4K for the finish...

[David Battistella]

Just to throw this out there.

My last film a 26 minute micro budget documentary was finished as:

4k dcp
2k dcp
4k prores
2k prores
1920x1080 prores

now. If I can, with the help of some friends and very little money finish a completely independant film in 4k then there is no reason this can not happen on a larger scale.

What is really happening is the war over who will distribute the stuff and how because its all doable now by anyone. The fact that its not being done is more an effort to retard or slow down progress and gain control over market share.

no one was ready for 4k, yet every user on this forum can easily do what I did with ready made tools.

The large question is why it's not already happening, when we can practically do it from our desktop now with extremely good results.

i hope RED is still planning to help the little guy with their distribution model, because things must change and now is the moment for that.

david

 

[Björn Benckert]

Just to throw this out there.

My last film a 26 minute micro budget documentary was finished as:

4k dcp
2k dcp
4k prores
2k prores
1920x1080 prores

now. If I can, with the help of some friends and very little money finish a completely independant film in 4k then there is no reason this can not happen on a larger scale.

What is really happening is the war over who will distribute the stuff and how because its all doable now by anyone. The fact that its not being done is more an effort to retard or slow down progress and gain control over market share.

no one was ready for 4k, yet every user on this forum can easily do what I did with ready made tools.

The large question is why it's not already happening, when we can practically do it from our desktop now with extremely good results.

i hope RED is still planning to help the little guy with their distribution model, because things must change and now is the moment for that.

david

You are right.
Making 4k masters is laptop stuff.
There is no reason for youtube, vimeo, netflix and the others not to jump on this. But the longer they wait the bigger impact will it have when it actually happen. I see huge profits for the once that are early adopters, kind of like when the iphone app thing was fresh then anyone that had an app available for download made a killing. So finishing a film now and do it in 2k is not so smart. Having a 4k master and be quick on the trigger when the floodgates for 4k distribution open will ensure big profit even for mediocre films. Basically people that jump to 4k will download what there is to download wich will not be much in the beginning.

It's quite amazing it has not happen just yet.

I'm hopping for a pro vimeo account where I can charge for 4k views via paypal or such.

 

[Phil Holland]

There's really two questions for distributors/service providers who want to release streaming/downloadable UHD content.

1. How to do it?
2. How to do it right?

Some have answered the first question. Whoever answers the second has everything to gain.

That's what the game is at the moment.

Combined with our current state of technology, 4K represents a new future for visual entertainment. Many are trying to figure out how to avoid complications and simplify the user experience. I can see so much that can be improved on in terms of the GUI and general user experience of current streaming services. Whoever can tackle that, make it reasonably affordable, and have the content to draw in the audience will do well for themselves.

Netflix was a given. They expressed interest in 4K two years ago. DirectTV coming to the floor is going to really heat things up. By the end of the year we'll see "most of everybody's" UHD plans. Cinema is still an open mine field. Again, a place where so much can be done differently and better. REDRAY/ODEMAX/CRIMSON/.RED can all greatly effect that. .RED in particular for security's sake. The interesting thing for .RED is it can make an impact in theaters and the home.

 

[M Most]

The large question is why it's not already happening, when we can practically do it from our desktop now with extremely good results.

With all due respect, and with no knowledge of your project, I would say that it is unlikely that you had 1000 or more VFX shots, most of which contain CG animation, you also likely had no delivery deadline, studio or international QC or both, or multiple language and subtitled versions. In short, just because you can do something doesn't mean you're doing what the higher end projects need to do, nor are you doing it in the time at to the quality level demanded in those situations. Yes, "anyone" can make a DCP. But most individuals cannot make a DCP that would be accepted unchanged by a studio or a serious distributor and satisfy all of the delivery and QC requirements that come with it.

BTW, if you have a 4K DCP there is no reason to make a 2K DCP. All DCI compliant servers are capable of playing a 2K stream "live" from a 4K DCP. And at full quality thanks to the choice of wavelet compression.

 

[George Butterfield]

We true 4K does finally make it to our living rooms, I wonder what that is going to mean for your common cineplex?

