Revising scanning procedure, cataloging fix-ups

Today I finished (I think) revising the scanning procedure. I also started working on fixing up some of the records I had talked to Mark about. Still have more of that to do tomorrow, and this weekend I hope to get some work done on the SMC metadata stuff.

Hours today: 3 (4:15-7:15pm)
Hours this week: 3
Total hours completed: 80.5

More scanning

Today I looked a little further for the answer to the perplexing problem of not being able to change the default resolution setting for the scanner. I DID discover, though, that once you set the Custom Settings to Grayscale, 300dpi, all you have to do is click on Custom Settings again the next time you scan and it remembers. I kept clicking on the Grayscale and THEN Custom Settings and it would reset back to 150. But, if you go straight to Custom it remembers. Good news! I still need to edit the procedure.

I scanned a few more of my pieces and fixed something in a bib record. Next week I'll focus on fixing up the things Mark and I talked about and I'll also need to focus on some metadata stuff, once I get clarification on what my next step is (I feel really stupid that I can't remember what to do next!).

Hours today: 2 (4:30-6:30pm)
Hours this week: 7
Total hours completed: 77.5

Exporting Balmer & Weber records

Tuesday, Sept. 19

On Tuesday I spent some time running lists to try to get only the Balmer & Weber records that have URLs (meaning the digitized versions have been linked). I came up with 218 records which seemed low to me, but I think I just had in my head that about half of the B&W stuff was digitized (over 600 total, so I was guessing about 300 had been digitized). I exported the .out file and sent it off to Cassandra who will run it through MARCEdit to turn it into MARCXML. Then comes the fun part of piecing things out and comparing and making sure we have what we need and have it in the right fields.

Hours today (Tuesday): 1
Hours this week: 5
Total hours completed: 75.5

Metadata, cataloging, scanning...

Monday, Sept. 18

On Monday I met with Andrew and Cassandra and Mark about metadata options and how to get from here to there. I finally feel like I have somewhat of a better idea of what needs to happen next. I went ahead and created a list of Balmer & Weber records that have URLs and it came out to about 218 records (fewer than I expected, but that might turn out to be a good thing!). Cassandra will run it through MARCEdit and turn it into MARCXML, and then I will do some comparisons and see how things match up with DLXS's BibClass metadata (which appears to be very minimal) and with DC that is required for OAI harvesting. I just realized I never did hear back from Stephen Davison whether or not they'd be willing to harvest such a small number of records. If not, at least things will be set in motion for the future and I will have learned a heckuvalot in the process!

I also met with Mark to go over some cataloging questions. There are some things I need to look up and fix, especially with the "Taming of the Shrew" pieces.

I also met with Brad to go over scanning standards and procedures. We agreed that if it doesn't take too much more time to go with 300dpi, we might as well. I can't figure out, though, how to have the default resolution settings at 300dpi - the way it is now you have to manually change it every time you scan a page (it won't even remember your manual setting from one scan to the next) - major pain! I'm hoping Brad can find the user guide for the scanner, but I'm not sure even that will help. I went all through the Photoshop Help to find an answer but found nothing. As far as I can tell, it has to be changed manually. I thought I had found a way around it (using the other 'Import' option, which DID for some reason have 300dpi as the default), but it automatically saved the file as a bitmap file that looked absolutely terrible.

Hours today (well, Monday): 4 (2:30-6:30pm)
Hours this week: 4
Total hours completed: 74.5

More cataloging

More cataloging today. Definitely starting to go faster now! Nothing terribly interesting to report. I need to find out more about this "Royal Musical Repository" that seems to be stamped on a lot of title pages (and sometimes it shows up in the imprint area and/or with the pl. no. on each page).

Hours today: 3 (1:45-4:45pm)
Hours this week: 8
Total hours completed: 70.5

Cataloging

Interesting quote from the The King's Theatre Collection book I found:
"The King's opera companies, even during the best of times, staggered from season to season, surviving on box subscriptions and marginal trading, which included running coffee rooms and selling publishing rights to librettos and "Favourite Songs". The latter, which occupy the lion's share of this catalogue, were fodder for the fashionable private and semi-private musical soirees of the gentry and connoisseurs. In the late 18th and early 19th centuries, the favorite song market was fuelled largely by the rapid development of the domestic piano for middle class music-making. Opera songs constituted a significant proportion of the total London book trade. A popular aria from the latest opera or a perennial such as La Buona Figliuola could sell thousands of copies. This potentially lucrative market led to frequent and bitter copyright disputes among composers, publishers, singers and opera companies" (x-xi).

Gives interesting insight as to where these things I've been cataloging came from and how they were used. I'm sure it was quite common for people (such as E.A. Foster, who was presumably the owner of the bound volume of songs I've been cataloging) to have their own "Favourite Songs" bound together in convenient volumes. I can't imagine it would have been very convenient to actually play or sing from these bound volumes, though. I wonder how that worked?

