Sensor Split Fix Opens Up New Camera Possibilities?

[Zack Birlew]

Hi, everybody, well, it seems RED finally got a fix implemented for the V-Raptor sensor split with the new version of REDCine X coming out and it appears to be in line with previously discussed methods to fix it. Now, getting technical, if a sensor split can be fixed with two sensors, couldn't this be expanded for a potential medium format, large format, or other camera with four or more sensors? If so, what do you all think could come from that and/or what would you like to see? (Disclaimer: I don't work for RED, I'm just using a basic naming theme for the cameras, also multiple choice is fine)

 

[Les Hillis]

The question of RED making cameras with larger sensors came up on a live-steam and the answer was basically (my paraphrasing) "We'd like to, but who pays for that?"

RED could probably make any or all of those cameras on that list, but they'd have to be made to order.

It's like the DSMC2 Firmware requests, if I were serious about getting anything like that made by RED, I'd contact them directly and find out if they're willing and how much money they'd need to see up front before they invest in the R&D, then I'd start a Kickstarter campaign or something to try and raise that money.

It's a big ask though imo. Bigger sensors would appeal to a fringe market, would have lower yields and be harder and more expensive to make than the 'bread and butter' sensors RED already makes.

Despite their origins and original aims, RED is a commercial company that needs to make sound business decisions. I don't think there's anyone here or in RED who wouldn't like to see them making large-format motion-picture cameras, but the numbers still need to add up (millions of them).

Nothing wrong with a poll to gauge or raise interest, and nothing personal intended in my respose. But yeah...what would it actually take for it to happen?

 

[Phil Holland]

I've asked as early as 2017 and somewhat frequently since. There are real market challenges for a proper larger format digital cinema camera. And yes, the question as to who realistically pays for that came down to me about about 6 other filmmakers realistically, though I do think rental houses would get behind it nominally as well.

There's two approximate trajectories to do that right. None are listed above however. And though the novelty of a 4x5 and 8x10 motion picture camera is enticing, that's not what anybody who is going to pay for such a beast is after. Is this camera $50K, $75K, over $100K?

Even with my desires to go bigger in format and resolution, the VV format size is really the sweet spot. Even from an FOV and perspective, erm.., perspective. And S35 remains the workhorse and backbone still.

Not surprisingly the market trends on this front also have applied to motion picture film over the years.

I think one thing RED does well is a bit of refinement in all directions that leads to new sensors, new bodies, etc. VV was their big "we are turning left while industry is going in a straight line". It is/was a big move. And I appreciated that enough to buy multiple VV bodies.

One thing to watch is what occurs with the Middle Format sensor size, the 44x33 ones, as there will be continued progress in development with several players utilizing that now. None make digital cinema cameras, but who knows what's around the corner on that front.

The market is very complicated and yet incredibly simple when you graph it out. Hard truths are higher end professional cameras sell far less than the sub $10K market. That is the current ledge threshold in our industry. And of most interest is when companies reveal their sales at that tier when it comes to professional versus hobbyist/enthusiasts buyers. And professionals aren't in the majority there, not by a long shot. Which puts a heavy burden on what professionals are looking for, where it should be priced, and what it should be able to do.

And the absolute hard reality of that is professionals need contemporary tools to tackle the modern filmmaking landscape. Just the features in DSMC3 that are geared towards Virtual Production and even Broadcast application show that RED is aware of the wider scope while still aiming at things like Autofocus when tends to be attractive to a whole different world.

 

[Zack Birlew]

Perhaps, however, Alexa 65, Fuji GFX series, Hasselblad 4K, Kipon Baveyes, and even Phase One would disagree. Making a new age medium format and large format sensor could have major developments on those fronts, of which there are plenty of pre-existing film camera bodies and lenses to take advantage of something as simple as a sensor in a box digital back setup.

What the theory proposes is that RED wouldn't even need to make a new sensor, just stitch together more of what they have to meet the coverage levels as needed. If it takes four sensors to meet 645 then that would be at least $50,000 since V-Raptor is $25,000. If it sakes six to meet 6x6 or 6x7, then $75,000 and then we get into 4x5 and 8x10 which could be in the $100,000 to $150,000 range.

Granted, 4x5 and 8x10 large format are a very particular market but it's a matter of what can be done with what is available and if those were possible, why not at least try it and see how the market reacts? RED already has made cameras for NASA, imagine the clients and artists that could utilize a new selection of specialty cameras? The real point of it all is that RED could easily be making the biggest product(s) nobody realized they needed.

