G3 - Image quality (In practice)

While looking through the G3 botanical images in this period no particular problems came to mind. And on the plus side I think the availability of wider apertures expanded my creative opportunities. (At full wide angle the SX10 ranges from f/2.8 to f/8, a range of three stops. This decreases as focal length increases finishing up at f/5.7 to f/8 at the telephoto end, a range of just less than one stop. In contrast the 45-200 and the 45-175 range from f/4 to f/22 at their (not very) wide end, a range of 5 stops, going down to f/5.6 to f/22, or 4 stops, at the telephoto end.)

I was not so happy as I looked through the invertebrate images from this period because a lot of the images did not look very sharp to me (the ones I picked to include in the following posts are of course amongst the sharper ones). I think this problem was mainly a post processing issue. I reprocessed the following three images (all from JPEG originals) using my current software and workflow. For invertebrate images like these the workflow is based on batch processing with DXO PhotoLab, Silkypix and Lightroom, with final adjustments in Lightroom. In this case I did the batch processing and almost nothing else apart from cropping to match the cropping of the originally processed versions. I think the recently processed ones have (variously) better clarity, sharpness and/or detail; the exact look in terms of lightness and tonality is of course a matter of personal preference - reviewing the recent versions as I draft this post the recent versions look a bit too light overall for my taste, but don't think that detracts from the point about sharpness/clarity/detail. 

I have not marked up these animations to show which version is which but I hope that will be obvious. You may see some banding/posterisation in the backgrounds. I believe these animated gif files use a restricted colour palette which is presumably the cause of the banding/posterisation, which I cannot see in the originals.

Click on an image to see a larger version

I had not noticed this issue with the S3is or SX10 images I had looked through. This made me wonder if the issue was G3-specific. I picked the following three SX10 invertebrate images and reprocessed them from the original JPEGs. I saw a similar improvement and so I suspect I simply missed the issue when rushing through the S3is and SX10 images picking out ones I liked and not really thinking about the others.