Ursa Cine - Facts and findings

If film or cameras are a means of communication, the tools for a mode of expression,

Sure, the same way a pencil is a means of communication. Just don't ask it to write a novel or express itself.

But this dream based in equipment is apparently deathless, so let it spool on....

This is a boffins camera with much less if any input from consumers. The last time this happen was the Big Ursa and we all know what happen to that one. It will be £7k by year's end if BMD is to recover losses.

What you are forgetting is that Black Magic has been gaining ground in the streamer market. It's barely made even a vague hint of inroads in the Hollywood market admittedly, but I have personally met DoPs who chose and were very happy with Black Magic cameras for Netflix shows. Hell if I remember which one now since there must be 47 Ursa Mini models floating around by now, but still...

And forget not that Black Magic developed the OLPF for the 12K SPECIFICALLY in response to people testing it out in virtual production, where the ludicrous resolution is useful in post.

I would just add to this there are only a few ALEXA 65 available in the World for rent only and they dont sell...
This really means democratisation.

Further, I recently when casting for low budget short film had actress from Hollywood answer my call.

I did not end up casting her for un-disclosed reason, however you ll be shocked how many great actors even from Hollywood are keen to work with up and coming directors that have a project sounding interesting...

You advertise your film audition with words like shooting film on IMAX format and you as director have a few credits or short film awards, music videos and sounding like you are legitimate creative, there will be
lots of talent chasing you. Most actors and cast want good exposure more then a money...
my 2 cents ...

IMHO, it would be necessary to offer service contracts for the new Cine too, not only the future ‘65’.
Image-wise it’ll be up there with the best, but for the kind of productions that may want it, fast and reliable service is the key.
I’ve been talking with folks from Arri, and they confirmed that a very substantial part of their prices is financing the service infrastructure. From my own experience, even Sony doesn’t compete in that field.

I think that's part of why Arri makes cameras that no one but rental houses can buy. It is after all the largest movie rental house in the world, and the only one that you can go to for renting camera gear pretty much anywhere. Sunbelt (new owner of William White) is big and has great customer service, but not nearly as extensive a network as Arri.

Irony is that I believe Coppola made an Indie Production with Megalopolis and that used the Alexa 65.

Real irony is that part of why Arri introduced a 4K Alexa is that so many DoPs rented Alexa 65s in order to satisfy the Netflix 4K acquisition requirement... and windowed the sensor to 4K, turning the 64 into an overgrown Super35 camera.

I really can not wait for CINE 65 to come out,
will be renting it almost immediately once released.

Not sure what purchase price to expect, I would imagine it will be north of $50K ??

I think that's part of why Arri makes cameras that no one but rental houses can buy.

This is what Panavision always were in the film days; they lost their way in the digital age. Once they were the only game in town, at the top; Arriflex were very much in the shadows then. So it can always change; nobody stays top forever. The demands of large feature or broadcast production, make it increasingly impractical again for owner/operators, particularly where more than one or two cameras at a time are needed.

As truly impressive as BMD's new Cines seem to be, and I am very impressed; they will have to seriously up their support and service strategy, for these models, if they are to be taken seriously there, no matter their feature set. They can't have a simple consumer 1 year guarantee return to base/replace if defective plan on them any longer, IMV. This kind of professional support will cost some too; so it will be interesting to see what develops, if anything. Possibly some kind of 3rd party extended warranty/support?

Again, I used the math above to determine that the URSA Cine is far more affordable for Indie Filmmakers to shoot VistaVision than it was with motion picture film. That's a great democratizing of the medium.

The comparative cost of shooting on film and digital, no longer has any relevance; almost anywhere. Virtually no small films will be shot on 35mm; and only a few diehards, who can afford to, such as Speilberg and Tarantino, on major films. Whilst media costs have always been and still are often a significant cost to the low-no budget films, they are no where near the major part of the overall budget on most features and TV programmes; even if they were still on celluloid. It's slightly misleading Tim, to suggest huge savings that way or that one particular brand's card/drive media offers those huge production savings, over others either now.

VistaVision was simply 35mm on it's side and apart from running through the gate at a faster, therefore more costly rate, it was never branded in negative film cans as such. It simply was Paramount's answer to rival studios new Roadshow formats in the 1950's, such as Cinemascope. Although not anamorphic; the larger negative area combined with Technicolor imbibition release printing; made it superior in many ways, particularly in the hands of directors like Hitchcock. The approximate frame size is the only similarity with FF digital; since pixel count/resolution can be identical at different sensor sizes, whereas that is impossible with film grain - the bigger the negative, the better the resolution.

As well, the overriding main reason for, higher than 35mm Academy negative frame sizes was increased resolution, for larger huge widescreen cinema screens and/or greater fidelity with reduction printing down to 35mm print release; the added complication, often poorer quality at first, and expense of large format lenses; and greatly reduced DOF, were not seen as real advantages then; these formats were not chosen primarily for a 'look', as a lot of younger filmmakers think now, about larger sensor sizes. It was customary to light everything up or down to one often, favoured f stop, and often that would be around f5.6, where the lenses then performed best; to have consistent and relatable DOF across shots; and to maximise as much DOF possible, unlike the fashion today.

In any case, I very much doubt if any low budget independent, anywhere at any time, has shot on VistaVision ex Paramount cameras.

I would just add to this there are only a few ALEXA 65 available in the World for rent only and they dont sell...
This really means democratisation.

Further, I recently when casting for low budget short film had actress from Hollywood answer my call.

I did not end up casting her for un-disclosed reason, however you ll be shocked how many great actors even from Hollywood are keen to work with up and coming directors that have a project sounding interesting...

You advertise your film audition with words like shooting film on IMAX format and you as director have a few credits or short film awards, music videos and sounding like you are legitimate creative, there will be
lots of talent chasing you. Most actors and cast want good exposure more then a money...
my 2 cents ...

I won't respond point by point -- experience will teach otherwise -- but the biggest "indie" of the last 50 years was shot on Hi8.... The runners-up were mostly mini-DV or early consumer versions of HDTV. Camera developments of the last 10-15 years which supposedly "democratized" the medium -- 35mm quality within reach of everyone -- have made absolutely no impression on Hollywood, arthouse exhibition or the festival circuit. Or on audiences, for that matter.