 

[Phil Holland]

I'm on location at the moment, but reread my post please as I highlight the 44x33 world. You're correct mentioning the A65, but in your poll you outline a great deal of sensor technology and format sizes that wouldn't be applicable or desirable to even fringe cinematographers.

If you're curious I have approached several manufacturers to build a larger format camera, there is even a production using such a beast currently which is a custom solution, but we're talking a sum of cash that is likely way outside the general commercial filmmaking market. There's a few players aiming to do "the next big thing", all of which is fueled by significantly high end budgets or backed by much larger companies. About all I can say about that at the moment really.

4x5 and 8x10 are about 10,000 miles away from being desirable for motion picture production, still camera, different conversation. Anything up to about an 100mm image circle I would say is in bounds however when it comes to a tool most and the few would be after. And in that case optics are there or somewhat there to play ball.

 

[Les Hillis]

Some numbers from the other thread where larger formats were discussed (pre-Raptor S35 and DSMC2 demise) -

645 Format Film
9216x6830 (1.35:1AR)
63MP (62.95 megapixels)
56.00x41.50mm (69.70mm image circle)

To match that with the RED sensor that has the maximum resolution per pixel you'd end up with -

Helium 16K 645
16000 x 11850 (1.35:1AR)
189.6 MP
58.4mm x 43.3mm (72.7mm image circle)

For a closer match to the RED sensor that has the most similar pixel size to 645 Format Film you'd end up with -

Gemini 10K 645
10000x7406 (1.35:1AR)
74 MP
60mmx44.44mm (74.67mm image circle)



Was looking at it from a resolution angle before, 16K & 10K sensors and all the data and memory and processing that would have to go with it, but what's more apparent now is that 645 1:35 AR.

That's effectively a special presentation AR that you'd either be cropping the top and/or bottom of by default or basically shrinking the image by letter-boxing the sides to make the image fit a 'regular' screen.

And if you're going that far, why not just go for a 70mm IMAX presentation sized sensor?

Phil's handy sensor chart shows what that really means though in relation to current sensor sizes - https://phfx.com/tools/dpThing_phils...ensorChart.png

Also from Phil's Format Compare tool -

70mm 15-Perf
11586x8658 (1.34:1AR)
100.31 MP
70.40x52.61mm (87.89mm image circle)

So roughly speaking you'd need something like a 12-16K sensor to cover that.

Funny though how going bigger than 645 and calling it IMAX-sized actually makes it sound like a more viable proposition, for those who could invest in the 60K+ price tag.


Edit - Removed a reference to an interesting post above mine that disappeared.

 

[Nick Vera]

Honestly would just be happy with RED adding more height to the Raptor sensor for tall video capture and 36x24mm anamorphic lenses. If they made it 40mmx26mm it would still be the largest FF digital cinema camera out there. 44x33mm is interesting, though you would have a limited amount of lenses to choose.... and would we be going to back to lower readout speeds, less DR, and cameras costing $50K-80K? If anything internal NDs in smaller bodies is more important than going to 44x33mm.

 

[Phil Holland]

Honestly would just be happy with RED adding more height to the Raptor sensor for tall video capture and 36x24mm anamorphic lenses. If they made it 40mmx26mm it would still be the largest FF digital cinema camera out there. 44x33mm is interesting, though you would have a limited amount of lenses to choose.... and would we be going to back to lower readout speeds, less DR, and cameras costing $50K-80K? If anything internal NDs in smaller bodies is more important than going to 44x33mm.

Bigger sensor would = bigger body for internal ND solutions, but unknown if that's even the direction to go.

I don't think RED will go slower than 8ms readout from hear on out. The faster readout and even global shutter has ben very popular industry wide. Shooting some long lens stuff on Raptor this week and it's very nice and different to last gen's readout on those triple digit pans.

My hope, without divulging specifications is wider and taller while maintaining what's good of the now or improving on a few things. But the camera I'm after is likely 4-8X more expensive than DSMC3 at the moment realistically. And power draw and power options are very much a real concern as well as data rates and media.

 

[Björn Benckert]

Even monstro and LF has stitched sensors. So the tech is nothing new.

