dambusting.com

Sometimes it’s useful to know the width and height of the thumbnails that Cumulus creates from your assets, and maybe even be able to search and sort on those dimensions.

The way to achieve this is not well known, and doesn’t appear to be documented, but it is very easy to do.  You only have to add two extra metadata fields called Thumbnail Height and Thumbnail Width then when Cumulus creates or updates the thumbnail, it will automatically store the dimensions in those fields for you.

Here is how you add the extra metadata fields:

1. Open your Cumulus desktop client
2. To make sure you have the necessary privileges I suggest you log in as the Cumulus administrator
3. Open your catalog
4. From the main menu select Edit->Preferences
5. Select Catalog Settings in the left hand pane of the Preferences window
6. Make sure your catalog is the one showing in the Catalog pick list at the top
7. Choose the Record Fields tab and click Add Field
8. Open the Fields tab in the Add Field window (it may already be open for you)
9. Select both the Thumbnail Height and Thumbnail Width fields using ctrl+click
10. Click OK in the Add Field window

Possible gotcha: It’s really important to select the predefined fields from the Fields tab in the Add Field window. If you simply create some custom fields with the same names it won’t work.

Remember that Cumulus only sets the values in these fields when creating or updating the thumbnail, so if you already have records in your catalog you will need to “update” the records as described below:

1. Make sure all of your records appear in your current collection by turning off any filters and selecting Find->Find All from the main menu.
2. Click just one of the records to select it and then select all the records using Edit->Select All from the main menu
3. Select Metadata->Update records->Update now (incl. Contained Assets) from the main menu
4. If you have multiple Asset Handling Sets configured you will then get a pop up window offering you a selection. You will then need to select the one that you want to use and click OK

A very powerful, but much misunderstood, feature of Canto Cumulus is the concept of categories.

In my most recent blog post on categories and records I spoke about how categories resemble folders. How for example they are organised like folders into a hierarchy that Canto calls the Category Tree and how you can associate Cumulus records with them. I also showed how thinking about categories as being like folders is ok when you first start out, but will very quickly cause confusion and is to be avoided. I also spoke a bit about what Categories are in a technical sense, but I didn’t discuss what they can actually be used for, so I’ll make up for that now!

Let’s start with a working definition:

The word “category” in Cumulus refers to a named entity that can be associated with one or more Cumulus records for the purpose of tagging them for future search and retrieval.

Don’t worry if this doesn’t make complete sense just now as I’ll talk more about it below.

One way to get a “bead” on categories is to think of them as structured keywords. So let’s talk a bit about keywords.

You may already be familiar with the concept of keywords, which are essentially words and phrases that describe a digital asset. Typically a digital asset will be described by a number of keywords. You may even know that a keywords metadata field, courtesy of standards like IPTC and EXIF is embedded in various asset file types and can be edited with asset authoring tools like Photoshop. To save you from having to copy it manually, Cumulus can extract that information from the asset and put it into a special “Keywords” record field for you, and that you can then edit that field to add, remove or amend the keywords that it contains. You can then perform searches on the Keywords field so you can find records in Cumulus, and hence your digital assets.

For many DAM systems, this is as far as keywords go, and new Cumulus users sometimes think that being able to edit and search for them is all they need. But there are problems with this approach, as expressed by the following questions:

• What if you misspell a keyword, or decide that it should be replaced with another keyword? Using a simple Keywords field as described above you would need to visit every record containing that keyword and manually edit each keywords field.
• What if you want to use a “controlled vocabulary” so that only approved keywords are used?
• How do you differentiate between homonyms, for example “ball” the sphere and “ball” the dance?
• How do you capture and represent relationships between keywords, for example how do you express that keyword “Accord” refers to a type of car, specifically a model of car by Honda?

Cumulus categories solve these problems, and more. Let me show you how.

Open up a test catalog in your Cumulus client, for example the Sample Catalog that comes with Cumulus. We aren’t going to make many changes, but may want to make a backup of this catalog in case you want to restore it back to it’s original state later.

Take a look at the Category Pane, which is usually visible on the left hand side of the Cumulus client. If you can’t see it you may need to make it visible by selecting View->Workspace->Category Pane from the main menu.

Unless your Cumulus administrator has deliberately hidden them away from you, you will notice a number of tabs at the top of the Category Pane. These tabs represent what Canto calls the Master categories. If you click the All tab you will see them all listed underneath the catalog name as $Categories, $Keywords and $Sources. If you are using the “Enterprise” version of Cumulus there could even be more of them. I’ll write another blog article to explain fully the significance of these Master categories, but it’s not really necessary for you to understand them now, and for the purpose of this exercise we will just ignore them.

