their explanation is likely outdated now, and I would bet that now most movies are 4K but still... Also of interrest, a part of the 4K content sold to you like some 4K blurays are coming from 2K master. And people see the difference. Because of different compression level and also because the algorithms to upscales they use are good.

Did you notices people complaining of the low level of details of the blockbusters at cinemas? Often I see more people complaining of the lack of good history but they come to see the special effects and high quality experience you have at cinemas. Still until recently, it was most often 2K.

But there one thing. Most broadband content video is heavily compressed. If the 2K quality content you had was good you shouldn't see a difference between 2K and 4K on a 2K monitor. There a visible difference of course between 2K and 4K, this is visible, but a good share of the difference people see is that their prior content wasn't really the best 2K could offer and when they see an increased difference on their new 4K TV/project, a part of it is the added quality of 4K vs 2K, but also because their 4K display is also better. Typical example is if you upgraded to OLED... The difference is huge. I heard testers explaining that a 2K oled was typically offering better quality than a 4K LCD for example...

[clackers]

In defence of the videographers, their point is that you shoot 4k even when you intend 1080 output - you improve noise and dynamic range by downsampling.

[surfar]

4K resolution is here to stay,....8K?....not sure if the human eye will see the difference.....but at this stage i

havent seen 8K,so I'm content to wait for the K3iii and judge the 8K then!

[beholder3]

Most broadband content video is heavily compressed.

In Germany blockbusters with the most ads in them (so it is the most valuable mainstream stuff they have) are broadcast with 6.000 kbit/s = 750 KByte/s. Bandwidth is not going to change soon, since that would cost A LOT for the broadcasters.
Simple math:
FPS is 25 at minimum. FHD is 2 MPx. 4k is 8 MPx.
Currently the quality maximum is 25x2= 50 Mpx of information compressed into 750 kBytes.
4k quality maximum then is 200 MPx of information compressed into 750 kBytes, or 30 kBytes per 8 MPx frame.

[RonHendriks1966]

In Germany blockbusters with the most ads in them (so it is the most valuable mainstream stuff they have) are broadcast with 6.000 kbit/s = 750 KByte/s. Bandwidth is not going to change soon, since that would cost A LOT for the broadcasters.
Simple math:
FPS is 25 at minimum. FHD is 2 MPx. 4k is 8 MPx.
Currently the quality maximum is 25x2= 50 Mpx of information compressed into 750 kBytes.
4k quality maximum then is 200 MPx of information compressed into 750 kBytes, or 30 kBytes per 8 MPx frame.

On glasfiber overhere I have 10 channels that Stream at 12,75 Mbit/s. On my 2560x1440 screen I can see the difference with the normal HD broadcast in at (around) 7,5 Mbit/s. In Norway there is a glasfiber company streaming some channels at 25 Mbit/s.

[Nicolas06]

In Germany blockbusters with the most ads in them (so it is the most valuable mainstream stuff they have) are broadcast with 6.000 kbit/s = 750 KByte/s. Bandwidth is not going to change soon, since that would cost A LOT for the broadcasters.
Simple math:
FPS is 25 at minimum. FHD is 2 MPx. 4k is 8 MPx.
Currently the quality maximum is 25x2= 50 Mpx of information compressed into 750 kBytes.
4k quality maximum then is 200 MPx of information compressed into 750 kBytes, or 30 kBytes per 8 MPx frame.

Sure, but there no 4K broadcast widely available. There 4K from netflix and Amazon, but that's not a broadcast technology. I tried both in 4K, this is good but obviously not the best you could have. I still have ADSL and you can't have much more than broadband number you given. Maybe the double, but not all the time and not for all people.

With fiber becoming more and more comon, this is different. The broadcast could afford 4K bluray bandwidth. But they would do it only if there enough demand for that. Give it at least 10 years.

[beholder3]

I have ... there is a ...

Sadly, that is another irrelevant information. We could as well discuss the hundreds of thousands who are not even getting a 1000 kbit connection and who are not relevant to the adoption as well.

The technology adoption is not driven by anything else but what mainstream does. In Europe I only talk about the single most important market, which is Germany and here the vast majority of TVs are fed by satellite with the paltry bandwidth of 750 kByte/s or 30kByte per 8 Mpx frame that I explained above.

48% of people here still have not upgraded even to 720p capable TV ...