And even if it's likely possible to stick together more than 2 sensors it also comes down to the fact that larger sensors does not only come with positives. Not many understand that a larger sensors / backplanes actually needed more light or higher sensitivity or you can not create the same image as with a smaller sensor. Sure the higher resolution counter weight that a bit, you get less noise when downscaling it etc. But that only goes to a certain limit, the bigger sensor will hit its floor before the smaller sensor.

Simply put. as the light from the given FOV is spread out on a larger area, each pixel gets less light. It's the basic math / physics that makes large format lenses slower than normal lenses. The difference just between s35 and FF is about F1.2 and F1.8 and when the size goes up that number is exponential.

But sure there are plenty of cases where a large sensor would be great to have but if let say red makes a 4x raptor sized sensor it means your need 4x the light not to hit the sensor floor. And to create the same DOF for the same FOV you need even more light.

Example if you shoot a s35 sized sensor using a 50mm at 1.2 with focus set to 1.5m... then to get the same result on let say a Alexa 65 sized sensor you would have to use a 90mm at F2.2. The difference in F-stops would make it so that the larger sensor of the two would need to be 4 times more sensitive or given 4 times the amount of light not to hit underexposure before the s35 sensor.

 

[Les Hillis]

Some interesting points there in regards to the lenses, sensor sensitivity, readout times and power requirements once you get into the larger formats.

Apart from the business side of it, there are some pretty big technical limits and boundaries RED would have to overcome in order to start making larger format cameras.

It's not just a matter of stitching sensors together and adding a bigger cooling fan that's for sure (not that anyone was saying it was).

Could get a bit awkward too venturing into the realm where actual 65mm motion picture film is still being used.

I'm sure if there were enough reasons to do it RED could come up with an awesome camera to cover larger than 65mm formats, but all things considered it would be a huge project, in more ways than one.

 

[M Harvey]

What would you hope to get out of a system built on a medium format+ sensor? What aspects would be better (for a given purpose) in a larger format?

In the days when film dominated, larger formats were often used to improve resolution for larger exhibition formats-- standard formats looked soft when blown up bigger. (And add in the fact that with photochemical post-production you were losing much more resolution along the way to your theatrical print.) My simplified understanding is that improvements in film stock and post-production processes eventually sidelined many of these larger formats, because the studios and theaters felt that the smaller improvements in perceived quality were no longer worth the considerable extra expense of the additional film and specialized projection equipment. 70mm and Imax remained, of course, but as niche products.

With digital sensors from Red and others already so incredibly high resolution at the super 35mm and vistavision/full frame sizes, we already have the equivalent of extremely fine-grained, extremely sensitive film with very wide latitude, and with post-production processes that preserve that. If anything, it was your average lens that needed to improve to match the capabilities of the sensor. And at the same time, that has driven many of us to also seek out softer vintage lenses! So while it may be true that bigger will often be better in some sense, the original logic for the larger film formats of yore has been weakened.

So what would we be after with larger digital sensor formats? There's nothing wrong with enjoying the look of larger formats, but is that primarily what it's about?

 

[Zack Birlew]

Yes, the look of larger formats is exactly the reason to make a bigger sensor. You can't say that putting a Pentax 67 105mm 2.4 on a 35mm camera will give you the same results as it gets when on a Pentax 67 camera, no matter how many megapixels are involved, it's "The Look". Funkier lenses like the Kodak 178mm 2.5 4x5 lens also lose much of the point when being used on smaller formats, medium format being the bare minimum for somewhat of an effect. Arguing about using longer lenses on 35mm to "match" medium or large format is also a moot point because of the strikingly different ways those lenses look on their respective formats. We already see this when trying to use full frame lenses on Micro Four Thirds, it's just not the same at all.

As far as cooling and such, while I make the case for using the exact sensors of the Raptor, RED could opt to use lower resolution sensors to save on power and price for now, such as 4K squares instead of 8K ones. Looking at things like the Achtel 9x7 camera though, I think if it were planned out right there may not need to be much of a bigger camera at all. Even if it were, look at how small RED cameras are as it is, even going back to RED One MX size to start for medium format or large format wouldn't be the end of the world.

For distribution purposes, why do you all assume we're going to only ever have standard movie theatre projection at the most? We could have large venue hologram displays that fill the sky, wall sized 32K televisions, immersive video media event venues, skyscraper size video displays ala "Blade Runner", etc. Just as it is now, having a unique "look" or eye grabbing twist to your images could be even more of a key to success if such systems come into place in the near future and you have to be prepared to have your work displayed on such other known and unknown new technologies to come in the future.