Create a couple of categories by right clicking the catalog name in the Category Pane and selecting New Category from the pop-up context menu. Name them for some common colours, say “Red” and “Blue“.

Now find some records of images that have something red in them and associate them with the category “Red“. Remember you can associate records with categories very simply using drag-and-drop. Then do the same for the “Blue” category. If you have images that contain both red and blue you can associate them with both of the categories.

Did you notice how easy and fast it was to tag the records with the “keywords” Red and Blue in this manner and that regardless of how often you tagged a record you only had to type each keyword once? Isn’t that better than typing the keyword in every time you wanted to make a tag. You’ve also avoided making typos, and what’s more you have a ready-made controlled vocabulary that helps to ensure data entry consistency.

Now what if you later decide that you shouldn’t really start your keywords with capital letters unless they are proper nouns? With Cumulus categories it’s really easy to make this change because you are only changing it in one place. Feel free to give it a try by changing your new category names to “red” and “blue“. To do this just click the name once, wait for a second and click it again, then you can then edit it.

What about creating relationships between “keywords”? This is where the “tree” part of Category Tree comes in, meaning that you can create sub-categories under categories, and even sub-categories under sub-categories. Let’s give it a go.

Create a new category called “color“, then make your red and blue categories become sub-categories of color. You do this by simply dragging and dropping the categories, a bit like rearranging folders on your computer. By doing this you will have created a simple tree that looks like this:

color
-> red
-> blue

In this case the color category is used to indicate what is meant by red and blue, and in doing so can help you differentiate between keywords that are homonyms, being words that have the same spelling but different meanings. For example the word “blue” could also mean “sad“, and you could differentiate the meaning by having them in separate branches of the category tree, for example the other branch could be:

feelings
-> happy
-> blue

But what if you wanted to represent a synonym, being a word that means the same but has a different spelling? For example what if you wanted to have both “sad” and “blue“. Well, the category tree supports that, but it involves an advanced concept called “Related Categories” and that I’ll cover those in a future blog article.

Arranging your keywords hierarchically like this not only helps you during data entry by indicating the correct categories to associate with your records, it also helps you when searching for records by letting you browse the category tree for the right keyword before performing your category search.

If you are already familiar with Cumulus or if you have read my earlier blog post you will know that a category search can also find records in associated categories. By default if you perform a search on a category, e.g. by simply double-clicking it in the category tree, the search result will include all records associated with its sub-categories. For example if you wanted to fine tune your keywords to

color
-> red
  -> magenta
  -> puce
-> blue
  -> indigo
  -> sky blue

This way when you do a category search by double-clicking red you will also^† get the records that are associated with the sub-categories  magenta and puce as well as those that are associated directly with red.

† note: This assumes you have not changed your User settings in Cumulus to make the category search not include the records that are attached to the sub-categories.

I think that’s enough to think about for now. There’s more to learn about categories, so watch this space in the New Year.

In the meantime – all the best for the festive season!

Further reading:

• In your Cumulus “Client User Guide” check out those sections covering “Categories” and “Collections”
• Look on the Canto web site for tutorials about Cumulus categories. At the time of writing there is a good one here

It’s a common misconception for new Cumulus users to equate Cumulus categories with folders, and Cumulus records with files. This is because at first glance they seem to look and work like the folders and files in your computer’s file system. Here are the similarities that tend to make you think this way:

(note that the references to “file manager” below can be taken to mean either the Windows File Explorer or the Mac File Manager)

1. Cumulus categories are arranged into a hierarchy, or “tree”, in the same manner as folders in the file system
2. The category tree is depicted in the left hand pane of the Cumulus client much like the folders in your file manager
3. You can re-arrange the categories using drag-and-drop similar to rearranging folders in your file manager
4. The Cumulus records (representing the assets) appear in the right hand pane of the Cumulus client – just like files do in your file manager
5. You can attach a Cumulus record to a category using drag-and-drop in the same way that files are attached to folders in your file manager
6. When you double click a category in the category tree the records that are attached to that category appear in the right hand pane, just like files do when you double click a folder in your file manager

Canto probably made it work like this because it’s very intuitive for anybody who has used a file manager, and these days that’s pretty much anyone who has used a computer. This file/folder metaphor certainly means you can be up and running with categories and records right away, but therein lies the trap!

