Why 4K Might Not Be the Best Way to Attract Consumers

starcentral said:
No I would disagree it is "basically" what DSLR's have done.

DSLR's do not scale down video from the larger sensor the same way shooting 5k on the RED Epic scales down to 2k/HD. DSLR's pixel-bin or line-skip depending on the DSLR.




Again DSLR's do not downsample the higher megapixel sensor using "scaling" and the reason they can get higher ISO's and cleaner looking images is because of the pixel-binning or pixel-grouping effectively turning 4 surrounding pixels into 1. This drastically improves sensor sensitivity but at the expense of aliasing and moire.
Click to expand...

Yes I know it isn't a perfect scale but it was a step in the right direction. This is why I said it would be nice to have a DSLR that does a "true" clean down sample to 2k or 1080p. What is true is both situations start with one higher value and give you a smaller value. Scaling, binning or whatever process they used is still making the image smaller with some level of benefit. The same is true with true compositing software that let's you pick which formula is used when material is scaled such as bicubic, or Gaussian. They all do it in a different way. I'm not disagreeing that binning is the least optimal way of doing it in terms of quality. To say however that it is nothing like the Ender's Game example isn't exactly true either. Same concept just a lower quality form of it with artifact tradeoffs. I still stand by a DSLR that did a true downscale in camera and recorded a RAW 2k image would be a great tool. In the meantime however we already have a tool that sort of does the same thing, albeit with more artifacts and maybe not as high of quality. Maybe because I am a GH2 and GH3 user and don't have as many artifacts I view the binning process not as bad as some other DSLR users. Of course the aliasing and moire are still there but what Panasonic does is a heck of a lot closer to an image scale than other DSLR manufacturers.
 
snowblindfilm said:
budgets and equipment don't mean squat
Click to expand...

Well, THAT'S not true. But . . .

and all the picture quality will not make up for lack of acting in a bad story line.
Click to expand...

This generally is. Crap writing and acting will always be that.

But budgets and equipment very much will make all kinds of difference. You don't make, say, Children of Men without "budget and equipment." If Blair Witch wasn't SUPPOSED to look like it was amateur VHS footage, I doubt you'd even be aware of it today.

And a crap script and acting are infinitely easier to watch when they're well-produced than they are done on a micro-budget (or no budget). One need only sit through a few local film festivals to appreciate that.

Yes, you can produce good stuff for not a lot, IF your script, acting, and technique are solid. But to say budget and equipment "don't mean squat" is ludicrous. If you take all of that skill and give it a bigger budget, odds-on, you will have a much better final product.
 
snowblindfilm said:
Waterworld was a total flop,...
Click to expand...

Not my favorite film, but it certainly wasn't a flop. As Kevin Costner is fond of pointing out, Waterworld was profitable. It cost $175M to make, a record at the time. It made back $264M worldwide. How is this a flop?

The NA critics didn't like it, and the US audience didn't flock to it in the theaters as a result, but today it has a 5.9 rating with over 100K voters on IMDB. How is this a flop? I'm just sayin' that there's plenty of movies rated lower (and plenty rated far lower) than this in IMDB.

Universal studios has three parks (Hollywood, Singapore, and Japan) running the attraction "Waterworld: A Live Sea War Spectacular" still, in 2013, 18 years after the movie was released. How is this a flop?

If this is a flop, I'll take it any day.
 
Thomas Smet said:
This is why I said it would be nice to have a DSLR that does a "true" clean down sample to 2k or 1080p.
Click to expand...
That's true in principle - but don't underestimate the task. What you are referring to means reading all the photosites individually at framerate, then deBayering, then downscaling (also all at framerate). And good downscaling is difficult to do. It's all possible - but likely to need a lot of processing, and that means expense and also high levels of power consumption.

So possible - but "DSLR" means a camera primarily used for still photography. And incorporating what you want would seriously compromise (in terms of weight, cost, powering etc) the product for it's primary purpose. I suspect stills photographers would say that if you want such a thing - primarily for video - then buy a video camera, not a DSLR.
What is true is both situations start with one higher value and give you a smaller value. Scaling, binning or whatever process they used is still making the image smaller with some level of benefit......... I'm not disagreeing that binning is the least optimal way of doing it in terms of quality.
Click to expand...
Far more complicated than that. Better to differentiate between outputs derived by full deBayering /downscaling or alternatively a much simpler form of readout. The former will give good quality whatever the sensor dimensions, the latter depends on the form it takes and sensor dimensions. The DISADVANTAGE of the former is that it will likely be expensive, power hungry etc.

Simplest form of "simple" reading out is as used in the C100/300 - form a R,G,B pixel directly from a 2x2 Bayer block. If the geometry is right, that CAN give extremely results - as the Canon cameras show. But it relies on a 3840x2160 sensor (so 1920x1080 2x2 blocks), and such isn't likely to be well regarded on a stills camera - only just over 8 Megapixel resolution.
Maybe because I am a GH2 and GH3 user and don't have as many artifacts I view the binning process not as bad as some other DSLR users. Of course the aliasing and moire are still there but what Panasonic does is a heck of a lot closer to an image scale than other DSLR manufacturers.
Click to expand...
The earliest attempts were originally intended just to provide a "liveview" facility on still cameras and generally used various simple forms of line skipping. Quality was then not important - it was just intended to give an image on the camera LCD. Then someone had the idea it could provide a cheap video facility in a still camera and the quality aspect became important. I understand the GH cameras refined the concept by using essentially what the C100/300 do - use a Bayer block to directly give an R,G,B pixel - hence the lack of coloured aliasing.

But they are first and foremost still cameras and hence the sensor of much more than 8 Megapixel. It's believed they therefore have to use a 4x4 block (not 2x2) but otherwise do a similar trick as the Canon cameras. Hence the limiting resolution is lower. The photosite count is higher, but because the output is based on a quarter rather than a half, then overall it's well under 1080 lpph. But in other ways (lack of colour aliasing and moire) performance is far higher than other DSLRs. That's why they are generally liked.

It follows that an ideal result (optimal for stills and video with low processing required) could theoretically be achieved with a sensor measuring 4x1920 horizontally and 4x1080 vertically. That's 7680x4320 or about 33 Megapixels - so obviously more than enough for stills. But it could also be SIMPLY read in 4x4 blocks (so low, cost, power etc) to directly form 1920x1080 pixels simply and with no scaling needed.

Do that, and binning all the R,G,B sites within each block is likely to be a good thing.
 
snowblindfilm said:
.......on the other hand Blair Witch, which was made on a fraction of the budget of Waterworld turned to be a hit of the season, making millions.
Click to expand...
And unfortunately, some of that money was mine. I went to see "Blair Witch" at the cinema and came out wishing I'd saved my money.......

All credit to the team though. I thought the film was poor (and not because of the low technical quality, though that didn't help) but it's real success must have been down to skilful use of internet hype, and building up a belief as to whether or not it was true. As such it was an innovation and an internet milestone - but I still find it hard to believe that anyone could really have taken it for real.

Perhaps the earlier you saw it, the more effective it was?
 
harddrive said:
That's true in principle - but don't underestimate the task. What you are referring to means reading all the photosites individually at framerate, then deBayering, then downscaling (also all at framerate). And good downscaling is difficult to do. It's all possible - but likely to need a lot of processing, and that means expense and also high levels of power consumption.

So possible - but "DSLR" means a camera primarily used for still photography. And incorporating what you want would seriously compromise (in terms of weight, cost, powering etc) the product for it's primary purpose. I suspect stills photographers would say that if you want such a thing - primarily for video - then buy a video camera, not a DSLR.

Far more complicated than that. Better to differentiate between outputs derived by full deBayering /downscaling or alternatively a much simpler form of readout. The former will give good quality whatever the sensor dimensions, the latter depends on the form it takes and sensor dimensions. The DISADVANTAGE of the former is that it will likely be expensive, power hungry etc.

Simplest form of "simple" reading out is as used in the C100/300 - form a R,G,B pixel directly from a 2x2 Bayer block. If the geometry is right, that CAN give extremely results - as the Canon cameras show. But it relies on a 3840x2160 sensor (so 1920x1080 2x2 blocks), and such isn't likely to be well regarded on a stills camera - only just over 8 Megapixel resolution.

The earliest attempts were originally intended just to provide a "liveview" facility on still cameras and generally used various simple forms of line skipping. Quality was then not important - it was just intended to give an image on the camera LCD. Then someone had the idea it could provide a cheap video facility in a still camera and the quality aspect became important. I understand the GH cameras refined the concept by using essentially what the C100/300 do - use a Bayer block to directly give an R,G,B pixel - hence the lack of coloured aliasing.

