Jump to content


Photo

Mike Curtis on RED


  • This topic is locked This topic is locked
97 replies to this topic

#1 Mark Allen

Mark Allen
  • Basic Members
  • PipPipPipPip
  • 591 posts
  • Director
  • Los Angeles

Posted 06 August 2006 - 02:52 PM

If you're curious to hear a first hand opinion on the the first images off the "mysterium" sensor - check out Mike Curtis's latest blog entry:

http://www.hdforindies.com/

direct link: http://www.hdforindi...-mike-sees.html
  • 0

#2 Arni Heimir

Arni Heimir
  • Basic Members
  • PipPipPipPip
  • 326 posts
  • Other
  • Reykjavik/Barcelona

Posted 06 August 2006 - 04:35 PM

If it materializes, I think it could be the DV revolution all over again. But most filmmakers are doggedly conservative.
  • 0

#3 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 06 August 2006 - 05:25 PM

It's not the sensor so much that is the challenge, it's recording the data, especially if it is 4K. Look how many years it took for 4:4:4 HD to become (barely) practical -- and I have yet to be able to do it on any of my HD productions.

With indie people having troubles wrapping their heads around recording the 720P output of the HVX200 to P2 cards or Firestore devices, imagine them handling 4 to 6 times the data, every day, day after day. All-4K workflows aren't even common in major Hollywood studio productions yet. I suspect most indie people will be using the RED camera as a 1080P camera given the chance.
  • 0

#4 Mike Rizos

Mike Rizos
  • Basic Members
  • PipPipPipPip
  • 330 posts
  • Other

Posted 06 August 2006 - 08:34 PM

I especially like the part where the authur says he examined the picture at one foot, on 30 inch cinema display, and saw no trace of gain/grain/noise.
That link is nothing more than a solicitation for more deposits.
  • 0

#5 Jim Jannard

Jim Jannard
  • Basic Members
  • PipPipPipPip
  • 361 posts
  • Cinematographer

Posted 06 August 2006 - 10:10 PM

It's not the sensor so much that is the challenge, it's recording the data, especially if it is 4K.


David... you are too funny. Wasn't it you who said earlier that the sensor was impossible? If it wasn't you, please forgive me. I heard it 10,000 times. It is so hard to keep track.

We still have our hands full. David is right. This is a tough project. Just the way we like it. But the biggest hurdle WAS the sensor. And we are excited to show what we are seeing from it.

Jim

Edited by jannard, 06 August 2006 - 10:11 PM.

  • 0

#6 Mark Allen

Mark Allen
  • Basic Members
  • PipPipPipPip
  • 591 posts
  • Director
  • Los Angeles

Posted 06 August 2006 - 10:35 PM

Knowing David from the forums and once in person, whatever he said was probably very conservative and polite - at most cautionary - unlikely inciniary.

I actually think that the resolution people choose to shoot at will depend on if there is a super nice codec that can be handled by the post production process - that said - I think if the 1080 looked great, the indies would still be thrilled.


arnih - I'm hopeful and excited about the camera's future - but I don't know if it would be "the DV revolution all over again" because for most people $17,500 plus lenses is still going to be way out of the realm of possiblity. And most of the people doing DV movies are really not in the "rental" mindset. I don't think most of them realize how much more there is to a movie than the camera.

For the people who shoot regular enough though the price is just in the realm of a purchase instead of rental.

My reservation number is 89 (would have been lower had I been at NAB and not working) - and everytime I have to rent a camera now I shake my head thinking "Boy - if I only had this camera already." (And then I think... "Damn, I'll have to buy a tripod and a fluid head.... and should I buy some track... geeze... what about...?" ;)

I think if they achieve their goal they will for a time own a huge segment of the professional demographic (professional meaning a production where people are being paid for their work). I would imagine that many TV shows that had a meeting about switching off film or that are already on HD would convert. I would imagine nearly every indie movie with a remote budget will convert. I don't think, though, that it will be in the hands of filmmakers who are doing movies like "Tarnation" - but definitely the choice of people doing films like "Me, You and Everyone We Know" (and all the InDigEnt movies if they are still making them.) So more of a professional's revolution than a cultural revolution.
  • 0

#7 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 06 August 2006 - 11:06 PM

I don't know whether I said the sensor would be hard to make or not, but I don't think I said that.

