Olympus-OM
[Top] [All Lists]

Re: [OM] (OM) digital imaging and anti-aliasing filters

Subject: Re: [OM] (OM) digital imaging and anti-aliasing filters
From: Moose <olymoose@xxxxxxxxx>
Date: Wed, 27 Jul 2011 21:03:29 -0700
On 7/27/2011 6:08 PM, Brian Swale wrote:
> It looks as though for landscape photography anyway, it would be an
> advantage to remove any anti-aliasing filter from the camera.
>
> See here
>
> http://www.maxmax.com/nikon_d200hr.htm.

This is not as simple as you might imagine. Look at the vertical moiré 
patterns. Assume you have shot something without 
an AA filter, and such patterns afflict it. How are you going to fix it? A 
whole lot of hand work will improve it, but . 
. .

Then how about the circular one? How will the details of your feathery foliage 
look with no AA filter? Obviously, 
different makers make different compromises. I'm guessing Oly shots of those 
tests would show no visible moiré. 
Anywhere, including foliage, where fine lines run parallel, even for a short 
distance, moiré can rear its head.

But wait, there's more! Instead of looking only at the things he points to, 
look at the whole images.

Look closely at the letters below the first moiré test pattern. I can see it , 
but you may need a glass or to enlarge 
it. The AA letters are softer edged, but the others have little busy artifacts 
around them that are hardly visible in 
the AA version. It would appear that the AA filter is also covering up pixel 
level failings of the sensor and/or 
demosaicing algorithms.

So a weaker or non-existent AA filter is going to be a mixed blessing, no? 
Also, deconvolution algorithms are 
specifically designed to deal with the kind of OOF that AA filters cause - not 
by intent, perhaps, but by nature. So 
take a look at this. <cid:part1.04030301.06010407@gmail.com>

Deconvolution, in this case in the form of Focus Magic, has at least leveled 
the playing field, if not given the AA 
image a slight edge in sharpness/detail definition. But look at the thin black 
on white lines of letters and scale. FM 
has also enhanced the artifacts around them to about the level of the non AA 
image.

If this were a really important image to me, I could do some masking and use 
other tools to get the best of both worlds, 
but it admittedly takes time. But I sure prefer dealing with the AA softening 
to trying to correct moiré problems.

So think before you leap; is there a fire outside the frying pan? I know you 
want images direct from the camera that are 
as close to finished as possible. In this case, I think you are better off with 
the AA filter than without. Yes, there 
are cameras with weaker and/or different AA filters and/or sensor systems that 
may hit a better compromise, but that's 
another story.

I'd stick with the camera you have, and learn to correct the problem is those 
cases where it is necessary. If that 
doesn't do what you need, then it's time to consider a different camera or 
camera system. There are real reasons why 
some of us decades long OM shooters went with other digital systems.

I'm afraid that my advice may run you into a different problem. Deconvolution 
is computation intensive. Assuming you are 
still running computer equipment of modest power, memory, etc., you may find it 
tediously slow. No free lunch. :-(

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