But they are first and foremost still cameras and hence the sensor of much more than 8 Megapixel. It's believed they therefore have to use a 4x4 block (not 2x2) but otherwise do a similar trick as the Canon cameras. Hence the limiting resolution is lower. The photosite count is higher, but because the output is based on a quarter rather than a half, then overall it's well under 1080 lpph. But in other ways (lack of colour aliasing and moire) performance is far higher than other DSLRs. That's why they are generally liked.

It follows that an ideal result (optimal for stills and video with low processing required) could theoretically be achieved with a sensor measuring 4x1920 horizontally and 4x1080 vertically. That's 7680x4320 or about 33 Megapixels - so obviously more than enough for stills. But it could also be SIMPLY read in 4x4 blocks (so low, cost, power etc) to directly form 1920x1080 pixels simply and with no scaling needed.

Do that, and binning all the R,G,B sites within each block is likely to be a good thing.
Click to expand...

Again I never said the same quality just a process that does help our 1080p video. Without a doubt a true scale would be better but we do enjoy some of the benefits. At this point we are really starting to split hairs.
 
harddrive said:
All credit to the team though. I thought the film was poor (and not because of the low technical quality, though that didn't help) but it's real success must have been down to skilful use of internet hype, and building up a belief as to whether or not it was true. As such it was an innovation and an internet milestone - but I still find it hard to believe that anyone could really have taken it for real.
Click to expand...

The success of the viral / grass roots marketing of The Blair Witch project always felt twisted to me because it meant a whole lot of people wanted to see a snuff film.


-Andrew
 
Funny how people perceive the "success" of The Blair Witch project which comes up all time in terms of a "low quality film that did well". I mean "if they did it, surely YOU can too right"?

The truth is that the film didn't cost a lot to make and the film makers did not get "rich" or make a "killing" off the success of the film. When Artisan bought it for cheap, then spent hundreds of thousands of dollars to market the film so it would do well as a theatrical release, they really could have done that for practically any low quality film not specifically just "The Blair Witch Project".
 
Depends on your definition of "get rich". As I remember it, the guys behind Haxan Films struck a profit-sharing deal that, last time I checked in, had paid them $12 million. Sure, it's pennies compared to the total worldwide gross of BWP, but I'd still say they did pretty well and, yes, got rich and made a killing...
 
As with many of these discussions, I don't think there is a right or wrong answer...

Most will shoot with the highest quality they can afford, and each person may determine differently what factors will deliver the highest quality. If I remember correctly, early HD cams lost in many areas to their SD counterparts. Some wanted HD resolution and others went with cameras that were SD, but had better color, better sensitivity, etc. It's the same with 4K...4K alone will not make the most pleasing image to many people. For some, a high DR 1080 cam shooting to a very good codec will be better for them.

As far as budgets go, I believe one can do a lot by simply being very aware of what content is possible for a given amount. Sounds simple, yet I constantly see local indies shooting projects that are simply above the budget they are working with. I'd personally much rather see a well executed small scale film, than one that is super ambitious...but falls short on production value.
 
Bruce Watson said:
Not my favorite film, but it certainly wasn't a flop. As Kevin Costner is fond of pointing out, Waterworld was profitable. It cost $175M to make, a record at the time. It made back $264M worldwide. How is this a flop?

The NA critics didn't like it, and the US audience didn't flock to it in the theaters as a result, but today it has a 5.9 rating with over 100K voters on IMDB. How is this a flop? I'm just sayin' that there's plenty of movies rated lower (and plenty rated far lower) than this in IMDB.

Universal studios has three parks (Hollywood, Singapore, and Japan) running the attraction "Waterworld: A Live Sea War Spectacular" still, in 2013, 18 years after the movie was released. How is this a flop?

If this is a flop, I'll take it any day.
Click to expand...

That's a flop. This is what a successful film looks like. It's easy to spot them by comparing production budgets to world wide gross. Water World is so far behind behind in comparison any accountant or investor would consider it a flop.

http://www.the-numbers.com/movie/budgets/

You are also not taking into account the marketing budget, the cost of prints for distribution, or the fact that producers only get a PERCENTAGE of world wide gross, the rest goes to theater operators.
 
pulpfiction007 said:
As with many of these discussions, I don't think there is a right or wrong answer...