If we had left things camera-wise as they were during 2/3" HD times, imagine how things would have looked now that we have 8K televisions for the home? It would be like how we see television shows filmed on tape now, soft, blurry, low definition, and disappointedly locked in forever. Even simpler, what if we didn't move to full frame and let things top out at S35? All the benefits of FF35 just gone and any sense of "Large Format" just being gone, no Canon 5D Mark II revolution, just S35. That feeling of a big image and the added resolution that comes with it can change not just the visuals of a story but even the story itself. That's why I'm saying we move forward and look at bringing medium format and even large format into the picture and see what can be done because if we don't do anything and "lock things in" and "settle" for what we have then we're just pointlessly stifling any progress we could have as storytellers.

 

[Les Hillis]

It depends how large the larger sensor would be and what you want to do with it.

For me personally, it's all just theoretical.

For others, it would be a day one purchase or rental, or enable their making plans for future uses.

Regardless of how good the current S35 and VV sensors are, there would still be advantages to down-rezzing to those formats from a higher resolution.

Like-wise, not having to stitch together images from the current sized sensors to get higher resolutions would be an advantage.

There would also be advantages to having a larger format sensor for 360° and VR capturing as well.

And as much as I wouldn't want to see 65mm and IMAX film actually disappearing, a better digital alternative is something I imagine a few people would put to good use.

There's also the flow-on effect, similar to how the development of the Komodo had an influence on the development of the Raptors, where the development of a larger format sensor, along with the camera body to fit it in, could have a beneficial influence on future RED cameras with standard sized sensors.

That's all aside from the creative aesthetics of the larger format that would appeal to people, just like it does in the still photography realm.

And it's not like others haven't been catching up and pushing the limits already.

Apart from all the other 4K and 8K cameras around now, on the resolution side BMD have their 12K S35 camera.

There's also that 9X7 9.3K camera - http://24x7.com.au/9x7/

Again, just looking at it theoretically and to explore the idea, it's not hard to imagine how RED could potentially 'own' a big share of (and play a role in the developing and advancing of) that current and future larger format/higher resolution market.

Edit - Didn't see what Zack posted while I was writing, but we make some similar points.

 

[M Harvey]

For distribution purposes, why do you all assume we're going to only ever have standard movie theatre projection at the most? We could have large venue hologram displays that fill the sky, wall sized 32K televisions, immersive video media event venues, skyscraper size video displays ala "Blade Runner", etc. Just as it is now, having a unique "look" or eye grabbing twist to your images could be even more of a key to success if such systems come into place in the near future and you have to be prepared to have your work displayed on such other known and unknown new technologies to come in the future.

I actually assume we WILL have big, exciting jumps in display technology. The point I just wanted to make is that the relationship between capture medium (film/sensor) and resolution of that medium has perhaps changed. And after all, you could make a billboard in 2005 with a 12-megapixel APS-C Nikon. Our own Phil Holland would be on of the world experts on what goes into preparing images for ultra high rez, ultra large displays like your theoretical 32k TV-- and as we know, he'll use arrays of Red cameras for some projects.

And for what it's worth, I happen to really like larger still photography formats-- I even still have some precious 5x7 Portra in my freezer from one of Kodak's occasional custom runs, and I've been following the LargeSense project for years. For me, a lot of that has to do with the working method and engagement with photographic tradition as much as the technical quality of the final result. Still, I would encourage anyone interested in a larger format look to try substituting an 85mm lens for a 35mm lens on their regular camera and seeing if it gives them any of the visual qualities they enjoy. Technical experimentation with the tools we already have can yield some pretty amazing results...

 

[Phil Holland]

Our own Phil Holland would be on of the world experts on what goes into preparing images for ultra high rez, ultra large displays like your theoretical 32k TV-- and as we know, he'll use arrays of Red cameras for some projects.

I have lots of personal and professional reasons to desire both higher resolution and larger format digital cinema cameras, but as I've mentioned I think S16, S35, and VV (LF & FF35) are "the sweet spot" when it comes to format size for general production work.

I don't mind doing array work, it's a good challenge and always fascinating to create those 20-32K images. Several projects going on now that are beyond that even. But arrays are limiting when it comes to many types of shots and filming people want to actually do. I've now been capturing and finishing in 8K for 8 years in 2023. Which is wild. And I'm loving that.