I've generally held the position that the real problem is real-time processing and recording of large amounts of high-resolution data.

Of course you have to get the sensor to work too.

But for years now, the question you always hear is "hey, how come I can take a 8-megapixel (or whatever) picture with my digital still camera but they can't make a video camera with that sensor?" And the answer is generally "try taking that 8 MP picture 24 times a second and record it in real time... and then record & store 86,400 frames at 8MP each for every hour of footage in the course of a day of shooting. And then do that for 20 or more days in a row..."

I guess one of the questions on everyone's mind is how can you get such a sensor, which tend to be rather expensive to make, to be made cheaply enough to keep the cost of the camera at $17,000? I suspect the sensors in the Genesis, D20, and Dalsa cost a lot more than that. From what I gather, there is a high defect rate in sensor-making that requires economies of scale to keep the prices reasonable, and film cameras don't follow those sorts of economic models generally.

And then the other question is how do you develop a 4K recorder that holds any decent amount of footage and make that cheaply enough to be sold to someone who only wants to pay $17,000 for a camera?

I don't think anyone's questioning your ability to ultimately build a 4K camera, Jim, just that you could make and sell such a thing at those prices and see a profit.

I also question the current tubular shape of the camera from an operator's point of view, but until you design all the internal elements, I assume that the outward shape is in flux so for now, I'm not sure it matters.

Well, best of luck to you. Whatever comes of this venture, I'm sure it will push the industry forward.
  • 0

#8 Alan Lasky

Alan Lasky
  • Basic Members
  • PipPip
  • 55 posts
  • Other

Posted 06 August 2006 - 11:35 PM

I don't know whether I said the sensor would be hard to make or not, but I don't think I said that.


You may not have said it, but I certainly did. Making digital cinematography sensors is hard. Really goddam hard. Good on 'em if they have a working 4K sensor, take it from me, they ain't easy to make.

And you are correct: recording, storing and manipulating 4K data isn't easy either. Again, take it from me, I do it EVERY DAY.

Alan Lasky
DALSA Digital Cinema
  • 0

#9 Jim Jannard

Jim Jannard
  • Basic Members
  • PipPipPipPip
  • 361 posts
  • Cinematographer

Posted 07 August 2006 - 12:18 AM

With the RED camera you have lots of options for the data stream. RAW 4k if you choose the best and most difficult trick, 2K (easier to deal with) and various compression schemes and sizes that make the data much easier to handle. You are not stuck with the hose in the fully open position.

Jim
  • 0

#10 John Mastrogiacomo

John Mastrogiacomo
  • Sustaining Members
  • 210 posts
  • Other
  • Las Vegas, NV

Posted 07 August 2006 - 12:51 AM

With the RED camera you have lots of options for the data stream. RAW 4k if you choose the best and most difficult trick, 2K (easier to deal with) and various compression schemes and sizes that make the data much easier to handle. You are not stuck with the hose in the fully open position.

Jim


Jim,

Good luck with your camera. I really hopes it works and makes all the other manufacturers take note.

There is nothing like competition to drive the industry forward.

:)
  • 0

#11 Alan Lasky

Alan Lasky
  • Basic Members
  • PipPip
  • 55 posts
  • Other

Posted 07 August 2006 - 12:57 AM

RAW 4k if you choose the best and most difficult trick, 2K (easier to deal with) and various compression schemes and sizes that make the data much easier to handle. You are not stuck with the hose in the fully open position.


Yes, useful compression for production is a very difficult problem indeed. When I was a student at MIT's Media Lab many, many years ago we saw what the fourier transform, discrete cosine transform, and wavelet transform can do to images. Great for transmission, not so much for production and manipulation.

These are all really, really difficult problems and there is no magic leprechaun that can solve them; just hard work and sweat.

