A proposal for metal type nick identification

Hi everyone,

I am the Faculty Advisor of the Shakespeare Press Museum, a working printing museum on the campus of Cal Poly University in San Luis Obispo, California.

We have 18 operating presses and about 400 metal and wood type fonts.

Our never-ending problem is cataloging and indexing our collection of type, some of which is so rare that we don't know what it is, where it came from, or what to call it.

I have been working for several months to develop a method for cataloging the nicks on the type we have, which I plan to use to help identify type that is not distributed correctly, or at all.

Acknowledging the potential weaknesses of the system (different nicks in the same font, for example) I would like to put this out to the Typophile forum for comment.

My idea is to scan the profile of a character in a font, then to make a digital record of the profile. Then, I would like to encode a series of numeric values for the size, position and shape of the various type nicks.

With this in place, it might be possible to "simply" look up the nick pattern, using the font database, and have the database respond with a list of fonts that qualify.

WHAT I WANT FROM YOU:

Please tell me if I am completely crazy (I am), and please provide constructive ideas if you can.

I have not begun to implement this yet. I am in the planning stage, asking if the typographic public thinks this is a good idea.

Any input from the members of Typophile would be appreciated.

Thank you,

Brian P. Lawler
San Luis Obispo, California

Bleisetzer's picture

I thought about the same problem, when I started to print and scan my Font Collection of lead letter fonts:
http://www.bleisetzer.de/schriftensammlung.html

Here I show a full character set with the following information:
font name
foundry
year of first produing this font
developer
Groups, using DIN 16518.
I upload a picture of the full character scan.

The next problem, I saw, was to find a solution for an internal database. This was, what the website archive could not solve.
And I found another way, which might help you.

I took Excel.
Included name, foundry, year of production, developer (and could, of course, add sizes, which I own, archive numbers of cabinets and case). And in the last column of the Excel file I wrote a link to the picture of the scanned characterset, together with a small picture of this it into the column.

The big advantage now is:
I can sort the Excel file in every way I want.
I can sort by developer, by groups. Or (in your case) e.g. sorting by all 14 p sized fonts or whatever. From the sort file I can generate a PDF, what I offer to my visitors on my website for downloading. The PDF (this is not ready yet) can be generated, so the link with the picture of the character set, can be choosed.

For my needs its a very satisfying solution.

Georg

Gus Winterbottom's picture

>My idea is to scan the profile of a character in a font, then to make a digital record of the profile. Then, I would like to encode a series of numeric values for the size, position and shape of the various type nicks

Have you considered adapting biometric identification (fingerprint, palm print, face, iris) software and hardware? All of those basically identify lines and curves and convert them into numeric values. A USB fingerprint reader might work just fine.

brianlawler's picture

I appreciate both comments. Thank you!

I think the idea of a fingerprint scanner is great. I wonder if it can be adapted to work with type?

The other thing I might be able to do is to have some brilliant computer science student write an edge-identification program for me, then we could just feed it scans, and let the database find similar edges.

Once again, thank you!

Brian

Syndicate content Syndicate content