How do you plan to use collections?

[mgm1965]

IMatch 5 Collections are very interesting.

I'm re-thinking my workflow with the unique features of IMatch 5.

My objective is to minimize personalization and use all possible the standard feature of IMatch.

So I need your experience...

Thank a lot.

Marco

[JohnZeman]

Since dots, pins, and flags are new to IMatch I wouldn't be surprised if few beta testers have come up with good uses for those yet.  Speaking for myself only at this point I normally keep all my collections empty, then I'll use one or more of those collections to hold test results.

For example I have a script which will compare the descriptions of the selected images, all images that have the exact same description as the focused image will be assigned to the Set Flag collection while all others are assigned to the Unset Flag collection.  Likewise for another script I have that will import metadata locations into your category tree.  Any images with incomplete metadata location information are assigned to the Red Pin collection.

In IMatch 3 I created a temporary category to hold flagged images like that, I find collections to be not only more versatile, but they don't affect my categories at all.

[cytochrome]

Likewise for another script I have that will import metadata locations into your category tree.  Any images with incomplete metadata location information are assigned to the Red Pin collection.
...

Do you maintain a "where" category outside of the data-driven one included in IM5? Has it benefits? or is it just to track incomplete Location data? Could this not be done with the data driven one?

I ask because I am still trying to define the most simple AND efficient way to handle my "classic" metadata (Location, Headline, and so on..)

Francis

[JohnZeman]

Yes, I also have a standard "Where" category as part of the commonly used 5 W strategy.

Where
When
Why
What
Who

As useful as they are I don't want to rely exclusively on data-driven categories for locations because some locations need a little special attention.   For example here in the USA National Parks can sometimes cover more than one state.  Yellowstone for example.  99% of it is in Wyoming but a little bit of it is in Montana and Idaho too.  In my manually configured Where category I list all Yellowstone photos as being in Yellowstone National Park, not spread across 3 different states.

[cytochrome]

I understand this if you resort often to automatic metadata assignation via geonames or the like.

I populate the Location metadata during image ingestion (from SD card to an non-Imatch  import folder) while I remember best where it was and all details. The data-driven cat picks it up there when I transfer the files to IMatch, so no problem.

The down side is that I have sometimes the same location under different names. I have to find a way to detect these. By the way is it possible to relax case checking so paris and Paris are not 2 different towns? I know of aliases but this is proper to IMatch, so if I export all the cats one day it wont be simple.

Francis

[Mario]

Check out the Unify option available for data-driven categories. They allow you to unify spelling to:

Paris
paris
PARIS

This will not modify the data in the file or the database. It will just use the unified name when creating the child categories, thus neatly folding all spelling variations of Paris into one child category.

You can also 'fold' multiple values together using a replacement Replace Mask.

See the examples for these options in the help for more details.

[jch2103]

The down side is that I have sometimes the same location under different names. I have to find a way to detect these. By the way is it possible to relax case checking so paris and Paris are not 2 different towns? I know of aliases but this is proper to IMatch, so if I export all the cats one day it wont be simple.

Given that you may not know where your metadata will ultimately end up, it may be worth the investment of time to clean up your metadata from time to time. Fortunately, IMatch makes it very easy to spot 'paris' and correct to 'Paris', for example (select all 'paris' and drag and drop to 'Paris'.)

I see Mario has already answered your capitalization question.

[mgm1965]

I was studyng collection types:
Flag
There are three states for the flag collection: No Flag, Flag On, Flag Off

At this moment I've not use of this

Dot
A file can have either a red, green, blue or no dot
I was thinking to use:
- Red      Just arrived from my photocamera
- Green   No use at this moment
- Blue   No use at this moment
- No dot   Complete catalogation (5 W filled)

Pin
A file can have up to three pins, in red, green, blue

I was thinking to use:
Pins red   Photo Book - Collection of photo printed or can be "dysplayd" with a table or similar
Pins green   Screensaver for my Computer
Pins blue   no use at this moment

Bookmark
A file can have a bookmark

I was thinkig to use Bookmark to "Bookmark" image that must be quickly found

Rating
A file can have a rating between 1 and 5, no rating or a reject rating. IMatch uses the XMP rating of the file as the basis for this collection.
1 and 5   
no rating    
reject rating in Imatch 3 I was usig the red label to do that!

Label
A file can have a text label, optionally with an associated color. IMatch uses the XMP label of the file as the basis for this collection.   

At this moment I've not use of this

I was also thinking to fill xmp:
Author                          to use with data driven category
Location, County, City    to use with data driven category
Event                           to be use to catalogue like Christmas, Even, Birthday, Marriage, and other

Regards.

Marco

[cytochrome]

..
Given that you may not know where your metadata will ultimately end up, it may be worth the investment of time to clean up your metadata from time to time. Fortunately, IMatch makes it very easy to spot 'paris' and correct to 'Paris', for example (select all 'paris' and drag and drop to 'Paris'.)

I see Mario has already answered your capitalization question.

Yess, this I do sometimes, I go down the tree for Country, Province, City and Location  but especially in France there are zillions of composed location names Moustiers Ste Marie = Moustiers Sainte Marie = Moustiers Sainte-Marie = Moustiers-Sainte-Marie and so on. And all the éèê !!

What I would like is setting up a "Location view" showing all the Location tags of the database so I don't have to explore all branches of the tree. There is certainly a way, there is always (almost) one in IMatch!

Francis

PS I have now set up an All_Locations Data-deriven category, i see all the used Locations in alphabetic order, makes it easy to spot alternate spellings

[sinus]

IMatch 5 Collections are very interesting.

I'm re-thinking my workflow with the unique features of IMatch 5.

My objective is to minimize personalization and use all possible the standard feature of IMatch.

So I need your experience...

Thank a lot.

Marco

I am also in the progress of looking for a good workflow. At the moment I tend to not use the collection or only use the collection at the beginning of a workflow, because then an dotted (for example) thumb gets my attention.
But because in my case these collections icons are often INTO the image (I am sure, i could create a layout, where these icons are not in the image, but then I must have more space), I tend to create an image so, that at the end of the edition of an image (when it is ready for archiving), this image has NO collections, means, this image is "through" the workflow.
Say at the beginning it has a red flag, and a blue dot, and a bookmark, at the end these icons should be deleted, this means "this image is finnished".

But what I use, like in IMatch 3.6, is the coloured categories, mostly 5 cats, and also here, I look, that I do have not toooo many colors, like above, at the end of an image-editing, the colours should turn more to grey.
Also the label. Rating I did not use, but I think, I will use them.

At the end this means, an image has in my workflow at the end of its editing a almost grey-coloured category-pane under each image and no collection-icons.
A "clean" image means: this image is finnished.
An image with colors (collections or categories or labels) means: this image is still in the middle of my workflow.

But as you can imagine: there are a lot of different use of collections, labels, cats and so on.