We've used the Gpixel camera head and recording module that is the 9x7 and it does indeed have stunning resolution, but lacks in a variety of directions from an IQ perspective. But there are decent "big" sensors out there now. But likely only appealing to a small market and similarly all have their quirks.

I think the thing RED hit square on the head was being early to the larger format game in digital with that VV format size as well as 8K. They did it early, they did it well, and they've continued to build forward on that format size. The industry as well as gotten behind that format with companies like Panavision and other rental houses that outfitted that direction. Even lens manufacturers supported that effort. But it did take a couple years from announcement for that swing to start after announcement.

I'm patient when it comes to this stuff however. We're filming things on Raptor right now that we've literally waited for the correct hardware to exist to tackle. Similarly whenever we go bigger, I've got stuff geared towards that.

The hard bit is it needs to be done right. And in this case it's going to be a weird dance of a variety of factors which even includes likely another lens mount standard, but maybe we build off of existing stuff. RED's still got the edge technology-wise to do a lot of the things I'm hoping to achieve on the camera side. Raptor as a camera platform is still by far the most powerful camera on the market when you look at all it can do. So lots to be excited by here currently and in the future. Very literally the only gripe I hear now is the omittance of a 1st party or modern EVF. Some grumbles still about the UI on the Touch LCD, but I haven't found anything there that impact productions, but would say there's stuff to improve upon.

The industry as a whole right now is very focused on smaller and lighter weight products mainly as well as lower price points and pricing tiers. Won't say much, but behind closed doors Arri has expressed interest in making less expensive cameras, but with their current methodology likely difficult to achieve. They are also pursuing the high end still which will be evident soon enough. RED arguably is very much competing or exceeding what we consider high end priced bodies right now at about 25-33% the price. 50% if you consider XL. As I've said before though, if RED goes back into the larger price point realm, I'm very curious what they could achieve and what that would look like.

I say all of this with a needle pointed that all formats are valid and all that goodness. I am still of the mindset that when it comes to format flexibility, slowly but surely, we are getting to a place where film excelled. Outside of the Alexa 65, which isn't available for purchase, that means two major formats of interest exist to be covered. And I think only one of those two formats is something to even aim closely at for so many reasons.

One thing I like to point out about Raptor 8K VV in particular, though the format size somewhat different, what we consider the modern take on a digital VistaVision format actually exceeds the maximum frame rate from the film camera itself. Raptor VV tops out at 120fps. VistaVision 72fps. Not the most important metric when it comes to everything, but it's the first time that happened this way.

We'll see how this pans out. We have a whole generation or two of cameras to go through before I think we'll see a big swing towards any of this. S35 and VV/LF/FF35 is where the party is currently.

 

[Björn Benckert]

I think the race will be more about pixel density than sensor size. Arri kind of went back to the S35 format and Pavels 21k camera shows quite well that the pixel density race is far from over. Sure a 65mm camera with Pavels density would be something really cool. But still very few are able to show something bigger than 8k and most post pipelines are not even at true 4k yet.

I'm down in Prauge on a production for 14 days where they shoot Arri S35's and I have the raptor as my side / plate camera. So I could not resist but to shoot qutie a few of the setups side by side with the s35 that has optimos and summilux-c´s while I had my rehoused FD´s on the raptor... I know which look I like better to put it kindly. Being able to crop in to get a cleaner edged frame, or go full FF to get more character of the lens if that's what I'm looking for all while still having plenty of resolution and framrate Raptor is great in that way. All while the S35 and the summilux-c almost looks kind of dull not the most arty of setups... I kind of forgot how it looks with perfect glass and S35 with full coverage, been so many years of Venice, LF and Monstro.

Then also the ease of having proxies in camera. Playback... people complain about the red playback, well try go shoot something overcrancked on the s35 and play it back to client... I did not know it was that bad, the camera stands and buffer and can't really play back the clips straight of the bat and then it's even cropped 16:9 100fps captures.

Also start digging the new dsmc3 touch. Sure some things in the menus are not there yet, but then there are also plenty of new things, the pinch zooming while filming is a god send. and the swipe to different screen setups, histograms, focus etc is super nice just as all the web/cloud stuff. Then battery time and everything else, the S35 looks like a generation late in comparison as I see it.

Will be interesting to get back to office and start to compare Noise etc.

 

[William Long]

Hej Björn, wouldn't mind seeing some comparisons if and when it's possible?

 

[Les Hillis]

One avenue of future development that I think has potential is the design of sensors and lenses that are much more dependent on computational processing to get to the 'captured' image.

