Olympus-OM
[Top] [All Lists]

[OM] Re: CCD's versus CMOS sensors

Subject: [OM] Re: CCD's versus CMOS sensors
From: Moose <olymoose@xxxxxxxxx>
Date: Sun, 01 Apr 2007 01:12:06 -0700
John Morton wrote:
>                                                >>>><<<<
>    
> Thanks for the heads up on that, Moose; I dug out my Auto Bellows (I tend to 
> use the telescoping auto extension tube much more, for field work) and I can 
> see how the tab you mention would be a problem with that overhang on the 
> pentaprism housing in those photos of the 410/510.
>    
> .......
>    
>   It is interesting to see that this unit will have a CMOS sensor; 
Not really. It is some sort of NMOS sensor, different from any others to 
date except for the E-330 and Panny and Leica versions of that.  A 
different sensor design which can scan continuously in real time is 
required for live view. So it acts that way in live view, then as an 
ordinary sensor when capturing an image.
> this technology seems to have come of age for digital imaging. Flipping 
> through my copy of the 5th edition of John Russ's "The Image Processing 
> Handbook", I note that the trend seems to be away from CCD's and toward CMOS 
> sensors.
>   
Depends on where you are looking. The conventional wisdom was that CCD 
was the future of high quality sensors, until Canon went their own way 
with their own designs and foundry and set new standards for performance 
using CMOS. Nikon went on with CCD and lowered noise with smarter 
amplifier/channel design. Has Nikon gone with CMOS on a recent model? I 
don't remember.

At the lower end of cameras, CCD still rules the roost.
>    
> The more compact pixel spacing of CCD's might tend to give a better 
> resolution, but the higher energy requirements of a coupled charge sensor 
> also increases the noise generated by electrical interference in the 
> circuitry. While CMOS sensors have lower energy requirements, they also have 
> more circuitry (the signal amplifiers, for instance) right on the chip... 
> which again increases the noise from electrical interference. The added 
> circuitry of the CMOS sensor increases the spacing between pixels: this 
> decreases resolution, but, also serves to decrease the 'blooming' effect 
> (trans-pixel contamination from strong light sources) that CCD's are 
> sometimes prone to.
>   
I saw an announcement that somebody (Canon?) was developing a dual layer 
CMOS imaging chip which moves all the support circuitry below he sensor 
level, bringing sensor coverage way up. CAn't find the reference at the 
moment.
>    
> So, there seem to be a number of trade-offs when choosing between the 500 (8 
> megapixel CCD) and the 510 (10 megapixel CMOS); but, since CMOS sensors have 
> unique registry values for each pixel (whereas CCD's offload entire rows of 
> pixels to their signal amplifiers), it is apparently possible to design 
> software that can (in theory) eliminate the noise characteristics of each 
> individual pixel... and in the end analysis, this will give CMOS technology 
> an edge over CCD's.
>   
You'll have to redo this analysis, as the 510 isn't CMOS. Or just wait 
for the dpreview tests of the 410, which should be coming soon. I prefer 
looking at the results to theory.
>    
> Thus, I am now leaning toward the 510 (with its CMOS sensor); and, I am glad 
> to hear that my bellows and favorite lenses will work (well enough) on that 
> particular model.
>   
Bellows, 65-116 and the macro lenses work great on 4/3s and Canon EF 
mount DSLRs.

Moose

==============================================
List usage info:     http://www.zuikoholic.com
List nannies:        olympusadmin@xxxxxxxxxx
==============================================

<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz