[Tig] Digital data storage (was something else)

Martin Euredjian ecinema
Wed Mar 9 23:53:06 GMT 2005

This is purely based on uneducated, uninformed and grossly irresponsible

  What does it cost to store film for 100 years vs. 
  keeping data alive for the same period of time?

Let's say we are talking about a non-trivial library requiring 30,000 square
feet (about 3,000 square meters) of environmentally-controlled space.  I'm
going to guess that this means 10,000 titles, about two hours of film on

Let's say that this storage is not located in prime real-estate geography
and you can get a $1 per square foot rate, all up.

Here are the cumulative costs, in USD:

Year 10  $9.2M
Year 25  $26.9M
Year 50  $71M
Year 75  $143.5M
Year 100 $262.3M

According to this, if you do absolutely nothing but store the material (with
minimal staff to run the operation) it'll cost $262 million to keep it alive
for 100 years.  This assumes a 2% annual increase in all costs for that
period, which might be conservative.

Compare that to this scenario:

In ten years scan all 20,000 hours of film to a high-res (both spatial and
color) losslessly compressed data format.  Do this at $2,000 per hour of

Year 10, cost of transfer: $40M

Ten years later, spend $5M to copy this library to a new format.  Automation
should be a part of this in order to reduce costs.

Twenty years later, spend another $5M for yet another transfer.

Fourty years later, THE format is available and you spend another $5M for a
transfer that will survive, say, 200 years.  By now --80 years have gone
by-- the 20,000 hours of material will probably fit in a small filing
cabinet, or less.  Five million for this transfer might be a grotesque
exaggeration.  I'd guess that this would be a trip to the Fry's Electronics
of the future and a simple copy command to be executed overnight or over a
couple of days if the source media is slow.

However, the $5M figure per replication is constant because, while it will
probably be cheaper to run the dubs as time goes by, inflation might make
the cost numerically equal.

Total bill:  $55 million.

If I was off by a factor of two:  $110 million.

Savings:  Over $150 million.  I could be off by a factor of three and it'd
still make sense.  

Presumably, at the end of the 100 year period one would have to start
thinking of getting the film duplicated at rather significant cost.  Keep
that in mind.

Food for thought.  

I'd be happy to email anyone an admittedly brain-dead Excel file with
figures for the film case.

Martin Euredjian
eCinema Systems, Inc.
voice: 661-305-9320
fax: 661-775-4876
martin at ecinemasys.com
ecinema at ieee.org
NAB booth SL1210

More information about the Tig mailing list