|
Adobe Photoshop Elements 09/2001 Review for PCUGR R. Frazier Let's jump to the punch line first. I think Adobe Photoshop Elements (a.k.a., PSE) is a great image-editing program. Lots of powerful features, good help and tutorials and it runs well, even on moderate equipment. If you don't own Adobe's Photoshop and can't justify its high price, then seriously consider this program. It is a significant, useful subset. Ditto, if your only image editing software is something that came bundled with some hardware like a scanner (e.g., Adobe's Photo Deluxe) and you are looking for more capable software. It lists for $99 and I've frequently seen it advertised for about $50 (via rebates, etc.). The version reviewed was 1.0.1 (but I must say, this was not so easy to determine). System Requirements:
(Note these requirements are what are on the box and on the flyers available at PCUGR meetings. Various Adobe web documents were different from this, generally indicating a newer CPU or OS as minimum. What's on the box is correct.) Actual test system was a 166 Mhz, Pentium (no, not even a MMX), 64 MB RAM, Win98 (1st ed). Matrox Millennium graphics adapter with 2 MB, 16 bit color (a.k.a., "High Color") and 1024x768 resolution (96 dpi). Box contents:
Comparison to Photoshop:
Although a subset of Photoshop, there are still a ton of features in this program. Some make the process of image editing efficient while others address the results. For example, on the efficiency side, there is the Navigator Palette. With it you get an image view/zoom assistant. You can adjust the zoom via buttons and sliders or select where the view is by pointing or dragging a view rectangle on a thumbnail. As a side benefit, it is also a way to interactively play around with cropping ideas. Once I discovered this existed, I found myself using it all the time to reposition or resize my view while I was using some other tool. As an example of a "results side" feature, there is the "clone stamp". This tool lets you select one spot on the image to clone (copy) onto another by just painting with the cursor (or brush as it is referred to in PSE). The brush-cursor can be round with hard or soft edges or act to spatter the area copied. And you have total control on the size, hardness and roundness. And so what use is this? Removal of unwanted stuff from an image. Anything from wires cluttering up the beautiful landscape to Aunt Minnie in the family portrait. Of course, it takes a bit of practice to use this effectively, but the effect can be dramatic. And having a multiple undo feature doesn't hurt either. Fire away and, if you goof or go too far, just undo it!
When starting up the program, the first thing that greets you is the "Quick Start" screen. This provides immediate access to creating a blank image,ng an existing image, pasting from the clipboard, acquiring from a TWAIN device (e.g., scan an image or import from a compatible digital camera), going to the tutorials orng help. This automatic display can be turned off or on as desired. The basic work area is composed of the usual menus plus:
The shortcuts and options toolbars can be docked or moved around and closed up as desired. The toolbox can be moved or closed up, but apparently not docked to an edge or another toolbar. Palettes are a feature of Adobe applications. Essentially these are specialized dialog boxes that display in re-sizeable windows. In use, a palette can bed on the work area and, therefore, always remain accessible while editing. Or it can be stored away in something called the "Palette Well" - basically like a set of file folder tabs, similar to how worksheets are "tabbed" in a spreadsheet application. In the well, a tab can be clicked toclose a palette as needed without having it take up space on the desktop. The available palettes are:
To try out a recipe, I picked a photo that had been taken in the shade with a bright background under a colored awning. The person in the picture was too dark and had a greenish cast due to the awning. First I picked the Color Correction slate and then "Remove a Color Cast". I followed the steps, selecting a T-shirt as an object that should be white. The colors definitely changed, but the image was still dark so it wasn't clear if things had improved much. Next, I tried the "Quick Fix", but I didn't care for the increased contrast, so I used the undo feature to reset things. Since the person in the shade was dark, the "Fill Flash" looked appropriate. I could lighten up things, but now the person's face looked rather pasty. Next I tried the recipe for "Tonal Range". This displays a histogram showing the distribution of color (or all colors) across the lightest to darkest pixels. With a bunch of fiddling with various sliders on various colors I got a somewhat better contrast and color for the person although the image in the background clearly had it's own issues. None-the-less it seemed better over all. Outside of the recipe, I used the "magnetic lasso" to select around the person (this is a really neat selection tool, facilitating custom tracing around irregular objects) and then apply a further color/brightness correction to the background. This was the best image overall for a small picture displayed on a monitor (like might be done on a web page) but all the image manipulation had taken it's toll. No doubt, it's better to have a good image from the start, but if you've got to salvage something, you can. This was a pretty difficult image to fix, but I felt the help afforded by the recipes made a lot of difference in understanding what things could be done and, having picked up some ideas, I redid the editing on my own. This time I did the selection around the person first, then applied the Color Cast and Fill Flash corrections to just the selection. A little judicious use of the Blur tool to remove hard edges from the selection and of the clone tool to fix some areas inadvertently selected made a pretty reasonable image quickly.
