Zoom vs Sensor Size - More Zoom == Smaller Sensor?

In another thread there's been some debate about zoom and sensor size.

With the current cameras,

More Zoom == Smaller Sensor

OR

More Zoom == Bigger Camera Body

For example, the s95 has a larger sensor, but less zoom -
1/1.7" Type CCD sensor (larger than the 2 below)
Dimensions - 100 x 58 x 30 mm
Zoom - 28-105mm (35mm equiv), 3.8x optical zoom



sd4500 is a smaller "superzoom".
1/2.3" type back-illuminated CMOS sensor
Dimensions - 101 x 59 x 22 mm
Zoom - 36-360mm (35mm equiv), 10x optical zoom



The Canon sx30 is a bigger superzoom -
1/2.3" Type CCD sensor (same sensor size as the sd4500)
Dimensions 123 x 92 x 108 mm
Zoom - 24- 840mm equiv, 35x Optical zoom *



I admit this is a little hectic, but I don't have more time to put into it as the moment - it seems like when you want a lot of zoom, you get either a smaller sensor or a larger camera body. Thoughts?

Well the claims have always been that it's a matter of real estate but I can't imagine there would be much issue with fitting the sensor and zoom into a camera similar in size to the SX30. I also don't think anyone would complain if their SX30 was ever so slightly larger if they could get the bigger sensor. It seems more like a matter of money to me. More zoom plus bigger sensor might equate to a price higher than they think they could get within a given market. That's just speculation of course - surely the real reason isn't so simplistic.

Correct assesment.

harrison2 said:

Correct assesment.

Click to expand...

Almost. You can add "or a slower lens" to "smaller sensor" or "larger body".

--
Lee Jay
(see profile for equipment)

larger zoom range == larger lens
and larger image circle == larger lens

The sd4500:
f = 6.3 - 63.0 mm

The s90:
f = 6.0 - 22.5 mm

And then DSLRs are also huge because of the mirrors and prism, and the fast electronics.

Why not solve the problem by having a large sensor with lots of megapixels and just cropping? Diffraction limits? This page has a calculator:
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm

A 40 MP 4/3-type sensor would be diffraction-limited to under 12" print size at f/11.

Maybe electronics, then, or lens sharpness. Don't know.

--
Not waiting for anything.

Michael Barkowski said:

Why not solve the problem by having a large sensor with lots of megapixels and just cropping? Diffraction limits? This page has a calculator:
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm

A 40 MP 4/3-type sensor would be diffraction-limited to under 12" print size at f/11.

Click to expand...

Shoot - just answered my own question. Cropping a 12" print in order to zoom? Not gonna work. Plus, cropping reduces depth of field further.

--
Not waiting for anything.

ljfinger said:

harrison2 said:

Correct assesment.

Click to expand...

Almost. You can add "or a slower lens" to "smaller sensor" or "larger body".

--
Lee Jay
(see profile for equipment)

Click to expand...

Hmm...that might be a good point...

If they came out with a superzoom with awesome IQ, hence a bigger sensor then they would lose a lot of entry-level DSLR business.

I may be crazy but isn't Fuji out of the DSLR business...if so, maybe it will take someone like them to come out with a high-quality superzoom and take the business away from everyone else.

Another contrasting example: the Canon Pro1.

7.2 - 50.8mm f/2.2 - 3.5 (28-200mm equivalent)
Sensor slightly larger than G12/s95.

On the large end of the scale (for compacts, anyway), due to the slightly larger sensor, the lens being both brighter/faster and longer range than the G12. Check out the cross-section:
http://www.dpreview.com/reviews/canonpro1/

Some proposals have been made for bending the shape of a lens using a prism, to make the camera more compact. Don't know of any practical application in use, but I would imagine it would first be used in cellphone cameras:
http://www.opticsinfobase.org/abstract.cfm?uri=JOSK-13-2-206

--
Not waiting for anything.

Some of the Panasonic TZs use "folded optics" with a prism.

--
Lee Jay
(see profile for equipment)

so this is saying that cropping softens the final picture. but to what extent is that noticeable?

i've been trying to decide whether to keep the 24-360mm Samy HZ30W i have (12mp 1/2.3 CCD 80-3200 f3.2-f5.8)...

... or sacrifice that reach for the better IQ of the 28-140mm G12 (10mp 1/1.7 CCD 80-3200 f2.8-f4...

... hoping that i can crop G12 shots when i can't get close enough.

so if i cropped half the shot because i couldn't get closer, perhaps due to a fence at a zoo or a monument, is it possible to "calculate" (or just estimate) if the loss in IQ will be noticeable at prints A4 or smaller?

ajaxblue said:

so this is saying that cropping softens the final picture. but to what extent is that noticeable?

i've been trying to decide whether to keep the 24-360mm Samy HZ30W i have (12mp 1/2.3 CCD 80-3200 f3.2-f5.8)...

... or sacrifice that reach for the better IQ of the 28-140mm G12 (10mp 1/1.7 CCD 80-3200 f2.8-f4...

... hoping that i can crop G12 shots when i can't get close enough.

so if i cropped half the shot because i couldn't get closer, perhaps due to a fence at a zoo or a monument, is it possible to "calculate" (or just estimate) if the loss in IQ will be noticeable at prints A4 or smaller?

Click to expand...

The G12's pics are 3648 pixels long, 2736 pixels wide. Here's a really blunt rule of thumb: as long as you have 80 pixels per cm, you should be okay. 80 to 120 ... depends on how picky you are, so experiment.

So, for example, if you cropped in half length-wise you'd have 1824 pixels. Divide that by 80 and you get 23 cm (9"). Divide by 120 instead and you'd get 15 cm (6").

In other words, with a 10 MP camera you can make up for some of the reach, but only for small prints.

This is all assuming that you need to see fine detail in your photo. A lot of times that is not really necessary, in which case it's fine to print much, much larger than what the above rule would indicate ... just as long as you resize the file properly in an image editor.

All I was originally trying to highlight, was that there are other optical limits which make increasing the number of MP basically pointless, at least until there is some major technology breakthrough.