Here's how to use GraphicConverter to simply and quickly create small digital images suitable for identifying azaleas.

GraphicConverter is a shareware image editing program for the Macintosh, available at www.lemkesoft.com (I think Irfanview, a shareware Windows® image editor available at www.irfanview.com, offers similar capabilities, although the menus and commands will be different). This is also true for PhotoShop.

The steps I follow are to:
- take the pictures (images);
- get the images from the camera onto my hard drive;
- open the folder of images with GraphicConverter; and for each image:
    - (maybe) rotate it;
    - (maybe) adjust its white/black levels
    - resize it
    - sharpen it; and
    - rename and save it.

The notation to describe GraphicConverter commands is eg, File > Open to mean "choose the 'Open' command from the 'File' menu". In a few cases I give some choices and explain why one choice is better than the others. When I don't say why, the generic reason is it works well for me, usually because I read about it and tried it, and it worked better than variations I had also tried.

To preserve as much flexibility as possible, set your camera to take as large (high quality) a picture as it can. While you won't need the quality for the web, you may care to print some of your pictures, and you will need it there. To use less space I set my (2 MP Casio QV2900UX) camera to store the images as high-quality JPEG images rather than TIFF or RAW images.

My camera has a Flower Closeup setting, which puts it into macro mode with "high saturation". I use that setting for all flower images, whether near or far, to get the "high saturation" effect. That looks more natural to me, which may be completely unique to me and/or my camera.

My camera uses a CompactFlash card to store its images, and my computer has PCMCIA card slots. To put those two together, I bought an inexpensive CompactFlash to PCMCIA adapter. To download the pictures from the camera:
- take the card from the camera and put it into the PCMCIA adapter
- put the PCMCIA adapter into the slot on the computer
- in a few seconds, the card appears on the desktop as a disk drive named untitled
- open the drive (mine has two folders: Scene and DCIM, with DCIM containing the images)
- find and open the folder with the images
- select them all and copy them to your hard drive (I use a folder per each six months, with a subfolder per each download in those six months -- think up an organizational scheme that fits your long-term retrieval needs, and follow it faithfully)
- drag the images out of DCIM into the trash
- empty the trash (very important)
- close the folders, close the drive, and drag it to the trash to eject it
- the card is then ready for use in the camera
If I trash the images but forget to empty the trash, my camera tells me the card is empty if I try to review the images, and tells me the card is full if I try to take a picture.

Other adapters for USB or serial ports will act similarly. The significant point is removing the card from the camera to download the images, rather than using the camera software.

You can set an OS preference so GraphicConverter is invoked to open any JPEG file upon a double-click, or upon a click and File > Open. If not, open GraphicConverter and choose File > Open... > browse to find the file to open. If GraphicConverter has saved the file, it will always open it again upon a double-click, regardless of preferences.

If you have opened the file before, it may be listed in File > Open Recent.

Graphic Converter will open a folder and display all its files as thumbnails. To do that:
- choose File > Browse Folder...
- browse to and select the name of the folder of interest
- click Choose

That shows a small image of each file in the folder and, the first time only, creates a thumbnail for each file. From then on, only the thumbnails are shown. In either case it shows the number of images in the folder along the bottom margin, and it shows your folder's parent folder along the left margin. To work with one of the thumbnails, click it to show a larger thumbnail to the right, or double-click it to open the full size image (full size is set as a GraphicConverter preference: Edit > Preferences... > General > Display Content - I use User Defined Zoom > 50%).

You can change the size of the active image via a pulldown menu at its bottom left, either by choosing one of its preset size percentages, or by typing in a different percentage. It also has increase/decrease buttons to go up/down the preset percentages, or you can press cmd+ or cmd- keys to go up/down the presets. It usually shows the image at your preference-set size after you perfom some editing operation on the image, even if it was some other size before that operation.

You can open as many images as you wish, and choose the one to work with from the Window menu. I usually open one, edit it, save it, and close it, one at a time.

