Re: Experiences with digitizing material not suitable for aflatb

From: Lossau (lossau@MAIL.SUB.UNI-GOETTINGEN.DE)
Date: Wed Dec 31 1969 - 17:59:59 CST

  • Next message: ELIZABETH RODERICK: "Re: Wavelet Compression & Digital Networks"

    Message-Id: <200004070950.CAA20170@dns.ccit.arizona.edu>
    Date:         Fri, 7 Apr 2000 11:48:10 GMT+0100
    From: Lossau <mailto:lossau@MAIL.SUB.UNI-GOETTINGEN.DE>
    Subject:      Re: Experiences with digitizing material not suitable for aflatb
    To: mailto:IMAGELIB@LISTSERV.ARIZONA.EDU
    

    <pre> At the Digitization center (GDZ) of Goettingen State and University Library in Germany we made some experiences with digitizing unique and fragile material. In February we finalized e.g. the digital conversion of our Gutenberg Bible (a vellum copy). We scan with 300 dpi at 100%
    (as our colleagues from Australia) and use a professional digital scanning back "Picture Gate 8000"
    (http://www.anagramm.de/de/produkte-picture-gate.html, Anagramm company in Munich) which offers a maximum resolution of 8000 x 9700 pixels.

    From the preservation point of view the astonishing short exposure time with 1,5 to 2 minutes for maximum resolution was convincing for us. The large format camera is a Linhof Kardan-E, a special lense (Rodenstock Apo-Sironar-N) is used related to the digital capture.

    File sizes are up to 440 MB, for our Gutenberg Bible we got about 380 MB for the illuminated pages and 180 for the text pages (altogether about 350 GB for the preservation master=48-bit color depth , stored on CD-Rs).

    To handle the enormous mass of data we use a harddisk system (140 GB) during the daily production and store the data (as backup and for preservation) on CD-R. A CD-R jukebox enables us to burn 35 CD-Rs over night.

    We use so called "cool light"-area lights (5000 Kelvin, flicker-free, efpe-company), important again from the preservation point of view. Internal color depth is 36-bit but we store the images at 48-bit for the preservation master and scale them down for further work to 24-
    (our "access master") and 8-bit.

    Our computer is a Macintosh G3 with 756 MB RAM, Apple provided us with a G3 and 1 GB RAM for the digitization of the Gutenberg Bible.

    Color calibration is the same as at the National Library of Australia
    (CIE D65 and profiles ICC, Software "Colorblind").

    At the GDZ we work with a special table (including moveable cradle, vacuum to hold pages down and a laser light to enable optimal horizontal and vertical position of the cradle).

    You see some pictures of our equipment under http://www.gutenbergdigital.de/technik.html. Our conversion process is described (in English) at http://www.sub.uni-goettingen.de/gdz/en/gdz_main_en.html.

    Regards Norbert Lossau

    On 6 Apr 00 at 8:58, LeMar,Greg wrote:

    > Date: Thu, 6 Apr 2000 08:58:59 -0400
    > Reply-to: IMAGELIB <mailto:IMAGELIB@LISTSERV.ARIZONA.EDU>
    > From: "LeMar,Greg" <mailto:lemarg@OCLC.ORG>
    > Subject: Re: Experiences with digitizing material not suitable for aflatbe
    > d
    > To: mailto:IMAGELIB@LISTSERV.ARIZONA.EDU

