About big sensors...

[simonfilm]

One of the things that I less understand in electronics are the image sensors (CCD, CMOS). I see how Processors, Memory, Hard discs and many other things advance in giant steps.

But what happens to sensors?. It seems that we are stuck in the same sizes. Yes, thay have grown in pixel density in the last years. But we are at the 1/3 inch limit for prosumer cameras.

The super expensive cameras like Panasonic Varicams, Sony F23 and many pro ENG cameras use a 2/3 inch. This is 8,8 x 6,6 mm, ridiculous compared to the DSRL or RED ones.

Wouldnt it be possible to have a camcorder with one (not three) 35 mm sized CMOS, or at least a CMOS the size of the one that comes in DSRL cameras like Canon EOS400 (22 x 15 mm). I mean a sensor not of 4K, nor 2K, nor even 1080. 720p would be enough, and for many indie filmmakers a standard definition (PAL 720x576 25p), like the Panasonic DVX100 with that big CMOS would be wonderful. Imagine no more mini35.



Thanks.

Simon.

[Kevin Halverson]

Part of what effects the size of any silicon product (sensor, memory, micro, etc) is the die yield. There are two factors at work here. One is the number of finished products that result on a given size wafer. This is a simple equation and is based upon the percentage of wafer area. The second concept is a bit more complex and has to do with defect density. All wafers have defects, so if something is large enough (in relationship to the wafer), the usable yield will approach zero as a defect will nearly always be present. This partially explains the reason for fewer large sensors in the market; its simple economics.

As for why not three sensors. Well this requires additional optics which in itself has 'issues' which effect performance.

The size of the sensor in the RED ONE matches that which is used in digital cameras so current optics have the same behavior (field of view and hence depth of field).

[jbeale]

The size of the sensor in the RED ONE matches that which is used in digital cameras so current optics have the same behavior (field of view and hence depth of field).

Probably meant "...matches that used in [35mm motion picture] film cameras...".

As far as what's available on the market vs. what's possible, apparently, a lot of people are happy with the image from existing compact video camcorders. The rapid advances in computer chips and hard drives has been supported by the global consumer commodity market for computers. Red One is an amazing piece of technology but I think some big companies that might have the technology to do something like it, haven't been interested because they believe the market for 35mm digital cine equipment is too small.

[Chris Kenny]

A large chip is not necessarily what you want in a prosumer camera video, and it's almost certainly not what you want in a consumer one. Large chips make focus much more of an issue, which consumers don't want to deal with. They also require much more substantial optics. All consumer and most prosumer video cameras have a fixed zoom lens with a fairly long range -- 10-20x. Have you seen how big a even a 10x zoom that covers 35mm is?

[David Mullen ASC]

Yes, you have to factor in that a larger sensor needs a larger lens, which runs counter to what most people want in a consumer camcorder, which is to have something that fits in the palm... but has a zoom lens on it.

And generally a consumer doesn't want to deal with 35mm depth of field for moving subjects (still images are easier to deal with, focus-wise.)

But there are also manufacturing and marketing reasons why a company like Sony wants to make 1/3" CCD's for their consumer cameras and 2/3" CCD's for their professional cameras, and not put those 2/3" CCD's into their consumer cameras.

As for why their professional cameras don't use a 35mm-sized sensor like the RED, one reason I've been told is that they don't want to piss of their customer base who has invested millions of dollars in professional 2/3" optics over the years, by now making a camera that needs 35mm optics.

[dvpixl]

It seems like trend among indie shooters are going backwards now. DV used to be easy and fast. now everyone wants the real stuff- without the hassle.

[Chris Nuzzaco]

I think "trickle release" tactics might also be a work to some extent. Why sell the best camera in the world to date, when you can sell all kinds of lower end versions of it that slowly lead up to the best one? To maximize profits...

[Craig W. Bickerstaff]

Straight off the f23 FAQ

Why did Sony adopt a three 2/3” CCD optical path for this of camera instead of using a single 35mm imager?
There is a wide range of 2/3” lenses available for purchase or rental. These lenses are designed to work with a prism optical system, and are affordable, rugged, and match the image making capabilities of the F23.

Is Sony planning to introduce a 35mm CCD version of the F23 like Panavision's Genesis™ camera?
No.


Personally I don't think 35 mm lenses/imagers are all that important for cinema I'd be more than happy to shoot a film on Sony's F23 camera or the new XDCAM EX.
It's just a nice option to have Sony doesn't need to make a 35 mm camera.

[deepcode]

You have to see also most of the professional cameras are used for ENG. The reality look wants to have as much as possible information in the picture. The more DOF the better, the more details are in focus. It is the complete opposite to scenic work / storytelling.
On ENG you often have to deal with unpredictable situations, often in bad light, you have moving objects and people but nevertheless you have to get a proper focus.
So 2/3 is the best for this because of a relatively high DOF for this and high light sensivity.
Also the 2/3 glass is tolerable in weight and size and the 2/3'' chips don´t need the power of a larger chip, so the batteries last longer. You simply can´t carry a 35mm sized camera with a bigger and heavier zoom glass comparable to a 2/3 all the time.

:: www.deepcode-one.com ::
sounds, vision, interaction