Archived posting to the Leica Users Group, 2000/03/10

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: RE: [Leica] Re: Leica and the digital future
From: Austin Franklin <austin@darkroom.com>
Date: Fri, 10 Mar 2000 19:34:44 -0500

> I was talking to someone this lunchtime about why it should be that 
digital
> camera design should have pursued cramming more pixels onto small chips,
> rather than enlarging the chips up the size of 35mm frames, thereby
> facilitating the union of the two technologies.
> The reply was that the manufacture of large chips is a most complex and
> difficult one, whether this will ever be overcome, I've no idea.

[Austin] That is correct.  One issue is the size issue, both going smaller 
and larger.  It is uneconomical to make the cells larger, therefore making 
a larger chip to cover the larger area the lense projects to.  The smaller 
they can be made, the more that can be made at once, and therefore, the 
more economical each one is.

Secondly, it is expensive, and though getting better, there is a 'yield' 
problem when you put more cells in a larger area.  If cells are bad, and 
the odds of them being bad increase with the more you have.  Given the same 
process, the entire chip can be of no use.

Given a 1/4" CCD is, say 640 pixels square, and a 35mm film plane is 1" x 
1.5", that would be a sensor of 24 times the size...that's a lot of pixels. 
 2560 x 3840 pixels.  Do-able, but QUITE expensive given current day 
technology.  That would be an array of 9.8M pixels, and say 8 bits per 
color, so a single picture would be around 30M....and storage technology 
isn't quite there to make this really work if you want more than 10 or 20 
pictures per 'roll' (read as 340M mini hard disk).

Also, there is the time it takes to scan an array of that size.  Say it 
runs at 10ns per pixel (all three colors at once), so it would take 300 
milliseconds to scan the array...  That may or may not be a problem...

It's something that we will probably see in 3-5 years or so.  There is 
quite a bit of new technology that has to be 'accomplished' in order to 
make this economically feasible.

- -------------