    > Still more. Many solutions, but still no mention of an archival
    > intermediate.
    >
    > -----Original Message-----
    > From: Andrew Stawowczyk Long [mailto:mailto:anlong@NLA.GOV.AU]
    > Sent: Thursday, April 06, 2000 12:39 AM
    > To: mailto:IMAGELIB@LISTSERV.ARIZONA.EDU
    > Subject: Re: Experiences with digitizing material not suitable for
    > aflatbed
    >
    >
    > --------------201CB6430B73D0AB615741D5
    > Content-Type: text/plain; charset=us-ascii
    > Content-Transfer-Encoding: 7bit
    >
    > Mark,
    >
    > Here in the National Library of Australia we did couple of projects that
    > involved digitising at high resolution. We aimed at 300dpi at 100%. As we
    > were working with unique, fragile and oversized material
    > (rare maps, manuscripts) we couldn't use a flat bed scanner. Instead we
    > worked with high-end digital scanning back combined with a Sinar large
    > format camera. It worked extremely well for most items
    > achieving the required resolution. The back was Phase One with 8,400 x 6,000
    > pixel resolution and 36-bit internal colour depth. Some of the maps we
    > needed to shoot in tiles, later stitching them together
    > on a computer to get 300dpi resolution. It can be avoided by using a new
    > Phase One back called Power FX with resolution exceeding 12,000 x 10,000
    > pixels.
    >
    > Some of the important things to be aware of:
    >
    > - Required light levels are very high - can be dangerous from the archival
    > point of view.
    > - Power conditioners (stabilisers) have to be used for lights as light
    > levels have to be maintained very acurately during 5-20 minutes of exposure.
    > - High resolution lenses (designed for digital capture) have to be used.
    > - Powerful computers with large hard disks and a lot of RAM (500Mb - 1Gb)
    > are needed.
    > - Scanning back and computers need to be colour calibrated to a known
    > standard (we use CIE D65) and colour profiles need to be used (ICC) to
    > maintain colour accuracy throughout the system chain.
    > - It is a common practice to capture at 36 or 48-bit colour depth and then
    > reduce it to 24-bit for storage. It gives best results.
    >
    > Also, we evaluated an option for capturing images on film (4" x 5"
    > transparencies) and scanning them later on but the results were not
    > satisfactory. It was due partially to the much lower dynamic range of
    > a photographic film than those of digital devices (typically for film it is
    > 6 f-stops and at least 11 f-stops for high quality digital devices). Also,
    > it is much more labor intensive and more expensive.
    >
    > I hope that'll help. Let me know if you need more info.
    >
    > Regards
    >
    > Andrew Stawowczyk Long
    > Public Affairs Officer
    > National Library of Australia
    > Phone: +61 2 6262 1382
    > EMAIL: mailto:anlong@nla.gov.au
    >
    >
    > >
    > > >>> Mark Jordan <mailto:mjordan@SFU.CA> 04/06/00 11:02 >>>
    > > Hi,
    > >
    > > We've been doing some research on the costs of equipment suitable for
    > > digitizing material such as books, oversized posters, drawings, etc. that
    > > we wouldn't want to (say pages from a rare and fragile book) or couldn't
    > > (because of size) scan with a flatbed, even one with a 12x17" platten. To
    > > get 24-bit color at a 300 dpi or higher, we'd have to spend quite a bit of
    > > money.
    > >
    > > Does anyone have any experience with alternative methods of digitizing
    > > largish-size material not suitable for scanning with flatbeds into 24-bit
    > > color, high resolution images? For example, could photograph the items,
    > > create high-quality prints or slides, and then scan the photo? Or is this
    > > kind of method just not practical?
    > >
    > > Thanks for sharing any experiences you may have had with this kind of
    > > material,
    > >
    > > Mark
    > >
    > > Mark Jordan
    > > Librarian / Analyst, Systems Division
    > > W.A.C. Bennett Library, Simon Fraser University
    > > Burnaby, British Columbia, V5A 1S6, Canada
    > > Email mailto:mjordan@sfu.ca / Phone (604) 291 5753 / Fax (604) 291 3023
    >
    > --------------201CB6430B73D0AB615741D5
    > Content-Type: text/html; charset=us-ascii
    > Content-Transfer-Encoding: 7bit
    >
    > <!doctype html public "-//w3c//dtd html 4.0 transitional//en">
    > <html>
    > Mark,