Hollywood is already lobbying Government to curb the Internet. Free Internet for everyone (controlled by the Government of course) is how they're going to do it, and regulate it. The Government will decide what the content is, it's Hollywood's last hope to stem the Independent Tide.

 

[M Most]

And where exactly did you get that information?

 

[Mark L. Pederson]

In that case, I would also point out that in all of these studios, restoration projects are also commonly producing film separation masters as the actual archival element. The digital elements are used for current deliverables such as any theatrical play and BluRay, and are also used in some studios to source IMF masters for current HD and other deliverables worldwide. But the archival need is usually filled by the film seps, because in today's world it is still the most effective archive method.

Thanks for bringing up 4K IMF. Watch how fast this becomes the standard final deliverable on feature films and high-end, episodic television. It offers many advantages over DCI's current DCDM spec.

As for long-term archival, of course the accepted industry standard is film separation masters, for a good reason. But it's not cheap and it won't get cheaper given the state and economics of motion picture film stock manufacturing. So ... if the studios want to archive in film separation masters from films delivered in 4K ... that's certainly their choice. I don't that practice being common is a few years. If you look at the number of titles that the studios have created film separation masters for vs. how many titles they release - I don't know that I would even call it common now.

Back to the 4K finish ....

 

[David Battistella]

Im not taking away from anything you do Mike. But you are talking about the top1% of content. I think smaller amounts of the type content you are talking about, not more will be produced as we move forwards.

My film has an Italian subtitle stream, I have deadlines like anyone else.

if I had to change my dcp , that would be a good problem to have and would not be hard to alter.

One whole original idea of red was to even the playing field a bit. Ten years ago I would not have shot this project on 35mm, now I can capture on EPIC. I might be a shit filmmaker, I might not be in your league, but I'd defend the right for anyone to be able to create a high quality master because films will emerge from these places and it won't be the difference between super 8 and 35mm anymore.

These distribution models are from a pre Internet universe and I don't think they are sustainable because controlling what films are seen where, like DVD region codes, is retarded. It's like limiting free speech.

So I am just saying I hope 4k breaks other barriers besides resolution and dynamic range.

david

With all due respect, and with no knowledge of your project, I would say that it is unlikely that you had 1000 or more VFX shots, most of which contain CG animation, you also likely had no delivery deadline, studio or international QC or both, or multiple language and subtitled versions. In short, just because you can do something doesn't mean you're doing what the higher end projects need to do, nor are you doing it in the time at to the quality level demanded in those situations. Yes, "anyone" can make a DCP. But most individuals cannot make a DCP that would be accepted unchanged by a studio or a serious distributor and satisfy all of the delivery and QC requirements that come with it.

BTW, if you have a 4K DCP there is no reason to make a 2K DCP. All DCI compliant servers are capable of playing a 2K stream "live" from a 4K DCP. And at full quality thanks to the choice of wavelet compression.

 

[David Battistella]

And where exactly did you get that information?

Its the whole sopa/pipa thing....

david

 

[Axel Mertes]

With all due respect, and with no knowledge of your project, I would say that it is unlikely that you had 1000 or more VFX shots, most of which contain CG animation, you also likely had no delivery deadline, studio or international QC or both, or multiple language and subtitled versions. In short, just because you can do something doesn't mean you're doing what the higher end projects need to do, nor are you doing it in the time at to the quality level demanded in those situations. Yes, "anyone" can make a DCP. But most individuals cannot make a DCP that would be accepted unchanged by a studio or a serious distributor and satisfy all of the delivery and QC requirements that come with it.

BTW, if you have a 4K DCP there is no reason to make a 2K DCP. All DCI compliant servers are capable of playing a 2K stream "live" from a 4K DCP. And at full quality thanks to the choice of wavelet compression.

Hi Mike,

while you are correct that one can play a 2K DCP from any 4K DCP stream on any player is valid, I have to disagree with your recommendation.

