Lesson 3: The basic workflow in Digital Photography

These are the basic steps when editing an image taken with a photocamera -- or to be precise, with any device capable of producing digital image files. Let us start exploring the GIMP's tools.

Rectangular and mirroring transformations

Rotation

The simple use of the Rotate Tool on one layer. The commands exposed in the previous video act on the whole image, that is to say on all layers and the canvas altogether. This tool acts on one specific layer or a group of layers. Now we will see the one layer case, and in a later lesson about layers we will see how transform tools affect layer groups.

Cropping

Cropping simply is to keep a part of your image dismissing the rest.

Fixing the Red-Eye

Sometimes in photographs it is practically necessary to fix the red-eye effect. This is how they often call it, but we might call it as red-eye defect. Children have big pupils, so sometimes the defect seems really bad, making them look like devils. I will show here a bad case, using only a part of a family portrait photo.

You may find various tricks for red-eye reduction on the web. Even camera manufacturers have taken this into account and modern photocameras feature automatic functions for red eye reduction during shooting. If best quality is your intention, just deactivate this function in the camera once and for all. The reason is that all colors tend to be affected more or less and not only the red inside the eyes. As it has been explained in theory, red, being a color component, takes its share actually in every pixel in an image, so reducing red in the eyes may mean reducing red in relevant areas, thus affecting color in unwanted areas. Even if your high-end modern camera theoretically detects faces and applies some extra special function, however this means sophistication through extra tasks left upon the cameras' software. For the best result, the preferable logic is that you are the master who has the control manually at post-processing with your software. GIMP comes with a special tool to help you with this. No more devils around.

Save it!

I am probably satisfied with the image now. I have cropped a little, i have corrected it to be perfectly horizontal, i have corrected a red eye, so it is the right time to save, to make sure that i will not lose my work so far. In the future, at anytime, i will be able to open the file again and continue editing it, or to use it to produce a working image file. I will save my work as a GIMP project, that is to say a XCF file. So i can have two files that represent:

1. the original file that the camera created, and
2. my work on that file.

"File" > "Save As...":

Select XCF as file type, browse the file system, type a filename, click "Save", done. I have it. When working, every time i am satisfied with an action, it is a good idea to save: File > Save (or by hitting CTRL+S).

Taking it logically: what the intention is

So the basic work may seem to be done at this point. I shot it, i made the basic post processing, now my image seems acceptable. Now or later i will want to use my image. To sum up the steps:

1. open the XCF in GIMP,
2. change the image size if needed,
3. do the sharpening,
4. export to the appropriate file format.

There are two cases for using the image:

• case 1: to use digitally: email, webpage, internet, or just seen on digital devices,
• case 2: to print.

In case 1, as we saw in theory, resolution does not matter, since in digital devices it is an easy job to control the projection size when showing an image on screen and the only thing that counts is the image size. By principle an image is shown on a "pixel-per-pixel" basis on a screen, but we easily control the projection size through the software (like we have seen in GIMP for example). If the projection size is too big for a given number of pixels per unit of measure, the image gets "pixelated" losing its consistency. The same natural principle of optical detail holds true when printing. So in terms of optical quality and as for the number of pixels per length unit, there are limits to the physical size we can print. So here the relationship between the image size and the resolution comes into play.

Case 2: Getting ready for a print

Case 1: Getting ready for a usable image file

Let me say that i want to use the image on the web, or to email it to someone to see. Normal monitor sizes do not exceed 1900px wide or 1200px high, hence a width of 1200 pixels would seem right to enjoy viewing a photograph. A bigger size would be unnecessary, and still a bad way to fill up mailboxes and servers. So i just have to change the dimensions of my image: i have to scale it down.

Sharpening

Sharpening is an issue that must be understood. The relative technical terms are rather confusing even in english (sharpening, unsharp mask etc.), but we are interested in practical results. Sharpening is a necessary step in the editing of a file which came out of any digital device through a process of digitizing of optical data - that is to say, conversion of an optical image to digital data (digitization, sampling).

Hence this holds true for digital photography as well, including image data that came out from a photocamera or a scanner. On the other side instead, sharpening is not necessary on an image that we create from scratch - for example a drawing or a painting or digital art with a drawing tablet and an application like the GIMP with its painting tools. When an image includes data that fall in the category of digitized optical data, on those data normally sharpening must be applied. In other words, if i paint or draw in GIMP and i want to use in my image any image data from my photocamera for example, on this specific data only typically i must execute some sharpening.

What is "sharpening" in reality? It is generally thought to be a fix for an unfocused photo, that is to say to help a photo that is out of focus to be in focus, to an extend that this may seem possible. This is partly right, but it is more than that. It is inherent to the digitizing procedure itself as a frequency step problem. Please refer to the GIMP's help file on the "Unsharp Mask" topic for an explanation. Theoretically every digitized image needs a sharpness correction as a countermeasure.

Sharpening must be the last action on the image data just before exporting to a usable image file. That is to say you must have finished all the work with the image already and you must have scaled the image to the exact size of the exported file (width/height in pixels). This is so because sharpening acts on the pixel level. In other words, the same parameters of the sharpening function have a different result on the image at different image sizes.

Modern digitizing devices (like photocameras and scanners) have sharpening functions. If you want quality, just deactivate this. If you do not, some sharpening will be done by the device's software, after the sensor yields the natural scene, using some algorithm. You do not know this algorithm - actually nobody knows but the manufacturer. So here you have an issue about something being out of control. You should control the process instead, at post-processing. With GIMP you have a reason to trust an open algorithm elaborated and being used by people who know exactly what they want. It is well stated in the GIMP's Help: Some imaging devices like digital cameras or scanners offer to sharpen the created images for you. We strongly recommend you disable the sharpening in this devices and use the GIMP filters instead. This way you regain the full control over the sharpening of your images. So leaving theory, just wanting to do the job, let us see how a decent sharpening is done.

Finally, we come to the last step: saving to a usable digital image file.

Saving for quality: creating a TIF image file

This specific file format is ideal for applications where no loss of quality is acceptable. As said in theory, TIF is lossless, so hey, there is one more nice thing to note: you can open the TIF later in GIMP and export the image to any other file type, if needed.

Saving for file size: creating a JPG image file

Saving for transparency: creating a PNG image file

If transparency is needed: for importing image data in other programs, multimedia and video applications, desktop publishing applications, etc. etc.

Other file types?

Of course there are many more of them. These file types appear to be the most basic ones, widely supported on all platforms, rather enough to get your job done, whatever this may be.

Final note:

At closing your image, GIMP will ask you if you want to save. You might not want to do so! If you have properly saved at the right moment after finishing the basic work - that is to say before scaling and sharpening, if needed - that was enough. If you do save again, this will be a "scaled-down and sharpened" version of your image! You might want this only in case you have the specific intention to do so -- and of course you still have the option of saving an extra XCF.