Olympus-OM
[Top] [All Lists]

Re: [OM] Mega Pixel equivalent of medium format negatives

Subject: Re: [OM] Mega Pixel equivalent of medium format negatives
From: Chuck Norcutt <chucknorcutt@xxxxxxxxxxxxxxxx>
Date: Wed, 05 Nov 2014 08:14:57 -0500
Using Occam's Razor, the simplest explanation for f/9 producing results as good as larger apertures is that the photographic "system" doesn't deliver the maximum that theory would predict at larger apertures. I doubt that you'd disagree with that. In fact, I think it's your main point. I also agree that measurement trumps theory every time but I still find theory useful to avoid chasing unlikely results.

I suppose I should spend some more time studying deconvolution but can't get excited about it. I know it's useful in astronomy but that problem is much more tractable than trying to solve the unknowns of general (especially hand-held) digital photography.

Chuck Norcutt


On 11/4/2014 6:59 PM, Moose wrote:
On 11/4/2014 5:08 AM, Chuck Norcutt wrote:
Not so easily determined.  Resolution is ultimately limited by
diffraction.  But diffraction varies by wavelength with blue being
least affected followed by green and then red.

If you shoot a 4/3 image at f/8 the diffraction limits leave the blue
at a max of 16 MP but cut the green to 8 MP and the red to 5 MP.

Remember analog TV? The color was at a much lower resolution than the
luminance.

To get the theoretical maximum out of a 16 MP 4/3 sensor you'd have to
limit aperture to f/6.3 or maybe even f/5.6. Only at these or larger
apertures would you then be able to consider the effects of reds and
blues having fewer sensels.

Well, phooey. I was trying to avoid this ... As with contrast, in my
last post, the kind of loss of resolution you describe may also be
changed in post. Deconvolution increases effective resolution, partially
recreating the 'signal' before it went through the convolution of lens
defects, including aperture size.

In theory ( that tricky thing, again), I understand that with full
information on what the lens does to light passing through, the original
may be completely reconstructed. General purpose, non lens specific
deconvolutions can't do anything near that, but it can recover some
resolution.

A lot of things that were true for film and optical prints are not the
same anymore. Unless adjustments are made, they are often inaccurate for
digital. It's my sense that deconvolution and contrast adjustments,
local and global, add about a stop, so f8 is not, in effect, diffraction
limited on 4/3. Even f11 gives away very little I can see, with the
right tools and ability to use them. I've been using f9 as my default
lately. Maybe because I like odd things? Or bumped it and didn't change
it back? ;-)

You make me think it might be interesting to run deconvolution at
different intensities on the color channels. Likely too crude for much
use. Oh dear, am I channeling Mike G? :-)

That reminds me. I still haven't re-compared DxO's lens/sensor specific
deconvolution to FM's generic version. OH well, another day.

Convoluted Moose

--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/

<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz