Oh yes, you may have actually got one of the later offshoot copies of the K7 ... can't remember what its called, k7IIs, something like that. Its no wonder you didn't notice any difference - they look identical.

I'm holding off with the K7 for the next release, hoping for some new options for Fine Edge Detection - looking for 3-stage custom options : 1. hair threads 2. brush bristles 3. fluff

Very 'pin-sharp', impressive captures, and the little creatures all look very engaged (No 5 focus is fractionally off isn't it?)

I like the way you have limited the contrast range, despite the conditions, which is right for this type of subject I think. Only thing is, I wonder if they are a little over-exposed, just 1/3rd - 1/2 a stop?

Great work there. You're getting some real mileage out of that atmospheric location. 1 is nearly perfect - does it need a slight nudge clockwise?

No 2 has energy in the sky and water, and I especially like the perfectly pitched shadow levels on the underside of the jetty. Maybe there's just a little too much room on the right side?

No doubt, the latest K5II/s models are a lot of camera, and image-making performance, for the price.

We certainly will - if you can try and get some bits of fluff, hair threads and brush bristles in the shots that would be ideal to ensure you're getting the most out of it. And we look forward to hearing your reports on the PP settings applied and resulting image degradations incurred as a result.

I agree up to a point - up to the level of the first images above, and for standard types of output, yes, that's effectively my conclusion. But the last image C just above represents about the limit I felt I could push this image to achieve razor sharp large output before degradation kicked in too much. The question is, which of the camera's took it - and could the other give us more, or less ?

It was always meant to be about the real world for me ... the real world of image processing and output. I can confirm that the quality and detail of the 'original' processed files, as viewed directly on my monitor in Fullscreen Photo Viewer and iMatch's side-by-side windows, has translated perfectly well onto the forum - I don't see any significant loss through having output to 'web-size'.

Rob, I agree that there are no discernible differences in the inherent image characteristics, and I accept that the signs of processing are there. Lets remember that its not just +1 sharpness in PDCU, there is additional default work going on under the hood which goes a long way towards the equalisation. But at this simple level of processing and for normal, average output demands, then clearly these tiny processing signs are of no significance.

They certainly became more significant further along - my IMAGE C here has been subject to much greater detailing, with inevitable signs. But I took it up to the point that I saw as acceptable - ie: no significant image degradation but extracting the maximum clarity the image could give. And this image could provide larger printed output more easily.

Let me test you some more .... this time I'll put up an image from further through my tests - this one (I'll call it Image C) was taken from the same PDCU TIFF as before, but this time also processed in PS Elements using the TopazDetail plugin - this provides sliders to control sharpening within 3 specific ranges - small detail, medium detail and large detail. The plugin first runs and examines the photo and then allocates what it considers an optimized range of pixel radius settings to each of the 3 ranges. So when using the 'large' sliders you're only affecting smoother, more uniform-toned areas. Its similar to manually controlling radius in Photoshop, just easier to apply and more intuitive.

I applied as much detailing as I considered was normal for me when trying to extract the best detail for large print output - and I stopped when I felt the IQ degradation price was becoming to high.


For comparison purposes first I'll put the same image in its earlier version, as posted above (minus the label!) ....









Now here's the same shot from the later version with the extra processing ....

IMAGE C




Now who can tell what camera this one is from ?

I would like to clear up any possible misunderstanding which is suspect may have crept in ..... it was not the purpose of my testing and all the image processing to try and produce identical images. I did not intensely scrutinize the shots & tweak all the controls to try and make them indistinguishable. I don't even know if this would be possible.

What I was doing was running the shots through a series of 'real world' sharpening and detailing approaches, a complete range of techniques that would be used by people for differing output needs. I made standard adjustments and progressively increased the strengths until I saw the K5II image 'equalise', and I reported these settings at various stages.

Up to the point that I considered the K5II image 'equalised' without significant degradation (for practical photography purposes) then I reported a recommendation that effectively said 'you can get enough without the 's'. Once the visible degradation's exceeded this then I reported that you'd be better with the 's'.

So I am not at all surprised that some tiny differences are being seen .... I commend Rob on his eagle eye. I was aware of some of these tiny differences, but my own assessment on them was that they varied between the shots, so tended to cancel each other out. That was enough for me though, for my definition of 'equalised' ... they surely are for any practical purpose. It was never the aim to produce indistinguishable images, only to explore the practical limits to aid personal choice.

OK - which is which? I commend the analytic brains here on this forum - IMAGE A is from the K5II, and IMAGE B the K5IIs. Remember, the 'non-s' image has only +1 of sharpness increase over the 's' version - nothing else is different and the processing through RAW conversions and JPEG outpu is identical. There is a latent slight extra contrast in the 's' image (I saw it on the histogram when testing), and I believe this is causing the slight extra saturation.

But, what I was saying with the images at this early stage was that if your processing followed this kind of simple global approach, and the images were not highly critical or intended for very large output - (this is the sort of simple sharpening I'd use just for family shots say) then there's no real benefit to the 's'. The K5II image has been 'equalised' (made equally effective) for the intended purpose, with minimal effort.