Most will shoot with the highest quality they can afford, and each person may determine differently what factors will deliver the highest quality. If I remember correctly, early HD cams lost in many areas to their SD counterparts. Some wanted HD resolution and others went with cameras that were SD, but had better color, better sensitivity, etc. It's the same with 4K...4K alone will not make the most pleasing image to many people. For some, a high DR 1080 cam shooting to a very good codec will be better for them.

As far as budgets go, I believe one can do a lot by simply being very aware of what content is possible for a given amount. Sounds simple, yet I constantly see local indies shooting projects that are simply above the budget they are working with. I'd personally much rather see a well executed small scale film, than one that is super ambitious...but falls short on production value.
Click to expand...

My suggestion for an indie film is a romantic comedy. Its' all about good writing and good acting. If done right, you don't need any FX, no special zombie make up, no green screen, no digital monsters or set extensions, no big crowd scenes with tons of extras, no stunt men, car chases or crashes, no gun fights with bullets and blood effects, no uniforms for soldiers, etc. A lot of your scenes can be things like people talking in a kitchen, or sitting on a park bench. Most of the wardrobe will already be in your actors closets. The story, what they are talking about in teh kitchen, why they are sitting on the park bench, is your main event not an explosion.

It's still going to be low budget because you can't afford to hire a Meg Ryan and a Tom Hanks, and you can't shoot in Manhattan city streets with major crane shots, but if you have some solid acting and a good story, a nice shot list it can look really good, compared to trying to shoot Apocalypse Now in your neighborhood park or something.
 
This thread is about why or why not the end consumer will care about 4k, not movie sales.

At the end of the day, why will or won't consumers care about 4k?
 
Consumers will buy 4K cameras and UHD TVs by this time next year with 4K smartphones and low cost 4K UHD TVs with HDMI 2.0 .

 
The question isn't (or shouldn't be) "will anyone buy one?" but the question is (or should be) when will there be enough market saturation that it becomes a viable option for delivery? As for 4K HDTVs, surely we're a decade away from that.
 
Barry_Green said:
The question isn't (or shouldn't be) "will anyone buy one?" but the question is (or should be) when will there be enough market saturation that it becomes a viable option for delivery? As for 4K HDTVs, surely we're a decade away from that.
Click to expand...
For home entertainment maybe. But for commercial apps like digital signage and medical imaging it is here now. Depends on what market you are shooting for.
 
Well stated. You remind us again that the appropriate tool is all dependent on the job being done, and for some markets it may already mandate 4K acquisition...
 
Barry_Green said:
Well stated. You remind us again that the appropriate tool is all dependent on the job being done, and for some markets it may already mandate 4K acquisition...
Click to expand...
Absolutely. And apart from the markets Razz16mm has stated, don't anyone forget industrial, scientific, and defence - quite apart from more obvious ones like higher end cinema and theatre screening of live events.

And these are the applications with an end to end 4K pipe. Further reasons for 4K *ACQUISITION* (even if almost certain 1080 delivery) include the extra flexibility it gives in post, the ability it gives for such as lossless reframing, better chromakey etc.

As for 4K in the home, then it's not likely to be long before the sets lose their "special" label (and pricetag) and start to become the most common type in the shops - think back to how "fullHD" displaced "HDready" in just a few years, at least for the bigger sets. If they aren't that much more expensive, it becomes less a question of why to buy one, far more why buy anything else.

And yes, I also think 10 years is a reasonable guess as to how soon before even the major broadcasters start 4K transmissions, but in the meantime 4K sets may have an advantage for home use for other applications - data display, gaming, still image display, multiple image displays, etc.

I don't see any reason to replace my very nice home 42" fullHD plasma with a 4K set any time soon, at least, not solely because the new one is 4K - but if it dies in a couple of years time, then I think it's highly likely it's replacement will be 4K.
 
I think the acquisition end of the pipeline will adapt 4K very soon. It's always an advantage to have a higher acquisition format than the delivery format. Just think about framing, stabilizing, greenscree work - and your footage is future proof. But the day 4K is a common delivery format, we will probably produce in 8K.
 
You must log in or register to reply here.
Back
Top