Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: "Film" lenses with small sensors  (Read 2281 times)
sergio
Sr. Member
****
Offline Offline

Posts: 661


WWW
« on: June 29, 2003, 07:11:33 PM »
ReplyReply

Exactly.
Logged

Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #1 on: July 01, 2003, 05:02:04 AM »
ReplyReply

Not exactly.

Cheap lenses are still cheap lenses. Cheap lenses have poorer center performance than good lenses, its just the magnitude of difference is greater at the edges. Cheap lenses also have slower AF motors than good lenses. Why put a Yugo engine in a Ferrari?
Logged

BJL
Sr. Member
****
Offline Offline

Posts: 5060


« Reply #2 on: July 01, 2003, 09:58:20 AM »
ReplyReply

Whatever you achieve with lesser lenses through the cropping done by most DSLR's, you could also do by cropping the same amount when you print from 35mm film, so you should expect similar trade-offs: less problems at the edges, but less sharpness overall due to greater enlargement of the part of the image used.

I have not heard of anyone consistently cropping all film images in order to improve image quality from lesser lenses, so I doubt it is an optimal approach with digital either, now that the pixels are small enough to reveal the resolution limits of even quite good 35mm format lenses.

So I would expect that getting the best out of DSLR sensors smaller than 35mm format requires very sharp lenses, even more so that with most 35mm film: either top quality 35mm format lenses, or lenses specifically designed to give more sharpness over the smaller image circle involved. Perhaps Nikon DX or Olympus E-series lenses will achieve this, but I have seen no solid evidence yet.

That is not to say that "advanced amateur" lenses like the Canon 24-85 often paired with the 10D are hopeless; they just sacrifice a fraction of the sensors' full potential in most tests I know about.


About larger maximum apertures, the industry trend is exactly the opposite: new pro zoom lenses at f/4 rather than the traditional f/2.8, perhaps taking advantage of the higher ISO's now usable both with pro transparency film and DSLR's to reduce cost, bulk and the abberation problems at larger apertures.
Logged
BJL
Sr. Member
****
Offline Offline

Posts: 5060


« Reply #3 on: July 02, 2003, 12:59:47 PM »
ReplyReply

With all the new formats (sensor or film sizes) produced by digital, it would be nice to have some clear distinctions between (a) resolution in the traditional sense of lp/mm on the image formed in the camera, and (b) measures relating to resolution in resulting prints, taking account of the different magnifications used in printing from different image sizes.

Perhaps the simplest measure, used on some digital equipment reviews, is "lines per picture height" (lines rather line pairs seems to be common these days, if only for the sake of possibly misleading comparisons to pixel counts).

For example, the 30lp/mm often used in MTF measurement of 35mm lens performance gives 1440 lines per picture height for 35mm format, but with the 10D it is only about 900 lines/height, probably a bit too low for reliable estimates of print sharpness under normal viewing conditions.

Canon and Nikon should be pushed to publish MTF curves at more appropriate scales for lenses that they advertise as suitable for use with APS format DSLR's; perhaps the new standard for MTF sharpness measures should be something like "MTF at 1500 lines/height".
Logged
Tony Collins
Guest
« Reply #4 on: June 29, 2003, 05:37:23 PM »
ReplyReply

Because cameras like the 10D don't use the whole image circle does this mean that lenses with poor edge definition look better than when used on film because their weak area is off the edge of the sensor? Likewise is barrel distortion which is apparent on the edges of a picture less apparent when used on a 10d than on film?
Logged
Tony Collins
Guest
« Reply #5 on: July 01, 2003, 01:00:24 AM »
ReplyReply

So I can use "cheap" lenses on a 10d and get acceptable ?? quality.
Logged
Tony Collins
Guest
« Reply #6 on: July 01, 2003, 06:22:16 AM »
ReplyReply

As a non pro who buys his own equipment rather than claiming it against tax (In UK to boot with inflated list prices and VAT... fade in violin music) I am interested in finding out real world information on how lenses perform. The consumer magazine "tests" make some bland comment on how edge sharpness is affected aperture wide open, but don't say whether a film or cropped sensor digital camera was used.

On an associated thread, less expensive lenses start at a smaller aperture presumably to keep edge softness within acceptable bounds. If used on a small sensor, the same lens design could be used with the aperture pushed a bit wider? If so this could be an opportunity for the third party lens manufacturers to market larger aperture lenses that work with the 1.5/1.6 multiplication DSLR bodies
Logged
Ray
Guest
« Reply #7 on: July 01, 2003, 07:38:32 PM »
ReplyReply

One should also bear in mind that many lenses, especially the cheaper ones, do not have particularly good performance at full aperture even in the centre. In fact, some of the cheaper third party lenses with impressively big apertures (the Sigma 20mm f1.8, for example) are quite awful at full aperture, but okay a couple of stops down.

The other issue is one of different formats. Cameras such as the 10D and D60 are really smaller format cameras than 35mm. They just happen to use the same lenses as 35mm. When comparing different formats, it's not considered polite to crop the larger format to the same size as the smaller format because then one might arrive at a conclusion that the smaller format is in some ways better than the larger format - which might confuse some people. For example, the D60 is not considered to have quite as high resolving power as 35mm film. However, if one were to crop the 35mm film to the same size as the D60's sensor, an area less than half the size of 35mm, the D60 would without doubt show superior resolution. On that basis, one could argue that better quality lenses are needed with the DSLR's.
Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad