Image management solution for a medium sized personal scanning project

I'm looking at doing a largeish (3000+ image) scanning project [2].
For various reasons I'm probably going to hire a student to do the
scans and buy a Nikon V ED negative scanner (hard to find btw, most
vendors are sold out).

The image acquisition part is relatively straightforward. I'll be
keeping the negatives of course. I'll image at about 2000 dpi and
store as 99% JPEG. [1]

My main questions are about image mananagement. I've looked at a few
reviews of lower end software (iView MediaPro, Microsoft Imaging
Suite, Adobe PhotoAlbum, Picasa, ACDSee, etc), and I can't tell if
anything does what I want. I've used iPhoto extensively on my OS X
machine, but this project will probably be PC based. Here's my short
list:

1. I want all the metadata to be accessible, ideally stored using a
commercial database structure. So I can directly manipulate image
identifiers, image paths, image titles, descriptions, roll
information, catalog/album names, etc. I'd be happy with an Access
database, a FileMaker database, or an open source database.

2. I'd like very good support of embedded EXIF tags. So the album
software should be able to write data to EXIF tabs within the JPEG
headers -- such as image title, description, data of acquisition, etc.
This data will mirror what's in the image management database.

3. I want the album software to manage unique identifiers, ideally
also within the EXIF header. I want to be able to go from any image to
its metadata. The album software also needs to manage filename
collision. I'd be just as happy for the album software to name every
file with a unique identifier and blow away the original file names.

4. I'd like to be able to set a prefix or suffix applied to images in
addition to the album maintained image unique identifier.

5. The solution needs to scale to tens of thousands of images and to
manage image migration to external media catalogs.

6. I'd like to be able to define a subset of the catalog and burn it
to a CD along with a local catalog.

7. I'd like to be able to edit images in an external editor
(Photoshop, etc), and have the image software handle versioning
(retaining the original).

8. Indexing, searching, keywords, etc are nice, but not the main
thrust of this project.

I think these requirements are more typical of high end professional
solutions. I hope to cobble something together from a few packages. I
wonder too about some of the less familiar open source image
management solutions, including some that are web based. I'd guess
they'd be more likely to meet my needs.

I don't care as much about integrated image management tools.

Any thoughts from experienced users -- esp. Pros?

Thanks!

john


meta: jfaughnan, jgfaughnan, image management, photoalbum, database,
photo album, metadata, scanning, imaging, home, personal

[1] In 10 years I may do this again with 2014 technology. Then it will
be lossless. I am also consider JPEG 2000 for the better color
management.

[2] I have thousands of unfiled family photos. I plan to image the
negatives and then manage digitally. I'll discard the prints and keep
the negatives. After imaging everything, I expect to delete at least
half the images over time. To help fund the project the student will
do similar work for others. I plan to sell the scanner after a few
months of use -- I don't do ongoing film work.

 

If you're willing to do this operation separately, I don't
think you can beat Lupas Rename:
http://www.azheavymetal.com/~lupasrename/lupasrename.php
I've used it for just the purposes you describe, flawlessly.

Also, I have used ACDSee and IMatch. Both work fine
for most of what you want to do.

 

You don't mention Thumbs+ or IMatch, the best two packages in what I
consider low-end. I suggest you start with them. I use IMatch, so most
of my answers will deal with that.

IMatch at http://www.photools.com

Thumbs+ at http://www.cerious.com

Thumbs+ uses commercial DBs. The default is Access, but it can be
configured to use various other DBs, including SQL Server.

IMatch does not use a commercial DB. However, you can access all the
metadata using it's built in scripting. Also, you can export the data
to XML (or CSV) and re-import it. So you could export to XML, use
various XML tools to make your changes, and then import the data back.
In my opinion it would be easier to use the internal scripting (object
oriented, quite powerful), but whatever works for you...

IMatch can import data from EXIF or IPTC into the DB, and export data
from the DB to EXIF or IPTC fields. They are not, however,
automatically kept in sync -- you must explicitly export again after
you have changed the DB.