There are two reasons why some may still decide to make a seperate 2K DCP:

1. Data rate for 2K is typically higher.
A 4K DCP is quite often done at 250 MBit/s. That is for four times the picture data of the 2K signal. In turn that means you average on 1/4th of 250 MBit/s, so roughly 67.5 MBit/s per 2K quarter. 67.5 MBit is quite poor level for 2K DCPs. Most of the encoding systems default to ~125 MBit/s for 2K content, which is double the data rate your 2K strip from a 4K DCP might contain. However, this is not exactly true, as we would need to check the data rate of each of the resolution levels to see if we have really so much data in the last 2 wavelet layers above 2K to create the 4K. But from what I can tell, it must be close, because its the final detail level.

I have not even brought up the arguement one might decide to make a true 2K 250 MBit/s DCP. That will look significantly better than the 2K strip out from the 4K DCP...

2. DCP size and copying time.
The 4K DCP is typically 250 MBit/s is about 2 GByte/min. Over normal USB2.0 this copies at roughly 1 GByte/min, so it takes twice as long to copy the DCP than its runtime. And again, a 4K DCP on a 2K only server will not look better, see above.
The good 2K DCP at 125 MBit/s will copy from USB2.0 in about the playback time, so in half the time of the 4K DCP. Further, it will have nearly double the data rate per pixel on screen in a 2K playback scenario.

One should always think twice!

Cheers,
Axel

 

[Brad Allen]

Hi Mike,

while you are correct that one can play a 2K DCP from any 4K DCP stream on any player is valid, I have to disagree with your recommendation.

There are two reasons why some may still decide to make a seperate 2K DCP:

1. Data rate for 2K is typically higher.
A 4K DCP is quite often done at 250 MBit/s. That is for four times the picture data of the 2K signal. In turn that means you average on 1/4th of 250 MBit/s, so roughly 67.5 MBit/s per 2K quarter. 67.5 MBit is quite poor level for 2K DCPs. Most of the encoding systems default to ~125 MBit/s for 2K content, which is double the data rate your 2K strip from a 4K DCP might contain. However, this is not exactly true, as we would need to check the data rate of each of the resolution levels to see if we have really so much data in the last 2 wavelet layers above 2K to create the 4K. But from what I can tell, it must be close, because its the final detail level.

I have not even brought up the arguement one might decide to make a true 2K 250 MBit/s DCP. That will look significantly better than the 2K strip out from the 4K DCP...

2. DCP size and copying time.
The 4K DCP is typically 250 MBit/s is about 2 GByte/min. Over normal USB2.0 this copies at roughly 1 GByte/min, so it takes twice as long to copy the DCP than its runtime. And again, a 4K DCP on a 2K only server will not look better, see above.
The good 2K DCP at 125 MBit/s will copy from USB2.0 in about the playback time, so in half the time of the 4K DCP. Further, it will have nearly double the data rate per pixel on screen in a 2K playback scenario.

One should always think twice!

Cheers,
Axel

Some good points here. Hopefully we will get some new SMPTE standards soon that allow for higher bitrates @4K. Especially for 4K combined with 3D and HFR.

One thing I have considered also is aliasing in a 2K DCP playback from a 4K DCP source. I understand that wavelet compression makes a great deal of sense for compatibility between the differing performance capabilities of servers in theatres. However when once pulls a 2K source from a 4K master it is essentially line skipping, correct? Or is there some secret sauce going on in downsampling methods? For the same reason we would prefer to downscale a full 4k R3D when making a 1080 master (vs a 1/2 debayered R3D > 1080) it would seem best to properly downscale a 4K source to 2K before creation of a DCP if it will only be played back at 2K. Thus creating a 4K and 2K DCP seems to be a better option if the best image quality is demanded.

 

[Mark L. Pederson]

) is it not also preferable from an image quality point of view to downscale a 4K DCP to 2K if only a 2K media server is installed?

If the media server can only play 2K, you can't get a downscaled image from 4K - only a 2K extraction. Only way you can get a downscale from 4K is from a 4K server.

Further, you want a less compressed 2K DCP. As Alex & I have pointed out in this thread - current limit on DCI spec is 250 Mbs - so ... 2K 3D is more compressed than 2K 2D - and 4K 2D is more compressed than 2K 2D - so a 2K extraction from a 4k DCP is more compressed.

based on what I have seen with my own eyes - I'd choose a 250Mbs 2K DCP over watching a 4K 250 Mbs DCP downscaled to a 2K projector.

