"Club Penton" - Ask questions to Hollywood insider "Penton-Man"

[Penton-Man]

See -
https://forum.blu-ray.com/insider-di...ml#post4448356

Feel free to correct me if I’m wrong but, does the U.S. Secret Service still carry Sigs as standard issue?

[Penton-Man]

Originally Posted by sharkshark

Managed to see the TAXI DRIVER "print" tonight - wasn't a humungoid screen, bu there was a decent sized audience, and the restoration looks...

AMAZING.

Completely, staggeringly good. Even the end stuff, it's never been clearer, even as it's, uh, "unclear" due to the process used at the time and discussed above.

Last time I saw it on screen was with Paul Schrader in attendance, got to dig up the interview I did with him.

So, congrats to Grover and the team at Sony, such a pleasure to be able to see something like this maintain completely its film look, yet be presented using some excellent digital projection. Brilliant.

So glad to hear everything went well at your local venue. Did Dennis hook-up with you to enjoy this rather unique cinematic experience?

Keep in mind, despite the fact that all the 35mm film prints that will be exhibited anywhere will be off the digital neg, and all screened by Grover before they leave the lab. In the end, on the screen, it is what it is. Personally, for me, in this modern age of technological alternatives, there is just too much resolution loss in going from 4K files -> negative stock -> show print for film projection as compared to 4,096 x 2160 pixel digital exhibition, not to mention the lack of a rock steady image like one gets with digital projection …

http://www.youtube.com/watch?v=RPPhy...e_gdata_player

Or, for a behind the scenes look at the digital cinema projection process from your student reporters on the ground, Kourtney and Kaleigh, in this now vintage (’07) YouTube video…
http://www.youtube.com/watch?v=BMZWw...layer_embedded

[Bobby Henderson]

Originally Posted by Penton-Man

Feel free to correct me if I’m wrong but, does the U.S. Secret Service still carry Sigs as standard issue?

Yes, US Secret Service agents carry the Sig P229 in .357 Sig. Some agents might carry the FN Five-Seven pistol, but I've never actually seen that in any media coverage. Just Sigs.

Lots of US government agencies carry SigSauer pistols: US Federal Air Marshals, the DEA, Dept. of Homeland Security, US Coast Guard, US Postal Inspection Service, etc. Navy Seals have used the SigSauer P226 and P220 since the 1980s, but they also use H&K and Colt 1911 .45 pistols.

The Beretta 92F is still (unfortunately) the standard pistol used by US military troops. The Army nicknames it the "M9". In the military's pistol trials the P226 beat the 92F in terms of performance but lost due to price. Over the past few years the US military has conducted more trials, this time looking to adopt a .45 caliber platform. But they haven't settled on anything yet. The US Army does use the SigSauer P228, nick-named the "M11", as an alternative to the Beretta.

I was kind of surprised to learn the New York Police Dept. used the P226 as a duty weapon. I always thought they just used Glocks. I guess I was just going by what I had seen in movies and TV shows.

Originally Posted by Penton-Man

Personally, for me, in this modern age of technological alternatives, there is just too much resolution loss in going from 4K files -> negative stock -> show print for film projection as compared to 4,096 x 2160 pixel digital exhibition, not to mention the lack of a rock steady image like one gets with digital projection …

The only way 35mm film projection can have an advantage over digital is if the imagery on the print has a higher native resolution than what the digital projector can show. That's not the case with movies post produced with 2K or 4K digital intermediates. An entirely film-based workflow using the full 4/35 frame, such as anamorphic can beat 2K projection but would have a tough time beating 4K at least in terms of detail. 4K projection hasn't been exactly perfect in terms of brightness issues.

70mm is still the best thing for giant screen use. And it can beat 2K and 4K projection handily if the print has imagery natively photographed in 65mm.

Some 4K versus 70mm "shootout" was recently conducted, but I consider the test to be unfairly rigged in favor of 4K. They scanned a certain 65mm special venue movie in 11K and created down-sampled 4K files for the 4K projector. They also used down-sampled 4K imagery for creating the 70mm print used in the comparison. Digital won. But a big, obvious question remains: why did they drop the 70mm print's source down to such a low resolution (4K) when the print could have easily held 8K or even 11K resolution? Not a very fair comparison at all I think.

[Deciazulado]

Originally Posted by Penton-Man

Anyway, for those motion pictures in which the VFX elements are done at 2K resolution, with the help of a high performance up-conversion algorithm such as Dynamic Element Expansion Protocol, they can be interpolated and then the VFX elements can be merged with the 4K live action elements so that all the imagery is more visually seamless…if a 100% native 4K production is not possible due to budgetary or other concerns.