Since a lot of image editing is for the web these days, PSE has a number of web oriented features. For example, when saving a file as a JPEG, you not only get the usual small-file to large-file slider, but the size is also displayed along with an estimated download time at selected bit rates (14.4 kb to 56.6 kb). And, when the size/quality slider is moved, the image changes to preview the effect. Thus you can determine just how strongly the compression is going to effect your picture. There is also a "Save to Web" feature that displays the original image side-by-side with the save choice. Here you can select GIF, JPEG at various compressions, PNG-8 and PNG-24 with the same preview-before you save both for image and size. If the GIF is an animation, you can preview the animation as well. You can have lots of fun with added text (or "type", as the PSE documentation calls it). Each set of text "lives" on it's own separate layer, so it can be manipulated individually. Special effects such as drop shadows, glows, warps and more can also be added. Again, recognition of the web is at play here. The only limitation is that the pre-done selections are all you get. You can't manipulate them or make custom ones (e.g., the drop shadows cannot be any other color than black). There is a grid that can be toggled on and off along with a "snap" function to help line up text edges, but no other automatic alignment features. As mentioned before, the layers features can be very useful. For example, making composite images is done using layers. Text is typically added as a separate layer, as well. Layers can be duplicated, rearranged in order, merged and grouped. By using a set of layers as animation frames, the "Save to Web" feature can combine the layers into an animated GIF file. For basic animations this facility works fine, but the controls are limited: a toggle for looping and a globally adjustable frame delay. Missing is any loop count (the animation either always loops or only plays once) and there is no provision for individual frame delays. Perhaps in the next version.
Another nice touch are the Adjust Backlighting and Fill Flash commands. Difficult lighting - bright backgrounds and dark foregrounds - can be easily tamed with just moving a slide bar, although one is advised to go easy with these, especially the Fill Flash, as it can definitely be overdone leading to some rather flat, pasty images. I found it is best to use these in tandem with the normal brightness/contrast controls. Along the same lines was the Color Cast command. In many cases this worked great to re-balance colors that were skewed due to some lighting anomaly (for example the commonly seen blue cast of a photo taken in deep shade). In others the result was to cast the color in some other direction. Overall, however, this feature is a definite plus. There is another command called Equalize whose purpose is to redistribute the color/brightness evenly across all the pixels. I found it to be a much more mixed bag compared to the Color Cast command. When it works it can have a marvelous effect making it much easier to finish up the photo. But when it doesn't, you can end up with a high contrast, posterized-looking mess. Fortunately, since everything is reversible, you can undo it easily so it is still worth a try.
I spent a bit of time exploring the Photomerge feature for making panoramas. Overall, I would characterize it as useful, but fussy. The really great aspect of this was the ability to stitch images horizontally, vertically or in both directions at the same time! Using this feature, I was able to put together a set of 10 images - 5 across the top and 5 across the bottom to create a composite image that would have been impossible otherwise. But almost all the images I combined exhibited some issue with balancing exposures, especially in even colored areas such as the sky. The detailed tutorial recommends using a tripod and holding the exposures constant across frames (e.g., engaging the "exposure lock" feature of some cameras). While this helped, it did not completely eliminate the issue and, as a result, the combined images required fiddling (sometimes a lot) to compensate. Ultimately, I concluded it was much better to adjust the brightness of each image that was to go into the composite first before trying to combine them. A stand-alone panorama program, used for comparison, did a much better job at blending the various frame exposures. But this comparison program did not do as well stitching the frames together and it could only do horizontal compositions. Other items regarding Photomerge:
Bottom line: In some ways, Photomerge seems like a work-in-progress. It has some nice features, but you need to work around some of the others to be effective with it.
On installation, PSE recognized a TWAIN scanner attached to the system and I was able to scan images in without issue. In one case, I digitized a severely color faded photo (it had a strong red cast), corrected the color and then removed the quilt pattern of the photo paper picked up by the scanner. I also tested the automatic straightening feature by purposely making a crooked scan. It worked fine.
