Olympus-OM
[Top] [All Lists]

[OM] Re: My Olympus DSLR rant

Subject: [OM] Re: My Olympus DSLR rant
From: Wayne S <om4t@xxxxxxxx>
Date: Fri, 28 Apr 2006 09:02:55 -0400
I also one of those who gave up on Olympus and the E-1, which is a fine
camera up to about ISO400. But the Canon 5D I have will do as good
or better at ISO1600. You couple that with an IS lens and the number and
types of shots goes way up. If you want to see what the 5D can do with
low light, take a look at this photo (not mine), which is a 1 minute exposure.

http://www.timecatcher.com/Marc/transformation.html

I bought into Olympus OM cameras because they could do low light
shots like this. The Olympus 4/3 sensor size is one of the smallest
reduced size sensors, and will always lag behind the larger sensors.
I'm sure they will improve the noise performance over time, but so
will the bigger sensors, so Olympus will always be behind the curve
on noise performance. And no matter what noise reduction software
you get, you can't create information from nothing. You can fix it
to look better, but noise is noise.

People also claim they like the E-1 colors better, and if you need
jpegs out of the camera ready to go, at iso 400 or lower, then they
may be correct. However, I'm beginning to think that is related
to the higher dynamic range of the Canon sensors. And I believe
the future will continue to increase the dynamic range. Given that
I think a new approach to post processing will become the norm.

I think I have figured out a nearly optimal workflow to work with this
higher dynamic range Canon sensor. The optimal exposure for
digital is to get as close to clipping in the highlights as possible.
But doing so tends to create a rather flat looking photo. To give an
example, here is a shot I took last night during low light (2 sec exposure)
with the Canon 5D. This is what the exposure looks like after
converting in PS raw converter and no enhancement adjustments:

http://www.airsprite.net/pub/flat_MG_5544.jpg

This is not what the scene looked like to me, but is what a higher
dynamic picture will tend to look like, a bit flat. This is what the
photo looks like with about 15-20 seconds of post processing in
the Lab color space:

http://www.airsprite.net/pub/lab_MG_5544.jpg

The critical things to watch when converting from raw is to back off
on the two end points, (black point and exposure adjustment) so you
preserve all the data, however the photo will look a bit flat. Now there
are two types of adjustments to do. The first is to use Lab color space
and then adjust the levels. Adjust the A and B channels symmetrically
on both ends to enhance the color separation, and adjust the L for
the black and white points. Next select the L channel and sharpen
only it. (if you are going to reduce the image size then do that before
sharpening) The second optional step to do *before* sharpening is to
use the shadow/highlight adjustment, which is similar to the LCE
(local contrast enhancement) step many of us use, but with the
shadow/highlight command you can be more selective on where
the contrast enhancement applies. (I'm not sure if this command
works better in Lab or RGB, I'm just now learning it) The shadow/
highlight command can bring back detail in either/both the shadows
and the highlights, and can also adjust contrast in the mid-tones.
It is a very powerful filter.

This revelation in Lab workflow and why it works so well with the higher
dynamic range digital sensor of Canon occurred to me after starting
to read a book by Dan Margulis titled "Photoshop LAB Color".
Check this link for the types of things it discusses:

http://www.dgrin.com/showthread.php?t=18203

The Lab color space has been a revelation to say the least.
Also, if you need to reduce noise in an image, working in
Lab is very interesting, since the AB channels are color only,
you can blur them with any number of blurs, surface blur, dust
and spec filter, and have very little impact on the sharpness of
the image. I'm also beginning to think this approach to noise
reduction beats most canned noise reduction programs, because
you have a lot more options and control over the process, so you
can adjust the procedure and customize it for each image.

Most of the above basic adjustments can be done very quickly in Lab
color, and the more I understand it, the more the mystery is coming
out of why there is such a difference in images people claim about
different cameras. I'm now much more likely to go for shots
in sub-optimal light knowing I can probably get the look the way
my eyes see it, even though the camera does not. The camera
just has to not clip, that is, expose for max dynamic range.

I am convinced that Dan's book is the new way to work with today's
digital cameras. I've always been skeptical of just anecdotal claims
people tend to make about this camera or that camera. I know
photography is an art form. But I'm an engineer by training so I
need to know why whenever possible. Check out the www.dgrin.com
site for the discussion on Dan's book.

WayneS


==============================================
List usage info:     http://www.zuikoholic.com
List nannies:        olympusadmin@xxxxxxxxxx
==============================================

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