Kind of like how the latest phone cameras have a lot of processing going on under the hood just to make a decent image out of the kinds of lenses and sensors the phone form-factor requires.

It's already happening to some extent, but I think there's more potential there with digital cinema cameras to do things sensor-wise (and also in combination with specifically designed lenses) that might not have made sense or even been possible before.

Analogous to modern fighter jets that are so aerodynamically unsound that, in order to just stay in the air, they need an onboard computer to make all the flight adjustments that a pilot just wouldn't be humanly capable of making in real time.

Basically, by design, deliberately allowing one or several elements in the data-gathering chain to be (or remain) 'compromised' in order to meet another criteria (resolution, DR, colour fidelity etc.), then 'adjusting' them, by in-camera or post processing computational means, in order to arrive at an overall less compromised, better result.

Not so as to lower the standard or precision of sensor or optics, or to implement something like what we know of as built-in noise-reduction or dual-gain sensitivity etc., but to push the limits of and integrate at a more fundamental level the image processing/data interpretation, and to let that influence the sensor (and ideally lens) design.

TLDR - Upping the coding capabilities and letting that influence and advance the hardware design, to get to a result that either alone cannot reach.

 

[Brian Timmons]

One avenue of future development that I think has potential is the design of sensors and lenses that are much more dependent on computational processing to get to the 'captured' image.

Kind of like how the latest phone cameras have a lot of processing going on under the hood just to make a decent image out of the kinds of lenses and sensors the phone form-factor requires.

It's already happening to some extent, but I think there's more potential there with digital cinema cameras to do things sensor-wise (and also in combination with specifically designed lenses) that might not have made sense or even been possible before.

Analogous to modern fighter jets that are so aerodynamically unsound that, in order to just stay in the air, they need an onboard computer to make all the flight adjustments that a pilot just wouldn't be humanly capable of making in real time.

Basically, by design, deliberately allowing one or several elements in the data-gathering chain to be (or remain) 'compromised' in order to meet another criteria (resolution, DR, colour fidelity etc.), then 'adjusting' them, by in-camera or post processing computational means, in order to arrive at an overall less compromised, better result.

Not so as to lower the standard or precision of sensor or optics, or to implement something like what we know of as built-in noise-reduction or dual-gain sensitivity etc., but to push the limits of and integrate at a more fundamental level the image processing/data interpretation, and to let that influence the sensor (and ideally lens) design.

TLDR - Upping the coding capabilities and letting that influence and advance the hardware design, to get to a result that either alone cannot reach.

I agree with this.
It is true that there is a good deal of image processing going on in a lot of these smart phones to help those small phone sensors
look and behave like much larger ones. They are steadily getting better at doing it.

RAW is a great starting point for additional image processing to help enhance an already strong image.
In a sense advancements like IPP2 are essentially doing a good portion of enhancement processing with REDcode already.

I always thought a more aggressive form of noise reduction (or perhaps texture control) would be a
welcome addition to RCX (and other editing software via SDK).

Perhaps some more aggressive forms of processing may be introduced.
As long as it doesn't permanently affect the original RAW capture what is there to lose?

Brian Timmons
BRITIM/MEDIA

 

[Les Hillis]

More processing that has no effect on the RAW capture should be possible.

What I'm imagining and kind of just spit-balling here really would definitely be happening before the RAW capture though.

Where you attenuate and otherwise alter chosen aspects of the light before it hits the sensor, the sensor that is designed to receive that attenuated light and alter it further, in such a way that the resultant 'signal' can be made sense of and translated into recognizable images by previously unused and unrequired methods of computation.

Like milk. They get the stuff from the cows, pull it apart into its constituent elements and then put it back together again in various ratios and sell it in its new forms like it's just been pastuerized and cleaned up and altered a bit.

That's what cameras are doing with light too, but with the aim of trying to get the 'whole milk' to survive the process as intact as possible. Nothing wrong with that. But with the computational ability now to 'make sense' of more complex problems, metaphorically pulling the light apart and messing it up in a way that suits a sensor designed to receive that, then re-ordering it however you want could have advantages.

You could probably even have your favourite lenses characteristics calibrated and applied to the image later, ray-tracing style, if there wasn't any room left to attach them in front of the sensor when recording.

Anyway...whatever...still amazed by, and challenged to achieve, the high quality images that can be captured by the currently avilable RED cameras.