I can tell you, I dislike these over sharpened images where halos and artifacts fights with each other being the most dominant part of the image. If you compare any image shot with an analogue camera to one achieved with a digital camera you will note the following differences:

Analogue image: more noise (film grain), less detail in the shadows, more micro details, crisp

digital image: smooth, more detail in the shadows, less micro details, flat

Get the most of your digital SLR

This action is optimized for a six Megapixel Digital SLR like Nikon D100, D1x or Canon D60, but you can use it of course even with 3 Megapixel cameras. It is optimized for the Nikon D100 at 200 ASA but it works also with other Digital SLR’s and ASA settings because it sharpens only where it is needed.

Digital images appears often a little blurred. This is the result of keeping down the noise level.
Every digital sensor either CMOS, CCD or the new Foveon produces some noise. So with higher sharpening setting you will get more noise and vice versa. Also high contrast settings will generate more noise.

The best thing is leaving all settings which concern sharpening and contrast to a low level, leaving camera in-processing as less as possible. Even better to use the raw file format and control all processing later on your PC.

Flat and blurred, conventional sharpened                                More crisp and visible micro details

To get the best „digital negative“ you should shoot with the following settings:

• RAW-Modus 16 bit
• Low sharpening
• Low contrast

I have seen a lot of methods how to sharpen images, but some lack in edge sharpening and the other produce a lot of noise without sharpening any details.

The better way to go

A correct processed image should look very natural, a wide range of tonality, accurate shadows and highlights and what’s most important: well sharpened mid-tones!

The most visible difference between an analogue and digital image is the amount of micro detail in the mid-tones which determines the sharpness of an image. And it’s exactly there where my action will work.

Well perhaps all pro photographer will remember how they achieved beautiful Cibachrome prints from those terrible crisp kodachrome slides where the big problem was to handle heavy contrasts, dark shadows and blown-up highlights. A real pro lab had to use several masks (pos. & neg.) to succeed in printing.

Now I am using a similar technique not to manage contrast, but to enhance sharpening in the mid-tones. As I mentioned at the beginning the mid-tones will determine how sharp and crisp an image is. Micro detail like a structure in a piece of wood or the texture of a cloth have to appear clearly visible and well defined.

Sharpening using a positive and negative mask

The action creates the two Alpha channels

That is my solution. Using a tweaked positive and negative mask, intersect both selections and you will sharpen only those parts and details that needs to be sharpened, without producing any halos around edges and no additional noise. While working only in 16bit-Modus you will get smooth curves without any gaps.

Of course there has to be done some tweaking and adjusting to the levels, contrast and color balance, but it’s all included in the action-action.

Try and enjoy a new way of processing digital images.

Roberto's action can be downloaded here.

How the Action works

This action works best when we use RAW-images in 16-bit mode.

1) Open image in 16-bit mode

2) Enhance contrast with a moderate „s-curve“

2) Apply usual tweaking in levels, color balance and saturation still in 16-bit

3) Convert to 8-bit mode and select Lab color, select Lab luminance channel and
copy it.

4) Go back 2 steps in the history to RGB, do not convert!

5) Add two alpha channels and paste the Lab channel into them

6) Inverse the first alpha-channel. Now we have two masking channels:
One is negative and one is positive

7) The dark areas of the negative channel will be tweaked using Levels to output value 20,
so that the shadow areas will be not that much enhanced.

8) First select the neg. channel and then by intersecting the pos. Channel -
we got a selection that protects the same way highlights and shadows. Only the mid-tones will be affected by mayor enhancement, so the USM-sharpening can be applied with very strong values like 500% _ 0,6 Pixel_2 Threshold.

Note: You will get a Photoshop warning that the selection cannot be displayed. Do not worry about this.

9) Now we switch back to 16-bit mode to apply the masked USM sharpening with the values
mentioned above. Working in the 16-bit mode gives smoother tonal curves.

10) This way we get more sharpened micro textures without producing artifacts and halos around the edges. The picture looks more natural and even more crisp.

About Roberto Casavecchia

Roberto is a graphic artist and photographer with over 15 years of experience with analogue and digital imaging processing, several exhibitions, book editions and writing for magazines covering the whole area of digital imaging.

First of all we find this method very interesting and like the results. We also experimented with some stronger sharpening radius settings (0.8-1.2) and found them useful too.

Roberto, thank you very much for your excellent contribution.