Why not make CCD a sensor / module option for the DSMC?

[Graeme Nattress]

Practically any way of using CCDs can be used with CMOS and the other way around too! Major difference between the two is the pixel design and A to D conversion, where on CCD AtoD is off chip, and in CMOS it's on chip.

You can do rolling shutter or global shutter with both CCD and CMOS! You can do 3 chip, or 1 chip bayer or 1 chip RGB stripe with CCD or CMOS!

Graeme

 

[Aaron Newsome]

My understanding was that film cameras also suffer from rolling shutter, albeit in a more "analog" fashion... i.e. As the shutter physically moves, partial areas of the film / frame suffer uneven exposure to light...
Is that not so?

I'm not aware of this. I'd love to hear the mechanics of how this phenomenon actually works. I thought the film was stationary in the gate while it is being exposed. The physical shutter, which is a wheel, exposes every piece of the film the exact same amount of time.

It would seem that film doesn't suffer from the issues associated with expose/read/store/reset that a CMOS sensor would.

I hadn't heard of motion picture cameras using film, "suffering" from a rolling shutter. I'd like to know more about it though.

 

[Michael Romano]

The rolling shutter now is a nightmare when you're shooting a wedding scene. Flash bulbs everywhere and halfed frames are abound. I am eagerly looking forward to a global (Or even a vastly reduced) sensor acquisition speed! I've been able to dig up only one other system with a global shutter, the WEISSCAM HS-2 which looks pretty incredible. Got to catch it in action for NBC at the Ironman competition this past year and the footage was pretty incredible.

Question: Obviously every camera flash is different, but what combination of RED shutter would help correct for people taking pictures in a scene? Would shooting it in high speed 2K then ramping it back up to normal be effective?

 

[Craig Parkes]

I'm not aware of this. I'd love to hear the mechanics of how this phenomenon actually works. I thought the film was stationary in the gate while it is being exposed. The physical shutter, which is a wheel, exposes every piece of the film the exact same amount of time.

It would seem that film doesn't suffer from the issues associated with expose/read/store/reset that a CMOS sensor would.

I hadn't heard of motion picture cameras using film, "suffering" from a rolling shutter. I'd like to know more about it though.

There used to be stills floating around from the movie 'Pitch Black' where you could see the strobes causing partial exposes in the frame because they weren't sync'd to the camera. It definitely can happen in a film camera.

 

[Aaron Newsome]

Yes, you are correct. The partially exposed frames due to unsynced HMI, flashbulbs, etc can be a concern. I guess I was thinking in terms of motion artifacts.

 

[Andrew Walker]

The read/reset speed on the new M-X sensor is double that of the original sensor. Its just a tad slower than film.

 

[Max Wendel]

Practically any way of using CCDs can be used with CMOS and the other way around too! Major difference between the two is the pixel design and A to D conversion, where on CCD AtoD is off chip, and in CMOS it's on chip.

You can do rolling shutter or global shutter with both CCD and CMOS! You can do 3 chip, or 1 chip bayer or 1 chip RGB stripe with CCD or CMOS!

Graeme

How hard or expensive is it to implement a global shutter on CMOS?

 

[Graeme Nattress]

Stereotype - you need to get your real name in your profile as per forum rules. In the meantime, just put your name in your signature file.

Answer is that I'm not a sensor designer, so I don't really know. But as far as I understand, global shutters have certain trade-offs in other areas that are undesirable, hence our goal of super-fast rolling shutter.

Graeme

 

[Stuart English]

Film cameras.

Film cameras.

I thought the film was stationary in the gate while it is being exposed.

That is true

The physical shutter, which is a wheel, exposes every piece of the film the exact same amount of time.

That is also true

The physical shutter, which is a wheel, exposes every piece of the film the exact same amount of time.

What's missing in your thought experiment is because the shutter is moving, parts of the film frame see light at different times.

All parts of the image have the same exposure TIME but they don't necessarily have the same exposure START or STOP times.

The difference is a function of the method used (rotational or linear shutter) and also the shutter (rotational or linear) speed.

 

[Craig Parkes]

How hard or expensive is it to implement a global shutter on CMOS?

Red were/are planning a mechanical viewfinder option, that would make it like film, but there really doesn't seem a point if the read reset rate is currently the same as film. If it was going to cause issues on film, then it would have to be fixed in post anyway, so with the new sensor read reset pace there really is no cause for concern.

 

[Matthias Hutter]

But as far as I understand, global shutters have certain trade-offs in other areas that are undesirable

I thought that a global shutter reduces the pixel fill rate (more on-chip electronic) and thus reduced dynamic range and high-iso performance. Another issue is cost.
I think if back side ilumination will become affordable in large sensors (not going to happen for a while), than trade-offs could be drastically reduced.

 

[Graeme Nattress]

I think you could be right Matthias - and perhaps power requirements go up too (not sure there) ? As I say, I'm not a sensor designer, but I do know that there are pros and cons.

Graeme

 

[Joe Calabrese]

The biggest problem would be cramming a lot of pixels onto a CCD. Ever notice why Panasonic can release a 1/3" CMOS camera that has full raster chips? The only other full raster cameras they offer is the Varicam 3700 and the HPX3000 (2/3" CCD).

CMOS is less expensive, better in low-light, and offers the ability to get more pixels into a smaller size. However, many people do know that there are some trade offs with this technology.

 

[Jannard]

No sensor technology is perfect... nothing in life is actually. But... our goal was to produce a film alternative camera. With the new M-X sensor, we have virtually matched a film cameras skew performance (4 vs. 5 milliseconds)... we have Peter Jackson successfully intercutting The Lovely Bones with film and RED, RED footage has been described by many as "filmic", we have more resolution than film and almost every other motion capture digital camera, you can shoot 5K @ 100fps, incredible low light performance, the color science is now spot on... and all at a price/performance ratio that can't be matched by anyone in the industry. So why exactly would we want to change to a CCD?

Jim

 

[Stephen Williams]

you can shoot 5K @ 100fps,
Jim

Hi Jim,

Did I miss something?

Stephen

 

[Steven Caesare]

Hi Jim,

Did I miss something?

Stephen

Perhaps you missed the context of the OP's question, which was a replacement sensor for the DSMC.

The DSMC lines includes Epic, for which the sensor capable of those framerates already exists.

-sc

 

[Stephen Williams]

Perhaps you missed the context of the OP's question, which was a replacement sensor for the DSMC.

The DSMC lines includes Epic, for which the sensor capable of those framerates already exists.

-sc

I asked as Red Ones original spec was 4K @ 60 fps, the 5k @ 100 fps is presumably also subject to change.

 

[Steven Caesare]

I asked as Red Ones original spec was 4K @ 60 fps, the 5k @ 100 fps is presumably also subject to change.

Then why are you asking now? Or about any of the other specs? Why single that one out?

 

[Dave Edsall]

Epic will shoot 5k @ 100fps

 

[Julio Quintana]

Just out of curiosity, what kind of media will we need to capture 5k @100fps. That sounds like an enormous amount of data.