Migration to a DAM can be a daunting task if your brand collateral images
are spread over a number of hard drives throughout your organization. There are
bound to be many duplicated items that have accumulated over time.
The first task is to gather all your images into one location. You will
need a large hard drive to do this. Ideally you will be able to gather all your
images onto an internal drive in a good workstation.
If you have many terabytes of data you could use an external drive but
make sure it is a USB 3.0 with a good speed – buy a new one, do not “make do”
with an old spare hard drive that you might have lying around.
The existence of metadata in a spreadsheet or database usually indicates that you will have very few duplicate issues described above because each image is likely to have a unique filename. The process with be simply to upload the assets and apply the metadata.
For collections of images being migrated to a DAM for the first time it
is quite common to find there is no available metadata – description,
photographer/creator, location, date, copyright status, etc.
Where do we start? Well, existing folder names are often extremely
useful. It is very common for folders to have notations like ‘must seek
permission’, ‘copyright expires next year’, or have the name of the
photographer. This is valuable data that might only obtain in the folder name –
apart from in the knowledge of a past or present staff member. You need to keep
When copying file to a central location try to preserve the folder
structure and this valuable information – we’ll show you how to mine this data
Are your images in
Hopefully you have named folders but how logical are they? Having
logical folder names is a good place to start.
Folder names are useful for migration but long term, folders are too much work to sustain.
We can use the folder names as smart data, then move away from the whole idea of folders.
With Lookatme you have a database that can be interrogated on many variable - you will not need folders!
Some of your images will be copyright restricted. For example a
photographer may have granted you a 5 year licence or somebody in your
organization may have downloaded images form a stock library. You must ensure
that the copyright information is available in the DAM. One way to do this is
to quarantine restricted images in a separate folder structure. When the images
are imported the copyright information can be added in bulk.
Metadata can also be embedded
before the import process. Once
images have been sorted onto folders it is easy to embed information in the XMP
metadata fields using Photoshop Bridge. If this step is omitted the
data can also be embedded in Lookatme DAM.
If you have images - TIFF or JPEG?
What is your key business objective?
Is your key objective to promote your brand efficiently?
Or is your key objective to preserve archival images to a standard that
will pass scientific analysis?
If you are an archivist at a cultural institution, your role is most likely to preserve the file in the format in which you received it.
For a brand manager, the answer is different.
To promote your brand efficiently, the best format for photographic
images is RGB JPEG, specifically with the .jpg extension.
Some photographers supply TIFF images which are very large, commonly 30
to 60 MB and sometimes over 100 MB. A TIFF file saved as JPEG is usually 75% to 90% smaller.
Photographers will invariably supply RGB images and it is wise to maintain that colour space but the selection of format is much more debatable.
TIFF is a very inefficient format for general marketing images unless
you have exceptional requirements. TIFF is really only justified for cultural
institutions for archiving purposes.
While JPEG does employ a “lossy” compression, it would take a forensic
expert to differentiate a JPEG from a TIFF in a printed publication or a
website. Print workflows are now almost exclusively PDF, which makes every
image into a JPEG, and TIFFs cannot be used on web pages.
JPEG files are generally 10-20% of the size of TIFFs, yielding massive
efficiencies in storage, internal bandwidth and external distribution
bandwidth.To some, the question of TIFF or JPEG is almost religious and when it comes to personal beliefs, Lookatme is very inclusive! We support both TIFF and JPEG but we do not generally make TIFFS available to your end-users because many people do not know how to handle TIFFS and many people are totally confused by CMYK images. Of course your graphic designers will probably know how to handle TIFFs, whether CMYK or RGB, and you could allow them access. You can also use Lookatme as your single source of truth, so, yes, upload all your TIFFS. Just make sure your average end-user can only access RGB JPEGs.
In summary, regardless of your TIFF/JPEG affiliation, if you want to reduce daily stress:
- If you are supplied RGB images (TIFF or JPEG) it is not advisable for you to convert them to CMYK before uploading to the DAM
- If you are supplied with JPEG images it is not advisable to convert them to TIFF
- Allow average end-users access to RGB JPEGs only, do not allow access to CMYK files and do not allow access to TIFFs
- If necessary, you might allow experts like graphic designers to access CMYK files or TIFFs you may have
- As a standard, Lookatme creates RGB JPEG derivatives of all JPEG and TIFF files.
Minimum and maximum size
Technically the smallest size image that we will process is 800 pixels on the shortest dimension. Very small images may have possible application limited to website usage, which might be OK.
Aim for over 2000 pixels, and ideally over 3000 pixels along the smallest dimension.
Yes, we offer de-duplication as a service. Typically you can expect to
find that up to 50% of your images are duplicated.
There will be exact copies with the same file name, cropped copies and
copies of different sizes. But there will also be files with the same file name
that are actually unrelated. This is very common because images are stored as
“File 1.jpg” for example – there will be lots of files with non-unique names.
You may have lots of high def files with names like 1234.mov and a low
resolution version of the same images called 1234.mov.
Both of these issues mean that purging duplicates is not a simple matter
of keeping one copy of each file name – that is not likely to work at all.
Duplicates can be flagged by software but often only a human can decide which
image to keep.
Be realistic about how thorough the de-duplication process can be – it is
far from an automatic process even with the best software. Duplicate locating software can look for
similar images based on content and similar file names but the results can
never be completely accurate. Human intervention and assessment is necessary.
If you upload a duplicate filename in the same upload session, Lookatme will alert you.