My Dragon Thoughts

[John Marchant]

hehe. You and I love a good diagram

 

[Dave Summers]

I'll bite. If your NLE has native support for R3Ds, as well as myriad other file types - and - your hardware can manage 1/4 res debayer without a noticeable performance hit, why not? I contend they only care if it slows them down. The real fly in the ointment is usually a post infrastructure and workforce that revolves around AVID.

So if Paul's point is that many of the major shops are AVID based, therefore any format that doesn't have robust native AVID support is a fail, that's fair enough - as long as you don't mind letting post dictate your camera choice ;-). BTW, they also want ProRes not ARRIRAW, now that they're in charge and all.

Cheers - #19

100% agree with you Blair,
Personally I feel that this "concept" that editing R3D's is somehow an editor's nightmare is just plain nonsense. Like you stated, Sony Vegas, FCP, Adobe, etc. all support R3D natively so there's really no excuse why a post team can't do it. Yes, the debayer is taxing, but the benefit of all that extra resolution is more than worth it. Or they could just shoot in 2K(same as Alexa) and debayering is no problem. And any post house that insists on sticking with Avid should think about how Panavision insisted on sticking with film.

Panavision assumed that their monopoly on the camera side meant digital would never become mainstream, yet look how that turned out. And now we have these post houses that insist on never budging from Avid and refusing to have a workflow that embraces R3D. It really amazes me that such a creative industry can be so resistant to change, especially when the change is a huge improvement. And with 4K TV's starting to make it into peoples homes, the demand for native 4K content is eventually gonna expose Alexa's limitations.

 

[Gavin Greenwalt]

Common misconception - HDRx is not like that. A frame comes first, X frame comes second starting immediately the A frame finishes exposing. Red have described it as 'motion conjoined' - there's no overlap, but if there's a gap I've had little luck finding it
Click for larger

If the exposures were concurrent you'd be able to have 360 shutter in HDRx, which you can't. Its also currently a reason why Motion mount and HDRx don't play together - the motion mount wants the camera's shutter at 360, to allow the MM shutter to control things.

Isn't the direction of motion icon in that graphic wrong? The xTrack is at the beginning, not the end of the exposure, no? So how is the x-track after the A track?

 

[paulherrin]

wow, thanks for the infographic, john. schooled me some...

i just now had a shot where i (finally) needed hdrx the other day, helpful to have! still figuring out the post options, though, and what works best/fastest.

 

[Paul Harb]

Sigh....

100% agree with you Blair,
Personally I feel that this "concept" that editing R3D's is somehow an editor's nightmare is just plain nonsense. Like you stated, Sony Vegas, FCP, Adobe, etc. all support R3D natively so there's really no excuse why a post team can't do it. Yes, the debayer is taxing, but the benefit of all that extra resolution is more than worth it. Or they could just shoot in 2K(same as Alexa) and debayering is no problem. And any post house that insists on sticking with Avid should think about how Panavision insisted on sticking with film.

Panavision assumed that their monopoly on the camera side meant digital would never become mainstream, yet look how that turned out. And now we have these post houses that insist on never budging from Avid and refusing to have a workflow that embraces R3D. It really amazes me that such a creative industry can be so resistant to change, especially when the change is a huge improvement. And with 4K TV's starting to make it into peoples homes, the demand for native 4K content is eventually gonna expose Alexa's limitations.

 

[David Mullen ASC]

Panavision assumed that their monopoly on the camera side meant digital would never become mainstream, yet look how that turned out.

That's a bit off-base... as soon as the first 24P HD camera hit the market, the Sony F900, Panavision bought over 100 of them and started building cinema zooms and primes that would work with its 2/3 3-CCD design, which was a big financial investment. They also developed the Panavision Genesis along with Sony, which was the first 35mm single sensor camera to draw big Hollywood productions and cinematographers away from 35mm. Hardly the attitude of a company that believed that digital would never become mainstream. They are not a big manufacturing company and yet they invested a lot of money developing digital technologies and support.

 

[John Marchant]

Isn't the direction of motion icon in that graphic wrong? The xTrack is at the beginning, not the end of the exposure, no? So how is the x-track after the A track?

Fixed, as even when I'm right I'm wrong Cheers Gavin

 

[Steve Das]

That's a bit off-base... as soon as the first 24P HD camera hit the market, the Sony F900, Panavision bought over 100 of them and started building cinema zooms and primes that would work with its 2/3 3-CCD design, which was a big financial investment...

Yes..actually looking back they were in very early on and were really an innovator in digital. As long as digital remained
expensive and proprietary it fit into their rental business model.
But lower cost digital tools really changed the rental business drastically.
Everyone either owns a camera or knows someone who owns one so rental companies throw in the cameras for
free in order to make money off the lenses..lights...and grip.