    > <p>Here in the National Library of Australia we did couple of projects
    > that involved digitising at high resolution.&nbsp; We aimed at 300dpi at
    > 100%.&nbsp; As we were working with unique, fragile and oversized material
    > (rare maps, manuscripts) we couldn't use a flat bed scanner.&nbsp; Instead
    > we worked with high-end digital scanning back combined with a Sinar large
    > format camera. It worked extremely well for most items achieving the
    > required
    > resolution. The back was Phase One with 8,400 x 6,000 pixel resolution
    > and 36-bit internal colour depth.&nbsp; Some of the maps we needed to shoot
    > in tiles, later stitching them together on a computer to get 300dpi
    > resolution.&nbsp;
    > It can be avoided by using a new Phase One back called Power FX with
    > resolution
    > exceeding 12,000 x 10,000 pixels.
    > <p>Some of the important things to be aware of:
    > <p>- Required light levels are very high - can be dangerous from the
    > archival
    > point of view.
    > <br>- Power conditioners (stabilisers) have to be used for lights as light
    > levels have to be maintained very acurately during 5-20 minutes of exposure.
    > <br>- High resolution lenses (designed for digital capture) have to be
    > used.
    > <br>- Powerful computers with large hard disks and a lot of RAM (500Mb
    > - 1Gb) are needed.
    > <br>- Scanning back and computers need to be colour calibrated to a known
    > standard (we use CIE D65) and colour profiles need to be used (ICC) to
    > maintain colour accuracy throughout the system chain.
    > <br>- It is a common practice to capture at 36 or 48-bit colour depth and
    > then reduce it to 24-bit for storage.&nbsp; It gives best results.
    > <p>Also, we evaluated an option for capturing images on film (4" x 5"
    > transparencies)
    > and scanning them later on but the results were not satisfactory.&nbsp;
    > It was due partially to the much lower dynamic range of a photographic
    > film than those of digital devices (typically for film it is 6 f-stops
    > and at least 11 f-stops for high quality digital devices). Also, it is
    > much more labor intensive and more expensive.
    > <p>I hope that'll help.&nbsp; Let me know if you need more info.
    > <p>Regards
    > <p>Andrew Stawowczyk Long
    > <br>Public Affairs Officer
    > <br>National Library of Australia
    > <br>Phone: +61 2 6262 1382
    > <br>EMAIL: mailto:anlong@nla.gov.au
    > <br><a href="http://timeframes.natlib.govt.nz/"></a>&nbsp;
    > <blockquote TYPE=CITE>&nbsp;
    > <br>>>> Mark Jordan &lt;mailto:mjordan@SFU.CA> 04/06/00 11:02 >>>
    > <br>Hi,
    > <p>We've been doing some research on the costs of equipment suitable for
    > <br>digitizing material such as books, oversized posters, drawings, etc.
    > that
    > <br>we wouldn't want to (say pages from a rare and fragile book) or couldn't
    > <br>(because of size) scan with a flatbed, even one with a 12x17" platten.
    > To
    > <br>get 24-bit color at a 300 dpi or higher, we'd have to spend quite a
    > bit of
    > <br>money.
    > <p>Does anyone have any experience with alternative methods of digitizing
    > <br>largish-size material not suitable for scanning with flatbeds into
    > 24-bit
    > <br>color, high resolution images? For example, could photograph the items,
    > <br>create high-quality prints or slides, and then scan the photo? Or is
    > this
    > <br>kind of method just not practical?
    > <p>Thanks for sharing any experiences you may have had with this kind of
    > <br>material,
    > <p>Mark
    > <p>Mark Jordan
    > <br>Librarian / Analyst, Systems Division
    > <br>W.A.C. Bennett Library, Simon Fraser University
    > <br>Burnaby, British Columbia, V5A 1S6, Canada
    > <br>Email mailto:mjordan@sfu.ca / Phone (604) 291 5753 / Fax (604) 291
    > 3023</blockquote>
    > </html>
    >
    > --------------201CB6430B73D0AB615741D5--
    >
    ************ Dr. Norbert Lossau State- und University Library Goettingen Head of the Center for Digitization (GDZ) Subject librarian for Finno-ugrian Studies Platz der Goettinger Sieben 1 37073 Goettingen, Germany Tel.: +49 +551/39-5217 Fax. +49 +551/39-5222 WWW: http://www.sub.uni-goettingen.de/GDZ http://www.gutenbergdigital.de E-Mail: mailto:lossau@mail.sub.uni-goettingen.de

    </pre>



    This archive was generated by hypermail 2b29 : Fri Apr 07 2000 - 04:58:02 CDT