Notes on the scanning of Red Cartoons

Notes on access to the original documents:

As with much of the rest of the Riazanov Library digital archive project, I strongly prefer to either own or have total control over the material I am scanning, so that I can as needed handle the originals, unbind the originals, and employ the scanning equipment of my choice, at my liesure, in order to get the very best possible scans.

In this case I was able to purchase original copies of the first three books (1926, 1927, and 1928) in the series of four, from Bolerium Books of San Francisco (Thank you, John!), from Bibliomania of Oakland (Thank you, Daryl!) and one from Lorne Bair Books (Thank you, Lorn!). Thus I was able to remove the staples from these volumes, and scan them at my own facility, with my own choosen flat bed scanner, in the comfort of my home, one page at a time in a fashion that was both convenient and involved no harm what so ever to the pages themselves.

However, try as I might, I could not acquire a copy of Red Cartoons of 1929. Nor did Holt Labor Library (with whom I have a special relationship, and which is often willing to allow me all of the access to its holdings I desire when I wish to make a digital record of them, to an extent usually unprecedented and impossible with any other library) have a copy of it.

In the end, I used the obsolete (to use the adjective Tamiment Library's then staff librarian used to describe it), creaky, cantankerous and poorly-maintained Minolta PS7000 (original, "Mark I") overhead book scanner at Tamiment Library to scan their copy of Red Cartoons of 1929. This meant being unable to flatten the pages so as to get perfect copies, and also that the images to some extent sufferred from the jitter that afflicted it's infrequently-serviced scanning mirror motor.

However, by persisting and being very careful and patient, I ended up getting marginally decent scans.

Notes on scanning technique:

The vast majority of the roughly 366 images in these four books are charcoal sketches. A few are block print type images (some of Gropper's and Gellert's art, for example), and a few are traditional line drawing ink cartoons. Only two or three out of the 366 images have areas of what originally must have been actual gray, now represented in the printed books using half-tone technique.

After some experimentation, I ascertained that it was not hard to find a single acceptable black threashold setting that would work for scanning nearly all of the charcoal sketch and line drawing and block print art using single bit scanning. To be sure, I occasionlly tweaked the setting a bit, to kill print thru in some cases, or to bring out blocks of black or line drawings, but for the most part I was able to use a single setting to scan nearly all of the material in these books.

Although for many situations it's wisest to use gray scale or color scanning, because that is far more forgiving and can be fine tuned later in post processing (single bit scanning is for the most part NOT able to be fine tuned after it is done), in this particular case of charcoal sketches and line drawings, single bit scanning both can easily give very accurate renditions AND produces a genuinely more authentic image: The charcoal and line drawings consist of lines and blocks of solid black, boardered by pure white paper background. If one scans these with a gray scale technique, the result (when magnified) is a fuzzy gray outline around the black areas... something NOT present in the original images. I used 600 dpi for the scans, the highest true ("raw") optical resolution my Epson GT-20000 scanner would permit.

I blew up scanned images 10 fold on the screen, and compared those portions to the same part of the image on the original paper, viewed thru a 10x triplet magnifier. The correspondence was extremely good at very fine levels of details.

It is true that sometimes my chosen settings to best render the art resulted in the rendition of the small amount of printed text with the images... especially where small and fine italics were used... to be just a bit thin, with some dropouts in the letters. Never to the point that in the slightest compromised human read-ability, but at times to the point that would compromise the ability of Optical Character Recognition to work well.

The images often tended in fact to look better on the screen, and even more so when printed out on a xerographic printer, than they did in the orignial. This was the effect of slightly heightened contrast, and blacker blacks, and whiter whites in the screen and printed out versions, compared to the slightly faded ink and less than pure white paper of the orignals.

I am, however, satisfied that these scans are extremely faithful to the original images on the pages of these books. So are a number of others who I asked to make comparisions of the scans and printouts with the original documents.

The two or three images that had sigificant areas of gray rendered via half tone technique, the covers, and one or two other samples from the charcoal sketches I did render in either 24 bit color or 8 bit gray scale in addition to rendering them as single bit black and white scans. So a comparison can be made.

At the time these scans were made, computer monitor screen resolution is (and has been for the last many years) around 100 dpi or less. The scans are 600 dpi. So to fully appreciate the nature of these scans, one should print them with a xerographic ("laser") printer that offers at least 600 dpi (many xerographic printers at this time offer 1200 dpi).

It really is only when printed out that the quality of these scans can be properly and fully appreciated.


Martin H. GOodman MD, Director Riazanov Library Digital Archive Projects associated with Holt Labor Library associated with Marxist Internet Archive

Sept 1, 2012


The scans of the 16 page book / pamphlet of Fred Ellis cartoons about the Sacco and Vanzetti case were made at Tamiment Library using their old PS7000 scanner.

I used 600 dpi 1 bit Black and White scan settings.

The scanner seemed in particularly good health the day I made the scans, free of the jitter that sometimes afflicted it. And tho their scanner has no means of flattenning the document with glass plate (as is provided in the Mark II version of that scanner), the pamphlet itself in this case decided to be quite cooperative about lying RELATIVELY flat.

Because the pamphlet was not COMPLETELY flat (as was the case with the first three Red Cartoons books, which I owned and could unbind and scan with a flat bed scanner) the scans are not quite 100.0% exactly perfect, but they are nearly so, and most folks will not notice the very slight distortion that resulted from the pamphlet not being pressed totally flat the way it would have been on a proper flat bed scanner. To the extent that distortion is noticeable at all, it can (perhaps) be seen in the less than 100.00% perfectly straight lines of text at bottom of these cartoons, but even the text really isn't that "off" at all.


Sept 5, 2012