You're also forgetting that in the film era, cameras were available for free from schools and institutes, and even film stocks could be obtained free or at low cost. And labs cut deals. The impediment was never the cost of the camera. For way less than the price of one year of film school, anyone could shoot a S16mm feature if all we're counting is camera, stock, processing, negative cutting and answer print.

And yet the medium was not "democratized". And persuasive dramatic illusion was only rarely achieved with the resources at hand. Equipment enthusiasm is dandy, but it goes nowhere.....

I think that's part of why Arri makes cameras that no one but rental houses can buy.

This is what Panavision always were in the film days; they lost their way in the digital age. Once they were the only game in town, at the top; Arriflex were very much in the shadows then. So it can always change; nobody stays top forever. The demands of large feature or broadcast production, make it increasingly impractical again for owner/operators, particularly where more than one or two cameras at a time are needed.

As truly impressive as BMD's new Cines seem to be, and I am very impressed; they will have to seriously up their support and service strategy, for these models, if they are to be taken seriously there, no matter their feature set. They can't have a simple consumer 1 year guarantee return to base/replace if defective plan on them any longer, IMV. This kind of professional support will cost some too; so it will be interesting to see what develops, if anything. Possibly some kind of 3rd party extended warranty/support?

Again, I used the math above to determine that the URSA Cine is far more affordable for Indie Filmmakers to shoot VistaVision than it was with motion picture film. That's a great democratizing of the medium.

The comparative cost of shooting on film and digital, no longer has any relevance; almost anywhere. Virtually no small films will be shot on 35mm; and only a few diehards, who can afford to, such as Speilberg and Tarantino, on major films. Whilst media costs have always been and still are often a significant cost to the low-no budget films, they are no where near the major part of the overall budget on most features and TV programmes; even if they were still on celluloid. It's slightly misleading Tim, to suggest huge savings that way or that one particular brand's card/drive media offers those huge production savings, over others either now.

VistaVision was simply 35mm on it's side and apart from running through the gate at a faster, therefore more costly rate, it was never branded in negative film cans as such. It simply was Paramount's answer to rival studios new Roadshow formats in the 1950's, such as Cinemascope. Although not anamorphic; the larger negative area combined with Technicolor imbibition release printing; made it superior in many ways, particularly in the hands of directors like Hitchcock. The approximate frame size is the only similarity with FF digital; since pixel count/resolution can be identical at different sensor sizes, whereas that is impossible with film grain - the bigger the negative, the better the resolution.

As well, the overriding main reason for, higher than 35mm Academy negative frame sizes was increased resolution, for larger huge widescreen cinema screens and/or greater fidelity with reduction printing down to 35mm print release; the added complication, often poorer quality at first, and expense of large format lenses; and greatly reduced DOF, were not seen as real advantages then; these formats were not chosen primarily for a 'look', as a lot of younger filmmakers think now, about larger sensor sizes. It was customary to light everything up or down to one often, favoured f stop, and often that would be around f5.6, where the lenses then performed best; to have consistent and relatable DOF across shots; and to maximise as much DOF possible, unlike the fashion today.

In any case, I very much doubt if any low budget independent, anywhere at any time, has shot on VistaVision ex Paramount cameras.

At the same time, in the 1970's we saw VistaVision return for VFX. ILM used it to great effect making Star Wars and other major films of the late 70's into the 80's. The benefits it added there are as you said: reduced grain. And, that was pivotal for compositing.

Today's young filmmakers however can learn from the benefit of shooting f/5.6 on VistaVision.

35mm Film 35mm at F4

20240411-KODAK-GOLD-200-05.jpeg (429.45 KiB) Viewed 5891 times

The Above image is 35mm film shot at f/4 on a 35mm lens.

35mm Film 58mm at F8

20240323-KODAK-PORTRA-400-R2-021.jpeg (365.39 KiB) Viewed 5891 times

And, the above is 35mm film shot at f/8 on a 58mm lens.
I shot some comparisons digitally going to f/1.4 and easily people think the film looks better because embracing the higher f stop and greater depth of field you have a better sense of the environment.

I honestly can't wait to try shooting that with an URSA Cine 12K. I may even rent the camera and some of the Ancient Optics Minolta Rokkors in order get that classic feel of the old Minolta glass I love shooting with my Minolta film camera.

At the same time, in the 1970's we saw VistaVision return for VFX. ILM used it to great effect making Star Wars and other major films of the late 70's into the 80's. The benefits it added there are as you said: reduced grain. And, that was pivotal for compositing.

Today's young filmmakers however can learn from the benefit of shooting f/5.6 on VistaVision.

There were only a handful of these VistaVsion cameras ever made. Early ones were adapted Mitchell cameras on their side with enlarged gates. Later 'Elephant Ear' ones were made by or for Paramount. Later after the format had been largely abandoned, ILM and others used them for VFX plates because of the greatly enlarged negative area and higher precision registration, essential for good quality composite matte work. Nowadays with 'perfect' digital that is totally unnecessary.

Aesthetic concerns weren't really the foremost reason for shooting at a target, f5.6. As I say, it was the very narrow sweet (or any) spot of these early large format lenses and for non-jarring consistent DOF shot to shot; regardless of focal length; which was also very limited in choice. Added to which the very slow early tri-pack film stocks, of 25-50 ASA, that needed a huge amount of light, brought huge restrictions. Today's digital cameras and lenses have massive exposure flexibility in dynamic latitude and sharpness, even closer to full open aperture, that they then could only wish for.

Apart from B&W Noir and horror, most big studio productions then were high-key lighting; often mandated by the studio: MGM musicals for example; outside of the full control of DOPs or directors. Today most films seem to favour a sort of same looking low-key; even though we have HDR and brightness that could never hit a cinema screen then.

A while back I posted some stills from Vertigo, that showed interior shots with quite remarkable deep focus, which is quite common throughout that film; a remarkable achievement then by Robert Burks and still to me today, one of finest lensed movies I have seen. In that thread 'What is Cinema?', it would be worth taking into account such work, in light of what some might consider necessary to be cinematic now. You hardly ever saw, despite such slow stock and large frame sizes, exterior backgrounds out of focus, except for CUs, that would logically happen, for our eyes too.