These editing comments are in time sequence. By the time you get to the bottom of this page, you will have a web-ready image.

Choose Window > Information to open a window with sizes and other information about the image, with a menu to show its size as pixels, inches, etc. It also has tabs to show the EXIF information provided by your camera, which gives the date, shutter/aperture settings, etc.

GraphicConverter has one level of Undo. To undo the last operation, either press the cmd+z keys, or choose Edit > Undo: (it names the operation it will undo after the colon).

Rotate
I almost never rotate a flower image, although I commonly rotate an image with structural elements or other straight lines. To rotate an image:
- choose Picture > Rotate > Other Value With Line...
- find an element in the picture that should be either horizontal or vertical
- press at one end of the element and drag to the other
- when the line it draws looks right, release the mouse button to show a dialog box with the amount of rotation in degrees
    - first time only: click the button for Use Quicktime
- click OK to do the rotation.

You can instead rotate the image by trial and error. To do that:
- choose Picture > Rotate > Other Value... to show the same rotation dialog box as above
    - first time only: click the button for Use Quicktime
- type in a guess on the number of degrees of rotation
- click OK to do the rotation
- if it doesn't look quite right, undo it and try again -- do not rotate an already rotated image a little more or less.
It shows the rotated image in a window sized to show the image corners. Crop the image to cut off all the white space and leave as much image as possible (see below for how to crop).

You can also rotate the image in 90° or 180° increments by choosing that from the Picture > Rotate menu. In those cases there is no white space to deal with.

Levels
Check the image histogram and possibly adjust the white/black endpoints and the midpoint of the image. The histogram is a bar graph of the distribution of the 256 possible levels of lightness in the image (see Using Histograms). For many well-exposed images it will appear somewhat like a centered bell curve, or have values which extend completely from left to right. Overexposed images will appear to be crowded to the right (too many white pixels = "blown" highlights), and underexposed images will appear to be crowded to the left (too many black pixels = lost shadow details). Many cameras let you see the histogram immediately after taking a picture, so you can retake it with a different exposure when the histogram shows an exposure problem.

To see and use the histogram and the level controls in GraphicConverter:
- choose Picture > Levels...
- (usually) click Cancel to do nothing if the histogram extends to both the left and right edges of the axis
- drag the left triangle to the right to darken the image and its highlights, but (usually) no further than the leftmost intersection of the histogram with the axis
- drag the right triangle to the left to lighten the image and its shadows, but (usually) no further than the rightmost intersection of the histogram with the axis
- drag the center triangle to adjust the midpoint level of the image

There are also two Output triangles which change the image contrast. While I do not use them, moving them has the opposite effect of moving the input triangles, and seems to always reduce the contrast.

On my computer, the triangles move slowly if the Full Screen Preview box is checked on. I leave it off and watch the effect of my changes in the small After picture at the top right, and occasionally toggle it on and off to see the overall effect of the changes and original image (while the Reset button does the same thing, it loses your current settings).

When it is done well, adjusting the levels will brighten a dull image without affecting the color.

When you are happy with the changes, click OK to apply them to the image and close the window, or click Cancel to close the window without any change to the image.

Crop
If you intend to crop the image, do that next. Try to almost fill the frame with the flower when taking the picture, as being quicker than cropping it later. To crop an image:
- choose Window > Show Position to show a current size and ratio box at the top right of the image
- choose Window > Show Toolbox to show the tools window to the right of the image
- click the selection tool, which is a dotted box icon at the top right of the tools window
- I point to the top left of the area to be kept in the image window (the pointer will show as a cross), drag down and to the right, and release when the selection encloses what I want and the size box shows a ratio of 1.33
- the selection will appear as a dotted line box of "marching ants" with small drag boxes at each corner and in the middle of each side
- you can adjust the size and shape of the selection box by dragging any of its corner or side drag boxes -- when the pointer is on a drag box, the pointer changes to a double-ended arrow
- you can move the selection box without changing its size or shape by moving the pointer just outside the selection box -- the pointer changes to two crossed double-ended arrows when it is in a good spot, at which point you can drag the selection box
- when you like the selection, press the cmd+y keys or choose Edit > Trim Selection to throw away all of the image outside the selection box