Cataloged a few more pieces in the bound volume today. Starting to get a few lithograph drawings on the title pages which makes things a bit more fun. One was a duplicate of something already cataloged in the regular catalog (in a music spec. bound volume), not in the suppl. catalog, so I'll need to ask Mark how to deal with that.


Hours today: 2.5 (4:30-7:00pm)
Hours this week: 5
Total hours completed: 67.5

Scanning standards, more cataloging

Today I spent some more time looking at scanning standards and experimenting a bit with different resolutions. Here is a synopsis of what I discovered about other institutions' (a select number and with varying degrees of specificity!) sheet music digitization standards:

1) Duke (Historic American Sheet Music) - all scanned at 150 dpi with 75 dpi access images and thumbnails produced later by an automated process

2) Detroit Public (E. Azalia Hackley) - 300 ppi (ppi is more accurate than dpi, since we're really talking about pixels per inch on a monitor instead of dots per inch when printing, but generally people just use dpi anyway), 24-bit RGB color; access image 72 dpi, 600 pixel width; thumbnail with standard height of 150 pixels generated using IrfanView (I'd like to look at IrfanView further at some point, just to see what it can do)

3) 19th Century California Sheet Music - 400 dpi

4) Brown African American Sheet Music - 300 dpi master; high-res download (JPEG, 1650 p tall); printing (JPEG, 860 p tall); page-turner (JPEG, 400 p tall). Brown claims that 300 dpi (as opposed to 150 dpi) made a significant difference in detail and color fidelity, particularly in the printing version.

5) LC Music for the Nation - bitonal TIFF first at 300 dpi and then later switched to 400 with Group IV compression; 200 dpi grayscale - "tonal capture was able to suppress some of the print-through" (print-through was one problem I was coming across in scanning the pieces from the bound volume, but I had trouble finding more information about this "tonal capture")

6) Sheet Music from Canada's Past - 300 dpi, 24-bit RGB for color covers; 600 dpi for Line Art (all pages, including cover, scanned as bitonal images) - this one didn't really make sense to me, need to find more info about bitonal images

7) Indiana University Sheet Music Collections - 300 dpi, 24-bit RGB; for interior pages and covers printed in b&w - 8-bit grayscale; IN Harmony: Sheet music from Indiana collection scanned at 400 ppi

8) UC Boulder - 400 dpi, copied as JPEG for web site, music converted to grayscale for web (better image and print copy)

So, the general consensus seems to be that 300 dpi is sufficient for sheet music purposes. In my experimentation, though, I failed to see a huge difference that would make switching to 300 dpi worthwhile. Though I did not experiment with printing the PDF images (I didn't want to waste too much ink!) - doing so might change my mind, but I still can't image that the difference would be THAT remarkable. I did see a big enough difference, with the particular things I'm scanning, between black & white and grayscale that I highly recommend using grayscale for older, discolored items that have some bleed-through. The black and white scans would certainly print better (more quickly and using less ink), but I think the quality was really not acceptable in this case. I saved examples of each type that I can show Brad and we can compare. Again, this is just for one type of sheet music - I don't know that I'll have time to experiment with any other things (like color covers, etc., but I believe the procedures for that already call for scanning at 300 dpi, so I expect that that is quite sufficient). Brad had mentioned that if at all possible he prefers actual music to be scanned in black and white since it will print better, which I agree with. The problems come when you have older paper with lighter ink, bleed-through and such.

I looked again at Howard Besser's (Getty) Introduction to imaging - he talks a lot about the need to find the right balance between getting a high quality image and the file size. The better the quality, the bigger the file size and the harder it is to manage (store, load, display, print).
Though of course one usually creates a large, high quality, uncompressed master file (usually a tiff file) and then one can create smaller, more manageable access files from that master. One can create access files for viewing, for download, for printing, thumbnails, etc. He goes into great detail about image reproduction and explains resolution and compression quite clearly. The details and issues surrounding color management are very complex.

I like this quote in the section on Image capture:
"Digitizing to the highest possible level of quality practical within the given constraints and priorities is the best method of "future-proofing" images to the furthest extent possible against advances in imaging and delivery technology. Ideally, scanning parameters should be "use-neutral," meaning that master files are created of sufficiently high quality to be used for all potential future purposes. When the image is drawn from the archive to be used for a particular application, it is copied and then optimized for that use (by being compressed and cropped for Web presentation, for instance). Such an approach minimizes the number of times that source material is subjected to the laborious and possibly damaging scanning process, and should emerge in the long term as the most cost-effective and conservation-friendly methodology. "

I also did a bit of cataloging today. Note to myself to remember to pick up aforementioned book at WC tomorrow.


Hours today: 2.5 (4:15-6:45pm)
Hours this week: 2.5
Total hours completed: 65

(I need to get in about 6 hours per week in order to finish by the end of November (shouldn't be a problem), so I'm starting to keep track of hours per week as well as per day and total)