Very nice shots though BTW.

Today's young filmmakers however can learn from the benefit of shooting f/5.6 on VistaVision.

What you're also overlooking is the relationship between audience expectations and budget. "Indie film" to the extent it has any market these days (and is actually "indie") depends on allowances made by audiences. Lack of professional sheen usually plays to the advantage of such movies. Create a disconnect between the apparent quality of image (mainly resolution in this case) versus performances or lighting, and you've got a problem.

What an audience might have regarded as an interesting if flawed movie if it looked like a bargain basement effort is suddenly just a bad movie.

VistaVision for the typical American indie? I can't believe anyone who's actually familiar with the genre would see any hope or value there.

As Rodney Charters , ASC , once said , " If you tell your Story Well , at Any Resolution " .....

Words of Wisdom ...

Maybe we have to Cool Down ...

Regards , Mike ..

… Create a disconnect between the apparent quality of image (mainly resolution in this case) versus performances or lighting, and you've got a problem...

Excellent observation and summary of the challenge of superior technology only working well as other tools and techniques and talent keep up.

I suppose that most ‘successful’ indie movie referenced earlier was the Blair Witch Project. Talk about low tech and low technique and talent off the chart (in a bad way), it’s viral success may have been the perfect match of low quality components all around with my reaction to the effort as ‘crap’. Tells you how little I know! Found footage films certainly appear to have a place, but there are many non-technology factors that make something go viral. It might be shampooing your hair with toothpaste. It might be knowing an influencer who owes you a favour that kickstarts interest among their million followers. Whatever it is, it’s more about manipulation of the viewer than an appreciation of fine art.

So I don’t see much of a future for the URSA Cine 65 in found footage films. But it could have a significant role in bringing great stories to the screen for a great number of creators and their audience.

If they can get this next gen camera out by 2025, in significant quantity, with key early adopters (who could have easily chosen to rent an ARRI 65) which adopt the role of social media influencers, shooting with Panavision LPL lenses, well… talk about disruption of an industry.

With respect and not trying to belittle anyone’s sincere efforts in the military, in days gone by when you had an adversary with a great navy and want to challenge their authority on the high seas, you built the Bismarck and hoped for the best. Today when you don’t really have a serious navy, you launch underwater drone attacks and try to defeat the adversary’s vastly superior navy.

ARRI is that adversary now, really. A great company, great business plan, wonderful support, and so on all of which helps them control a certain market. I think RED and Panavision thought they could disrupt that, but with the right partner relationships, it looks like BMD will be more successful. It’s not just technology that wins the day, but it is one tool in the kit.

URSA Cine 12K is the edge, however URSA 65 is pure revolution...
We are confusing here creative directions and ability with capability.
Most of the great directors started with short films and basic tools and small budgets.
You cant win the Olympics without going thru the ranks, training and sweat and takes also failures.

Camera is an object not a creative direction.
Attractive object with huge capability , not creativity.
That is all in hands foremost of director and cinematographer and other departments.
Directing is art form, cinematography is also am art form so do others departments.
Camera is important tool, I see no harm in having more capable tool.

Rant over.

I just can’t shake the background of Grant Petty in figuring the future of these two new cameras and where they are going tomorrow.

Let’s remember DaVinci Resolve for a moment. Didn’t that product (mostly hardware) suite cost $100,000? Today it’s given away (the software at least) with optional supporting hardware at less cost. With astounding capabilities continually enhanced at no cost. This leader thinks differently. His motivations may be unique among successful CEOs. He wouldn’t last a day at any of the Magnificent Seven dominant companies. He’s clearly a madman. He now finally has his focus set on a new objective but he still has the same motivations.

We are flowers that come from tiny seeds, that sprout from dirt, that bloom in the sun, and that wither away after producing new seeds for future generations. What do you really want to accomplish on that journey so many are fortunate to share? Where do you place your efforts? What do you find enriching? What gifts will endure? How will you be remembered, if at all? We should all be blessed if we are able to answer any one of those questions. An answer isn’t in a paragraph in a blog or a deposit in your bank or a cup on your mantle—an answer is in your good work.

I just can’t shake the background of Grant Petty in figuring the future of these two new cameras and where they are going tomorrow.

Let’s remember DaVinci Resolve for a moment. Didn’t that product (mostly hardware) suite cost $100,000? Today it’s given away (the software at least) with optional supporting hardware at less cost. With astounding capabilities continually enhanced at no cost. This leader thinks differently. His motivations may be unique among successful CEOs.

Yes Rick, in the nineties, I could only dream of owning such tools; in the noughties the gulf between prosumer and high end was a chasm too far. We were almost pleased as punch with 1/3" chips, compromised tape codecs and workarounds - and those cameras cost more than an Ursa now too - the JVC HD101 I bought then, for example; nowadays the pure picture quality between even a Pocket and an Alexa is much closer; and whereas my first Avid Adrenaline I/O cost 25K, without the HD board mind; now the US4K Mini is vastly better and was a fraction. BMD disrupted that market and made that possible, more that anyone and I am grateful for that

But it takes more than a camera and some great software to make a movie; no matter how good they are. As I said, these things are not the major part of either the budget nor talent in making those great movies. You can still rent a camera and in many ways, if all you really care about is making the best films ever; that often still makes more sense. Manufacturers know though, that a lot of people just want to own their own camera; more than they do perhaps actually making a film; that they might never do in any meaningful way; not talking about anyone here, of course and looking forward to seeing your great work, all

I honestly can't wait to try shooting that with an URSA Cine 12K. I may even rent the camera and some of the Ancient Optics Minolta Rokkors in order get that classic feel of the old Minolta glass I love shooting with my Minolta film camera.

So, you're a fan of those too? Wonderful glass! A Minolta was my first camera when in arts university, while my very first was a two-eyed box at the age of five from my uncle…
I got about all of the Rokkors when they were really cheap, since most could not be adapted to EF and nobody wanted them. The 35mm f1.8 or the 58mm f1.2 are such gems, and the 85mm Varisoft is still a beauty for dreamy portraits, even if you can fake it to some degree these days.