Resize
The next step is to resize the image for the web. The preferred size is 6.7x8.9 inches, which is around 640x480 pixels. That is big enough to clearly show the subject, yet small enough to avoid having to scroll to show all parts of the image even on a small screen. To resize an image:
- choose Picture > Size > Scale to open a size window
- make several settings the first time only -- they stay until you change them:
    - click Keep Proportions on (put a check in it)
    - click Scale Picture on (put a check in it)
    - choose Algorithm > Mitchell
    - choose X-Resolution > Pixel/Inch
    - choose Width > Inch
    - choose Height > Inch
= type 72 into the X-Resolution box
- type 6.666 into the Height box (which it will round to 6.667)
- in a moment, it will show the width in the Width box that corresponds to the 6.667 inch height (it should be 8.889 unless you've cropped the image)
    - when I crop, I try to keep the ratio of width to height at 1.33, which can produce a size of 6.667x8.889 inches -- see above
    - if the width is greater than 8.889 inches, you can crop it later to cut it back
    - if the width is less than 8.889 inches, click Cancel, choose Edit > Undo: Trim which shows the old selection box, adjust the selection (either a tad wider or a tad shorter), and choose Picture > Size > Scale again

Sharpening
Digital images tend to be "soft", with indistinct edges. A slight amount of sharpening improves most images, and it may salvage a slightly out of focus image. Too much sharpening makes the image grainy and unnatural. I use two different approaches, sometimes both on the same image. The recommended (and slow) approach is:
- choose Effect > Unsharp Mask... to open an Unsharp Mask dialog box with lots of controls
- make several settings the first time only -- they stay until you change them:
    - Zoom: your choice; I have the slider about 1/3 from the left
    - Radius: type 1 into the pixels field
    - Amount: type 60 into the % field
    - Threshold: type 0 into the levels field
    - Full Screen Preview: click it on (checkmark in the box)
- the Picture image contains a very small box to drag around to see its portion of the picture in the Before and After windows
- you can see the effect of the sharpening on the image itself, a moment after making the settings -- with the settings above, the effect should be subtle
- click OK to do the sharpening
- to sharpen it some more, repeat the above process (as a learning experience, do it over and over until you see the image becoming very grainy -- and then just close the image window without saving it, and open it again and start over for real)
- after much practice, now I choose Effect > Unsharp Mask and press the Enter key without even looking at the dialog box -- if it's enough, I'm done, if not I repeat it, and if it's too much I undo the last one.

A much quicker approach is to:
- choose Effect > Sharpen Edges... to open the Sharpen dialog box
- make several settings the first time only -- they stay until you change them:
    - Zoom: your choice; I have the slider about 1/3 from the left
    - click the Normal radio button on
    - Full Screen Preview: click it on (checkmark in the box)
    - adjust the slider as desired --I've always used the recommended 90%
- the rest of the comments above for the Unsharp Mask apply here as well.

On some occasions I have used both Sharpen Edges and Unsharp Mask, in that order, on the same image.

Gamma Adjustment
Gamma is an elusive concept (at least to me) involved with how images appear on different screens. The standard gamma for PC screens is 2.2, and is 1.8 for Macintosh. I change the gamma only if an image appears unusually light or dark. To change the gamma:
- choose Effect > Gamma Correction... to open its dialog box
- make several settings the first time only -- they stay until you change them:
    - Zoom: your choice; I have the slider about 1/3 from the left
    - type 1.8 into the Standard gamma values... field
    - Full Screen Preview: click it on (checkmark in the box)
- drag the slider as desired to make the lightness of the image look better (I used to set the gamma to 2.0 as a compromise, and now let it stay at 1.8 unless the image is unusually light or dark)
- click OK to change the gamma

Color, Contrast, Saturation
I have almost no experience with changing either the color, contrast or saturation of an image, and have had very little success with it. Thus, I can only comment that:
- GraphicConverter has powerful commands for these purposes; and
- different monitor settings and printer settings make it hard to predict how it will look to others, whether you change them or not.

