|Because of the apparent need for image browsers, there is a strongcommercial and shareware market; Cumulus Network, Corel Mosaic,CompuPic-32, and ALLVUE were analyzed due to their close relation withour experiment. While some of these packages include image editors, ourinterest is solely in the image browsers and thus that will be the topicof our study.|
The tools discussed here all conform to a set of criteria based on orexperiment; therefore, this is what each was evaluated on. The criteriaused to evaluate the browsers include:
The fastest browser was CompuPic-32 due to its quick screen refreshesand excellent load time -- Corel Mosaic was not far behind, however. Infact, all of the browsers provided similarly adequate speed except forALLVUE. ALLVUE was considerably slower than the rest leaving the userwith a sense of frustration. And while sometimes a decrease in speed ismet with an increase in abilities, that is not true in this case.
There are many images that are too big to display on any given user'shardware; and, there are two methods of circumventing this problem:scale the image or crop the image; all of the browsers reviewed usedscaling. While cropping maintains picture quality, part of the viewablearea is actually not displayed. But, while scaling reduces picturequality, it allows large images to be displayed on hardware that cannotsupport it. Therefore, because it is unacceptable to arbitrarily discarddata, scaling is the best method.
The ability to display multiple images is perhaps the most importantdetail of an image browser. And, along with its configuration, thisability defines whether or not a browser is of good or poor quality. Allof the browsers studied here support multiple images; the followingcriteria were studied for each.
Corel Mosaic is the image browser for the popular graphics package CorelDraw. It provides the user with a similar vertical scrollbar interface,but does not provide the aforementioned window-snapping option. Thismeans that the number of rows and columns that are used to display thethumbnails is based on the window size. But, if the user resizes thewindow to a size that is not equal to the number of rows or columnstimes the thumbnail size, the browser window has empty space - thiscauses it to waste desktop space. However, while Mosaic does notprovide window snapping, it does allow the user to modify both thumbnailsize and thumbnail spacing. It allows the user to specify almost anywidth and height at which to space the thumbnails, and provides the userwith a choice to two thumbnail sizes, 100 pixels by 100 pixels and 200pixels by 200 pixels. Mosaic provides the user with a search enginesimilar to that of Cumulus Network, but because Corel Draw is a graphicspackage, it includes a database of thousands of images that already havedescriptions associated with them.
Of the systems reviewed, CompuPic-32 is the most configurable. Itimplements two different browser controls and has an option similar to aslideshow; so, while it may have an awkward file searching interface,the browsing interface is quite excellent. The first browserimplementation provided is a scrollbar. CompuPic-32 allows the user tochoose between a vertical and horizontal scrollbar. And, unlike all ofthe other scrollbar implementations reviewed, CompuPic-32 uses a smoothscrolling plane allowing for a more pleasant visual display. Similar toMosaic, CompuPic-32 allows the user to define the viewing matrix byresizing the window, and does not provide the window-snapping feature. The second browser provided is a button-based browser. And, thisbrowser, as well as the scrollbar, can accept any thumbnail size andspacing. While ComputPic-32 does not provide a search engine, it doesprovide an interface called "maxi-show". Essentially, maxi-show is aslideshow; but, it goes beyond the functionality of a normal slideshow. It provides a multi-picture slideshow that displays the selectedpictures in a user-defined matrix. So, although it does not allow foruser activity during its execution, the maxi-show does provide the userwith a configurable timed browser implementation.
The final, and perhaps most disappointing, browser studied is ALLVUE. It provides the user with the same vertical scrollbar that was seen inMosaic and CompuPic-32, but at half the speed. It provides the userwith a 3 by 5 viewing matrix and does not allow the user to changethis. One feature it does offer is thumbnail resizing; however, theuser cannot change the thumbnail spacing, and no search engine ispresent. Because of its slower speed and lack of configuration options,ALLVUE is simply a poor browser.