For those new to these lenses, this is a great resource: www.artaphot.ch.

I honestly can't wait to try shooting that with an URSA Cine 12K. I may even rent the camera and some of the Ancient Optics Minolta Rokkors in order get that classic feel of the old Minolta glass I love shooting with my Minolta film camera.

So, you're a fan of those too? Wonderful glass! A Minolta was my first camera when in arts university, while my very first was a two-eyed box at the age of five from my uncle…
I got about all of the Rokkors when they were really cheap, since most could not be adapted to EF and nobody wanted them. The 35mm f1.8 or the 58mm f1.2 are such gems, and the 85mm Varisoft is still a beauty for dreamy portraits, even if you can fake it to some degree these days.

For those new to these lenses, this is a great resource: http://www.artaphot.ch.

Yes, I'm a Minolta fan. I inherited my Minolta SRT-102 from my father along with his Soliger 35mm f/2.8, Minolta Rokkor 58mm f/1.4, and Vivitar 135mm f/2.8. That is still the setup I shoot film stills with today. I'd love to have picked up the Rokkor 35 f1.7, 85 f1.8, and 135 f2 but sadly missed doing it when the lenses were cheaper. The above two shots were done with the Soliger 35 and Rokkor 58mm f/1.4 on the SRT-102. I am a big fan of the Victor 135 however, as it has gotten me some of my favorite shots just like the Rokkor 58mm f/1.4.

My grist for the mill:

- Fully kitted out, we're probably closing in on $20K. At that price, many owner/operators will want to rent out the camera. Will there be a healthy market for that? Traditionally, there has not been rental demand for BMD cameras. Is this camera good enough to change that? Somehow I doubt it. It's a meaningful advancement for BMD, but not really compared to the more respected competition in the rental market. Then there's the whole issue of proprietary media and rentals. Given all of that, you can understand why it might be worth it to spend the extra $1 on a V-Raptor and still make back your money quicker.

- The last time I bought a camera, the OG 12K was in consideration. After testing it, none of the DPs who participated in the test were impressed. It offered LESS real world resolving power than some of the 8K cameras we tested, and the much ballyhooed RGBW sensor offered no perceptible color improvement. (In fact, it was rated worst in the blind test.) On top of that, no one liked the form factor: Too big for run 'n' gun, and too small for a big production. Battery life was poor, and media was unreliable. The Ursa Cine has a lot to make up for.

None of the above precludes me from renting and testing it. I always root for BMD. I'm such a fan of their ethos. But this camera has to really perform in some special way to make sense for purchase.

Well judging by the downloadable footage, it does perform in a special way. It looks spectacular out of the box and when tweaking it...phew.
If you haven't yet, try it out!

Well judging by the downloadable footage, it does perform in a special way. It looks spectacular out of the box and when tweaking it...phew.
If you haven't yet, try it out!

Define "special"?
- The dynamic range is good, but not special compared to Venice 2, V-Raptor, Alexa 35.
- The color accuracy and total gamut don't seem to be "special" (certainly wasn't with the original 12K sensor) either.
- The resolution? Meh. 12K is a solution looking for a problem.

To be clear...I would LOVE for there to be a picture quality advantage to BMD's in-house RGBW sensor...I just haven't seen anyone empirically demonstrate it with the original 12K, and I didn't notice anything in the sample footage. (Which is understandable as there's nothing to compare it to.)

What do you need to be empirically demonstrated?

I downloaded it, opened it in Resolve and was blown away by the colors, how they react, the whole feel of the footage.

Imo it's by far the best quality BMD ever produced.

What do you need to be empirically demonstrated?

I downloaded it, opened it in Resolve and was blown away by the colors, how they react, the whole feel of the footage.

Imo it's by far the best quality BMD ever produced.

That's a vague, indefensible reaction. The "feel" of the footage could be xx% cinematography. Someone could say the same thing about well shot Arri Alexa 35 footage. For that matter someone could say the same thing about well shot iPhone 15 footage.

How are the colors better? Better than what? If the difference is real, shouldn't BMD be able to explain why it is better? Shouldn't we be able to measure it?

What is the actual advantage of the RGBW sensor?

Joe,
I can't tell you all that, it's not something I care too much. I only know it's the best footage I ever played with from a BMD camera. Try for yourself if you haven't yet.

It's a fair question. When the 12K was first announced, there were technical explanations of why the sensor was an advance but there was no (at least to my knowledge) rigorous comparisons.

In the age of high bit rate footage from all major manufacturers, it's not unreasonable to ask if subjective differences or preferences amount to no more than grading choices or a particular debayering/normalization process.

Ideally, a colorist (a real one) accustomed to working with the best would take a good look and opine. But if that's happened, it's not been announced.

It's a fair question. When the 12K was first announced, there were technical explanations of why the sensor was an advance but there was no (at least to my knowledge) rigorous comparisons.

In the age of high bit rate footage from all major manufacturers, it's not unreasonable to ask if subjective differences or preferences amount to no more than grading choices or a particular debayering/normalization process.

Ideally, a colorist (a real one) accustomed to working with the best would take a good look and opine. But if that's happened, it's not been announced.

I guess almost 3 years ago now I took part in a shootout with every 8K camera plus the Ursa 12K, but it wasn't shared publicly. When comparing apples to apples in a blind test, the 12K image scored last by a clear margin. Once I had the footage back in Resolve, I tried to find a reason to like the 12k, I really did, but I just couldn't find an objective PQ advantage.

Now the new Ursa Cine has a new sensor, with what should have meaningfully better dynamic range and supposedly "better color", so I'm looking forward to being able to test.

If the question is, which camera is better for me (given my approach to the work, my limitations, time-constraints, etc.), than what you propose makes sense. Any faults in the testing would represent one's own workflow, which is all there is in the end.

But I'm interested in ultimate reality! I'm not a colorist. Whether I can exploit the technology or not, I want to know what's possible.

And the kind of comparative testing you propose ("shoot-outs") is notoriously difficult to pull off. So much can go wrong, to give false or arbitrary results. Even when actual engineers are present, there's always something......