The appearance is obviously not as fine as true 4K but it is better than native 2K and it has a pretty nice texture with no jaggies. I can go into the mechanics of *how* the up-conversion algorithm works, later if desired, as the terminology is probably new to many readers and would take a bit of time to explain/translate.

desired

[sharkshark]

Originally Posted by Penton-Man

So glad to hear everything went well at your local venue. Did Dennis hook-up with you to enjoy this rather unique cinematic experience?

No, alas, he bailed... BAILER!!!

Personally, for me, in this modern age of technological alternatives, there is just too much resolution loss in going from 4K files -> negative stock -> show print for film projection as compared to 4,096 x 2160 pixel digital exhibition, not to mention the lack of a rock steady image like one gets with digital projection

Agreed, although I'm coming strictly from a "gut" level. But I worry about above, "gaming" 4K presentations and thinking they're the equivalent of, say, IMAX. But, yeah, dead horse time...

Bring on the nerdiness, we want details!!

[Penton-Man]

Originally Posted by Deciazulado

desired

You scoundrel you! And here I was hoping I could turn this thread into something like daytime TV.

Hey, since you dropped in, did this post give you any ideas (i.e. *digital squish*)….
https://forum.blu-ray.com/insider-di...ml#post4507580

Because unless me eyes are deceiving me, it seems my Penton avatar has been squeezed into a smaller version of its former self and my good friend Jeff Kleist made that avatar specifically for me.

Nevertheless, the details of the process will have to wait till tomorrow as I’m trying to get out of the office early (by 5) today.

[ZoetMB]

Originally Posted by Bobby Henderson

Some 4K versus 70mm "shootout" was recently conducted, but I consider the test to be unfairly rigged in favor of 4K. They scanned a certain 65mm special venue movie in 11K and created down-sampled 4K files for the 4K projector. They also used down-sampled 4K imagery for creating the 70mm print used in the comparison. Digital won. But a big, obvious question remains: why did they drop the 70mm print's source down to such a low resolution (4K) when the print could have easily held 8K or even 11K resolution? Not a very fair comparison at all I think.

Hey, Bobby, where was this shootout? Haven't heard about it before your post.

[Bobby Henderson]

Check this web page:Digital Cinema Society Newsletter and scroll down to the piece on the Galveston Digital Symposium.

[PeterTHX]

Originally Posted by Bobby Henderson

Check this web page:Digital Cinema Society Newsletter and scroll down to the piece on the Galveston Digital Symposium.

No comment on my single LightPeak per color channel idea for uncompressed digital video?

[sharkshark]

Comment: "Sound rad".

(but you didn't need to hear that from me... and I'm pleased that "rad" is making a comeback of sorts.)

Meanwhile, for all you bandwidth junkies:

http://newenterprise.allthingsd.com/...e-but-of-data/

A tour of the Lucasfilm/ILM computer room.

*Warning: Hypernerdy

[Penton-Man]

Originally Posted by sharkshark

No, alas, he bailed... BAILER!!!

Hey, don’t be too tough on Dennis as I love his sense of humor. He cracks me up -
https://forum.blu-ray.com/insider-di...ml#post4422935

[Penton-Man]

Originally Posted by Deciazulado

desired

Well, you run your 2K motion picture into the magic box that contains the processing for the high performance UP-Conversion algorithm. The structure of the original image (every frame or every signal) is then broken down basically into 3 bands or elements. Namely, one for diagonal information (think edges), another for noise (think the texture of the image) and a third for low-pass frequencies.

Once the image structure has been broken down into these multiple bands, the edges are then analyzed and then each one of these bands is interpolated….in other words, doubled horizontally and vertically, so, from a 2K image you create a 4K synthetic equivalent. But then you have multiple bands of 4K which must be multiplexed together (structure merging), which is the final step.

Bottom line is that the images don’t look fuzzy and have rather discrete textures with jaggy-free edges. It kicks butt compared to something like a Lanczos solution. Like I said, for productions which can’t do true 4K, this is a high quality alternative. Real world applications would be for content production, i.e. to merge 2K VFX elements with 4K live action elements in order that the lower rez images blend more seamlessly with the natively acquired 4K footage and also for content distribution, i.e. to create a 4K DCP from a totally 2K motion picture (VFX as well as live action).

[Penton-Man]

Originally Posted by sharkshark

....*Warning: Hypernerdy

Thee previously mentioned as yet un-named Sony Beyond HD camera –
https://forum.blu-ray.com/insider-di...ml#post4507603 indeed contains a 20.4 Megapixel 8K CMOS sensor as the article suggests. To be super nerdy, what the piece doesn’t reveal is that the sensor has an ‘active’ pixel count of 18.7M, 5,882 (horizontal) x 3179 (vertical).