The problem is, that if you keep thinking of categories as folders and records as files you will:

• Not understand the subtle relationship between records and the assets they represent, and…
• Not understand the full power of categories, which are one of the great design features of Cumulus, and worse still…
• You will get frustrated with the product when things don’t work quite the way you expect

For example, the record=file misconception causes great confusion when it comes to deleting. In your file manager you will be used to using the delete key to delete files, but if you are new to Cumulus and try this you will get a very perplexing “surprise”. Here a simple scenario that shows what I mean.

1. Open a Cumulus catalog that has some records that are already associated with categories. The Sample Catalog that comes with Cumulus will do.
2. Double click on one of the categories so that the associated records appear in the right hand pane. (A new user will think of this as opening a “folder” with the records that are “in” that folder appearing in the right hand pane.)
3. Select one or more records and hit “delete”
4. Notice that the records have disappeared – in the same way that files disappear when you delete them in your file manager

At this point the new user will erroneously think that they have successfully deleted the records, they may even think they have deleted the actual asset files that the records represent. Neither of which is right.

Continuing the above scenario, double click once more on the category you chose before, and see that the “files” (actually records) have reappeared. A new user would be very perplexed by this and may even be inclined to think that there is something wrong with the product.

Another possible source of confusion, though for some reason new users don’t have the same “what the heck?” reaction to this, is that when you double click on a category it doesn’t show only the records that have been attached to that category. Let’s try another scenario to illustrate this:

1. Open up a test catalog. Use the Sample Catalog again if you don’t mind the fact that you will be messing with it.
2. Create a new category – call it “A”, and then a sub-category under it – call it “B” (A new user will think of this as creating a folder and a sub-folder)
3. Associate some records with category “A” and then associate some different records with category “B” (A new user will think of this as putting the records “in” those categories)
4. Double click category “A” and notice that not only the records “in” that category appear, but also the records “in” category “B”

This does cause some confusion, though most users pretty quickly work out what is happening without having it explained. The folder/file metaphor is so strong however that it seems to survive this experiment unscathed. Probably the new users simply think of this as being like a “super” file manager that shows all of the files in all of the sub-folders.

The problem is that what is really happening is that Cumulus is performing a search on the category and displaying the results, and it’s possible to change the meaning of a “category search” in Cumulus such that different sets of results can be returned. For example you can actually configure your Cumulus such that it finds only the directly associated records, though the default behaviour is to show those in subcategories as well. To make things really confusing for new users you can even configure Cumulus such that the search returns all directly associated records, plus those associated with the higher level categories.  To add to the confusion, you can also configure it to include or exclude what are known as “related categories”. Configuring the category search and “related categories” deserves a separate blog entry, but the point I want to make here is that these more advanced concepts will be hard to grasp if you get stuck in the folder/file metaphor.

Some folks get so caught up in the folder/file metaphor that they miss out on one of the most powerful features in Cumulus, being the ability to associate not only multiple records with a category, but also multiple categories with a record. Stated like that it seems so simple, but many folks struggle to “get” it. All too often I see people creating convoluted category trees with the same sub-categories repeated over and over again, simply because they don’t take advantage of the many-to-many nature of the category/record relationship.

So here are some tips for new users on how to avoid thinking that categories=folders and records=files.

1. Be aware that there can be many Cumulus records created for a single asset file. For example Cumulus can create separate records for the slides inside a Powerpoint presentation, even though there is only one actual Powerpoint asset file.
2. Don’t be tempted to use the word “folder” when referring to categories.
3. Don’t think of records as being “in” or “belonging to” categories, instead think of them as being “attached to”, “assigned to” or “associated with”
4. Don’t think you are “opening” a category when you double-click it, instead think of it as doing a “category search” and be aware that Cumulus can be reconfigured to perform category searches in different, potentially confusing, ways.
5. Try creating some many-to-many relationships between categories and records. Experiment with multi-selecting several categories and dragging them onto the records, rather than vice versa.
6. Experiment with searching for records in multiple categories by multi-selecting several categories before double-clicking.
7. To get out of thinking that the records in the right hand pane of the Cumulus client represent the files in a folder try running a “Quicksearch” and notice that the search results also get placed in the right hand pane. In fact this right hand pane is is really just a collection of records, and there are many ways in which you can modify this collection. That is why in Cumulus terminology it’s called the “Collection Window”.

Further reading:

• In your Cumulus “Client User Guide” check out those sections covering “Categories” and “Collections”
• Look on the Canto web site for tutorials about Cumulus categories. At the time of writing there is a good one here

Until next time!