If the question is, which camera is better for me (given my approach to the work, my limitations, time-constraints, etc.), than what you propose makes sense. Any faults in the testing would represent one's own workflow, which is all there is in the end.

But I'm interested in ultimate reality! I'm not a colorist. Whether I can exploit the technology or not, I want to know what's possible.

And the kind of comparative testing you propose ("shoot-outs") is notoriously difficult to pull off. So much can go wrong, to give false or arbitrary results. Even when actual engineers are present, there's always something......

True, no test is perfect, but shooting the same scene designed to push the limits, with objective data (color charts, etc.) with the same lens, by operators familiar with each camera and then reviewing the footage "blind" to which camera is which, is the best way to objectively compare performance. Given how good ANY modern camera is today, looking at beautiful demo footage in isolation tells us almost nothing about actual performance.

Outside of that, publishing actual data about a technology goes a long way to giving users confidence. For example Arri's relatively surface level, marketing-infused, info about their new-ish REVEAL color science still gives more detail than BMD has ever published. https://www.arri.com/en/learn-help/lear ... tab-301424

If RGBW sensors offer a material advantage, BMD should explain and demonstrate that advantage.

Hey Joe,

Do you know why it is that you can’t turn reveal off and choose to not have it as an option?

I’m sure it’s in their not marketing white paper…

JB

(Edit. I meant to say “Textures”, not reveal.)

Let’s remember DaVinci Resolve for a moment. Didn’t that product (mostly hardware) suite cost $100,000?

That was the price of the Power Window card that gave you two additional power windows for each shot

What is the actual advantage of the RGBW sensor?

Color rendition + dynamic range.

The 12K has gorgeous color rendition, and in spite of having a ridiculous resolution, it has great dynamic range.

Also, it has a stacked design; that is what enables the sensor scaling that makes the 12K also a great 4K or 8K camera without cropping. That's not a sneeze feature, it's actually quite brilliant and IMO follows shortly behind the color rendition as the 12K's biggest advantages.

If RGBW sensors offer a material advantage, BMD should explain and demonstrate that advantage.

If you want an Arri Alexa Super 35 but you don't have $80,000 for the basic kit, you buy a 12K.

I guess almost 3 years ago now I took part in a shootout with every 8K camera plus the Ursa 12K, but it wasn't shared publicly. When comparing apples to apples in a blind test, the 12K image scored last by a clear margin. Once I had the footage back in Resolve, I tried to find a reason to like the 12k, I really did, but I just couldn't find an objective PQ advantage.

It was never an open shootout that you took part in. You're mischaracterizing that.
You crafted the event yourself. Whatever it was, whomever the supposed DP's participated that you imply weren't shared, you were in control of.

Doing some testing this weekend.
Will the camera that costs 670% more look 670% better?
We're gonna find out.

When Captain Hook learned of this, he said he didn't know it was being presented that way on social media and would not have gotten involved. Accordingly, he asked you not to post any video here.

And you had stated to me you would try and do a better job "presenting the fact from editorializing," but that you are going to editorialize because you do have an agenda.

John Brawley was trying to help you! He said to you, "Every time you make these kinds of statements that speak with the authority of a test, you devalue your own status."

Here you are 3 years on, announcing the result, lecturing on objective methods and data, testing to the limits and giving users confidence, while chiding another poster for "a vague, indefensible, reaction." This is classic narcissism, with hubris. Sorry Joe, really am.

Ideally, a colorist (a real one) accustomed to working with the best would take a good look and opine. But if that's happened, it's not been announced.

Well this is one of my main hats, John; as a longstanding editor first, then a finishing/grading editor in UK TV only; who also has shot a lot, over the years, and now owns a Broadcast G2. I wouldn't claim ever to be a top colourist nor work with the best footage by any stretch; nor do I work in features. But I do have very good post gear and monitors here. I should say too that I don't think even the best colourists could necessarily talk meaningfully about the technicalities of sensors and cameras neither - they only know the end result and how that footage stacks up.

Having played with a fair bit of it; I can say whatever the proprietary advantages or goings on of BMD's RGBW sensor there is no way it has 4x the resolution of a 6K Bayer sensor; from what I can see. The colour is very beautiful but then so is Gen 5 and Braw in the others; and the 12K is not hugely better; and all of them can fall apart quickly - not just BMD cameras - in the grade if you're not sensible. And it has to be said also, to my mind, there is no way that footage from an Alexa 35 is 16x better than an Ursa Mini neither, nor a Pocket, for that matter.

But I don't think that's the point of RGBW neither. I think it goes hand in hand and was planned over many years with Braw, or maybe the other way round. And Braw is a joy both in shoot and post; is all I know.

Shootouts are not the best way to test cameras; not the end result anyway; IMV again. The only real way is how Alan Roberts used to do them scientifically, for the EBU. They stopped doing that because it is pretty meaningless now that most cameras have gone well beyond being much of a concern; or having any real control on who shoots on what; though I think Netflix have adopted a subset of those principles. But if you truly wanted to see what is going on with these sensors, that is still the best and perhaps only definitive way to do it. And more important things like poor OLPFs or the absence of, and moiré/aliasing; not just in the raw footage but across codecs to transmission; are still very much with us; in spite of or because of often, super hi rez.

If you want an Arri Alexa Super 35 but you don't have $80,000 for the basic kit, you buy a 12K.

I think this is the bottom line when comparing these cameras; as I said my JVC HD101 cost more than the 12K Ursa, in 2005; what you get now anywhere, is almost beyond credible. BMD, like all the rest, use a lot of marketing speak in and amongst the science; 'The same tools they use in Hollywood' etc; but all the cameras, hardware and software are absolutely great value and they have been the most disruptive force in the business, since we've been able to afford any of our own gear at all.

A BMD guy at NAB, in one video I saw defined how they pitch the new Cine: They realise they are not currently A cams in high end work; they see the Cines as aspirational up for filmmakers now on Ursa Minis; they also think DOPs might consider them for Owner/Operator 2nd or personal cameras too; they realise it's only the start of a new move 'up-market' and I think that's a totally a fair assessment too.

