Re: Digital vs 35mm camera optics
From: Stephen H. Westin (westin*nospam_at_graphics.cornell.edu)
Date: 08/19/04
- Next message: Henk van der Kraan: "glueing plastic optical fibres"
- Previous message: Helpful person: "Re: re:Gaussian beams - ZEMAX"
- In reply to: Bob Monaghan: "Re: Digital vs 35mm camera optics"
- Next in thread: Bob Monaghan: "Re: Digital vs 35mm camera optics"
- Reply: Bob Monaghan: "Re: Digital vs 35mm camera optics"
- Messages sorted by: [ date ] [ thread ]
Date: 19 Aug 2004 10:19:07 -0400
rmonagha@engr.smu.edu (Bob Monaghan) writes:
> yes, but the problems with aliasing and other flaws of the Kodak high
> density sensor camera offerings are well known, and if nothing else, have
> helped increase the appreciation for anti-aliasing filters in digicams.
Ah, but when it works, it's breathtaking.
And what about the Sigma/Foveon cameras, also without AA filters?
You claimed 2 things:
1. Digital cameras all have optical AA filters, and
2. These filters remove all detail beyond 50 cycles/mm.
The first isn't true, and I suspect the second isn't, either.
> Lens resolution limits won't be much of a resolution limit, at least with
> most 35mm camera lenses, esp. at the faster f/stops. That's because lens
> aerial resolutions often run 200 to 300 lpmm and up; I have some pentax
> lenses which are in the 400+ lpmm aerial resolution range in 35mm, some
> zeiss and kowa lenses that are similarly stellar optics, and a number of
> leitz/leica lenses in 35mm are rated over 600 lpmm aerial resolution.
So what's the contrast at these resolutions? If it's at the visible limit,
then any film MTF less than 1 wipes it out, doesn't it? Not to mention
projection optics.
> In
> these cases, above f/16 diffraction limits anyway, the lens is unlikely to
> be the system resolution limiting factor, rather, that 40 to 50 lpmm
> anti-aliasing low pass filter does those honors ;-)
Actually, one workaround for when your Kodak shows moire is stopping
down to f:22. So it looks like the sensor Nyquist frequency is pretty
close to the resolving limits of many lenses.
> Yes, I supplied some references in my posting, the AFIPs paper has
> evidently been dropped/deleted/lost, drat. Still, you can believe kodak or
> pop-photo's real world tests.
Sorry, but I seem to have missed the Kodak reference. As to Pop Photo, I
would be quite skeptical. When was the last time they tested a product
and said it was lousy?
> Or zeiss' tests, which at least show what is
> possible with the better films and lenses and good technique, but without
> requiring special processing or other techniques.
Execpt, of course, we can't apply those results without knowing what
the tests were.
> Or you can calculate it out easily enough, 24x36mm, at 100 lpmm = 200
> pixels/mm, we get 4800 * 7200 = 46+ megapixels, 125 lpmm =-> 54 MP and so
> on. So isn't 200 lpmm = 400 pixels/mm -> 138+ Megapixels, per zeiss for
> ektar 25 ;-)
Again, I don't know what that 200 means. What difference does it make
in the image?
> I used Tmax 100 last 3 weeks ago for some of my tree "portraits" around
> White Rock Lake (many of which trees got damaged in recent floods and
> storms, some washed out into the shallow lake itself ;-( I just bought a
> dorm sized fridge full of short dated fuji film from some overstocked
> local stores, one of the positive side effects of the more rapid than
> expected switch to digital ;-)
>
> The problem with all the comparisons you cited is quite simple; they were
> all scanned imagery, right? When you put film on a scanner, it gets
> trashed;
And when it goes through an enlarger or projector, all is wonderful?
I think there's some superstition about scanners. They certainly vary
in quality, but the best are quite good. After all, their fundamental
limitations are the same laws of physics that apply to projectors and
enlargers. See <http://www.clarkvision.com/imagedetail/scandetail.html>
for a comparison of various scanners.
> for one thing, lots of scanners have Nyquist limits close to 50
> lpmm or worse too, right?
Wrong. Many of the scans on the page I pointed out were at 4000 dpi or
more; that would be 2000 cycles per inch or 78 cycles per
millimeter. And when we see film grain, doesn't that suggest that the
scanner isn't the limiting factor.
> ;-) When you do laser drum scans, you get a
> closer experience of what film does.
So what did you think of the drum scans I pointed you toward?
<http://www.terrapinphoto.com/jmdavis/> shows scans from Aztek,Howtek,
Scanmate, Heidelberg, and ICG drum scanners.
> Or you can do what I do and project
> your slides really big and glory in the colors and details you can see.
What is the resolution limit when you project? BTW: projection limits
contrast to about 100:1, by my measurements.
> And when I whip out my 15X loupe, I am always amazed at being able to read
> tiny print or see details that I never saw with my eyes when taking the
> photos.
Which seems pointless, if your viewer needs a 15x loupe to see it.
> I don't get that kind of fine detail in digital shots, rather the
> smoothed and interpolated images which some viewers obviously like a lot
> (their privilege ;-), but which I don't, probably due to years of looking
> at MF quality images, I'm just spoiled for digital ;-)
Well, I can't argue with you, as I don't know what digital images you
have seen. The Kodak SLR forum at DPReview.com has a number of
examples,
e.g. <http://forums.dpreview.com/forums/read.asp?forum=1026&message=8853929>
or <http://www.warrensearle.com/Kodak14n/Examples/>.
-- -Stephen H. Westin Any information or opinions in this message are mine: they do not represent the position of Cornell University or any of its sponsors.
- Next message: Henk van der Kraan: "glueing plastic optical fibres"
- Previous message: Helpful person: "Re: re:Gaussian beams - ZEMAX"
- In reply to: Bob Monaghan: "Re: Digital vs 35mm camera optics"
- Next in thread: Bob Monaghan: "Re: Digital vs 35mm camera optics"
- Reply: Bob Monaghan: "Re: Digital vs 35mm camera optics"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
