A Detailed Description of the Use of Portfolio and Other Software
Paula Hardin
Curator Visual Resources Library
University of North Carolina
10/05/2001
Text from DIGI-mar.com
http://www.DIG-Mar.com/Commentaries/port_guest.html
I use the client version 5 of Extensis Portfolio on multiple machines. I have not had the opportunity to use the server version, though I would love to have it and may do something about it this fiscal year. Though I use Portfolio constantly and on a daily basis, I have not exploited all of the features it offers, so this text is not to be considered comprehensive. I will mix and match my experiences using the product from version 3 on, and from my stint at LSU as well as in my current position here at UNC. Hopefully, this will not cause any confusion. I am using Filemaker Pro (FMP) server version as my text database. Filenaming is a particularly important issue that I have written and spoken about before and won't get into in here except to say I highly recommend using a sequential number of 6 digits. Do not use periods (some software automatically reads that as the delimiter for the extension and this can limit your options) or other special characters or spaces and certainly not names like Picasso1 or Monet's Water Lilies. Make sure you have one master source for generating the numbers and do not ever reuse a number. If there is some confusion, skip the problem number(s) and move on to a clean one. Don't get hung up on the numbers having meaning, e.g. sequentially or accession year or whatever. Keep these other data elements in fields in the database. The filename should be a finding aid only.
Daily Use in Acquisitions - Mundane but Helpful Methods
We scan every new item from the books on flatbeds, including a tabloid sized scanner (11x17) for large books. We begin with data entry in Filemaker, obtaining a 6 digit number that is automatically generated by the serial number feature of FMP. This number gets written on a form in a row that shows the page and figure number. This number is written on the Shooting Record form we use during copy photography so that we are able to determine if something has been shot, is in the camera, or on a roll at the lab. This number also becomes the filename during scanning, which is the next stage. By the way, each stage we complete is marked with the date on the acquisitions form.
When using the flatbeds to scan, we can assign the proper number for the filename then and there. (Scanning from slides to be mentioned later.) Files are stored on a University server under the path name of Image Master\New Acquisitions\OrderNo\Source. Orders consist of multiple books/originals and each order and each source within the order is assigned a unique object identifying number in related databases in FMP. When an order is completely scanned, we create a new catalog in Portfolio under the OrderNo folder. There are several ways to get the items into the new catalog. We usually use the click and drag method, so we have Explorer (or the Mac equivalent since it is totally cross platform) open, select all the source folders (you can select an option to automatically include subfolders too), and drag the selection into the new empty Portfolio catalog. I won't go into the details of the options in this document, but as a general point of information there are choices to add items that ignores or overwrites already existing images. There are options to include portions of folder names and the path as keywords. This can be exploited to do a bit a keyword assignments if the folder structures were named "painting\Italian\Raphael" or similar, such as in retrospective conversion activities where there is not an acquisition process to form the structure of subfolders.
Once the items are in a Portfolio catalog we sort them by filename and do a printout. The printout comes out with 30 thumbnails per page. Even as black and white printout they are very useful and are extraordinarily detailed. Color thumbnails are extremely compelling and informative. Now when we get the slide film back, we can easily match the slides with the proper identifying number, easily recognize reshoots or other items out of sequence so to speak, and can make sure that the slides are not reversed or mislabeled in orientation during mounting. This also provides a crosscheck that all slides were shot for the order and vice versa, that all scans were made. Every once in awhile, we discover that either the wrong picture on a page was shot or the wrong item was scanned. If you have 257 images of Buddha and you know you are supposed to have 258, the printouts can be helpful to figure out what is missing or to clarify numbering problems. I lay the printout on the light table and proceed to lay the slides on top of the thumbnails until I find what is missing.
When scanning slides on my batch feeding fabulous Minolta Dimage scanners, I can either use the scanning program directly or use Photoshop's import. The difference is significant in terms of image management. When using the scanning program directly, the scans are written to disk. This is great if you set it off just before you leave at night because it chugs along doing the 50 or 60 items (depending on if glass mounts or not) and they are safely written to the drive and ready for you the next day. The program automatically appends distinguishing numbers, i.e. 00, 01, 02 etc. I usually give the batch the day's date so the resulting image is 10-5-00, 10-5-01 etc. If you scan by importing through Photoshop, the images are saved to RAM and given no number, but then there is the danger of total loss of scans if the power goes off or the machine hangs up. I do use Photoshop during the day usually, when I plan to finished the images immediately. I use the scanning program at night because otherwise something could happen, such as I might forget the machine is already on and reflexively punch it "on" when I get in the next morning, thus losing it all. (I hate when that happens!)
