Thanks for the image samples and comments, Ian. I read that and the Forum posts a few weeks ago but haven’t had time to post until now.

I agree with your position that the important criteria for assessing the images is what you can print them to. I think a lot of people these days see digital cameras as semi-disposable electronic toys that can make all their decisions for them. To my mind the real test is what you can get through the whole photographic process. That used to involve the chemical darkroom. These days it involves Photoshop and the digital darkroom, with greater repeatability and improved quality.

I've had an FZ20 for about 20 months, using it primarily to takes photos of Blues bands. At first I thought that 400ISO was unusable but as my processing capabilities improved I found I could get quite good results even with underexposed 400ISO images. My workflow includes Photoshop CS2 (RGB and LAB), PK Sharpener and Neat Image, with a significant amount of individual processing for each image and at least potentially resulting in A3+ prints.

I didn't go for the FZ30 because it was slower than the FZ20 but the FZ50 is another matter. I now have an FZ50 and have taken shots with it at a couple of gigs. At first I was disconcerted by the noise at 1600ISO but after considering it for a while I am not so bothered. I think it is about equivalent to the 400ISO noise from the FZ20 and mainly applying to shadow noise so that it will be easier to control for images exposed to a “correct” histogram. On that basis I think it will remain effectively true that the FZ50 is one and a quarter stops faster than the FZ20 with the additional benefits of a larger sensor and 50% greater resolution. This latter is important because I think it is easier to control noise than it is to manufacture detail.

And then there is the 3200ISO setting. It does come with built-in tonal dropout and reduced resolution and that's OK if it lets you get a shot that was otherwise unobtainable - You just have to recognise it for what it is, perhaps as a starting point for further manipulation, and I don’t think the scorn of the empirical pixel-peepers is necessarily relevant. After all, in practical terms there is no objective reality in a photograph even where a perception of reality is a point of reference. What is more disappointing though is that it is only a scene mode so that the only possible control over what it does would be exposure compensation (I haven't checked but I presume that would still be active).

For the moment I'm not tempted by a DSLR. I'd probably have to spend 5 times the price of the FZ50 to get a technically superior DSLR of equivalent low-light and zoom capability. My main vehicle for landscape photography is a 6x17 panorama camera but no doubt I'll also be checking out what the FZ50 can do in the lower ISO ranges, especially by combining bracketed images.

Regards,
Murray

Postscript: With my first look at SilkyPix as a RAW Developer I was inclined to dismiss it. Partly since RAW development for the FZ50 is not yet available in ACR and Lightroom, I decided to give it another go – if only to see what bypassing the Venus 3 engine would do (at least I presume there’s no noise reduction in RAW). I processed an image I had already dealt with from a Jpeg and was alarmed to see the degree of noise that was presented. However, this effect was very much reduced with the file imported to Photoshop so apparently that’s just a weird artefact of SilkyPix presentation. Not that noise was absent, it was just exaggerated in SilkyPix.

RAW processing can presumably produce an extended tonal range but this wasn’t really relevant for this sort of image. I found it difficult to get the two versions exactly comparable, partly because the processes were different and partly because I didn’t save a curves definition file of what I did in LAB to the first image. The perception of noise is also affected by colour balance, contrast and luminosity. Still, in Photoshop the image from RAW had both more noise and more detail; it was more of a problem to control but I think had more potential and a slightly better result.