Re: Digitization standards

John Blunden (mailto:blunden2@LLNL.GOV)
Tue, 7 Oct 1997 09:22:31 -0700

Message-Id: <199710071622.JAA26064@dns.ccit.arizona.edu>
Date:         Tue, 7 Oct 1997 09:22:31 -0700
From: John Blunden <mailto:blunden2@LLNL.GOV>
Subject:      Re: Digitization standards
To: mailto:IMAGELIB@LISTSERV.ARIZONA.EDU

Anne has some good questions and I would like to hear more comments and
opinions on this matter also. We are presently beginning to evaluate
various image archiving packages and methods of storing images and some
questions pop up regarding "standards".

1. Should any standard remain fixed or should it change with technology? For instance, Photo CD was Kodak's first approach to image archiving that is now evolving (supported by many others such as Live Picture and HP) into FlashPix. My opinion is that we will need to support various evolutionary file formats over time and need a system that can support this kind of change. Migrating current files into new formats will be an ongoing burden over time.

2. Regarding the DPI issues, I wonder if it wouldn't be better to capture the images at the maximum that the original can support. Some photographic prints are fairly low res while other media such as 4x5 negatives could be very large. The idea is that we are creating an archive that will have to support future output devices and requirements. It would be smart to pay for the file size storage up front and have long term assets than just look at archiving as a short term approach. Maybe someone has a middle of the road approach here?

A lot of the answers will depend on what the long term usage will be for each organization. More than just hardcopy output is involved and the WWW is ever more important. Who are your customers, what are their needs and, for an archive to stay relevant over time, what WILL be their needs in the future? For our needs, we are looking at storing the maximum data for each image. We will be looking at some new compression schemes (Altamira's Genuine Fractals for example) to try to reduce the gross file sizes without compromising the image integrity. I really don't know where we will end up but want to avoid creating a database with a limited lifespan due to resolution issues.

John

>Dear ImageLIB,
>
>I, too, would like to know this. My main question is: isn't d.p.i (output
>resolution) irrelevant as to whether something is of archival quality? I
>would think the pixel heighth and width, and the scale, that you scan
>something in, are more relevant because (at least I think) those are the
>items that determine the amount of information that is going to go into the
>scanned file. I've been told to go by Kodak Photo CD standards (base, base
>64, etc.), without using their compression modes. Thanks!
>>
>>Has there been a standard established to define what the d.p.i.
>>for a "preservation quality" b&w photographic image should be if
>>digitized in 256 grays? I'm specifically asking about an uncompressed
>>archival version of the scanned file. I've read suggestions which
>>range from 300 to 600 d.p.i. and I'm interested in what others
>>are doing, specifically, those that do their scanning in-house.
>
>____________________________________
>
>Anne Troy
>User Services Specialist
>College of Architecture and Landscape Architecture
>University of Minnesota
>612.624.4080
mailto:>troyx003@maroon.tc.umn.edu

John Blunden mailto:blunden2@llnl.gov Phone (510) 422-4989 Fax (510) 422-6944 "My job is not all that difficult, but I do have to know the entire alphabet." -Vanna White