Scroll down to read about digitizing microfilm. Consulting the critical reviews of the Film-Kurier and including them in the database.
Recent ideas on a similar subject


6/18/2018 \\ Development


In order to better host the GSRM / DResSUP digitization project of Film-Kurier reviews, I decided to learn WordPress. I then proceeded to conceive a new “feel” for the Weimar Talkies Project.


Visualizing the Film Productions of Weimar Directors

3/2/2017 \\ Development

The most basic data having been entered into my database, I decided to try and visualize it in a meaningful way. The idea of creating a timeline was rather self-evident to me.

One of the things this shows us is the pace and the level of productivity the directors achieved during the Weimar sound period.

Having each film displayed as a spot allows us to view these productions from a distance. It also gives biographical information by telling us whether a director worked early on or started later, whether they were able to make a lot of films or whether, for one reason or another, they stopped working after one or a couple of talkies.

I downloaded my dataset from Airtable as a CSV file and plugged it into Aeon Timeline. The result could be exported as a PDF file (part of which can be seen in the screenshot above).

More than just giving us a clear picture of the director’s filmography over time, I started thinking about ways of enriching this visualization by further editing it.

Color coding the films in terms of Production Companies seemed important to me. This could show whether a director was working for more than one company or whether he may have been under contract with only one, such as Ufa. Every diamond shaped spot could therefore get a different color.

Each spot could then be made to link to an IMDb entry of the film. This would allow someone to actually know what film the particular spot stands for. But what kind of question could one ask the visualization?

Screenshot from the binary (Y/N) question in the “People” table of the research database relating to the projected visualization.

The research question at the forefront of my mind concerns “Jewishness in Weimar Cinema,” as this is the topic of a seminar I will attend in Fall 2017. For this question, religion plays a big part. I therefore thought about placing all the Jewish directors in bold.

Next there were certain questions that relate to giving a general idea about the directors in terms of the larger history, using different symbols/colors for marking:

1. Silent Period Activity

2. Exiled Filmmakers

3. NS-Regime Activity

Already, the visualization – with a couple of simple changes – would be starting to tell a story, giving you a distant reading about the directors and their place in a larger historical development. You could therefore click on one of the films with a certain preconception of the relation it has to the directors who made them.

One of the issues that I still need to solve is co-directors. When two directors work together, they get a separate section in the visualization above. This is not very helpful and I need to find a way of showing the co-directors without having them be mentioned separately. Maybe I can have the directors in question be featured next to each other and have the spot appear on the line between the two of them? Not ideal… But something to think about.


Failed Thick Mapping Attempts

3/1/2017 \\ Experiment

The idea was to map 1931 Berlin, paying close attention to where the theaters were located in which the talkies premiered. The list of films that premiered in the respective theaters can be found in the Film-Kurier-Index 1931 (Edited by CineGraph Hamburgisches Centrum für Filmvorschung and Deutsche Kinemathek. Hamburg: CineGraph, 1991). Then, I inserted this information into Google MyMaps and downloaded an exportable file from there. Background knowledge about the theaters which would later help in articulating specific questions about the theaters is provided in Janet Ward’s book Weimar Surfaces: Urban Visual Culture in 1920s Germany (Berkeley/LA/London: UC Press, 2001).

Before I could do this however I needed to get a map of 1931 Berlin. I downloaded it from HyperCities. The problems really started when I tried to put it all together. For some reason, no matter how often I tried or what kind of different tricks I tried to pull, all I kept getting were crosses on top of Google Earth indicating my failure to accomplish this task.

A thickmap of 1930 Berlin was created for the official website of Tom Tykwer’s TV series Babylon Berlin. Accessible online.


Beyond the Shadowy World of Microfilm:
The Media History Digital Library

2/27/2017 \\ Resource

A grand-scale DH project that started as an open access initiative led by David Pierce and Eric Hoyt (University of Wisconsin-Madison) now presents a rich array of digitized collections of classic media periodicals. Rather than scanning microfilms, MHDL scans the surviving periodicals themselves and makes them available to the general research community.

The added feature called Lantern allows you a detailed search through the collections. Everything is searchable, high resolution and in color. It is an incredible resource and the best thing about it: It’s constantly growing.

In my specific case, I would have been interested in seeing some periodicals from Germany issued between 1929 and 1933. Two sections may be of interest here: Universal Filmlexikon (1932-33) from Germany and Close Up (1927-1933) from Switzerland. Both of these are part of MHDL’s Global Cinema Collection.

