1.6 crop factor requires sharper lenses?

[Lenny_D]

Correct me if I'm wrong......
The resolution of the image at the sensor surface is determined by the lense. The crop factor of the 300D or 20D implies that only the central part of the image is recorded. For a print with a fixed size (say A5) the recorded image from a 20D is blown up more (factor 1.6) than the same picture taken with a full frame sensor (in combination with a lense with a lower focal length and the same frame on the print). This means that the quality of the lense is much more important for a 20D or 300D than for a camera with a full frame sensor......! Does this imply that you should always buy the highest resolution for these consumer camera's and that this is relatively important?
All 20D and 300D owners should buy high quality primes then! Or is the resolution of a 200mm f/2.8 prime higher than a 20D sensor can record?
In other words when does the sensor resolution match the lens resolution? Can somebody explain this to me?

Lenny

[Jon]

Just as in film, the resolution of an image is affected by both the lens and the recording surface. In the case of the 20D vs. the DR/10D/D60, the 20D is capable of recording a higher resolution image (since the individual sensor sites are smaller). The degree of enlargement doesn't enter into the equation except insofar as over-enlarging a picture will give you a bad picture. IIRC, a sensor's resolving capability is roughly 1/2 the linear sensor site density (a 20D, with 3500 pixels/22 mm can resolve about 80 lines/mm; a DR/10D/D60 does about 70 lines/mm). A good, fine-grained film can record up to 100 lines/mm, although some films are better than others. So I don't think digital users have a lock on needing high resolving power.

You also need to consider that resolving power isn't the only lens design element you need to be concerned with. Contrast and the various and sundry aberrations glass is heir to come into the picture as well. A lens with high lines/mm resolution and abysmal barrel distortion isn't going to win any prizes from its users (unless it's marketed as a fisheye :{)# ).

[Lenny_D]

Jon, I completely agree on distortions and contrast.
Regarding resolution it all boils down to the number of lines/mm that a lense can resolve, I suppose. Where do I find the number of lines/mm that the various lenses can resolve?

Lenny

[DocFrankenstein]

www.photodo.com (http://www.photodo.com)

and there are some other sites i don't remember off the top of my head

[Lenny_D]

If I look at the photodo site it reports a weighted MTF for several different numbers of lines/mm. To stay with the 200mm f/2.8 L USM its shows a MTF of 0.55 at 40 lines/mm. If I understand this number correct, the 80 lines/mm of the 20D is very high and the lens is limiting the quality of the final image.
Back to practical photography. I own a 20D with two lenses: the 18-55mm kit lense and the 50mm f/1.4 . At 100% pixel size a huge difference is noticable between the resolution of the two lenses (apart from the superior colors and contrast of the 50mm).
Also in print, the resolution is noticably better for A5 format.
Still makes me wonder.... is the sensor able to outperform the lenses?

Lenny

[CyberDyneSystems]

It's not so much the crop factor that pushes the res limits of the lens,. as "pixel density" .. a number which includes the sensor size as one of it's factors.

On the flipside,. a "crop factor" can make a "bad" lens look better...
Most times it is the outer edges of a lens that are the worst.. (actually pretty much allways) in some cases the "crop" factor can indeed crop off the parts of what would have been the distorted areas of an image .. making a lens in fact more usefull...