Unskew
Buildings, signs, and similar objects with parallel lines may appear to have those lines converge. You can reduce or eliminate this convergence with the Unskew command. To use it:
- note the size of your picture, in pixels
- choose Effect > Unskew...
- that opens a Note dialog box that tells you to click the top left, top right, bottom right and bottom left of the area to unskew -- do that very thoughtfully:
    - your clicks outline the area that will end up in the unskewed image, much like dragging out a cropping rectangle, so click close to the edges of the image to end up with as much of the image as possible
    - click such that the lines it draws between your clicks are parallel to the sloping lines in the image (when it is finished, those lines will form the outside rectangular bounds of the image)
- after your last click, it opens an Unskew dialog box
- make several settings the first time only -- they stay until you change them:
    - click the Manually radio button
    - type the original dimensions into the Original Width and Original Height fields
- click OK to do the unskewing.

Retouching
On occasion, a picture is marred by a fault such as a tiny hole in a flower petal. In many cases, that fault is easy to fix with the rubber stamp tool. In effect, you copy a nearby part of the picture and paste it over the fault. To use it:
- click the rubber stamp tool at the bottom left of the tool window -- if the tool window isn't visible, choose Window > Show Toolbox to show it
- click the tiny red dot at the bottom left of the rubber stamp icon to open its preference window
- type the size of the stamp into the Size field -- 10 is a good size, but it depends somewhat on the size of the fault
- set Always use unstamped picture as master on (have a check in the box) and click OK
- to load the stamp, point to a pattern and color suitable for covering the fault (usually found very close to the fault), then hold down the option key and click
- then click directly on and around the fault, as needed to cover it up
- when it doesn't look quite right and you've clicked a number of times, close/reopen the image and start all over, since you can only undo the last click

Organization
Because JPEG is a lossy compression scheme, a JPEG image loses a little quality every time you save it. To minimize that, consider the image from your camera as being your original -- open it only to look at it, and never save it back into itself. Instead, after you open and edit an image, save it with a different file name.

That means you need a filing scheme for naming and storing both your original images and your edited images. Some of the needs that scheme should meet include the ability to:
- identify the image subject
- find the original file, given the edited file name
- identify what use you have made of the edited image
- identify what edits were done to the image
- be quick and easy to update

What I did (and was not very happy with) was to set up two spreadsheets. The first is a list of the original images. Column 1 is a short identifier of the folder where I've placed an edited image, or empty if I haven't edited and saved it. Column 2 is the file name from the camera. Column 3 is the name of the subject. My camera names each picture in the form "100_mmddxxxxx: where "mm" is the month the picture was taken; "dd" is the day of the month the picture was taken; and "xxxxx" is the sequence number for that date, which it increments for each shot during the day.

When I download a card full of images, I first edit the folder name by changing that "100_" to be the last two digits of the current year, and then drag that folder into my "camera originals" folder, where it appears sorted by date. Then I use GraphicConverter Browse Folder... (described above) to show a window of thumbnail images for the selected folder.

With the GraphicConverter window of thumbnails showing and my spreadsheet alongside to the right, I add a new row for that date in the spreadsheet, and add subentry sequence number rows for each of the images. For example,

	04082500001
	2
	3
	4
	...
	37

An Excel shortcut is to add a couple of subentries (the 2 and 3 in the example), select those cells, point to the bottom right of the last cell of the selection (the pointer changes to a +), and drag down to create more cells with increasing numbers.

