While companies like Sony and Canon are just starting to offer their first affordable 4K cameras, RED is forging ahead to even higher resolutions. 5K is so 2012. Thanks to RED we're about to get 6K. But it's not all about resolution, we're also getting other improvements with their new Dragon sensor, including better sensitivity and more dynamic range (Dragon will likely surpass the best 35mm film stocks with the latter). With all of these advancements comes a potential data headache for any working professional -- what are we going to do with all of this data? RED and Sony both work in compressed RAW formats, but 4K compressed RAW is still a lot of data, and 5K pushes that even further. How will we deal with 6K RAW? According to CEO Jim Jannard, that's where the power of RED's wavelet codec comes into play.
RED's CEO Jim Jannard explained, in a post that was deleted after this article was completed:
The Dragon sensor has less than half the noise of the Mysterium-X sensor. That means that you can use approximately twice the REDCODE RAW compression settings to get the same quality noise results at 6K as you have been getting with 5K. Dynamic range increases are not affected by compression settings. It also means that you can raise the ISO of the Dragon by (more than) double to get similar results as compared to the Mysterium-X sensor.
In the end... 6K is 45% more resolution than 5K. With double the compression settings, the files are actually likely to be SMALLER than what you are currently shooting. And dynamic range goes up significantly.
More resolution. Much more dynamic range. Higher ISO performance. Smaller files. Win.
And you don't have to buy a new camera...
You don't have to buy a new camera if you are a current EPIC owner, that is (SCARLET owners will receive a yet-undetermined trade-in credit if they want to upgrade to Dragon).
Noise and grain generally fight against compression schemes since they are random fine details, which brings us to Jim's statement. It's unclear why the post was actually deleted, but it brings up an important topic nonetheless. Theoretically having a cleaner image will mean less random fine details, and so the codec can compress more information for similar visual quality. That all makes sense, but it also means that you're physically losing more information the harder the compression scheme is working. REDCODE is a lossy wavelet codec, but in many cases if you're working with one of the lower compression settings, the results can end up being visually lossless (though they are not in a mathematical sense).
My knowledge of codecs certainly isn't near engineering levels of understanding (not by a long shot), but I do know that when you're dealing with lossy video codecs, even REDCODE RAW, the more you compress it, the more information you lose that you'll never get back if you need to push it in post-production. If you're currently using REDCODE 6:1 on an EPIC at 5K, does that mean Dragon could get the same visual quality at 12:1 on 6K? If so, that would go a long way towards making 6K file sizes manageable... but the smart money is likely on a more modest increase in the compression's efficiency (more modest than 2X, that is).
I think this whole conversation brings up an even bigger point that many are just starting to find out: uncompressed RAW -- especially 4K RAW -- can be very unwieldy. For example, Canon's 4K RAW is 1 Terabyte per hour. The forthcoming Aaton Penelope Delta shoots 3.5K RAW CinemaDNGs at a similar data rate as Canon's RAW implementation. Using the same CinemaDNG codec, 2.5K RAW on the Blackmagic Cinema Camera is roughly 512 Gigabytes per hour. RED 5K at 5:1 compression (which is what was used for Prometheus and Spiderman) is a little over 256 Gigabytes per hour. Hard drives might be getting cheaper, but when you consider your footage needs to be backed up somewhere, the numbers add up very, very fast. For all of the misinformation that gets spread about the company, when it comes to actually having to go out and shoot something, REDCODE RAW is one of the best supported, and most efficient codecs out there. REDCODE may have its issues (like any compression method), but your workflow is as important, if not more important, than the camera itself.
Let's take a quick detour into what "wavelet codec" means, since we mentioned it above and since we receive comments all the time from people assuming that editing 4K RED files on a laptop or desktop computer without special hardware is difficult. Via RED's site:
This efficiency is achieved in part because wavelet compression encodes image features at different scales separately. A wavelet file therefore contains a low-resolution base image, plus progressively higher resolution components — all the way up to the final full-resolution image, which is the net combination of all these:
As a result,low-resolution previews can be generated without having to process the entire high-resolution file. For example, a quarter resolution preview could be generated from just the above base image plus the leftmost component. Wavelets therefore make it easier to directly view and edit large videos on desktop computers.
So if you're editing a 4K file at 1/8 resolution, it's basically like editing a 512x270 video file. This is one of the other aces up RED's sleeve, especially now that Premiere Pro, Final Cut Pro X, and other NLEs natively support the .R3D format.
When it comes to more efficient compression thanks to lower noise, we'll have to wait and see if Jim's comments are accurate, but there is another bit of good news from Jarred Land. It seems Dragon footage will work with the current RED Rocket cards, and he also mentioned in the forum that there will be a newer heavy-duty card at some point down the road:
Its inevitable... but it's not a priority right now. It wont really be a replacement, The Turbo Rocket is going to be more for the power mongers out there.. insane speeds with some expensive horsepower. Current Rocket will work just fine, and you can double current Rocket up if you want to go even faster.
He also mentioned that we probably won't see a price reduction on the current card, and the new Turbo Rocket will cost RED twice as much to make -- so it's not going to be cheap. As technology gets better, every piece of the workflow should see improvements, and it looks like that is no different with the Rocket card. He didn't mention when we might see this new card, so I wouldn't expect it for some time.
What do you guys think? If anyone more knowledgeable than me can add to the compression conversation, we'd love to hear your thoughts in the comments.