Alan Lasky
DALSA Digital Cinema
  • 0

#12 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 07 August 2006 - 01:06 AM

The trick is to not use so much compression that it negates the point of shooting & recording more resolution, plus screws up your post color-correction needs.

This is why I suspect most indie people will end up opting to work at 1080P resolution out of convenience's sake if they are trying to do the post themselves on a home set-up.

Most people who want the higher quality from shooting 2K or 4K are also going to want the higher quality from recording with a minimal amount of compression.
  • 0

#13 Jim Murdoch

Jim Murdoch
  • Guests

Posted 07 August 2006 - 01:28 AM

David... you are too funny. Wasn't it you who said earlier that the sensor was impossible? If it wasn't you, please forgive me. I heard it 10,000 times. It is so hard to keep track.

We still have our hands full. David is right. This is a tough project. Just the way we like it. But the biggest hurdle WAS the sensor. And we are excited to show what we are seeing from it.

Jim


Here's the $4,096 Dollar question Jim: What exactly do you mean by "4K"?
Do you mean your raw CCD sensor has 4,096 x 2,304 pixels or thereabouts?
(Which is the same as a 12.6 Megapixel still camera sensor but masked for 16 x 9)
"Raw" as in the sensor itself without any colour masking.

Given that in the world of film scanning "4K" means 4 thousand Red, 4 thousand Green and 4 thousand Blue samples across the width of the film, to be a true "4K" sensor it would need to have at least 12,288 pixels across its width.

But I get the impression you mean 4K Bayer masked, which in practical terms would be closer to 2K RGB.
(Which is still better than the bloody Dalsa which with most 35mm lenses is only going to be able to use around 3K of its touted 4K due to vignetting effects).

And please don't send us any bloody 30 megabyte .tiff files of images which are about the equivalent of a 4 megapixel digital still camera JPEG compressed to about 600K!

Don't get me wrong; an under-$20K 2K digital video camera would still be an incredible achievement; I just wish you'd tone down the hype a notch.

I don't know why you aren't pitching these as surveillance cameras, surely that would be a much more lucrative market!
  • 0

#14 Jim Murdoch

Jim Murdoch
  • Guests

Posted 07 August 2006 - 01:43 AM

I especially like the part where the authur says he examined the picture at one foot, on 30 inch cinema display, and saw no trace of gain/grain/noise.
That link is nothing more than a solicitation for more deposits.

Panavision's John Galt told me the same thing about Genesis images blown up to IMAX size!
Yeah, there's no grain or noise; you could also get that by fiddling with the focus knob!
No grain or noise, no fine detail either!
Bollocks!

Edited by Jim Murdoch, 07 August 2006 - 01:48 AM.

  • 0

#15 Deanan DaSilva

Deanan DaSilva
  • Basic Members
  • PipPip
  • 27 posts
  • Digital Image Technician

Posted 07 August 2006 - 02:24 AM

We still have our hands full. David is right. This is a tough project. Just the way we like it. But the biggest hurdle WAS the sensor. And we are excited to show what we are seeing from it.
Jim


Congratulations on you sensor results, Jim. Having been through the excitement phase before,
it only fair to warn you that first time sensor results are very much like beginners luck
in Vegas. It's sucks you in with it's beauty and then it hits you later with the a myriad
of problems to fix and work around (let alone the rest of the camera). Unless you're
super lucky (likely you are :), the sensor is really one of those never ending hurdles.

Welcome to the 4k sensor junkie club. Lot's of highs and lots of pain :)


Deanan
Dalsa
  • 0

#16 Jim Murdoch

Jim Murdoch
  • Guests

Posted 07 August 2006 - 02:52 AM

You may not have said it, but I certainly did. Making digital cinematography sensors is hard. Really goddam hard. Good on 'em if they have a working 4K sensor, take it from me, they ain't easy to make.

And you are correct: recording, storing and manipulating 4K data isn't easy either. Again, take it from me, I do it EVERY DAY.

Alan Lasky
DALSA Digital Cinema

I've heard that for every PV Genesis sensor that worked, they had to throw away 600 faulty ones.

So all else being equal, if your chances of getting a contiguous patch of 12 million pixels all working is one in 600, what are your chances of getting a contiguous patch of 48 million all working (which is what you'd need for a true 4K RGB camera)?

A: 600 x 600 x 600 x 600 x 600 = about 129 billion!
Ah exponential nature, she can be such a bitch!!

I've heard that for every PV Genesis sensor that worked, they had to throw away 600 faulty ones.

So all else being equal, if your chances of getting a contiguous patch of 12 million pixels all working is one in 600, what are your chances of getting a contiguous patch of 48 million all working (which is what you'd need for a true 4K RGB camera)?

A: 600 x 600 x 600 x 600 x 600 = about 129 billion!
Ah exponential nature, she can be such a bitch!!

Sorry, that's 600 x 600 x 600 x 600!
Not as hard as you thought :P
  • 0

#17 Deanan DaSilva

Deanan DaSilva
  • Basic Members
  • PipPip
  • 27 posts
  • Digital Image Technician

Posted 07 August 2006 - 03:48 AM

Given that in the world of film scanning "4K" means 4 thousand Red, 4 thousand Green and 4 thousand Blue samples across the width of the film, to be a true "4K" sensor it would need to have at least 12,288 pixels across its width.

But I get the impression you mean 4K Bayer masked, which in practical terms would be closer to 2K RGB.
(Which is still better than the bloody Dalsa which with most 35mm lenses is only going to be able to use around 3K of its touted 4K due to vignetting effects).


Since this is one of those continuing troll type questions, I'll take the bait. The lenses that tend
not to cover are the wides at about 24mm and below(say on the Ultra Primes).
Likewise, with a larger sensor, the FOV is different, so a 20mm at ~55deg in acad 1.85:1
is about equivalent to using a 32mm in on the Origin (52.8 deg at 1.85 subtended out of the 2:1).
With the 1.85:1 extraction you get 3789x2048, you cut off the sides which means you can
then go down to a 20mm lens (or ~76deg FOV which is something like a 16mm academy fov).
The MasterPrimes have even better coverage but we've been waiting for our sets
for 8 months now so we haven't been able to fully test the coverate. Of course, if you're
shooting for a 2:35 extraction, then the vignetting is even less of an issue because you're
chopping off even more of top/bottom.

The increased size of the sensor pixels also improves the dynamic range which in my opinion
is far more important than meeting arbitrary pixel counts or even resolution for that matter.
I'd take a 4k bayer sensor with more dynamic range any day over a 2k cosited (ala foveon)
sensor with less dynamic range and clipped highlights.

For those that want to shoot the full frame, we're building two different custom types
lens sets based on still camera lenses that have full coverage. Without mentioning
names, you probably already know who else has been building most of their prized
lenses out of stills glass for many many years.

Just curious Jim, do you have 4K bayer image and 4K film scan that shows that a 4K bayer
looks like 2k? Alternatively, how about a 4K bayer image next to a 2k film scan or a ~2k
3ccd Viper image that shows the 4K bayer image to look the same as the 2K image?

Am I on crazy pills or is it not obvious that a 4k "bayer" sensor is just that and that's why it's called
a 4k sensor? It's not called a 4K with three samples per pixel sensor, is it?
And what about importance of luminance information versus chroma information or
a good bayer reconstruction algorithm or the broad spectral response of each biased pixel
that is so easy to ignore when saying 4k bayer = 2k rbg?
Gonna have to update my crazy pills perscription...

BTW, I've always wondered why no one mentions the lenses in most "4K" scanners
adding yet another layer of lens degradation when talking about "true 4K".

Deanan
Dalsa
  • 0

#18 Deanan DaSilva

Deanan DaSilva
  • Basic Members
  • PipPip
  • 27 posts
  • Digital Image Technician

Posted 07 August 2006 - 04:00 AM

I've heard that for every PV Genesis sensor that worked, they had to throw away 600 faulty ones.