Now that I have the scans, I have multiple choices in renaming them. I can do so when I Photoshop them, leaving me to clean up the original named files. I can Photoshop them and rename them after using Explorer, but without being able to see the image I get a little nervous that I am changing the right one. Of course with Windows 2000 it does have the little images, but so far I have found them too tiny to distinguish a Buddha with a finger pointing up from one with a finger pointing down. Anyway, it just takes a few minutes to drag all the items into a Portfolio catalog and get a printout of the thumbnails in order or just reference it online. You can use the Portfolio to do an Item/Rename Original as you are looking at the image, so that is nice. Right now you cannot change the extension, and sometimes the extension gets changed to uppercase which is as aggravating as all get out, but I have put in a request for this "bug" to be changed and it may indeed be fixed in the next release (v. 6).
I also use a nice cheap (good value) program called Namewiz to do some renaming and filename manipulation. For example, kicking the names back to all lowercase, or just the extension to lowercase, to add leading zeros (It is essential to use leading zeros in filenames!), and to insert or delete characters in filenames. This program would enable you to take an entire folder of badly names files (making a Portfolio catalog to capture the image and the original filenames for reference) and convert them to a strictly numeric sequence. I'd do a printout of the thumbnails and annotate it with the new range of numbers so that you can go to your drawers and use a Sharpie to put the 6 digit number on the slides. I always try to fit it on the front of the slide so that when they are photocopied for circulation, that data is captured for my circulation records. This way I can easily see that the slides have been scanned -- we actually mark an "S" on the front side too though. We also mark a "D" for data entered since not all the slides are in the database. This way we can tell for sure what has been done without having to wrap our brains about "Gee, it is a number higher than 256000 so it must be in the database" or similar actions. I expect it will be helpful in the future too when people will be looking at the slides and will want to know if they are available online.
The Portfolio printouts of new acquisitions make for a nice accompanying record to the orders to show what was added to the collection, as well as a reference for the Professors ordering the material. I have used Portfolio to generate HTML pages of images for student review as well as for special projects and presentations. It could be used to generate pages of new acquisitions so professors can see what their colleagues have ordered as well. At UNC, since we have developed a nice digital library interface, I plan to use it for this purpose, but if it doesn't quite meet the need or is too cumbersome for a quick display, I will use Portfolio instead.
Other Nifty Uses for File Management with Portfolio
After struggling with various file management schemes and decisions, I still have not found the perfect system, but keep trying. One of the things I have used Portfolio for utilizes the search features for various features. I love the "contains" and especially the "does not contain" options in particular. One way I used this was to find thumbnail images. At LSU my preferred filenaming scheme appended the image names with a "t" for thumbnail and other letters for other sizes. Then I could could do a Find for filenames containing a "t" and do various things with the set. (Others use the same filenames but put the images in different folders. This makes it real easy to accidently override thousands of things with the wrong size image in my opinion.) For example, if I had 200 items and 100 were supposed to be the full size jpg and there were supposed to be 100 thumbnails, the search should result in 100 items. If not, some thumbnails are missing or there are naming problems or something. A sort by filename allows me to quickly peruse a couple of thousand images easily to try to spot "non-paired" items. (You can resize the window so you only see pairs of pairs and then cruise for a break in the pattern.) Often I will want a count of the number of unique images I have so I would want to exclude thumbnails from the set or maybe I just want TIFs and not JPGs. I can do a search for "ends with" to identify them. Once I wanted to get rid of all the thumbnails in a folder and because they were interfiled with larger JPGs I couldn't think of any batch sort of way to do it. I didn't want to have Control Click every other item using Explorer. So I realized I could select the set using Portfolio and then use the Delete Original (use with care!) to clean up the directory. I could go on and on about the ways to use Portfolio to do file management, but I will leave this brief description as it stands for now.
