sonomichele -
Thanks for your impressions about LX3 image-noise (which I assume refer to the in-camera JPG case?). I have, indeed, seen a fair number of images (real-world and test) that show reasonable results in the ISO 400 to 800 range . After nearly four years (and four previous Panasonic cameras), I have evolved to be "gun-shy" of their in-camera JPG NR, responding (instead) by limiting all to ISO=200 at most.
ISO=200 is certainly looking rather viable (in my own LX3 shots, as well as in a bevy of 100%-crop samples I have seen in LX3 reviews) in JPG recording. I (personally) see "perturbations" beginning to appear at ISO=400 in the JPG "crops" that I see might well send me to wanting to apply (further) NR in post-processing.This, however, seems to bode well for the ISO=400 possibilities in RW2 recording. The reported higher dynamic-range of this newer 1/1.63" diagonal image-sensor (as compared to the, typically, around 1/2.5" diagonal range) does appear to please ...
The LX3 image-sensor (in physical dimensions of the active-area, and the pixel-pitch) being very similar to that of my FZ50's image sensor, I am very interested in what this newer incarnation can provide in terms of (peak) signal to (average) noise via RW2 image files, relative to the (also) 10 Mpixel (in 4:3 aspect ratio mode) RAW image files from my FZ50, and its image-sensor.
Trying "burst-modes" early-on with my DMC-LZ5 with OIS active, the subsequent shots (usually) seemed to me to (more often, and for some reason(s)) be less detailed relative to the first shot (by virtue of "camera-shake" and/or or small changes in the camera-subject distance to the extent that such would focusing in such a "fixed-focus" case) .
Are you referring to applying your above suggested technique (with, or without) OIS being operational (or does that matter, in your observations)? I would be curious to know.
I realize that (in my previous brief experimentation with the LZ5), a number of elements relating to
what and how I was shooting may have contributed to these observed results - and the quality itself of the (once fixed)
auto-focus itself in the first place (from the humble LZ5 to the present day LX3) has hopefully been improved in product engineering over that time.
Of the efficacy so far observed (in the DMC-TZ4, anyway) Panasonic Venus IV Engine low-light auto-focus in general:
Note: I am still "gun-shy" on the lower light-level abilities of the Venus IV Engine to manage auto-focus as well as the FZ30/50 do. My previous TZ4 (after no less than 3 firmware updates) with the same Venus IV Engine did an absolutely anything but focus properly (except for at higher light levels, and with very coherent edges, to boot). The spatial frequency resolution of the portion of the image-sensor utilized for auto-focus was abysmal, and I found that I (literally) felt compelled to use (only) spot-focus mode in the case of all shots ... :cry: The LX3 appears somewhat better so far than the TZ4 auto-focusing in lower light-levels in the (normal-speed) single-area focusing mode - but (perhaps) not performing as well as did/does my FZ30/FZ50 in the same situations. However, the implementation of the vastly different in size (and thus in light gathering ability) lens systems in the larger FZs compared to the compact LX3 may perhaps favor the larger FZs - as they would appear to by their larger physical outer lens diameter to have more initial photons to "work with" in order to arrive at an auto-focus "decision" in lower light-levels.
Other factors that may well have been at play. Hand-held, (usually) unbraced shooting, doing macro close-ups, and
slightly moving subjects at times in the near and far-field shooting cases. Using my mono-pod, however, I think that your approach would be well taken in the case of my shooting - as an image stabilization system that has "settled-out" in it's servo-ing process of compensating for the jolt could (perhaps in some situations) be
worse than the natural damping and absorption of such a jolt as accomplished by the mono-pod and it's natural anchoring effects! However, at the long shutter times that I end up flirting with in LX3 Aperture-Priority shooting, I am not sure if my (further) movements in the subsequent shots might not need (further) compensation on an ongoing basis. Thus,
my lowly shutter-speeds may require that my own (predominantly 15 CPS and higher frequency) involuntary physical movements be "monitored" and "mitigated" (so to speak) during the course of the shuttering of all of the "burst" shots - by the "wonder-servo" ... :ugeek:
One sometimes little-considered limitation of the (agreed) excellent Panasonic Lumix OIS (and affecting to some extent all mechanically implemented stabilization approaches) is that the "servo-range" within which the stabilization system can "hold-in" (and thus effectively servo the errors) is fairly small (on the order of barely one degree or so in radial measurement total, from what I have read in published papers of other folks knowing more than I about these systems). I am fairly certain that I have physically exceeded that fairly narrow requirement on a regular basis more than a few times when shooting hand-held in lower light-levels!
So, it seems that with a fair amount of initial mechanical stabilization provided, and a lens-system with qualities weathering well on their own, effective Image Stabilization (in general) can really be a boon (and for me, approaching a necessity at the lower light-levels that I find myself shooting in using F-Numbers on the order of 3.2 or higher, in order to be able to achieve adequate DOF). It's remarkable how many levels of "massaging with technology" the optically projected image goes through (from glass to digitization to in-camera hardware to software comprising post-facto attempted compensation for correctable optical perturbations, all the additional post-processing, etc., in and through final re-sampling and display/print). From all this smoke and mirrors at times emerges beauty!
.