Having played with a fair bit of it; I can say whatever the proprietary advantages or goings on of BMD's RGBW sensor there is no way it has 4x the resolution of a 6K Bayer sensor; from what I can see. The colour is very beautiful but then so is Gen 5 and Braw in the others; and the 12K is not hugely better; and all of them can fall apart quickly - not just BMD cameras - in the grade if you're not sensible. And it has to be said also, to my mind, there is no way that footage from an Alexa 35 is 16x better than an Ursa Mini neither, nor a Pocket, for that matter.

But I don't think that's the point of RGBW neither. I think it goes hand in hand and was planned over many years with Braw, or maybe the other way round. And Braw is a joy both in shoot and post; is all I know.

Shootouts are not the best way to test cameras; not the end result anyway; IMV again. The only real way is how Alan Roberts used to do them scientifically, for the EBU. They stopped doing that because it is pretty meaningless now that most cameras have gone well beyond being much of a concern; or having any real control on who shoots on what; though I think Netflix have adopted a subset of those principles. But if you truly wanted to see what is going on with these sensors, that is still the best and perhaps only definitive way to do it. And more important things like poor OLPFs or the absence of, and moiré/aliasing; not just in the raw footage but across codecs to transmission; are still very much with us; in spite of or because of often, super hi rez.

12K is not 4x the resolution of 6K, it is 4x the number of pixels. The sensor does have that. The mistake is in equating pixels to resolution. A couple problems with that. Resolution is stated per picture height, in one dimension, so 12K resolution in terms of pixels is twice 6K. In terms of resolution with small pixels, as John Brawley has stated numerous times, lens diffraction is the limiter, diffraction that comes from the aperture, not the glass. See Rayleigh's Criterion for more precise explanation, but the practical implication is that for smaller pixels you need larger f/stop openings in order to not have resolution limited by aperture diffraction. Specifically, comparing the larger pixels of the BMCC6K and UMP4.6K to the UMP12K, diffraction begins reducing optical resolution on the 12K at f/3.3 where the other two don't start reducing until after f/8. So you ask yourself, who shoots landscape at f/3.3 and what lens is optically sharpest at f/3.3, and unless those things happen together, you can't focus a small enough beam of light onto the individual 12K pixels to achieve the resolution potential of the 12K pixel spacing.

Alan Roberts did excellent analysis, but he never said pixels were resolution. In fact, he correctly stated resolution in terms of horizontal "TV Lines" as it still should be. But no one tests like that anymore. Imatest can do it, with the slant edge resolution test, and I think Alan did do it this way. TV lines takes into account the lens and sensor as a system. CVP has also pointed this out, that to achieve the full 12k resolution of the pixel count, the lens has to resolve 227 lines per/mm. It's not impossible, you need a very good lens with an iris opening of f/3.3 or larger. But when you have larger pixels, you can have a smaller f/stops before diffraction starts limiting resolution.

12K is not 4x the resolution of 6K, it is 4x the number of pixels. The sensor does have that. The mistake is in equating pixels to resolution. A couple problems with that. Resolution is stated per picture height, in one dimension, so 12K resolution in terms of pixels is twice 6K.

The point is Tom, even if it were merely a Bayer sensor, it is not producing that kind of extra resolution over one; i.e lest you take me literally again, which is my fault admittedly: 12K 16:9 = 79,237,440 pixels; 6K/Broadcast G2 16:9 = 21,233,664; or photo sites, or whatever. I do realise RGBW is not in fact a Bayer sensor though. There was no intent, on my part, to make that a technical definition of RGBW, which I don't have knowledge of, nor care enough about. I know what Alan Roberts did, I've talked to him enough and he didn't use Imatest. He used standard resolution or zone charts that showed both horizontal and vertical resolution and aliasing and what that really meant when both colour and luminance real resolution of Bayer sensors was taken into account. Obviously that is not the same thing as the stated pixels would imply. He retired before the BMD RGBW sensor came about; I have no idea what he would have made of it.

Can we please try to set a clearer nomenclature here?

– A pixel is a single picture element stored in digital form.
– A photocell, which some like to call sensel (sensor element) now, is a single light sensitive area.
– Resolution should be measured over the whole system, from the lens to the display, and is usually defined in line pairs per picture height.

These are all different.

I guess almost 3 years ago now I took part in a shootout with every 8K camera plus the Ursa 12K, but it wasn't shared publicly. When comparing apples to apples in a blind test, the 12K image scored last by a clear margin. Once I had the footage back in Resolve, I tried to find a reason to like the 12k, I really did, but I just couldn't find an objective PQ advantage.

It was never an open shootout that you took part in. You're mischaracterizing that.
You crafted the event yourself. Whatever it was, whomever the supposed DP's participated that you imply weren't shared, you were in control of.

Doing some testing this weekend.
Will the camera that costs 670% more look 670% better?
We're gonna find out.

When Captain Hook learned of this, he said he didn't know it was being presented that way on social media and would not have gotten involved. Accordingly, he asked you not to post any video here.

And you had stated to me you would try and do a better job "presenting the fact from editorializing," but that you are going to editorialize because you do have an agenda.

John Brawley was trying to help you! He said to you, "Every time you make these kinds of statements that speak with the authority of a test, you devalue your own status."

Here you are 3 years on, announcing the result, lecturing on objective methods and data, testing to the limits and giving users confidence, while chiding another poster for "a vague, indefensible, reaction." This is classic narcissism, with hubris. Sorry Joe, really am.

To be clear, the "fail" of that test had to do with trying to match Super35 and FF sensors. In the time we had, we didn't have time to adjust to match well enough and it was obvious which was which.

What is the actual advantage of the RGBW sensor?

Color rendition + dynamic range.

The 12K has gorgeous color rendition, and in spite of having a ridiculous resolution, it has great dynamic range.

Also, it has a stacked design; that is what enables the sensor scaling that makes the 12K also a great 4K or 8K camera without cropping. That's not a sneeze feature, it's actually quite brilliant and IMO follows shortly behind the color rendition as the 12K's biggest advantages.