Since our operation is a multiple machine and multiple users creating a moving target of images, I may use Portfolio differently than it was intended to be used by doing all this ad hoc imagebase creation. I like to build new ones as I go based on specific purposes because it is fast and easy and then I know for absolute certainty that what I am looking at is reality and not some imagined past view. At LSU I did do a project, Images from Nature, that featured digital photography images. These were kept in a Portfolio catalog of several thousand images that did not have frequent changes in content or image state. They basically were put in the imagebase and simply had new ones added to the collection. We did not maintain the data for this in a Filemaker database, but did use the Portfolio custom fields for capturing about 20 pieces of information about the images. I experienced no difficulty with sluggishness or other problems at this level. I have routinely built imagebases of several thousand, up to about 5,000 I think. More commonly I have been using it to gather smaller collections together based on courses or acquisition orders or even functional categories like circulated slides or painting and so forth. I think, since the question was raised, that I will try to do some testing with the 16,000+ images and other files (e.g. sound, pdf etc.) I have online and I will let you know how I find the performance.
Spiffy Use of Portfolio with Filemaker Pro
I maintain my data in Filemaker, not in Portfolio. Filemaker has so many features that make for fast and simple data entry that it simply is the best tool for that job. It is so fast and easy to add fields to a Portfolio catalog (and there is not a limit on the number of custom fields that some other similar products have) that I can do that anytime the need arises, then I just import the data from the FMP database. Well, not directly from the database. I export the data I need to a text file and then import that to Portfolio using the filename as the match. This means that in my FMP database I must have a field for Filename. In FMP, since I use the accession number (or object identifying number as I prefer to think of it) as the filename, I simply have to do a "Replace" function to drop that information into the records I need if the name isn't already there. I use what I call a jointext function formula to specify that the field contains AccessionNo & ".tif" to create the match. Oh, and be sure that you do not use hard returns in your Filemaker fields because those are treated as end of record delimiters in a text file and that will goof up the data for the import.
Another of my favorite uses of Portfolio is to check that I have database records for each image. After creating a catalog of the images I want to check, I do a Select All function in Portfolio and then Export. I can choose to export any fields but for this purpose I only need to export the filename. This gives me a list of all the images of a particular set to check. Then I create a fast temporary database (call it Checking or whatever) in FMP (or reuse an existing one used for this purpose previously if I can remember where I had it). I just need a field for Filename, AccessionNo, and a Temp field is often useful. I import the text file list of names that I created from the Portfolio catalog into the Filename field. I make a relationship to my main database (named Compilation) matching on the Filename fields. Then all I have to do in the Checking database is a Replace formula in the AccessionNo field using the relationship to pull over any accession numbers that match from Compilation. A Find for blanks (= sign in the field) in the AccessionNo field after this process shows me which images are missing database records. I can reverse the process and check the database for records that are missing images.
I really like the ease with which one can assign keywords to images in Portfolio too. You can develop a limited access list to function akin to an authority list for term assignments. Make sure you pay attention to the "Allow keywords to be parsed" box to see if it is checked or not. I prefer not to parse keyword phrases since there are times when the meaning changes, i.e. a "Classical Greek" keyword phrase would assign the phrase plus the individual words "Classical" and "Greek" to an image and that may not be what you want. As I mentioned elsewhere, you can pull in folder names as keywords too. Once you have your list of keywords you simply click (shift-click for sequential groups, control-click for non-contiguous selections) on the images and then click on the "assign" button with the keyword palette showing. This is great when they have no particular field data in common in the text database. Then you can Find these as a set, create galleries within the catalog for slide shows, printing, or export the data to a text file and add it to a subject database part of your Filemaker databases.
Another feature of the slide show that I want to mention (since it was a feature I asked for and they implemented it!) is the ability to include a field of data beneath the image. This is helpful for those pesky foreign terms and hard to spell items for classroom use. Mostly the teachers that used Portfolio to teach with did not have any trouble resizing the images to fit more things. The pleasure of being able to more than the limits imposed by a slide carousel -- and the beauty of never having to pull another slide (which we then did not have to file I might add) was more than enough to motivate the use of digital. I find I cannot prepare a PowerPoint slide show without pulling the images into a Portfolio catalog first. Once they have been brought into PowerPoint, I lose my place and reference the Portfolio catalog to figure out what I need or what the order is that I am trying to achieve. So I think having Portfolio as an image management tool is helpful to create presentations with other software too.
10/5/2001