There’s nothing wrong with a little compression, especially in the cost of production and post production.
Early one afternoon some 20 years ago a jovial young boffin called Peter Swinson of Rank Cintel demonstrated how, by scanning images in ever higher resolution, it was possible to use ever higher levels of compression, generating smaller files.
It seemed a miracle: using Peter's technique we got smaller files, less data and yet higher quality pictures.
Certainly, his work stimulated many of us to imagine that compression would become rather important from then on, and, indeed, it did.
Since that demonstration, and based to some degree on it, we've applied what has now become known as “Super Sampling” in all of our successive visual effects and post production work, and it has served us well. This technique uses over sampling, or using more resolution than the end product requires, so as to reap both compression benefits, better aliasing and higher quality images.
Through our many experiences, we've also come to the conclusion that small amounts of compression can be used effectively for mastering, editing, grading etc. in a pro-active way to reduce costs, and reduce storage requirements and expense. What we mean by this is that uncompressed formats are usually several orders of magnitude more data than their lightly compressed versions and, as the lossless compression is indistinguishable from the “uncompressed master”, can be somewhat inefficient if used for post, as they will inevitably cost more to master. We have also noted that there are not issues with other, downstream compression schemes such as those used by internet streamers.
In the case of the Blackmagic 12k, for example, the uncompressed image of 12,288 x 6480 pixels needs 7.220 Mbytes/sec to playback at 60 fps in post, whereas the lightest compression of 5:1 requires a more modest 1.500 Mbytes/sec. The point is, since we can’t see any difference between the two, surely needing only one fifth of storage requirements and lighter processing in post are advantages. And, indeed they are.
What this translates into is that a cost per hour for a post facility working in uncompressed is often orders of magnitude more that a lightly compressed version of the same. True, fixed facilities in some major cities often also have high overheads to contend with, a scenario that springs to mind particularly with Soho, London and its whopping rate cards, but also the underlying technology in many of them can no longer be the “cutting edge" it was when originally acquired, and may indeed not be able to handle these new very high resolution cameras.
8K Technical notes
Some digital camera examples
From Red Digital Cinema's website: "Raw is the revolutionary wavelet based compression codec that unlocks unbelievable image capture potential and creative flexibility. allowing creators to capture anything from 4k to 8k raw with manageable file sizes and visually lossless image fidelity"
From the Black Magic website: "Black Magic Raw is a modern codec that’s easier to use and much better quality than popular video formats, yet has all the benefits of the reduced amount of data. It offers us visually lossless images that are ideal for high resolution, high frame rate and high dynamic range workflows. Incredible image quality, extensive metadata support and highly optimized GPU and CPU accelerated processing make Blackmagic RAW usable for acquisition, post production and finishing". Specifically, the Raw format uses full bandwidth colour information, and which makes it excellent for colour grading.
ARRI now offers a more post friendly workflow with Pro Res XQ available on the more recent camera models. The current maximum resolution is 6.5K.
Sony Alpha Raw
From the Sony website: "A file with the ARW file extension stands for Sony Alpha Raw, and is, therefore, a Sony RAW Image file. A raw image format means that the file hasn't been compressed or manipulated in any way; it's in the same raw form it was when the camera first captured it". Sony Alpha Raw files are leggible wth Resolve, so can be input without transcoding. The current maximum resolution is 6.5K.
Information on digital cameras from Netflix
Netflix specifies camera models acceptable for their productions, and the other streaming services seem to follow suit. These requirements mean that the ARRI Alexa, ARRI Alexa Mini, and ARRI Amira cannot be used on original Netflix productions. If you want to use an ARRI camera then you can choose between the ALEXA 65 or the recently added ALEXA LF.
This is because Netflix specifications state that a camera must have a true 4K UHD sensor (equal to or greater than 3840 photosites wide) and that 90% of the total runtime of a final program should be captured on an approved camera. The only exception to the 90% rule is for nonfiction content, where the threshold can be more flexible.
Also any cameras other than the primary camera (crash, POV, drone, underwater, etc.) must be approved.
Calibration and Steve Shaw's
Anyone mastering images for later viewing on any display must use an accurately calibrated grading display. There is no alternative to this, regardless of the expected viewing environment or relative accuracy of the final displays the material will be viewed on. This includes material expected to be viewed in a theatre; on home TVs; on the web; on mobile phones & tablets, with digital signage - in other words, anywhere.
Calibration is achieved by the use of a probe, which measures the amount of light in the environment as a first operation. Steve Shaw of Light Illusion states that the ambient light for Home Entertainment should be 10 nits (just under 3 foot lamberts). The room is set at 10 nits by regulating the lights and using black cloth.
Care should be taken to avoid as far as possible reflections on the grading screens, and the room itself should be preferably without windows to exteriors with their associated changing light. One fairly inexpensive solution to expedite set up time and the amount of ambient light is to use a pop up black gazebo such as the one in the link and make a "room within a room".
Once the ambient light level is ascertained and then set up in the Light Illusion software, then the probe is used to measure the screen's brightness, gamma and colorimetry with a sequence of specific colour patches displayed. Finally, the software elaborates a 3d LUT from the readings, and this is applied to the images sent to the Apple XDR screen so that it will display the colours, gamma and intensity we need for specific projects.
For theatrical deliverables, image calibration can be achieved if there is a room that can be darkened so that it can become as dark as a Cinema. The afore-mentioned pop up gazebo may work well for this too.
Calibration for digital signage needs to take into account whether this is going to be displayed in interiors or exteriors and the Apple XDR screen rotates easily to become a vertical display for this and other vertical formats. Events will also have a particular calibration depending on the lighting at the event itself.