IMSLP Forums

So, I've noticed lately that a few scores from Harvard's Loeb Music Library have been showing up on Google Book Search, eg:

http://books.google.com/books?id=QHARAA ... frontcover

As well as a bunch of full scores of Glazunov Symphonies, and a few other interesting pieces, most of which are not on IMSLP. There doesn't seem to be an easy way to browse or search for them, but I've found a couple dozen ones so far.

So my question is, what is the best way to go about putting these up on IMSLP? You'll notice if you download the PDF of the file on that link above, that the first page is a long explanatory note about Google books and how they want you to use the file, but since these scores are PD, do we even need to keep that page in the file? For my own use I have been removing that page, the color scans of the front and back bound covers, and the numerous blank pages scanned. I'll sometimes keep the color title page, depending on how well the color came out in the scan and on how interesting the engraving is. I'll keep the B&W title page and whatever text pages there are after it, esp instrumentation lists. So, could the file as I have edited it be posted here with scanning credit given to Google?

I'd love to upload what I've found, but a little guidance on this would be appreciated.

The same applies to google as applies to CD sheet music.

The quality is sorely lacking in the google scans. Many times also it looks fine on the screen, but the pdf is messed up. So beware. For example here's elgars enigma variations which is missing pages in the pdf:

http://books.google.com/books?id=ICgRAA ... frontcover

Here's a link to the Hollis Catalog at Harvard. Be sure always to make sure that the "Digital Resources" tab is highlighted when you search.

Google is doing something very strange with the scans here. I've been aware of this collection for several months. Instead of doing an entire score in 600 dpi monochrome, they have odd sections that suddenly appear as 150 dpi JPEGs in grayscale, or even color - often a single system at random intervals. The scores tend to print out very poorly because of this.

There's also the somewhat absurd letter stating that public domain items must only be used for "non-commercial" uses, which is probably boilerplate emanating from Harvard or the other libraries whose collections have been scanned. Public domain is, after all, public domain - meaning there are no restrictions of any kind whatsoever as to how it may be used.

We're looking at ways these files might be processed in some way to make them more printable. It's a nice collection, with quite a few rare and unusual scores. I uploaded a couple they had of the American composers Chadwick (Tabasco) and Parker (The Skeleton in Armour). I tried doing some processing of the files on those so they would print better and would appreciate any comments about how they come out from others.

Google is also doing something very strange with the access restrictions here. I can't access both of the examples mentioned in this thread - there's simply no link to the full text, only the cataloguing information. Is Google doing IP-based access restriction?

EDIT: Apparently it is: http://books.google.com/support/bin/ans ... tx=sibling

It took me about half a minute to find a working web proxy located in the US which would allow me to download that thing, if I wanted to (I'm located in the EU). Which goes to show how pointless such filtering is. It's about as easy to circumvent as a locked door with no walls around it - no need to break it, you just walk around it.

So, from what I've seen in this thread, I'm going to go ahead and upload some of the better scores that I have found. One of the positive aspects of these Google scans is that many of the scores that I have found would have been far too delicate for me to scan on my own scanner and most of them, though I appreciate having, I would not have taken the time and effort to scan on my own. Anyways, expect to see a few of these come up in the next few days.

and BTW, if you go to the Hollis link above, the best way to find all of these is to click the Digital Resources tab, then click the expanded search link, go to the Format drop down box on the bottom right and select score. Then either enter something you are looking for, or enter something like "music" if you just want to see everything. Some of the links go to Harvard's page delivery service, in which case you just click the Printable Version tab at the top and it will make a pdf for you. To get the best quality available, make sure the image size is on the largest square.

Though imperfect, it's nice to have another source, especially one that has access to such an extensive and unusual collection. Now if only there was a way to request what they scanned next.

horndude77 wrote:The quality is sorely lacking in the google scans. Many times also it looks fine on the screen, but the pdf is messed up. So beware. For example here's elgars enigma variations which is missing pages in the pdf:

You can always insert the missing pages into the pdf from the web version. The quality is not the same - http://imagebin.ca/img/rDOxVI4.png (web version) vs http://imagebin.ca/img/pTOYmk.png (img extracted from pdf). The process could be automated, of course.