So all else being equal, if your chances of getting a contiguous patch of 12 million pixels all working is one in 600, what are your chances of getting a contiguous patch of 48 million all working (which is what you'd need for a true 4K RGB camera)?


FYI, it's next to impossible to get a perfect sensor. There will always be dead pixels, hot pixels,
or weak pixels that have to compensated for (within reasonable limits of course).

Don't forget, your true 4K camera is now recording 3x times as much data as the 4K bayer
camera and 3x the data on your servers and 3x the data in your archives. In our
case, we go from ~400MB/s at 24fps to ~1200MB/s (16bit bayer vs 16bit RGB).

If we're lucky, we'll see a true rgb 4k sensor in 5 years or so and by then
shooting 1200MB/s will be a piece of cake :)

Deanan
Dalsa
  • 0

#19 Jim Murdoch

Jim Murdoch
  • Guests

Posted 07 August 2006 - 05:52 AM

Since this is one of those continuing troll type questions, I'll take the bait.


Excuse me; in normal Internet Forum parlance a "Troll" is somebody who posts statements he doesn't really believe, simply to get a rise out of more "serious" posters.
To me the mixing of the terms "4K" as in "4K scanned film" and "4K" as in "a 4K Bayer-filtered sensor" is at best misleading, and at worst downright deceitful. I believe that most emphatically.

ust curious Jim, do you have 4K bayer image and 4K film scan that shows that a 4K bayer
looks like 2k? Alternatively, how about a 4K bayer image next to a 2k film scan or a ~2k
3ccd Viper image that shows the 4K bayer image to look the same as the 2K image?


Well, as it happens, no I don't. But then, I'm not trying to flog the technology; you are. Over the last 20 years or so I've been subjected to numerous eye-rollingly-bad demonstrations of supposedly "better than film" HD demonstrations. They all said much the same things. As I've yet to see any Origin footage up on the big screen; I suppose you expect me to reserve judgement. Like I did with the Genesis, even though from day one I couldn't see what was going to make that camera so much better than its predecessors. Put up or shut up.

And what about importance of luminance information versus chroma information or
a good bayer reconstruction algorithm or the broad spectral response of each biased pixel
that is so easy to ignore when saying 4k bayer = 2k rbg?


OK, so why did Sony/Panavision go to so much trouble to avoid doing that with the Genesis? They could have had the same sized chip with a smaller number of much bigger CCD elements, which would have improved both the manufacturing yield and the sensitivity, or they could have had a much higher-definition camera using the same chip. Oh I see, their algorithms aren't as good as yours.

BTW, I've always wondered why no one mentions the lenses in most "4K" scanners
adding yet another layer of lens degradation when talking about "true 4K".


Well, could it have something to do with the fact that a telecine lens doesn't really go anywhere during its operational life, it doesn't need anything more than a fairly elementary focussing mechanism since it's only ever focussed on one thing, and it doesn't get banged around in a flight case or dropped by hung-over camera assistants, so its original precision manufacture tends to remain where it was set. Also it doesn't need to be particularly fast, and so it can be optimized for best image quality.

In any event, the losses in the lens would be pretty insignificant compared to what is lost in the optical low-pass filter that both telecines AND video cameras use.






FYI, it's next to impossible to get a perfect sensor. There will always be dead pixels, hot pixels,
or weak pixels that have to compensated for (within reasonable limits of course).

Deanan
Dalsa

I am well aware of that. A "workable" sensor means basically: "one that can be made to work" by using whatever lookup-table jiggery-pokery is available to the manufacturer. Have you ever seen what comes off even a standard definition sensor when all the "makeup" is removed?

You certainly can apply correction on a pixel-by-pixel basis but that limits the dynamic range of the sensor, as you will be well aware.

But it's nice to get all these dirty little secrets out in the open isn't it? Takes some of the mystery out of why Digital Cinematography hasn't exactly set the world on fire :P

Edited by Jim Murdoch, 07 August 2006 - 05:43 AM.

