See the difference we can make here?
ScanWizard shows the Before and After images, which are too small to
see detail, but we are not looking for detail. We are looking for the
overall tones in the image. To see detail, we simply scan the image and
You've got to admit this change is worthwhile, and it only takes a few seconds more. Here's how it's easily done.
Your image's histogram data won't look exactly like this, but the purpose of the histogram is to show you the details of the specific image, and after you have looked at 2 or 3 of them, you will see the many similarities. It won't be a problem.
Adjust the tones by sliding the Black and White triangle pointer symbols. Grab one with the mouse and slide it over to where the image pixels begin, as shown. See the previous histogram above for the "before" positions.
Here, we set the Black Point to be 15. All values, if any, from 0..15 will now be Black (0). The reason is that we set the black point to be at the value of 15. That point will become 0, and you cannot get blacker than that. Tones to the left of the black point simply become black too. We "clipped" the end tones off of the histogram data.
We set the White Point to be 224. All values, if any, from 224..255 will now be White (255). Same reason. I don't literally mean the clipped highlight tones will become "white". I mean white in a grayscale. The histogram data is representing these grayscale tones. But if a color image, one or more of the RGB components will be saturated (maxed out at 255). It is bright, but not necessarily literally white.
The image tones will be expanded or stretched to fill the total range, i.e., values from 15 to 224 will be mapped to be 0..255. Said another way, a tone range of 224-15=209 is expanded to the scanner's full range of 255. The stretched tone range has more shades of gray possible within it, and thus can show more detail. We have hardly discarded any tone values, we've just set the range that we do have to fill the entire range of the scanner.
And there is a second concept: Sometimes it's desirable to be more extreme and to "sacrifice" a few highlight and shadow points to increase midrange detail. Discard is the wrong term, we simply remap them to be the darkest (0) or lightest (255), to extend the detail of the rest. It depends on the individual photo, which is the most important. Generally the midrange is most important, but at times, retaining the maximum shadow detail is instead more important. We're describing the hammer here, you'll have to decide where the nails go.
The method used above was to set the Points to the first valley (small dip) in the histogram curve, that is, to a point slightly inside the actual extremes, moving the points to "touch" the existing tone values well, to "clip" off the extremes. This sacrifice of a few values is an arbitrary concept, but the notion is that there SHOULD always be some pure black and some pure white in the image. This full range gives it contrast. Spreading the tones we have to reach from Black to White gives contrast.
The above example timidly touches the end tones with the points, and that is probably correct this time, but I sometimes (certainly not always) like to move the points in more, to considerably clip one or both ends, much more than shown here. This increases contrast, but at the cost of shadow and/or highlight detail. It may or may not be important. There are always choices, mild, extreme, and the middle ground.
Do be aware that a more aggressive attack can sometimes be appealing. Often the colors in dull low-contrast or faded prints can be amazingly rejuvenated by clipping both Black and White points considerably. You can make sensible judgments about what information is being sacrificed (changed to white or black without detail), and how important that information is to the image.
Don't be afraid to experiment, that is how we learn. You can always change your mind, you can certainly discard scans, no law says you have to keep every one. Experiment, say "Oops!" some, and discover the possibilities. It will become second nature very quickly.
I think it all depends on the photo image, and the purpose. There is no one perfect answer. Even if you adjust the scanner image to perfection, whatever that is, you still have to deal with your monitor, and your printer, and your friend's weird opinions too. <grin> You may adjust the scanner to compensate for one of these latter problems.
Remember that different images can often benefit from different treatments, but this overall general technique (Setting the White and Black Points) is normally always well in the ballpark. There certainly are a few times when you must deviate, photos of the polar bear in a snow storm, the black cat in a coal mine, etc. But even if you're pretty sloppy with it, almost all of the time you'll likely still get better results than if you didn't do it (which is sloppiest of all).