In addition to the image rotation functions there are a variety of image geometry transformations available. These can be used, for example, to correct the tilted building syndrome that occurs when a photo is taken with the lens pointed up at the subject. (This is particularly evident with wide-angle lenses.) I found these tools relatively simple to use and could be applied to the whole image or just a selection. There was also a "liquify" filter that applies a distortion to an adjustable, circular region. Some of the effects, such as the swirls, bow-in, bow-out and line shifts work by holding the left mouse button down to progressively apply the change. Others work via a click-drag. Using this tool, you can surrealistically stretch things all out of shape. There is a handy "revert" button if you don't like the result. I tried to correct some barrel distortion (from a camera lens) using the ordinary geometry transforms, but found these tools didn't do what I needed (or at least I couldn't coax them into doing it). I was able, however, to use the liquify tool to make an OK correction, so this special effect can have a more practical application. Finally, there is a 3D transform tool that maps a portion of an image onto a cube or cylinder or sphere (kind of like wallpapering the shape). The resulting object can then be moved or rotated around before being plunked down as a new layer. There are probably some clever things that might be done with this, but I didn't find any so I mark this one special effect only.
Composite images (overlaying two or more images) is greatly facilitated by the "magic" background eraser. This tool has a "hot spot" that identifies the pixels to erase and then selectively erases only those that match within a larger radius. In this manner, one can selectively "cut" a section of an image by careful painting. Because of how it works, the image left behind will typically not have a color halo unlike methods that use object selection first. With the right images, it was pretty amazing to use.
PSE does not come with a standalone image viewer or slideshow program to distribute with an album of images. Instead, there is an "Automation" tool called Photo Web Gallery that automatically creates an HTML set of thumbnails with links to the larger size images. Thus you can use this to create an album that you share on the web or via storage media such as a CD-R. With the tool you can choose from a selection of thumbnail styles (nicely previewed), create a custom text banner for the thumbnail page, choose the thumbnail size, caption the thumbnail with either the filename or embedded file info, choose the size and JPEG quality of the larger images, and customize the HTML colors for the banner, thumbnail background, text, and links.
To facilitate working with multiple images, there is an Automation tool for batch processing files. This tool will do its work on all the files in a folder and, optionally, the files in all subfolders below it. Processing options are 1) converting image formats (e.g., BMP to JPG), 2) resizing all images (resolution and pixel dimensions) and 3) renaming files. The file renaming utilizes a prefix and a suffix. There is a slew of pre-done naming selections including serialization options. You can enter your own text as well. Another automation feature is a contact-sheet maker. Among other things, this applet lets you select the number of images to place on the sheet and the resolution to use. It basically works by reading in the images, resizing them and then merging them into a single layer on a larger background image. Once constructed, you can do further manipulations to this composite image such as add text, change colors, add other images, print it or save it to a file. The final automation feature I examined was the Picture Package. With this applet you can select from a variety of layouts (including some large format layouts - 11X17 - in the event you have a printer capable of printing such sizes) to best fit multiple copies of the same, single picture onto a sheet of paper. The end result, again is a composite image that can be saved or printed. On the plus side, this minimizes waste of expensive photo paper. The negative side is only one image is used. Being able to do multiple images is something lots of people asked about on the web forum, so it appears a likely candidate for a new feature in later program versions. In the meantime a work-around is to create a paper sized "canvas" and then manually pack different images onto it. Using the program is greatly facilitated by the excellent HTML Help, one of the more effective uses I've seen in awhile. I especially liked the tutorials as they were a nice balance of detail and brevity along with good use of HTML hyperlinking. It is well worth browsing around the help to pickup hints and shortcuts. And speaking of hints, another nice feature is the Hints Palette. Like other palettes, this one can be tucked out of the way or displayed as you wish. When visible, it is like a tooltip on steroids. Move the cursor over some tool in the Toolbox or Palette in the Palette Well and a description of what the item is and is used for appears in the hints palette. Need more detailed information? Just click the "More Help" button and you are taken directly to the specific help. Another location for usage hints was in the status bar at the bottom of the PSE's main window. Although this might be overlooked at first, it does often contain a useful one line, quick instruction. (Having said all these nice things, occasionally I found the item selected in one area did not "stick" in the Hints Palette. The workaround was to click on another tool, and then back to the desired item, but it was still a bit annoying while trying to learn about some feature and discover your help selection took you somewhere totally unrelated.) As I said at the beginning of this review, Photoshop Elements is great image editing software. Yes, there are some rough edges, but nothing that can't be worked around. The full Photoshop program has a reputation as difficult to learn and use, but I found PSE to be quite useable from the get-go. No doubt about it, there are aspects that need experience and skill to be used most effectively, but with the tutorials, recipes and printed manual, there is no reason anyone shouldn't be able to get productive right away. You will be pleased with the results. For those who do acquire PSE, there is a little fun built in by the developers. See if you can find the "Minnesota phats", the "Honorary Guest Engineer", Mr. Bullwinkle, Legal Mumbo-Jumbos, the various Goddesses, Orange Julius and Adobe's favorite customer. |