Hey Rakesh, I understand those technical details, and that all those things are touted as advantages of the designs...do you have something you can share that demonstrates them? Some situation in which the sensor performs in some way that is materially better?

I know I'm coming off as a hater, but I'm really not. I'm not even saying there aren't real world examples, I'm just saying I haven't seen any, and I'd really like to if they exist.

Hey Joe,

Do you know why it is that you can’t turn reveal off and choose to not have it as an option?

I’m sure it’s in their not marketing white paper…

JB

(Edit. I meant to say “Textures”, not reveal.)

When BMD released Gen 5, their marketing materials were able to show demonstrable improvements. It was real, and immediately apparent. (So much so I threw away my LUT.)

When BMD released the RGBW sensor they offered no demonstration of the improvements. (Well, other than the reach of 12K resolution.)

Now BMD has released the new FF RGBW sensor. 2 extra stops of DR will indeed be a welcome improvement, and we'll be able to test and see that difference. But still no explanation or demonstration of how the unique sensor design delivers the "ultimate in image quality".

"The unique RGBW architecture provides equal amounts of red, green and blue pixels." Great. You can cleanly scale a "RAW" image. We get that. Now show us how that yields a materially better color than 2 green pixels.

Can we please try to set a clearer nomenclature here?

– A pixel is a single picture element stored in digital form.
– A photocell, which some like to call sensel (sensor element) now, is a single light sensitive area.
– Resolution should be measured over the whole system, from the lens to the display, and is usually defined in line pairs per picture height.

These are all different.

Yes you are correct, Uli; but it's just what people use; our bad. The point is no sensor currently delivers it's quoted resolution, even in line pairs per picture height, because that's impossible. Other factors such as limiting lens resolution; diffraction; aliasing; codec etc. etc. make it quite difficult to make these definitive statements either and it's a sideline that I am to blame for, really. On a perfect day, with perfect exposure, with a perfect lens, who knows, who cares...

I was actually defending BMD products but seeing some of the logic Joe, Michel, Rakesh and others pointed to; and we have more than enough resolution now anyway; naturally you end up pleasing no one on the fence. All the BMD cameras produce great images and there can be a very real, theoretical at least, benefit in oversampling too, through post; though not all NLEs downsample properly, to take advantage of that. So RGBW and Braw are only going to get better too, I imagine.

Joe,
I can't tell you all that, it's not something I care too much. I only know it's the best footage I ever played with from a BMD camera. Try for yourself if you haven't yet.

Michel,
Didn't mean to ignore your admonition to download the footage. I have.

As usual, JB's cinematography is beautiful. The location is beautiful. The models are beautiful. The question I'm always asking myself, though, is a variation of this:

Would a $3,000 Nikon Z8 look demonstrably worse?
Would a $70,000 Arri Alexa Mini LF look demonstrably better?

Because if the answer is "no", then my primary factors for deciding what camera to use are not related to picture quality. And there are a whole host of other factors to consider.

I can't answer that question from that footage. When someone like CVP does an objective comparison I'll know more. And when I test it for myself I'll know a lot more. I just can't help but find it frustrating that BMD won't speak more plainly about the "color" advantages of the sensor that they casually toss about.

Joe,
I can't tell you all that, it's not something I care too much. I only know it's the best footage I ever played with from a BMD camera. Try for yourself if you haven't yet.

Michel,
Didn't mean to ignore your admonition to download the footage. I have.

As usual, JB's cinematography is beautiful. The location is beautiful. The models are beautiful. The question I'm always asking myself, though, is a variation of this:

Would a $3,000 Nikon Z8 look demonstrably worse?
Would a $70,000 Arri Alexa Mini LF look demonstrably better?

Because if the answer is "no", then my primary factors for deciding what camera to use are not related to picture quality. And there are a whole host of other factors to consider.

I can't answer that question from that footage. When someone like CVP does an objective comparison I'll know more. And when I test it for myself I'll know a lot more. I just can't help but find it frustrating that BMD won't speak more plainly about the "color" advantages of the sensor that they casually toss about.

There will be difference as we all know, subtle, but huge at the same time.
2 stops of DR in itself is huge, latitude also, you only appreciate this if you light in such way, or in fact don't light much at all and when to grade to extremes...

People say all cameras can be matched, no they can't, in some scenes one will fall apart completely.
They can only be matched if the whole scene is re lit with different DR, same goes for the colour and latitude...

This cost money and time and changes mood on the set of the actors ...
There is reason why they cost more, to fully appreciate it you need to get into complex lighting and grading, otherwise noone would be spending $150K on high end cameras.

If lighting for 4-5 stops scene, probably not a huge difference.

Ursa Cine - BTS

Looks like fun John. Lots of cool toys and a great location.

There will be difference as we all know, subtle, but huge at the same time.
2 stops of DR in itself is huge, latitude also, you only appreciate this if you light in such way, or in fact don't light much at all and when to grade to extremes...

Certainly. Superior dynamic range offers real world advantages and is easy to measure.

People say all cameras can be matched, no they can't, in some scenes one will fall apart completely.
They can only be matched if the whole scene is re lit with different DR, same goes for the colour and latitude...

Except for edge cases, it's not that difficult to match most any modern camera to a degree that they can be intercut.

Latitude differences certainly are real and demonstrable, but we're fast getting to the point where you REALLY have to mess up your exposure for it to be a deciding factor. If you are over exposing by more than 4 stops, then you might have bigger issues than your camera choice.

If lighting for 4-5 stops scene, probably not a huge difference.

I think you're being overly conservative. People are shooting beautiful HDR scenes on "consumer" cameras. It's the rare camera that doesn't offer 10 good clean stops and 12 usable stops if you accept a little bit of shadow noise.

Ultimately, we can identify a finite number of material, testable factors to judge a camera's PQ. These are not ineffable characteristics. The most obvious are:
1. (spacial) Resolution
2. Dynamic range
3. Latitude
4. Sensitivity / Noise
5. Bit depth (color resolution)
6. Color gamut
7. Color accuracy

Most of these factors have mostly been demystified. But color...oh color...this is the hardest to understand, so it is also where we are most tempted to fall back to superstition, and manufacturers are mostly likely to use hand-waving mumbo-jumbo in their marketing.