Then I use my notes, memory, looking at the images, etc. to fill in the subject names. After editing and saving an image, I add its disposition codes (ie, what folders I saved it in). I use an "x" to show I've deleted an image for being out-of-focus or otherwise not worth saving. This gives me, for example:

misc	04082500001	back deck
bark	2	a. griseum bark
	3	x
lvs	4	r. Don's Variegated leaves
	...
	37	bird feeder

My second spreadsheet was a list of the images I had edited and put into folders for putting onto the web. The website I'm using, pbase.com, lets you specify the image sequence by numbering each image, and sorting the images in ascending order by those numbers. To keep track of where an image is in alphabetic sequence by subject name, this spreadsheet has a column for sequence number, a column for subject name, and a column for file name. For example:

deciduous azaleas
species
100	alabamense x austrinum	0304190004
110	canescens	0304160001
120	canescens, Ruth Bryan	0305030048
130	canescens	0304160002
140	canescens - staminoid	0204060018
200	cumberlandense	0305300021
204	cumberlandense prostrate	0305250009

When I was about to add some images to a web page, I scrolled to that portion of the spreadsheet, eg, "deciduous azaleas", manually inserted the image names and file names into new rows (option-clicking the row number inserts a blank row above the row you click), sorted that portion of the spreadsheet alphabetically, and filled in appropriate sequence numbers. Then, in pbase, I uploaded the images and gave them those sequence numbers. I arranged the screen layout so the pbase page filled most of the screen, with the very narrow spreadsheet to the right, such that they were both visible at the same time to facilitate flipping back and forth between them.

That was an overwhelming amount of effort to keep updated, and required constant reference to see what image was what cultivar. I now simply rename the file by prefixing the subject name to it, and quit using the second spreadsheet.

Save
Never use File > Save -- that would overwrite your original.

Instead, always use File > Save As... or File > Save a Copy As... (after the save, the former shows the copy, while the latter shows the original). That opens a file save dialog box to let you browse to the desired folder and to name the file.

File Naming
My habit was to prepend the first two digits of the year to the file name, and to add an "s" before the ".jpg" extension. That tied it back to my original file name, but it didn't add much information, and it forced me to constantly refer back to the spreadsheets to see what it was.

I now rename each image file to encode more of the organizational elements suggested above. For example, "Ben Morrison.GD.b.480.ea.jpg", with "GD" for Glenn Dale, "b" for a closeup of the bloom (code from eg, landscape, plant, spray, bloom, foliage), "480" for a nominal 640x480 pixel image size, and "ea" for use in the pbase.com evergreen azalea gallery. Since you will run out of file-name room, choose those elements most important to you. The important point is to design a naming style that meets your needs, and stick with it.

The file save dialog box has other controls to be set as (first time only):
- Format - choose JPEG... if not already set
- Options - leave empty (you will encounter it later)
- Split - leave empty
- Create LOWSRC - no check
- Set Format from Extension - no check
- Save web ready - check
- Preferences:
    - General Settings - no checks
    - Custom Icon - no checks
    - IPTC > Text Conversion > None
    - Settings - select Use Graphic Converter, no other checks

When you click Save, a dialog box of options for JPEG files appears, with controls to be set as (first time only):
- Quality - a slider (or type in a number) to control the degree of JPEG compression. A higher percentage = higher quality = bigger file size. I've been using 30% to 40%, and sometimes change it from image to image. See JPEG compression for a discussion and examples of the effects of JPEG compression.
- Library - choose JPEG 6.0, check Progressive
- Calculate File Size - check it on, it will show the file size that corresponds to all the other settings (play with the other settings to see how they affect file size)
- Show Preview - check it on to see the result of the Quality setting
- Show this dialog before saving JPEG files - check it on
- Embed (if available) > EXIF - check it on (it will only be active if the file has EXIF info) - the EXIF info takes very little space and pbase.com uses it to fill in the shot date and camera info
- make the other controls unchecked

The settings stay until you change them. When you click OK the save uses those settings.

You've done it! And the first image you save is by far the hardest. With experience you will find it can all be done in a few minutes per image.

Questions? Comments? Contact Bob Stelloh.

azaleas (parent page)