[PADG:1143] RE: Scanning Specificiations for Brittle Books

["Carli, Alice" <acarli@xxxxxxxxxxxxxxxxx>]

Dear Jean Ann,

At Sibley Library we are scanning in-house on demand a lot of rare public domain music, and have applied for a grant to "just do" all the scores and books we hold that are held by two or fewer other libraries on OCLC, and that aren't vault items. (We're holding out for better technology, or at least an in-house cradle scanner, for those.) 

That had turned out to be a lot of material (roughly 13,000 titles), and the primary goal of the project is access, though we are also treating the scanned files as preservation files, in terms of formats chosen and how they are backed up (we're using Dspace).

That said, we are definitely not using the "high quality/access quality/thumbnail" model. In the end, we're only keeping one file for the text of any score or book that was printed after about 1870 (most of the materials we're doing are in the range 1870-1922). That is a bitonal (black/white) pdf scanned at 300 dpi and compressed sufficiently to keep a file of up to 64 pages under a size of 5Mb (so that requestors with modem access can get the downloads). I was unable to find a visual difference between 300 and 400 dpi on the printouts we make using the printer we've been using to make preservation photocopies, and the results we've been getting have been good. So far, for music of the 1870+ period, I'm convinced that a well-made bitonal scan does a better job of preserving the cultural content of the scores than a color scan that adds extraneous information about the paper deterioration. For earlier scores the impact of the rougher paper on the printing is such that the color information can be more important in reading the score easily; that's why we're waiting to do vault items.

For colored covers we use color pdfs (I like the way they store, view and print better than jpgs). I haven't yet come across anything among the materials we've selected that includes colored ink (like manuscript additions) in the score; we'll cross that bridge when we come to it.

A sample of a scan that a student made last week is at https://urresearch.rochester.edu/handle/1802/2942 . (In looking up that sample I discovered that for some reason some of the color cover scans aren't getting uploaded... have to look into that!) 

Meanwhile, we are outsourcing the scanning of vault items that we don't want to disbind for in-house scanning. For that, we will be ending up with a larger preservation file, generally color images, probably TIFFs, compression level not yet decided, and may involve some case-by-case, with a smaller access file, bitonal pdf when that gives a clean resulting score, otherwise color pdf. This latter group, which probably corresponds to the primary group of what lots of places are doing, is still on the back burner for me; I'm waiting for more options. I'm not yet satisfied with simple things like skipped pages and overcropping and over-despotting of images with our current vendor. I'd prefer to do these materials in-house, and do more work on specifications once we've got a cradle scanner.

Alice Carli
Conservator
Sibley Music Library   

> -----Original Message-----
> From: owner-padg@xxxxxxx [mailto:owner-padg@xxxxxxx]On Behalf 
> Of jeanann
> croft
> Sent: Tuesday, August 08, 2006 11:59 AM
> To: padg@xxxxxxx
> Subject: [PADG:1141] Scanning Specificiations for Brittle Books
> 
> 
> PADG,
> 
> I received a couple of responses to my question about scanning 
> specifications for brittle book/brittle replacement programs 
> and wanted 
> to thank those institutions.
> 
> If you have not yet responded and are willing to share your scanning 
> specifications for brittle books, please send this information to me 
> (jeanann@xxxxxxxx) by Friday.  I will put together a summary report, 
> redact any identifying information, and share with the entire 
> list next 
> week.
> 
> Many thanks!
> 
> Jean Ann
> 
> 
> Jean Ann Croft
> Preservation Librarian
> University of Pittsburgh
> 412-244-7522
> 
> 
> 
> 
> 
>