Regardless, any deliverables list I have seen that includes a 4K DCP - also demand a 2K DCP.

 

[Brad Allen]

Hey Mark,

I think you may have misunderstood what I posted. I'm pretty sure we're actually agreeing on the same point here.

When I say it would be preferable to do a downscale to 2K from a 4K source I mean for this to happen before it gets encoded in a DCP. And for exactly the reason you have mentioned, if the media server can only playback 2K, it doesn't downscale; it just extracts by line skipping the original 4K material.

Hence, it seems like it would always be preferable to provide both a 4K and 2K DCP master.

 

[Björn Benckert]

Hey Mark,

I think you may have misunderstood what I posted. I'm pretty sure we're actually agreeing on the same point here.

When I say it would be preferable to do a downscale to 2K from a 4K source I mean for this to happen before it gets encoded in a DCP. And for exactly the reason you have mentioned, if the media server can only playback 2K, it doesn't downscale; it just extracts by line skipping the original 4K material.

Hence, it seems like it would always be preferable to provide both a 4K and 2K DCP master.

Yes you do not want any scaling to happen in the server or the projector. Straight out files at highest bit rate and resolution is the way to go. Some say 2k looks better than 4k since it has less compression. Usually those guys also shoots with the alexa...

 

[Mark L. Pederson]

When I say it would be preferable to do a downscale to 2K from a 4K source I mean for this to happen before it gets encoded in a DCP. And for exactly the reason you have mentioned, if the media server can only playback 2K, it doesn't downscale; it just extracts by line skipping the original 4K material.

Hence, it seems like it would always be preferable to provide both a 4K and 2K DCP master.

Well sure, when you do a 4K DI and make a 2K DCP - you are doing just that - you downscale from 4K to a 2K DCP.

Note however, that line-skipping is not how a 2K extraction happens from a 4K DCP - it's a sine wave thing, just like a 1/2 debayer of an R3D

 

[Brad Allen]

Well sure, when you do a 4K DI and make a 2K DCP - you are doing just that - you downscale from 4K to a 2K DCP.

Note however, that line-skipping is not how a 2K extraction happens from a 4K DCP - it's a sine wave thing, just like a 1/2 debayer of an R3D

Could you link me to anywhere that explains the sine-wave thing. Would love to get the technical background of how this works nailed into my head

I've always advocated that folks use a full debayer and downscale from that rather than a 1/2 debayer as this has always seemed to be a better way to go from everything I have read and looking at real life differences - especially noise characteristics.

In my mind line skipping = Bad. Full res downscale + sharpening (if needed) = Good. But I don't know where wavelet 1/2 or 1/4 res extraction fits along this scale.

 

[M Most]

In my mind line skipping = Bad. Full res downscale + sharpening (if needed) = Good. But I don't know where wavelet 1/2 or 1/4 res extraction fits along this scale.

It basically fits into the latter. It is not line skipping.

There are a lot of papers available on the Web about the topic of wavelet image compression. This one is a bit geeky but may be a good start for you: http://www.stat.columbia.edu/~jakulin/Wavelets/index.html.

 

[Brad Allen]

Awesome - much appreciated Mike

 

[Axel Mertes]

It basically fits into the latter. It is not line skipping.

There are a lot of papers available on the Web about the topic of wavelet image compression. This one is a bit geeky but may be a good start for you: http://www.stat.columbia.edu/~jakulin/Wavelets/index.html.

I think I take some small time and explain it in more understandable words than the link provided:

Many of you may know the principle of mip-mapping in 3D graphic cards to provide smaller resolution levels for textures when a 3d object is more far away. To do this, a mip-map is being created by placing the R, G and B layers as three quadrants of a four quadrant image plane. The 4th quadrant is then split again into 4 sub quadrants. One sub quadrant is filled with a perfect downscale of R -> R' to 1/2 width and 1/2 height, and so on for G -> G' and B -> B' channels. The then again empty 4th sub quadrant is again split into 4 sub-sub quadrants and the game is repeated, usually until you reach some very minimal size or actually a single pixel (single pixels would require mip maps to be powers of 2, though).