Here is a comparison in terms of digitization quality between microfilm and MHDL’s scans:


Film-Kritik. OCR

2/3/2017 \\ Development

An exciting new development for the Weimar Talkies Project is the first digitization test of the Film-Kurier critical reviews. The good news is: it works. The bad news is: it takes time.

Outset: Filling in the database so that it can serve as a “digital notebook” for scholars of Weimar sound cinema was going to be a time-consuming task. There was no doubt about it.

Current state of the project: As of today, the basic information for the films has been filled out for about 30% of all movies concerned. This includes titles, release dates, cast & crew credits, production companies involved, etc. The bare minimum.

The next step: Once all the people involved in the films will be included in the “Film” section, it is a matter of one click to generate a list of these in the “People” table which would be linked to all the projects they were involved in and in what capacity. Then, the second wave of filling out the database would start with relevant information about these people, the essence of which will be extracted from the Cinegraph Lexikon zum deutschsprachigen Film. We will see how much relevant information we can find there to fill out all the fields that I have set up in the People table.

Meanwhile: Another section of the “Films” table is dedicated to writings about each talkie. This is meant to reflect the written conversation that has led up to the current state of scholarship about each particular film. This section is designed so that one can enter into a dialogue with the critical material by directly commenting on the various citations (arguments, points of view, …) from books, articles, essays, etc.

Where to start? My personal opinion, and I am aware that this point may be contested, is to begin with what critics thought of the movies at the time. The reason is two-fold: (1) It gives the scholar insight into the culture, society, politics and tastes of the Weimar era in various degrees; what may have been striking at the time but has now become increasingly irrelevant to a modern viewer, (2) It represents to my mind the first recorded reaction to each film in terms of critical thinking. Contemporary film reviews may not be up-to-date reflections or academic but all the more revealing and worthy of comment.

The critical reviews of the Film-Kurier: Arguably the most important periodical published about film in the Weimar years and geared towards the industry, the Film-Kurier featured articles about the development, production and release of films, focusing primarily on the domestic German market. In addition to such reports, it also published film reviews about the movies. Frequent contributors are sometimes still known today: e.g. Lotte Eisner.

State of the data: The issues of the Film-Kurier survive in the form of microfilms of which UCLA own a copy. The UCLA YRL also owns microfilm scanners which are available to students free of cost. It is therefore possible to sort through the issues and blow up the reviews and save them as JPEG, PNG and PDF files. Having thought about the prospect of digitizing these reviews for a year now, I have made use of the multiple volumes of the Cinegraph Film-Kurier-Index to locate the issue that contains each of the reviews making it easier to navigate through hundreds of these.

Step 1) Scanning: Digitizing microfilm requires that you manually find the right spot on a filmroll blow up the picture and focus it until you get a readable image. In the case of the Film-Kurier, one column of text requires 2 or 3 blow-ups which are then each saved as an individual scan. One review may contain up to 3 columns, making a number of up to 9 scans per review in order to be working with usable data in the next step.

Step 2) OCR: The film reviews are written in Fraktur (Gothic type) making them impossible to be decyphered by most commercially available OCR systems. Thankfully, there is an open-source method of getting a relatively good result. This tutorial shows you how. While OCR software such as ABBYY FineReader and Adobe Acrobat have become extremely efficient at extracting text from files, the method used here is far from foolproof. This is due to the quality of the microfilms but also the added difficulty of the font. Which brings us to…

Step 3) Proofreading: The text requires a lot of cleaning up, which involves reading it side-by-side with the original scan and editing it.

Here are the results from my first test for the very first all-talking film in the database of Weimar talkies:


FK 291123 Sa (279)

FK01 – Jäger, Ernst – Dich hab’ ich geliebt

(Walther-Fein, Rudolf 1929)

UA: 291122 – Capitol, Primuspalast

»Dich hab’ ich geliebt.«
(Capitol und Primus-Palast.)



So das Ende der Uraufführung:
Seine weißen Fahnen hißt das Publi-
kum — die Taschentücher.
Diese Rührung, diese Rührung!
Vater heult. Mutter schluchzt, Kinder
wimmern unter Tonfilm-Trümmern; wo
alles weint, muß Aafa mitlachen. Hoch-
konjunktur in Rührung — — was sich da
vorher abgespielt hat, alles zerschmilzt in
einem herzlichen hingebenden Geflenne.
Die Erfolgs-Chronisten stellen fest: Die
Höhe des Tränenniederschlags hat selbst den
sonny Jolson um eine Portion Millimeter
geschlagen. Die Theaterbesitzer weinten vor
Freuden mit: Damit erhoffen sie
das große Tonfilmgeschäft.
Und den Riesen-Familien-Operettenfilm
mit den trotz aller Anfangsmängel be-
zwingenden Mitteln der neuesten Technik.
Die Opapas des Films können beruhigt
sterben: man hört auch in der Tonfilm-
Wiedergabe die Birch-Pfeifereien.
Müssen wir uns des Genres schämen?
Lachender Film, Witzfilm, Possen,
Situations-Lustspiele, wenn man sie im
Kino sprechen hören wird, . . . welche Aus-
ruhstunden für Millionen werden da ge-
schaffen! Für nichts sollte man heftiger
Talente erkämpfen als gerade für das
tendenzlose Unterhaltungs-Genre. Die
besten Begabungen her dafür!
Die stummen Aafa-Leute haben in
solchen Unterhaltungsfilmen Erfahrung.
Merkwürdig, daß sie im optischen
Teil ihres ersten Tonfilms — Oberleitung
Rudolf Walther Fein — sich nicht ganz zu
ihrem Humor bekannt haben. (Der Dreh-
buch-Verantwortliche: Walther Reisch.)
Ein Ansatz und ein Beweis, daß sie es
können: die Gurgelszene eines Tenors beim
lever. Zum Kaputtlachen. Welche Possen-
Atmosphäre! Geräusch, Wort, Lied-Andeu-
tung — — wie das harmoniert. Völlige
Lebensillusion ist verblüffend geschaffen.
Daneben der »Ernst« des Films. Das
bedenkliche Schauspiel einer unverstandenen
anständigen Frau. Stumme Monolog-
Situationen, die man nicht begreift. Seelen-
kämpfe einer Ehefrau und Opernsängerin
in Großaufnahmen. Mit dem beim Ton-
film chronischen Zerdehungstempo.
Und so ungeschickt gezeichnete Figuren.
Ein Ehemann, der den Theater-Partner
seiner Frau kaum kennt, und die Frau —
und der Tenor — —; ein paar Ueber-
legungen, und alle diese Figuren wären
lebensmöglich gewesen. So sind sie nur
Denn man hört nur auf ihre Tonfilm-
Aeußerungen, läßt sich nur von der Tonfilm-
Bereicherung fesseln. die durch die Aafa ein
kräftiges Stück vorwärts getrieben ist.
Die Tonfilm-Elemente sind ein Fort-
Trotz der Zufälligkeiten der akustischen
Wirrnisse, die auch für dieses Tobis-Produkt
typisch sind. Das Ton-Geräusch- und Sprech-
durcheinander. Das ungleiche Mixen des
Herrn Hans Conradi. Ton-Produktionslei-
tung: Rudolf Schwarzkopf.
Die Dialoge. Kein Mitbestimmer des
Films hat besonderen Sprachgeist, ein künst-
lerisches Organ für Wortmelodie. Man läßt
Sätze durch, Schwerhörigen in genauer Arti-
kulation zugedacht. Läßt andere durch über-
kopierte Musik verschlucken. Andere Dialog-
teile bleiben ganz stumm. Die Lispelfehler
sind wohl Kopienmängel.
Die nachträglich zugebrachte Musik wirkt
harmonischer als die gleichzeitigen Bild-Ton-
aufnahmen. Den hohlen Klang solcher
Szenen vermeidet man nicht ganz. Der un-
verschleierte Ton ist noch die Ausnahme.
Das Knacken beim Ton-Schnittwechsel stört.
Aber man ist doch auf dem Wege.
Da spricht ein Kind mit berlinisch ge-
wachsenem Schnabel. Nicht einmal eine Be-
gabung, sondern abgehörte Natur. Wun-
derliches Wunder — es packt. Wie’s sein
Gebet hinschnaddert oder die Geschenke am
Geburtstagtisch zählt (Maria Canradi).
Da hat auch Mady Christians ein
paar ganz echte Töne. Wenn sie schmollende
Komödie spielt. Darauf sollte sie ihre Ton-
film-Karriere bauen.
Denn sie hat die größten Chancen: über-
sieht sie doch hier als erste deutsche Diva die
Klippen des Sprechens und Singens. War
sie im »brennenden Herzen« mimisch außer-
ordentlich gewachsen, so gibt ihr die Sprache
neue ungeahnte Bereicherung.
Walter Jankuhn befriedigt, wenn ihm
der Film gestattet. den Schritt vom Tenör-
lichen zur Parodie zu wagen.
Larmoyant und nuancenlos Hans
Stüwe. Kein Gegenspieler für die
Manche nennenswerte Episode. Carl
Platen, Trude Berliner, Andre Pilot,
Sophie Pagay.
Der musikalische Teil technisch an-
sprechend gelöst, wenn auch Schmidt-Boelke
nicht mehr zu geben versteht, als eine
brauchbare Zweckmusik. Der Hauptschlager
»Dich hab’ ich geliebt« von E. May — ein-
schmeichelnd und geschickt verwandt.
Sehr geschickt zwei zwanglos eingefügte
Kabaretteinlagen, mit dem Negerduo und
der spanischen Tänzerin. Ein ganzes Pot-
pourri solcher wirksamen Tonfilmreize
erhöht den Wert des Films.
Eine durchaus selbständige Art der
Ausnutzung amerikanischer Anregungen.
Bemerkenswert, wie richtig die Aafa-
Männer den Schaucharakter des Tonfilms
berücksichtigt haben, ohne das Schauspiel
selbst dabei zu ersticken.
Das Ganze: Die erste Aafa-Eroberung
im tönenden Film. Die bisher bunteste
Leistung im deutschen Tonfilm.
Ernst Jäger.
Länge des Films: 6 Akte, 2780 Meter.
Weiße Zensurkarte: Für Jugendliche
Verleih: Aafa.
Zu diesem Film wurde ein Illustrier-
ter Film-Kurier in der bekannten Aus-
führung hergestellt der von den Theater-
besitzern beim Verlag des »Film-Kurier«
bezogen werden kann.