So since all of Panavision revenues come from rentals that is what really hurt them.
It wasn't the initial change to digital it was digital so cheap people could own and not rent.
Genesis was amazing to see in 2004-6 but from their to now...
is to now people shoot Academy Award documentaries on I-phones on the bottom end
and at the top end an Epic Dragon camera costs only 30k.

 

[Rudi Herbert]

That's a bit off-base... as soon as the first 24P HD camera hit the market, the Sony F900, Panavision bought over 100 of them and started building cinema zooms and primes that would work with its 2/3 3-CCD design, which was a big financial investment. They also developed the Panavision Genesis along with Sony, which was the first 35mm single sensor camera to draw big Hollywood productions and cinematographers away from 35mm. Hardly the attitude of a company that believed that digital would never become mainstream. They are not a big manufacturing company and yet they invested a lot of money developing digital technologies and support.

...and are now developing the first, and so far only, 70 mm digital camera, including a set of lenses made specifically for it. Another huge investment. I would not call Panavision a retrograde, monopolistic company at all.

 

[Will Keir]

Link?

+1

that commercial rocks the house

 

[Will Keir]

While their camera's might not be on top of the world their lenses certainly are some of the best in cinematography.

Wish I could buy a set of Primos.

...and are now developing the first, and so far only, 70 mm digital camera, including a set of lenses made specifically for it. Another huge investment. I would not call Panavision a retrograde, monopolistic company at all.

 

[David Mullen ASC]

Panavision's success supplying cameras and lenses for studio movies wasn't due to a monopoly, it was due to historical changes in the industry. Panavision started out making anamorphic projector lenses for theaters and then anamorphic camera lenses, but the studios back then had their own camera departments because equipment ownership made sense to them as a way of controlling costs.

But the huge losses that the studios experienced in the 1960's due to losing their theater chains and loss of attendance to TV meant that they had to cut overhead, and they started taking apart their assembly-line factory model for moviemaking, closing their camera departments. What evolved was the current system where each production is a limited liability legal entity that rents everything -- phones, copy machines, production offices, stages, camera equipment. They even rent from themselves and pay back themselves. That way, when production is slow, so is their overhead.

Panavision stepped in with one of the newest quiet sync-sound 35mm cameras for rent (the Panaflex) just at a time when the studios were retiring their old Mitchell BNC's and whatnot. So the 1970's saw the rise of the rental house. It's still the model for most studio production. What has really changed is that smaller independent producers and filmmakers now can own most of the means of production and don't have to rent necessarily. But the studios aren't really interested in starting up camera departments again and buying cameras and lenses, plus putting people on payroll to maintain all of that.

So what has hurt the rental houses are two trends: one is smaller production companies that have started to own their own cameras, and the second, the high cost of digital equipment combined with its short life expectancy and somewhat fragile physical nature -- a 35mm camera can work fine for decades if maintained, and 35mm images look the same whether they were shot with a 1980 Panaflex or a 2006 Arricam. So it's harder to maintain profitability when you can't expect to keep renting the same camera for ten years or more.

 

[Hrvoje Simic]

Personally I feel that this "concept" that editing R3D's is somehow an editor's nightmare is just plain nonsense. Like you stated, Sony Vegas, FCP, Adobe, etc. all support R3D natively so there's really no excuse why a post team can't do it. Yes, the debayer is taxing, but the benefit of all that extra resolution is more than worth it.

Extra resolution in the edit is a less important factor for editing decisions, once there is enough detail to perceive important elements of the shot. Editor is more focused on overall impression of the shot and the cut, and speed and fluidity of editing are often pretty important because the unnecessarily burdened system tends to lag behind the mental process, intention and even hands. Mind speeds up to optimal speed and the system yanks it back. Thousands of times.

Good editor can accommodate to this and use different routes, but once getting used to fluid cutting, to many the pros do not outweigh the cons.

 

[Elan Gunos]

That was very enlightening David—thanks.

 

[Will Keir]

David, thanks for taking the time to enlighten and entertain with your writing.

It makes me think. Private owner operators have become what the studios once were, it's a pretty cool flip on itself. So the next question. Are films better now, worse or just different?

If you could pick a decade to preserve all films that were shot, which decade would you favor and what films?

Panavision's success supplying cameras and lenses for studio movies wasn't due to a monopoly, it was due to historical changes in the industry. Panavision started out making anamorphic projector lenses for theaters and then anamorphic camera lenses, but the studios back then had their own camera departments because equipment ownership made sense to them as a way of controlling costs.

