The last couple of times that I have attempted to include photographs in a text post I have been unable to upload the images. So this time I thought I’d try a different method, i.e. creating ‘photo’ posts which I didn’t know I could do! Now I find that there is an option here to include photos so I’m going to try that too. Sharpened photo overload!!
We were asked to choose one image that had been processed but not sharpened, save that, then copy the image and apply different amounts of sharpening to the copied photos so that we could compare them.
A portrait photo was suggested so that we could consider the image of sharpening on the eyes, where we might want some as well as on the skin, where we probably would not and I chose a photo that I had processed in part for when we were looking at adjusting skin tone, highlighting eyes, etc.
Although I printed the 4 images as was suggested, I actually found it easier comparing the results in Photoshop, using the ‘actual pixels’ view and Lightroom in the 2:1 view. I have cropped all 4 images to get a closer view of the eyes and those with ‘no’ and ‘some’ sharpening are posted below.
The more sharpening that is applied, the grainier the image becomes, particularly around the eyes and fine strands of hair. What is more noticeable though is the effect sharpening has on the skin, it becomes puckered and blotchy and every blemish becomes more pronounced, so the photos with less sharpening are certainly better for that reason. The DPP part 5 text suggests that one of the problems with over sharpening is that halos can appear around the sharpened edges although I think I have been able to reduce this to some extent using the ‘lens correction tool’ in Lightroom. One thing that I did discover though was that when you sharpen and image in Lightroom, one of the options is to add an amount of ‘mask’ I’m not sure exactly what this means but the effect is to smooth out some of the graininess created by the sharpening. This is clearly a tool that needs further exploration and I can only imaging that the best way to do this is by trial and error.