This example should explain that the sub channels are of perfect downscale quality. However, mip-mapping needs 1/4th more storage, which is not desireable in the context of distribution formats for video content.

So how do wavelets solve the problem?

Very simple:

You take the highest resolution level (say 4096x2160y) and then you create a first level of wavelet from it. This is being done by making a perfect quality downscale to say 2048x2160y resolution, actually mixing every 2 pixels into 1 horizontally. The resulting image will fill say the left side of your 4096x2160y frame buffer. As you stored always the medium value of the former original 2 pixels, all you need to know is how much this stored value is different from the original 2 pixels. As its the middle value, the distance to both values is identical, but with different signs. So we can say:

left original pixel = middle value - difference

and

right original pixel = middle value + difference


A simple monochrome example (for understanding the principle):

Left pixel is 50 and right pixel is 70.
The middle value downscale pixel will be (50+70)/2= 120/2 = 60
The difference to the original pixels is 60-50 = 10 or also true 70-60 = 10 and is always 10.
You can now reconstruct the left pixel by using the medium pixel 60 +/- 10:
Left pixel = 60-10 = 50
Right pixel = 60+10 = 70

A simple RGB example (and X'Y'Z' would be just a different naming on channels):

Left pixel is RGB(50, 100, 200) and right pixel is RGB(70, 180, 200).
The middle value downscale pixel will be RGB((50+70)/2, (100+180)/2, (200+200)/2) which results in RGB(120/2, 280/2, 400/2) = RGB(60, 140, 200).
The difference to the original pixels is RGB(60-50, 140-100, 200-200) or also true RGB(70-60, 180-140, 200-200) and is always RGB(10, 40, 0).
You can now reconstruct the left pixel by using the medium pixel RGB(60, 140, 200) +/- RGB(10, 40, 0):
Left RGB = RGB(60, 140, 200) - RGB(10, 40, 0) = RGB(50, 100, 200)
Right RGB = RGB(60, 140, 200) + RGB(10, 40, 0) = RGB(70, 180, 200)

So as you can see all you need to do is to store the middle average value of two pixels in the first half of your framebuffer and the difference value in the second half. That uses exact the same amount of space as the original bitmap and allows (almost) perfect reconstruction. Important to understand: You need a signed value range for the difference level, as they can be negative. And there is a round error of 1/2 bit in a worsed case.

When you now iterate that step once more on the left 2048x2160y perfect downscaled image to scale it down vertically to 2048x1080y doing the same steps, but vertically, you have the top left corner being a perfect downscaled 2K image, the down left quarter being the difference level to reconstruct 2048x2160y resolution and the two right quadrants the difference to reconstruct the 4096x2160y full resolution image.

However, JPEG2000 for DCI specs goes well beyond that and includes typically 5 wavelet levels for 2K content and 7 wavelet levels for 4K content, whih means the downscaled image part is iteratively dowscaled by 1/2 alternating vertically and horziontally (and ONLY the image part, not the difference part). So the lowest resolution is much smaller than the original content. To reconstruct the image, you have to process all resolution levels until the desired resolution is reached.

Its definetly NOT line skipping.

Btw, CineForm uses wavelets too. The BIG point here is that you can reconstruct a lower resolution wavelet with far less computing resources, as you need to read and process dramatically less data. In fact you can view and work with a smaller (but perfect quality) resolution. Something that you can NEVER do with ProRes, DNxHD, BMD or AJA codecs, nor DPX or TIFF files. Its time that poeple start understanding this important difference for their workflows.

Cheers,
Axel

 

[Patrick Tresch]

Hello Axel,

Reading the above post make me ask a question...

What people say: the bigger the picture, the better wavelet compression works.

4k DCP @ 250MBit/sec projected on a 2k server should "VISUALLY" not be equal to a 4 times more compressed 2k DCP (like 67.5 MBit/s) due to the nature of wavelet compression scheme?

Do you really think it's VISUALLY equal to a "roughly 67.5 MBit/s" 2k DCP?

Pat