Re: Continuing Research

[Virginia Tech]

At 12:08 PM 6/12/97 +0100, L.Finch wrote:
>Dear All,
>Thank you very much to all those who answered my last query. The
>information was most helpful. Subsequent to those mailings, I have
>decided to record the watermarks by hand tracing, as no specialist
>equipment will have to be purchased.


I would urge you to reconsider your intent to do hand tracings. You may
think that it is best because you can get right into it with no equipment,
but you will find yourself increasingly frustrated, and then you will have
committed a lot of time for an inferior and unsatisfactory product.  To be
more specific --

You will be frustrated because your project, as i understand it from your
earlier post, will involve comparing watermarks. Hand tracings will
inevitably be just different enuf to leave you wondering if you've got two
watermarks that are similar, or two that are identical. You will have
similar questions about identifying pairs of watermarks -- are tracings A
and B come from papers from a pair of matched molds, or was I slightly
inaccurate in tracing one of them?  You simply cant get an accurate enough
result with hand tracings to answer those kinds of questions with
confidence, and NO ONE else will be able to trust your hand tracings a for
answering those questions, either.

Hand tracings will only preserve what you notice when you are tracing them,
and there will be elements of the watermarks that you won't notice,
especially with unfamiliar watermarks, and at the beginning of your work.
But afterwards , as you find other examples of the same or similar (?)
watermarks, you will notice details you hadn't noticed at the beginning --
maybe an indication of a knot, for example, or a characteristic bend or
break in a wire line, or a small feature of the design -- but if your
earlier records were hand tracings, you wont be able to check them for
anything you didn't happen to notice the first time, and you won't have
detailed facsimiles of the chain and wire lines. With Dylux prints or other
methods of making contact prints, you can go back to your print and it will
have recorded things that you didn't notice the first time, not to mention
the chain and wire lines.

Not to mention also that you can scan and enhance these images to bring out
such details.

Any project like this will involve analyzing your findings after -- even
long after -- you made your original facsimiles. Hand tracings will not
meet your needs as you start asking new questions about what you saw or
didn't see earlier.

The difference between a hand tracing and a contact print is like a
doctor's looking at someone's interpretive sketch of a broken bone vs
looking at an x-ray. With the first, you're stuck with the interpretation;
with the second, you've got "raw data" that you can make judgments about
and find new information in. (Pardon the dangling prepositions!)

Better by far to take more time at the start and do it right.

I don't mean to sound preachy -- it's just that I've been there, and ended
up redoing my earlier work for the very reasons I mentioned above!


>
>1) Would it be possible to provide me with information on the use
>of a digital camera to record watermarks? For example, what
>equipment is needed? How does the digital camera work?

I can send you some materials circulated at the Roanoke Converence about
it, and you can contact the presenter directly. The problem with digital
prints is that the watermark-bearing paper will be curved at times, or for
other reasons not flat in the plane of focus, and thus the image will be
distorted. Even if you include a millimeter scale in the image, this
problem of distortion won't be eliminated. The best option is a contact
print, in which you KNOW that the print reproduces the very size and
proportions of the original watermark.

>
>2) It has been recommended that I record chain spaces and
>wirelines, along with the watermark; how should that be done?
>Can they be recorded seperately, or should they be recorded in
>their postion on the watermark?

By "record" do you mean, "include them in your facsimile?"  If you are
doing hand tracing this means more work, and is difficult when the wire
lines are very close and fine. If you are making prints, the chain and wire
lines are included automatically.

If, on the other hand, you mean "record dimensions and measurements,"
The IPH Standard calls for the following (I include their data field numbers):

3.1.6 height of watermark in mm to be measured as a rectangle parallel to
the vertical axis of the watermark using a rectangular sliding rule

3.1.7 width of watermark to be measured as in 3.1.6

3.1.8 smallest horizontal distance between the watermark and the nearest
chain line on the left in mm

3.1.9 same, but on the right

3.1.10 smallest vertical distance between the watermark and the bottom of
the sheet in mm (if sheet is whole, I presume)

3.1.11 same, to top of sheet

3.1.12 Number of the compartment (i.e., space between 2 chains) which is
touched by the furthesst left point of the watermark. (presumes you have
numbered the compartments from left to right. Also presumes you have a way
of determing which is left and right -- IPH isnt clear about this, but
wants you to make your facsimile and do your measurements with the
wire-impression side of the paper facing down (3.0.17)





>
>3) How well can hand tracings be converted into digital form? I
>have access to a Hewlett Packard Scan Jet IIc/ADF, which can scan
>images and translate them into digital form.

Very easily, No more or less effort than scanning a contact print. but you
don't gain as much from it, since you can't see any more by enhancing the
image.

>
>4) What standards should I be working to?

Do you mean for making digital prints, or in general? In general, you
should get hold of a copy of the IPH Standard. For prints, it depends on
how you intend to use them. To make them available on the World Wide Web,
you will want to save your scanned images as GIF or JPG format, at 72 dpi.
For hard copy, you will want a resolution at approximately double the
"screen" that will be used in printing your images. You will have to talk
to the publisher about that. You would probably be fine if you did your
scanning at 400 dpi. But on the other hand, if you are only scanning hand
tracings, you don't need any resolution to speak of -- since you are not
concerned about losing any detail from your image.
>
>5) Is there any literature (in English) which details the design
>and excution of watermarks survey? I have a few references, but
>wondered if there were more.

You might contact Mark Sosower of the Department of Classics at North
Carolina State University. He has assembled a Repertorium of Papers in
Greek Manuscripts in Spanish Libraries. The Repertorium has identified some
2,000 types of paper, and has been accepted for publication by Hakkert. It
will include matched
pairs of watermarks based on machine generated and computer enhanced
tracings. The project was funded by an NEH Grant, so his grant proposal
would have all the information you would need. I think I have his e-mail
address if you want to follow up on this.
>
>Thank you in advance for your help. It is greatly appreciated.
>I look forward to hearing from you all.
> Yours faithfully,
>       Lorraine Finch.


/Bob Allison

Robert W. Allison
Dept. of Philosophy & Religion and
Chair, Classical & Medieval Studies
Bates College,
Lewiston, Maine, USA 04240

E-MAIL:         rallison@bates.edu
TEL:            (207) 786-6307
FAX:            (207) 786-6123