  • 0

#20 Alan Lasky

Alan Lasky
  • Basic Members
  • PipPip
  • 55 posts
  • Other

Posted 07 August 2006 - 10:05 AM

Algorithms. "Put up or shut up." Whatever. Let's see what else might be "setting the world on fire," Shall we:
_______________________________________________
Kodak Posts 2Q Loss, Will Cut 2,000 Jobs

AP NEWS WIRE

Aug 2, 2006, 10:24


Eastman Kodak Co. posted its seventh quarterly loss in a row on Tuesday and moved to axe 2,000 more jobs as it continues to move away from its waning film business, reports the Associated Press. Its shares tumbled nearly 14 percent to their lowest close in 15 years.

Due to $214 million in restructuring costs, Kodak lost $282 million, 98 cents a share, in the April-June quarter, close to double its loss of $155 million, or 54 cents a share, in last year's second quarter.
Stung by a continuing rapid slide in sales of silver-halide film, Kodak's revenues fell 9 percent to $3.36 billion from $3.69 billion a year ago. Excluding one-time items, Kodak lost $54 million, or 19 cents a share.

With digital-imaging sales now outpacing those from film, paper and other chemical-based products, the 126-year-old company is trying to find ways to spin out larger profits from its dominance in the digital imaging market.

Kodak said Tuesday it is shifting manufacturing of digital cameras to Flextronics International Ltd. and transferring 550 employees to the Singapore-based company.

Those will be among 2,000 jobs that Kodak is aiming to eliminate by the end of next year ? on top of 22,000 to 25,000 jobs already targeted since January 2004.

Kodak did not disclose where the other layoffs would be made, but analysts think most are associated with its tightening of global sales operations and its $1.8 billion buyouts last year of Canada's Creo Inc. and Sun Chemical Corp.'s 50 percent stake in Kodak Polychrome Graphics.

Kodak's stock fell $3.05, or 13.7 percent, to close at $19.20 on the
New York Stock Exchange ? its lowest close since it finished at $19.17 on July 24, 1991. More than 50 percent of the shares are owned by long-term institutional investors.

In the quarter, overall digital sales rose 6 percent to $1.83 billion, while revenues from film, paper and other traditional, chemical-based businesses slumped 22 percent to $1.52 billion. Profits from digital businesses totaled $4 million, compared with a $25 million loss in last year's second quarter.

Kodak reaffirmed that it expects to post an overall operating loss of $500 million to $850 million in 2006 and earn $350 million to $450 million from digital operations. But it lowered its forecast for digital sales growth to around 10 percent from a range of 16 percent to 22 percent.

Kodak predicts that by the end of next year the majority of its restructuring costs will be behind them.

Kodak has already cut 20,500 jobs, including 1,630 in the quarter. The shutdown of film and other manufacturing operations looks likely to drop its global work force below 50,000, down from 75,100 in 2001 and a peak of 145,300 in 1988.

The company acknowledged in 2003 that its analog businesses were in irreversible decline and outlined an ambitious strategy to become a digital heavyweight in photography, medical imaging and commercial printing by 2007.
__________________________________________________

Let's enter the real world of shareholders, quarterly loss sheets, and 'institutional investors.' As my Jamaican friends say, Jim: "Soon come."

Alan Lasky
  • 0


Broadcast Solutions Inc

Abel Cine

Tai Audio

Aerial Filmworks

CineTape

CineLab

Rig Wheels Passport

Media Blackout - Custom Cables and AKS

Glidecam

FJS International, LLC

Paralinx LLC

Gamma Ray Digital Inc

rebotnix Technologies

Visual Products

Metropolis Post

Willys Widgets

Opal

Wooden Camera

The Slider

Ritter Battery

Technodolly

CineLab

Abel Cine

Technodolly

Gamma Ray Digital Inc

Aerial Filmworks

CineTape

Metropolis Post

Rig Wheels Passport

Wooden Camera

Visual Products

rebotnix Technologies

The Slider

Media Blackout - Custom Cables and AKS

Tai Audio

Glidecam

Paralinx LLC

Broadcast Solutions Inc

Opal

Ritter Battery

FJS International, LLC

Willys Widgets