You can also edit the EXIF/IPTC data directly, so if there is some
stuff you want in EXIF but not in the DB, you can manage that easily.

However, below you mention keeping images off-line. That will make
keeping the DB in sync with the EXIF information difficult (regardless
of what image management solution you use).

I suggest you do this before entering the images into the database. If
you don't care about the original filename, why not get rid of it
immediately. Create a unique ID from the current date/time.

There are various renamers that can do this, I use THERename
(www.herve-thouzard.com/therename.phtml).

IMatch and Thumbs+ both include built-in renamers that can do this.
Neither can generate a UUID in the technical sense, but they can
create a unique name based on date/time (at least IMatch can, and I
think Thumbs+ can).

In the filename, you mean? Easy. Though again I would do this before
putting the image into the DB.

Both handle large databases, and can manage images on off-line media.

In IMatch, this could be done in various ways. You could create a
subset DB, but personally I would export the metadata for the "local"
images to XML, and burn that onto the CD/DVD. Because XML is text,
that makes the data readable any time anywhere (eg, if you need to get
at it in 10 years).

Versioning could be difficult, depending on what you want. Versioning
can be handled by using a standard set of filename conventions, or a
standard set of directories. If you standardize your workflow, these
packages will work OK for this. But neither handles versioning
inherently. For example, I would like image management to link all
versions of an image to the same "entry", with one thumbnail. Under
that thumbnail entry, you would be able to get to different versions
(and different file types, raw, jpg, tiff, etc.). I don't know of
Image Management software that works this way, though I think we'll
see it in the next year or so in IMatch.

In IMatch you can use the structured categories to keep track of
versions, and this can be quite powerful. But every version has its
own thumbnail, so it's not my ideal versioning.

And *any* package will handle this.

On the other hand, if keywords and searching are just "nice" (not
required), why are you bothering with image management at all? What
will you be storing in the database, and when will you be searching
the DB for what?

If you plan to delete half the images, why not do this *before*
scanning, instead of after. That will save you 1/2 the cost and time.
You don't have to scan it, enter it in the DB, figure out its
description and keywords, back it up to multiple media, etc.

A final question: Is this worth the time and effort to you? Managing a
large collection of images takes time. And it sounds like you plan to
do significant editing (since you want versioning) = more time.

HTH,

Terry

 

Terry, you ought to win some kind of usenet award for thoroughness and
helpfulness. Maybe a vintage 1970s Stan Lee No-Prize !

I will investigate both recommendations, they sound much better than
anything else I've looked at. Your renamer suggestion is also much
appreciated.

A couple of responses to some of your specific questions (you've
addressed all of my primary questions):

I was originally thinking I'd eventually move all the images to my
Mac, possibly after I buy my next machine (currently I'm running on an
iBook that's straining under the image load). So I wanted an
application to park and manage the images, preferably an app that
would allow me full control of metadata and optimal export
capabilities. The plan may change, but having control over the
metadata gives me more flexibility.

It's a division of labor issue. I can't outsource the editing step,
but I can outsource image acquisition. The person doing the
acquisition will be managing the metadata (basically limited to roll
date and roll identifiers!), I'll be doing the editing. I'm much
faster at editing/deleting in a digital than analog workflow -- I have
a rhythm for it using ranking measures, sorts, side-by-side
comparisons, etc. The main metadata entry (comments, etc) will occur
after the editing stage.

Given my available time for this kind of work, the cost/benefit ratio
looks good. For a few hundred dollars I can get a task done that
otherwise awaits my retirement.

I've been all digital for acquisition for a couple of years, and
managing that library takes me about 1/10th the time and energy it
took to manage albums. Maybe 1/20 to 1/30! I don't think that's true
for everyone, but it works for me. Even more than the acquisition
benefits of digital (many pictures -> good pictures, darkroom, etc)
digital image management has been a big personal plus. (Even using
iPhoto - a very imperfect product that's less imperfect than it used
to be.)

Thanks again,

john


meta: jfaughnan, jgfaughnan, digital image management, photography

 

If one already has Photoshop (CS) is there any compelling reason to not
use the Image Browser to manage digital assets?