But the huge losses that the studios experienced in the 1960's due to losing their theater chains and loss of attendance to TV meant that they had to cut overhead, and they started taking apart their assembly-line factory model for moviemaking, closing their camera departments. What evolved was the current system where each production is a limited liability legal entity that rents everything -- phones, copy machines, production offices, stages, camera equipment. They even rent from themselves and pay back themselves. That way, when production is slow, so is their overhead.

Panavision stepped in with one of the newest quiet sync-sound 35mm cameras for rent (the Panaflex) just at a time when the studios were retiring their old Mitchell BNC's and whatnot. So the 1970's saw the rise of the rental house. It's still the model for most studio production. What has really changed is that smaller independent producers and filmmakers now can own most of the means of production and don't have to rent necessarily. But the studios aren't really interested in starting up camera departments again and buying cameras and lenses, plus putting people on payroll to maintain all of that.

So what has hurt the rental houses are two trends: one is smaller production companies that have started to own their own cameras, and the second, the high cost of digital equipment combined with its short life expectancy and somewhat fragile physical nature -- a 35mm camera can work fine for decades if maintained, and 35mm images look the same whether they were shot with a 1980 Panaflex or a 2006 Arricam. So it's harder to maintain profitability when you can't expect to keep renting the same camera for ten years or more.

 

[Ed David]

I hope the Dragon does take off. I'm still not really into the Alexa's look. But I like the color science of the Alexa over the Red's current color science. The motion and resolution I prefer the Red Epic...well actually I like the motion from the Red One MX - something I think may have changed with how the camera handles motion. Hopefully the new color science will come out and make pleasing colors easier to obtain.

 

[David Mullen ASC]

There's good and bad films, and technically interesting films, from every decade so I can't pick favorites. They all have their strengths and weaknesses, though with the factory model of the studios, there was simply more content being produced period, increasing the odds for something good to come along, plus there is a certain level of skill that directors back then achieved from shooting tens times as often as directors today get to shoot. In that sense, I'm partial to the 1940's as a decade of film for many reasons -- it was the height of studio style, it saw the emergence of film noir and Neorealism in Italy, 3-stop Technicolor was at its height, and you had World War II causing massive changes in the types of movies being made. But I also love the widescreen revolution of the 1950's, just that it didn't necessarily produce better movies.

 

[Christopher Probst]

Okay, I'm still on location with the Dragon and don't have time to do a massive analysis/write-up, but there has been little posted by way of comparisons between Dragon and Epic MX so I'll offer up this quickie...

This is a hasty, down-and-dirty test I threw together before embarking on this job. I did many more frame grabs at various iris/ISO settings to stress test the sensors and then analyze in RedCine... But I found this one wedge quite illuminating!

These comparison frame grabs from RedCine feature both the EPIC MX and DRAGON shot with a Leica Summulux-C at a T22 with no filtration... no IR protection, no ND... Naked camera/lens. I took a DIT also prepping at Camtec outside and stood her under the sidewalk canopy/tree shadows and had blazing high-noon sunlight behind her. Now, for simplicity/speed I'm posting just the T22 exposure wedge comparison so you can see where the highlights hold/clip on the two sensors. Pay particular attention when we drive the metadata ISO down to 50... Look what suddenly reveals itself on the Dragon sensor over what nastiness is unveiled on the MX... This was shocking to me since I absolutely love what I do with the EPIC MX... But what a night and day difference! Creamy highlight clip roll off is actually FULL of color information and detail, while on the MX, the skip is irreparably clipped and has a skewed bias toward magenta lurking in that clip.

(right click for a slightly larger pic size...)

ISO 50

ISO 160

ISO 320

ISO 800

ISO 200

This was not a very scientifically controlled environment test, I took all of about 8 minutes to conduct my test... I just wanted to get a number value in my head for what my gut already told me about the behavior of the highlights on the Dragon. This same characteristic is clearly displayed in the clips Phil uploaded... The shots on the balcony showed TONS of detail in the sky when brought down... And it was a revelation... But the significant difference here is the ability to compare the MX in the same condition. And it's a shocker.

You can all commence thanking me now!

 

[John Marchant]

I've not posted any comparisons like these, as some people will likely misunderstand them... but another significant factor worthy of exposition is what happens at and beyond the point of clip on Dragon vs MX.

Viewing in linear light is a good way to check it out.

With apologies to my long suffering editor Holly...

 

[Christopher Probst]

Nice John. Another example of the difference we are seeing in the above screengrabs but with motion and post thrown into the mix.

BTW, I've handed over some files to Phil to poke around with. Maybe we can get some better uploads going. I'm still out of country and have no time to do the leg work.