Should a lens be matched to the sensor size?

[xpatUSA]

For example, matching a Nikon 'DX' lens to an APS-C sensor?

For some years I've preferred full-frame lenses on smaller sensors because I read somewhere that image quality is better (all other things being equal) because the sensor sees the middle part of the lens i.e. less CA, vignetting and the like. But, does using a full-frame lens on a Sigma SD10 for example (1.7X "crop factor") bring other problems, with all that "spare light" doing nothing?

If there's a CiC tutorial mention of the subject, please point me at it.

Thanks,

Ted

[Andrew1]

As you suggest, I would think that an FX lens would be preferred as the resultant portion of the light being used is coming through the less distorted center of the lens while the excess on the edges is being discarded. It seems to me you would get a more consistent frame edge to edge result for resolution, distortion, aberration and focus. That's just my untrained opinion so I'm sure a couple of the local scientists can offer you a more technical evaluation. Nikon also classifies DX lenses as affordable consumer grade so I have to assume the limited number of DX offerings is to keep the price down. That would imply quality was not the prime consideration during design or manufacture.

[Manfred M]

You can use a full-frame lens on a crop factor, and in theory you are right, you are shooting through the "sweet spot" of the lens, so in theory, image quality should be better. I decided that I would eventually upgrade to a full-frame camera about 3 years ago, and picked up a f/2.8 24-70mm and a f/2.8 70-200mm lens a couple of years ago, when I was able to pick them up at a good price and shot them on my D90. They worked quite well and never had a problem with all that "spare light".

I'm not sure where you are going with the Foveon sensor on the Sigma. The sensor configuration is different than on on the standard Bayer array, but the image circle is still what drives the suitability of the lens. The baffling inside the mirror box will handle stray light. The image circle, even on a DX lens is, well, circular, so there will be spill-over at the top and bottom of the mirror chamber, regardless. There will be more spill with a larger light image circle, but the baffling seems to handle it quite nicely.

[xpatUSA]

You can use a full-frame lens on a crop factor, and in theory you are right, you are shooting through the "sweet spot" of the lens, so in theory, image quality should be better. I decided that I would eventually upgrade to a full-frame camera about 3 years ago, and picked up a f/2.8 24-70mm and a f/2.8 70-200mm lens a couple of years ago, when I was able to pick them up at a good price and shot them on my D90. They worked quite well and never had a problem with all that "spare light".

I'm not sure where you are going with the Foveon sensor on the Sigma. The sensor configuration is different than on on the standard Bayer array, but the image circle is still what drives the suitability of the lens. The baffling inside the mirror box will handle stray light. The image circle, even on a DX lens is, well, circular, so there will be spill-over at the top and bottom of the mirror chamber, regardless. There will be more spill with a larger light image circle, but the baffling seems to handle it quite nicely.

Thanks Manfred,

It wasn't a Foveon mention per se, just a good example because the sensor is even smaller than APS-C, at approx 21x14mm. I am much encouraged by responses so far

Best Regards,

Ted

[Glenn NK]

But, does using a full-frame lens on a Sigma SD10 for example (1.7X "crop factor") bring other problems, with all that "spare light" doing nothing?

If there's a CiC tutorial mention of the subject, please point me at it.

Thanks,

Ted

The "spare light" is not used.

I think that this is probably the best way to explain what happens with a crop sensor - the same light goes into the lens that goes into a FF sensor, but the "extra/spare" light around the edges is absorbed into the mirror box (which is black).

And of course this happens with a FF sensor also - the image that is cast by the lens is not rectangular, it's round. So the sensor just utilizes the light within its rectangular box.

Glenn

[DanK]

It's not the spare light that is the issue. It's the spare glass and housing. Lenses designed for crop sensor cameras are often smaller and lighter, because they don't need to illuminate as large a circle.

However, I think it is not helpful to try to come up with a general rule. Some lenses have a lot of fall-off in the corners, while others don't. It seems more sensible to me to decide what focal length you want, given the FOV of your sensor size, and then read reviews of competing lenses. In some cases, a full-frame-capable lens may be best for your camera, while in other cases, it may be a waste.

[rpcrowe]

I use both full frame Canon EF lenses and a crop frame Canon EFS lens on my 7D and 40D (1.5x crop cameras).

The EFS lens I use is the 17-55mm f/2.8 IS and there is no equal lens in the full frame lineup equal to this lens which has:
1. constant f/2.8 aperture
2. 17-55mm focal range
3. Excellent IS (image stabilization) capability

I also use a Tokina 12-24mm f/4 lens which is designed for 1.5x crop cameras but, which IS NOT an EFS lens. It "could" be used on a full frame camera but, will vignette at wider focal lengths. The good thing for me regarding the fact that this is not an EFS lens is that I can use it on my old Canon D60 (note: not the new 60D) DSLR which I converted to full-time infra red photography. The EFS mount lenses cannot be used on any Canon DSLR camera of an earlier model than the 20D (they can be used on all Rebel models).

[tinzel]

I use both full frame Canon EF lenses and a crop frame Canon EFS lens on my 7D and 40D (1.5x crop cameras).

The EFS lens I use is the 17-55mm f/2.8 IS and there is no equal lens in the full frame lineup equal to this lens which has:
1. constant f/2.8 aperture
2. 17-55mm focal range
3. Excellent IS (image stabilization) capability

No scientific background but I am wondering if a lens for a Foveon sensor needs to bring the different colors to focus on different distances because of a separate layer for each color

[xpatUSA]

No scientific background but I am wondering if a lens for a Foveon sensor needs to bring the different colors to focus on different distances because of a separate layer for each color

Interesting thought. However, the layers are very close together:



5 micro-meters is pretty small compared to even a fish-eye's focal length.

Ted