DSLR:biggest imagesensor for the buck???

[hentaitic]

This is what I know so far?

Sigma SD9 20.7x13.8mm
Canon 10D 22.7 x 15.1mm
Canon 1Ds 35.8x23.8mm
Fuji S2 23x15.5mm
Nikon D100 23.7x15.6mm
Kodak DCS 14n 36x24mm

But I am not too sure about the prices.

So which DSLR has the biggest 'sensor' for the buck?
Is it the Canon 10D?

 

[sulhan]

And the list will grow.....

 

[ckhaos]

Now that I think about it, what is the problem with manufacturing full size CCD's? I mean, is there a technical limiting factor? Or is the problem a business one, in that the manufacturing is not cost-effective?

 

[denizenx]

Originally posted by ckhaos
Now that I think about it, what is the problem with manufacturing full size CCD's? I mean, is there a technical limiting factor? Or is the problem a business one, in that the manufacturing is not cost-effective?

same as LCD monitors.. bigger the yield not as good.. it's a "seasoning" of the process factor...

 

[ckhaos]

Originally posted by denizenx
same as LCD monitors.. bigger the yield not as good.. it's a "seasoning" of the process factor...

This is what everyone says, that the yield is not as good for big CCDs. But if you assume that the probability that you get a defective pixel in the manufacturing process is x%, then it doesn't matter whether you are manufacturing a small or a big CCD --- you will still get the same yield. You get less defective pixels in smaller CCDs, but a defect is a defect, and if you have a defect, your CCD is bad, regardless of whether it is big or small. So, I don't understand why big CCDs are harder to manufacture...

 

[denizenx]

Originally posted by ckhaos
But if you assume that the probability that you get a defective pixel in the manufacturing process is x%,

here is your error. it's not x%.
changing density is very tough, but changing die size is not easy either... it's not as "digital" or linearly scaling as u think... not just modify X from 20 to 50 and voila big ccd...
even maths wise:
lessay it's good = gd x gd x gd x gd 4 pixels
ramp up to 5 pixels = gd x gd x gd x gd x gd increase in chance of failure liao even if each pixel is lessay 5% fail-prone

roughly like this, even if the maths is not so simple it's similar.

 

[ckhaos]

Right. Got what you are trying to say.

One more question: Is is easier or harder to manufacture a 2 megapizel small CCD compared to a 2 megapixel large full-sized CCD?

 

[denizenx]

Originally posted by ckhaos
Right. Got what you are trying to say.

One more question: Is is easier or harder to manufacture a 2 megapizel small CCD compared to a 2 megapixel large full-sized CCD?

erm... depends on what they were making all along... sometimes common sense dun apply here (really!)
imagine I only know how to make char siew bao... u tell me the mkt changed, now need pork bao. so my new pork baos might turn out to be char siew bao but without the red colour, until mths later when I can get the "std" pork bao correct...
(LOL weird example but true)
same for CCDs, it's a tech xfer from maybe webcams or something, so it might have started "small and cheap", then need time to go to "big and good quality".
so I like to use the word "seasoning", where it means to get the new process smoothed out, via tech or experience or costs or just plain luck...

 

[f55user]

Imagine something like a D100 or 10D or whatever with a full-frame sensor. Then the DCS14n or 1Ds, which costs like 3-4 times more, don't need to sell liao.