If you like to see some in depth laboratory tests made on the LX3 you can have a look at diwa-labs.com
LX3 - DIWA LABORATORY TESTS
If you like to see some in depth laboratory tests made on the LX3 you can have a look at diwa-labs.com
Subscribe to:
Post Comments (Atom)
2 comments:
Taking an arithmetic *average* of the reduction (in LOG Base 2 steps) of (actual) ISO Sensitivity (as measured by DxO Labs for (claimed) ISO 80, 100, 200, and 400, the result equals -0.6362 stops. See:
http://www.dxomark.com/index.php/eng/Image-Quality-Database/Panasonic/Lumix-DMC-LX3
Select the "ISO Sensitivity" results tab. A "mouse-over" (over the location of the data-points themselves) will display the precise measurement values at the various data-points.
It's interesting to note that (as widely reported, and consistently found by me) Panasonic has calibrated their Exposure Level (as well as their Live Histogram) on the assumption that the ISO Sensitivity is as claimed by Panasonic (and not as measured by DxO Labs).
Thus, +2/3 EV of Exposure Compensation regularly needs to be applied in order to correctly expose ...
Note: In some cases of a lot of highlights in the frame (when using the multi-area metering), only +1/3 EV is required.
This tendency to under-expose results in the widespread impressions that the shadow areas in (non-exposure-compensated) shots seem unusually (and perhaps excessively) dark ...
The (non-adjustable, and rather irritating) aspect of this that troubles me is that I have to push the Live Histogram *past* its maximum and *guess* as to "how much is too much" - and review each potential shot for the "playback" Histogram (which, as a luminance histogram composed of the sum of RGB is, in itself, not necessary reflective of the actual RGB separated histogram).
It is evident that the luminance (and/or chroma) elements of a recorded JPG image's (full) histogram (basically) *do* appear to begin to "clip" (go non-linear, "blow highlights", etc.) at the far-right edge of the Live Histogram (as it is calibrated on the camera).
The ".RW2" image file has (about) 1/2 to 2/3 stops more "headroom" available before the (full, actual) histogram components begin to "clip".
So, the (unfortunate) net result is:
(1) To get the ".RW2" image file exposure right, the user has to "over-drive" the Live Histogram and "guess" as to "how much is too much"; and, at the same time
(2) The luminance and chroma components of the (full, actual) histogram *will* be "clipping" at a lower level (1/2 to 2/3 stop lower) in the ".JPG" image file - at a level corresponding the maximum-right-hand edge of the on-camera Live Histogram.
This forces a very non-ideal for the user who wishes to shoot (both) RW2+JPG. One if forced to:
(1) Keep the JPG from clipping by restricting the exposure level on the Live Histogram, and end up with an under-exposed JPG (as well as an under-exposed RW2); or
(2) Push past maximum on the Live Histogram in order to properly expose the RW2 (thus causing a JPG with oft-over-blown highlights, which, at least, does not appear under-exposed).
There seems no straight-forward reason why Panasonic cannot calibrate their exposure level and the Live Histogram to correctly reflect the reduced (actual) ISO Sensitivity!
I suspect that this is an intentional Panasonic marketing decision in order to make the uninformed user marvel at how low the noise seems at an (alleged) ISO 400 (when, in fact as reported by DxO Labs, they are shooting at *only*) ISO = 259)!
Such marketing choices may boost sales, but they clearly make shooting (both) RW2+JPG on the LX3 problematic - in that only *one* of the two can be optimized for proper exposure.
Further, having to exceed the Live Histogram and "guess" (in order to optimize the RW2, as well as to avoid an underexposed JPG) leaves a lot to be desired!
This is one of my major annoyances about the LX3, and Panasonic should forgo this "trick of the light" in favor of allowing the user to (actually) optimize the signal/noise ratio of the images (by maximizing exposure levels when shooting).
I have, upon further investigations acquired a better understanding of these matters.
While a number of people have (as I did) become concerned about this data reported at the DxO Mark database (regarding the DMC-LX3 as well as other cameras) - I located more information that more fully explains the meaning and correct interpretation of this DxO Mark data.
Therefore, my previous comment posted above is NOT accurate and complete!
I have posted newer (and I believe accurate, more complete, and reliable) findings at:
http://ricohforum.com/phpbb/viewtopic.php?f=22&t=1556&start=30#p16548
and at:
http://ricohforum.com/phpbb/viewtopic.php?f=22&t=1556&start=30#p16621
Please refer to these posts for updated information that definitively explains the issues surrounding these matters.
Post a Comment