Olympus-OM
[Top] [All Lists]

Re: [OM] Circ Pols

Subject: Re: [OM] Circ Pols
From: Ken Norton <ken@xxxxxxxxxxx>
Date: Wed, 2 Feb 2011 10:44:52 -0600
> BTW, where did you learn all these design details?

Reverse engineering and insider information.

I had spent a long time testing and trying to understand the metering
system. I had encountered anomalies which then required further research.
The manual mentions the presence of EV evaluation, which then launched me
into some backdoor research.

In a nutshell, here is how the A1 (A2 and A200) meter the scene:

1. Focus is determined through wide-field analysis. The AF system will then
determine closest subject within the coverage area. If using single-point,
the camera immediately assumes that to be the subject to focus on. In a
similar manner to the Nikon's 3D matrix metering, the Minolta determined
foreground/background distances. As the lens and camera are a closed-loop
system, the lens passes the distance information to the camera.

2. The matrix metering examines the exposure of the subject (active focus
point) and the overall scene. The entire image area is averaged and the high
and low points determined. Based on a special-sauce algorithm, the camera
makes an assumption about the lighting and hence the scene. When in bright
sunlight, the camera KNOWS it is in bright sunlight! When shooting a sunset,
the camera knows it is shooting a sunset. When shooting interiors under
artificial light, the camera knows it is indoors and in artificial light. It
uses this knowledge to base certain rules on.

3. The subject exposure is weighted as the critical exposure. But based on
#1, if the subject to background distance is about the same, the camera
knows to not base exposure strictly on it, and in some cases may entirely
ignore it and fall back on it's assumed scene. If the overall average EV of
the entire image area is very high, the camera knows it is being fooled and
will weight the scene to Sunny-16. If the overall scene is too dark, then it
knows it is dark and won't attempt to up the exposure too much. Of course,
this requires a single point of bright or dark in the image somewhere for it
to base the evaluative on, otherwise it starts to fall back to dummer modes.

4. If using the flash, the camera will adjust flash exposure for the subject
and aperture/shutterspeed for background. Nothing unique there, as all
systems do that now, but Minolta's system can also include a pre-flash (if
desired) to determine if the subject reflectance falls within what is
expected and if not, it will offset accordingly.

5. If for some reason, the camera is not making sense of the scene. For
example, if you have a polarizer on the lens and you shoot a photograph of a
white car in the bright sun, the camera will likely underexpose the picture.
Without the polarizer, the camera knows it is in bright sunlight, therefore
ignore the really bright blob in the middle. With the polarizer, it no
longer understands it is in bright sunlight and will use the focus-point
(subject) and average to guess. The camera thinks that white car is a person
and the background a darkened scene. So it will base the exposure on the
focus point, underexposing the picture.

6. If the focus-point (subject) and average exposure are close to each other
and the distance calculation for subject and background are close, (provided
it was able to even determine distance--manual focusing or single-point AF
disable some of this), the camera will fall back entirely to averaging.

That's the clifnotes version. Olympus' ESP functions similarily, except,
like the Canon and Nikon systems, it baselines the scene on a relative
exposure basis because they don't always have knowledge of the nor the
distance.

Minolta lifted the algorithm from the Maxxum D7 film camera, which has
arguably the second-finest metering system in any film camera. (OM-4Ti
obviously being the best...). Also note, that this technology is
cross-licensed across the manufacturers so if it seems to be somewhat
familiar, it's because it is. Just like the Nikon F5 system, it is
spectacularily good, but when it is fooled, the failures are equally as
spectacular. Five exposures will be dead on and the sixth will be four stops
off.

A little case history of the A1, A2 are in order. Minolta hadn't released
the D7D yet and used the A1 and A2 as their test environments. The reason
the cameas are so powerful and feature rich is not an accident. These two
cameras were their pro cameras (minus the sheer performance) and even today
are unmatched in many areas. The A2 quickly arrived after the A1 (six months
later), but had certain trade-offs. Regardless, the professional flash
system that Minolta had (and is still in use by Sony) is not only
outstanding, but in some cases better than either Nikon or Canon's best.
Essentially, the A1 and A2 is interchangeable with the Latest/greatest Sony
DSLR and only now are we starting to see a few of the features (especially
the 3D matrix metering described above) being fully implemented in DSLRs.

For professional use in event photography, the A1 is a great camera--if you
stay within the strict confines of its strengths. AF is slow and noise is
high. The lens, though, is simply among the best in any camera ever. That GT
lens is mind-boggling good!!! And it has that 28-200 equivalent focal length
range, so I never have the desire to swap lenses mid-stride. There is some
distortion at some focal lengths, but overall, the lens doesn't possess the
nasty cross-over points which some zooms have. When shooting events with the
A1, it is my preferred Stroboframe camera! I manually focus the lens (no AF
delays), place the Vivitar 285HV flash on Yellow Auto and put a UltraBounce
modifier on it. In almost every case, I use ISO 200, which is a true 320 on
my A1) and that gives me a working aperture somewhere around F5.6. F5.6 on
that format gives me unheard of DoF, so focusing is a non-issue. Oh, and
since the flash syncs at nearly any speed, I typically use aperture-priority
auto-exposure (averaging mode), with around -2.0 exposure compensation.

The RAW processed files clean up very nicely (Thank you RSE), are
exceptionally sharp and the colors are good. Not Olympus good, but still
quite good.

My girls have inherited the A1 since I got the DMC-L1, but I still use it
quite a bit. If Minolta didn't have such quality-control issues with the
D7D, I would have gone that direction instead of the E-1. Overall, I'm glad
I stuck with Olympus, but if it wasn't for a concerted effort by Bill Barber
to keep me in the fold, I'd have definitely gone with that system. The
higher-end Sony cameras are Minolta cameras. The lower-end Sony cameras are
Sony cameras. Some cross-over exists, but you can tell where the
"consumer-electronics" and "PlayStation3) aspects of product design kick in.

AG
-- 
_________________________________________________________________
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