Shark, if you’re unfamiliar with what ‘active pixel count’ means, then ask Bobby or somebody else chime in because that last brainiac post of mine has given me a headache. I’m headed on over to “grandpop’s “ house to look in the attic and cellar to see if I can find any goodies like this…
http://articles.philly.com/2011-03-2...mpionship-ring

A clip – http://video.nytimes.com/video/2011/...l?ref=baseball

[Bobby Henderson]

That doesn't make it sound like a true 8K sensor, at least not terms of how some people measure sensors according to pixels. More like "5.8K" People often measure the pixel width and go by that, like 2048 X 1080 being "2K". People who use the Canon EOS 5D Mark II for time lapse work may have "5.6K" original files (the sensor is 5616 X 3744).

Active pixel count can mean a couple of things depending on how the person using the term intends others to interpret it, such as the pixel count of the acquired image although I tend to call that native resolution or native pixel count. That can be lower than the number of pixels on the camera sensor, especially if the camera sensor is using three separate red, green and blue pixels to put together one full color pixel in the finished image. Some image sensor pixels may be used for error correction or other color filtering functions to help other pixels. And then the sensor may have extra unused pixels that fall outside of the boundary/frame of what the camera lens exposes to the imaging sensor. This is sort of like printing "full bleed" in print publishing. You just print near the edge on a bigger sheet of paper and then crop into it.

Originally Posted by PeterTHX

o comment on my single LightPeak per color channel idea for uncompressed digital video?

I must have overlooked that. I'm not sure how three separate LightPeak RGB leads would work for uncompressed 16:8:8 4K data, particularly for real time playback. The data would have to be packeted and buffered somehow to keep the three streams in sync.

[Deciazulado]

Originally Posted by Penton-Man

Well, you run your 2K motion picture into the magic box that contains the processing for the high performance UP-Conversion algorithm. The structure of the original image (every frame or every signal) is then broken down basically into 3 bands or elements. Namely, one for diagonal information (think edges), another for noise (think the texture of the image) and a third for low-pass frequencies.

Once the image structure has been broken down into these multiple bands, the edges are then analyzed and then each one of these bands is interpolated….in other words, doubled horizontally and vertically, so, from a 2K image you create a 4K synthetic equivalent. But then you have multiple bands of 4K which must be multiplexed together (structure merging), which is the final step.

Bottom line is that the images don’t look fuzzy and have rather discrete textures with jaggy-free edges. It kicks butt compared to something like a Lanczos solution. Like I said, for productions which can’t do true 4K, this is a high quality alternative. Real world applications would be for content production, i.e. to merge 2K VFX elements with 4K live action elements in order that the lower rez images blend more seamlessly with the natively acquired 4K footage and also for content distribution, i.e. to create a 4K DCP from a totally 2K motion picture (VFX as well as live action).

ooh so it's like a multiband smart 4bx dbx expander instead of the 1bx

Does it allow to resize in different ratios, like 1.1x or 0.96x or 3x, or just 2x.
Since sinc / 256 / Lanczos is sometimes about twice as good as bicubic in some images, wonder how good DEEP kicks butt compared to L

Picture me cautiously

[sharkshark]

I'm saddened I understand the last several posts. This is a hobby, I shouldn't know this much about what these three guys are talking about...

Sigh...

[Penton-Man]

Originally Posted by Bobby Henderson

That doesn't make it sound like a true 8K sensor...

Never claimed to be.
Sony will be up-front with its specs for the active pixel count when things like Spec Sheets become available online.

For general advertising, I think all motion picture camera manufacturers list the total pixel count for their sensors. For example, Arri advertise a 3.5K sensor for their Alexa camera….with the active area actually being 2.8K, if memory serves.

[Penton-Man]

Originally Posted by Deciazulado

ooh so it's like a multiband smart 4bx dbx expander instead of the 1bx

Does it allow to resize in different ratios, like 1.1x or 0.96x or 3x, or just 2x...

For practicality purposes, you can also create a 4K DCP from a 1080/24P RGB master.

[Penton-Man]

Originally Posted by sharkshark

I'm saddened I understand the last several posts. This is a hobby, I shouldn't know this much about what these three guys are talking about...

Sigh...

Well, here’s something all forum members should understand, namely 3D and fighting words -
http://www.ft.com/cms/s/2/6ab79d54-5...#axzz1HWUIMTNc

Alas, nobody has sent me the YouTube commercial with the actor and the monkey yet so I’m unable to link it at this time.

[Penton-Man]

Originally Posted by sharkshark

I'm saddened I understand the last several posts. This is a hobby, I shouldn't know this much...

On a bubbly note, I see in Bobby’s link on the past page, in one of the news pieces, it mentions Howard Hall (http://www.howardhall.com/stories/breakingin.html)

Makes me think of diving again. You may have some reservations about James Cameron (and I’m quite certain we’ll hear yet more lobbying by him for higher frames rates for 3D in Digital Cinema come March 31st when he speaks at CinemaCon) but, be that as it may, did you know that Jim spent the same amount of time submerged as did the underwater film crew did when filming The Abyss?

Quite impressive for an amateur diver. I mean the man didn’t miss one dive.