1.  Click here to listen to this tune in MIDI format 
2. Click here to obtain the modern notation in pdf format

I. Genealogy of the project
 

Valuable manuscripts have been found in many places, but... Cuckney, in the Sherwood Forest?

First leaf of the MS.: Autograph signature of James Dowland, probably a descendant of John Dowland, the most famous English lutenist of the Elizabethan and Jacobean era.

In terms of material, and fairly quickly, the discussion started to focus on the Folger Dowland MS. for several good reasons: it is one of several important and well-known English lute manuscripts from around the turn of the 17th century that is not yet in print or in facsimile format (although it is available on microfilm). The nickname of the MS obviously adds an extra-aura of respectability in the lute world, but even without the gloss, that manuscript is truly one of the treasures of early music, replete with information on early music instruction and staple tunes from the period. Whether of not some of the pieces in the MS are in John Dowland's own hand, the first page carries the signature of a Mr. James Dowland, from Cuckney, Nottinghamshire. The hand-writing appears to be from the 18th century, or more likely the 19th century. You can find a description of the MS and an index in Julia Craig-McFeely thesis on the Dartmouth lute site. The coincidence of the name on the fly-leaf and the fact that many pieces are attributed in the MS to John Dowland seems to me overwhelming evidence that the manuscript was indeed at some point John Dowland's possession, perhaps a tutoring book for one wealthy patron's children.

The issue of color was also a factor - if not the primary factor - in focusing on the Folger Dowland MS, because of the multiplicity of hands that took part in the compilation and because of the prohibitive cost of traditional color facsimiles. It is clear for instance that many of the ornaments ("#" ) are in a different ink and therefore added either later or by another scribe.

B. The challenge

In the course of the discussion, I enquired with the Folger library about the cost of obtaining a reproduction and the legal feasibility of publishing a full reproduction of their precious manuscript. The costs involved seemed reasonable, and after a rough estimate of what was involved, I raised a subscription: if I could find enough subscribers to cover my estimated costs, I would attempt to get the job done... Within the first week, I had 25 subscribers. Not quite the 30 required to complete the subscription, but too many to back down honorably. Quite decently, the Folger library left the publishing responsibility on my shoulders. After 400 years, the MS is safely in the public domain as it should be.
 

C. Technical issues

From the first, the success of the enterprise relied entirely upon one crucial element: the quality of the reproductions I would get from the Folger. My mistake was to cut it cheap. I picked the slides over the transparencies, that cost about twice as much. Yet, the estimated cost of the project using transparencies would probably have required either a steep increase in the cost of individual subscriptions or an additional 20 to 30 subscribers, all of which would have made the enterprise impossible.
Two pages folio were photographed per slide, thus making every detail so minute that an extremely high resolution scan would be needed, resulting in huge files and some fuzziness at best. Worse perhaps, the two pages not being exactly level, one page was ever so slightly out of focus which magnified 500% means a very palpable difference.

Scanning is an art. That is to say the highest possible resolution is not necessarily the one that works best. Scan a fuzzy subject at 4000dpi and you get a very large fuzz. Drum-scanning is the best, but at $36 a slide, it's off the budget. You can get cheaper drum-scans, but the output resolution (what counts for printing) is ridiculously low. If you scan for a reasonable printing output - at 600dpi on output these days - file sizes go beyond the capabilities of an ordinary pc to handle. The art is: which corner to cut and how badly to cut it.

The actual scanning of the slides was done with an Agfa T2000XL at 2000 dpi on input and 400dpi on output. Settings were highest level of sharpening, zoom at 500% and high contrast. The resulting data was saved in compressed jpg format, resulting in an average of 11 megabytes per file, or 35 megabytes uncompressed.

An example of cutting corners:
This bit of tablature has been reduced from its original file size of 700K to just about 7K. This was done by reducing the size to 40% of the original and reducing the number of colors down from the infinite to 16. Further changing the format from jpg to gif - more efficient with a lower number of colors - took the final size down from 14.1K to 7.
Amazingly enough, the information is still legible on screen, including the crucial dots that indicate right-hand fingerings, but it will print very badly....

After three different attempts, including source input scans at 3000 and 4000 dpi (the latter one interpolated, i.e. the software imagines the last 1000dpi in the batch), I settled for the expertise of the University of California, at Riverside, department of photographic services, with scanning on input of 2000dpi and 400dpi on output. The files were generated in jpg compressed format, averaging 10 megabytes each on disk, and 35 megabytes expanded.
Clearly those files are unsuitable for Internet posting at this stage of the network technology. We must keep in mind however that technology evolves very fast, so that a few years from now those file sizes may well be downloaded within seconds rather than hours. For the time being, I am making the original files available to the subscribers on CD-ROM, and will produce smaller files with a smaller resolution for WEB purposes. Still, the average file will measure around 1 megabyte: not a problem for people with very fast Internet connections, but very, very slow for people with a simple 56K modem connection. I have already mentioned these points before, but they are worth stressing one more time: this project is an experiment, that takes into account a very fast changing environment. So I am projecting ahead  to the time when network connections will be very, very fast indeed.

 

Although the digital facsimile and accompanying documents are to be made freely available, some money needed to be raised to get it under way. I offered to raise a subscription with the following conditions: subscribers would get a copy of the facsimile on CD-ROM, thereby providing them with better quality files and cutting their downloading time to zero, as well as a printed copy in black and white of the manuscript. In exchange for this, they were asked to send a check for $30. I estimated that 30 subscriptions would be necessary for the project to take place. The following people (in no particular order) very generously agreed to lend their support

A. Rights on the manuscript

As I see it, the situation in this case is as follows: copyrights apply to the physical object. The Folger library owns the copyright on the manuscript, and they exercised that copyright by selling me the reproductions. The material itself - i.e. the music - is in the public domain, because the English law, which grants the heirs full copyrights, limits the duration of the copyright to as maximum of the lifetime of the author + 70 years. In the United States, the copyright duration may apply for the lifetime of the author + 120 years. Finally, there is a third level at which copyrights apply, and those are the copyrights I own over the reproductions. Please read the following statement concerning the limitations of use you may make of these reproductions:

B. Rights on the reproductions
I hereby grant you the right to use those digital reproductions for purposes of research, study, private enjoyment, and public performance, as long as those activities do not infringe upon any other person's right to enjoy the same and are lawful activities. You may print or post privately or publicly those images within the same constraints. It is highly recommended that any public use of the contents of these pages and images should be accompanied by a clear identification of their original source. The same restrictions apply to all secundary material, such as transcriptions of those pieces in pdf, MP3 or MIDI format.
I recommend the following links for more information on copyrights to those interested:
English copyrights law
American copyrights law

Julia Craig-McFeely thesis

My warmest thanks go to the subscribers who make this project possible in the first place.
I also wish to thank the Center for Bibliographical Studies & Research, at the University of California at Riverside, for helping host the files on their server.
My thanks also go to the U.C. Riverside department of photographic services, and particularly Steve for his patience and for allowing for constant communication between us during the scanning phase of this project.
I also thank the Folger library for safely keeping the manuscript for us and for providing the reproductions.