My core argument is that there is a "good enough" point for all of these characteristics after which technical improvements don't materially matter because they are either beyond what we can perceive or are useful to manipulate in post.

There will be difference as we all know, subtle, but huge at the same time.
2 stops of DR in itself is huge, latitude also, you only appreciate this if you light in such way, or in fact don't light much at all and when to grade to extremes...

Certainly. Superior dynamic range offers real world advantages and is easy to measure.

People say all cameras can be matched, no they can't, in some scenes one will fall apart completely.
They can only be matched if the whole scene is re lit with different DR, same goes for the colour and latitude...

Except for edge cases, it's not that difficult to match most any modern camera to a degree that they can be intercut.

Latitude differences certainly are real and demonstrable, but we're fast getting to the point where you REALLY have to mess up your exposure for it to be a deciding factor. If you are over exposing by more than 4 stops, then you might have bigger issues than your camera choice.

If lighting for 4-5 stops scene, probably not a huge difference.

I think you're being overly conservative. People are shooting beautiful HDR scenes on "consumer" cameras. It's the rare camera that doesn't offer 10 good clean stops and 12 usable stops if you accept a little bit of shadow noise.

Ultimately, we can identify a finite number of material, testable factors to judge a camera's PQ. These are not ineffable characteristics. The most obvious are:
1. (spacial) Resolution
2. Dynamic range
3. Latitude
4. Sensitivity / Noise
5. Bit depth (color resolution)
6. Color gamut
7. Color accuracy

Most of these factors have mostly been demystified. But color...oh color...this is the hardest to understand, so it is also where we are most tempted to fall back to superstition, and manufacturers are mostly likely to use hand-waving mumbo-jumbo in their marketing.

My core argument is that there is a "good enough" point for all of these characteristics after which technical improvements don't materially matter because they are either beyond what we can perceive or are useful to manipulate in post.

That is my argument:
3. Latitude ( Komodo offers 7 stops bs Alexa 35 13stoos)
those almost 6 stops are huge and cost additional 100k.

My core argument is that there is a "good enough" point for all of these characteristics after which technical improvements don't materially matter because they are either beyond what we can perceive or are useful to manipulate in post.

Yes but I do notice that I need to put more post work in some cameras than others to get them to a look I like. These are differences, as Darko excellently said, subtle and huge at the same time.

I don't think scientific tests can reveal those as well as personal tests, as we have different needs (depending on content) and different taste. And color taste is quite subjective.

Also for color, 'true' doesn't always mean 'better', especially in cinema it rarely is.

I don't think scientific tests can reveal those as well as personal tests, as we have different needs (depending on content) and different taste. And color taste is quite subjective.

Also for color, 'true' doesn't always mean 'better', especially in cinema it rarely is.

1. Which "look" you prefer is a matter of taste that no one can argue with. I certainly wouldn't. But BMD boasts "The unique RGBW architecture provides equal amounts of red, green and blue pixels. This means it is optimized to deliver incredibly rich colors at all resolutions and provide the ultimate in image quality and flexibility." Isn't it reasonable to ask what the "ultimate in image quality" looks like?

2. When you combine how we perceive a scene with how much we like to deceive ourselves, it becomes difficult to not bias personal "tests" if we're not comparing to a known standard. I'm not saying that we shouldn't get our hands on a camera and see how we like to use it in real world scenarios, but if a scientific test isn't revealing something, we need a better test.

For me, the key word in that quote of ad copy is “deliver.” Deliver is the final page in DaVinci Resolve that renders our deliverables from our original camera footage with a host of adjustments applied to video and audio to suit our taste and fit within industry standards. I haven’t 1% of the experience and technical expertise of those who are engaged in this thread. So I apologize in advance for the following remark that doesn’t resolve the essential argument being discussed, but it important to me to add for a moment’s reflection. This doesn’t mean that I’m oblivious to any of the technical leaning relevance of things like aperture, pitch, colour depth, dynamic range, and latitude. Before the URSA Cine was claimed to achieve 16 stops of dynamic range, I also questioned why BMD cameras did not emphasize larger photosites to achieve higher dynamic range. Looking forward to the results of latitude tests in the future. These individual components are important but they’re not the only or the best way to judge an image.

We all are familiar with the meme “you can’t judge a book by its cover.” There’s just so much within the pages of the book behind the cover that we can’t presume to know much at all about what we will learn from the book without doing the work of reading it. That meme applies literally to the book, but it applies metaphorically to most everything and everyone we encounter in life.

A book is meant to be read, not admired for its cover. A movie poster or a trailer might be a tease for a film, but really no one is going to be on a safe footing judging the movie solely from a poster or a trailer or even a film review by Pauline Kael. Now when it comes to evaluating contemporary digital cinema cameras in the context of this thread, the ‘cover’ in this metaphor is the image we see and the ‘pages’ are all of the many technical and artistic choices made to create the images. There are two places we can see this cover unfortunately. The ‘cover’ on my monitor in Resolve always is the best image I see, but when I share the image on a website, you see a different image inevitably. But I judge the image on my system and I judge the camera by the ease and degree I can technically vary a flat image without having an intimate knowledge of the real values that deliver the ‘cover.’ In this instance, you can only judge the book by the cover. When you focus only on the components, you can’t know “the sum of the parts is greater than the whole.”

For me, the key word in that quote of ad copy is “deliver.” Deliver is the final page in DaVinci Resolve that renders our deliverables from our original camera footage with a host of adjustments applied to video and audio to suit our taste and fit within industry standards.

I don't think I follow. What you say above is true for every camera that allows you to record in something other than the delivery space. Assuming the signal specs are good enough and there is an accurate transform available in Resolve, then we can manipulate the images from a great many cameras with a high degree of confidence and flexibility. I think Arri, RED, Canon, Sony and Nikon all have shipping cameras that record 12-bit (or more) log encoded RAW files, in camera, that can be opened directly in Resolve.

If your argument is that BMD controlling the color "pipeline" from sensor to Resolve offers an advantage (which it certainly seems like it could) then what is that advantage? If it's real and meaningful, then how do we demonstrate it?