Playing with Professor Unrat and Der blaue Engel in digital comparison

1/26/2017 \\ Experiment

Using the German DVD subtitles for Der blaue Engel (a comparison of the different DVD and Blu-Ray editions can be found here) and the Gutenberg edition of Heinrich Mann’s Professor Unrat, I decided to see what one could extract from the texts using the digital text analysis tool Voyant.

Sources: First, how did I extract the data? Using SubExtractor, I was able to create a .txt-file from the DVD. This was a time-coded document, which meant I still had to get rid of some text. Then, my “Der Blaue Engel.txt” file was ready. The next step was copying and pasting the entire text from Gutenberg without its editorial notes into a separate .txt-file which I named “Professor Unrat.txt.” However, I didn’t wish to stop there.

Extracting the dialogue from the novel: It makes sense to compare an entire screenplay with scene descriptions to a novel or a play. However, the subtitles of the film only give me the dialogue. This idea consolidated when the result of one of the most commonly used words in the novel was “sagte” (the German word for “said”) Consequently, I decided to compare the dialogues of the film and novel as well. I created a separate file “Professor Unrat dialogues.txt” and started cutting out all the non-dialogue sections from the text. It took me a while to figure out how to do that without going through the text manually. I finally ended up using RegExr with this command line.

Rendering: Once I pasted the texts into Voyant, I went to the settings button and put in German for the stop words list. But since Voyant did not get rid of all the German functional words, I had to put some of them in manually (words mainly from German slang “gehn” instead of “gehen” for instance).

Professor Unrat’s most frequent words were: unrat (60); fröhlich (40); professor (40); künstlerin (35); schüler (31). Der blaue Engel’s main expressions included: professor (36); lola (27); liebe (16); bühne (14); kikeriki (12). What is essentially doing is quantifying the text and then rendering it in interesting ways.

Using my corpus, the tool generated distinct word clouds for the film and the novel that we can now consider side by side (Unrat left, Engel right):


One of the most common words in both texts is “Professor.” Let’s consider how often that word appears throughout the progression of both narratives (word trend of Unrat in green and Engel in blue):


Commenting on the results and venturing interpretations: Another function of Voyant is that it can pull up key words in context. When we consider the sections of the two narratives therefore where the word “professor” comes up most, we could see in what way it is used in the novel as compared to the film:

– The section of the film in which the word “professor” is used most often is the morning after Rath has spent the night at Lola’s. First, Lola uses it extensively when they have breakfast, playfully making fun of him. Then, the students use it, adding to it the mockery “Unrat.” In a way, the word professor therefore links the two scenes, showing the audience that it is not used as a mark of respect but rather the exact opposite. Furthermore, the fact that Lola uses it in a similar but ambiguous seductive way but that experience being followed by the children’s blatant mockery suggests that Lola may have been just as ill-intentioned.

– In the novel, the word “professor” is used most by the artists when Rath comes to visit his flame repeatedly at the variété. They flatter his ego, all the while making fun of him slightly: “Er ist doch ‘n Doktor und ‘n Professor, und ich bin ‘n armes unwissendes Mädchen, was hab’ ich so ‘nem Mann denn zu bieten … Frau Kiepert, is es vielleicht nich wahr, daß ich das zu Ihnen gesagt hab’?” Here again the emphasis on the professor as opposed to the ironic self-portrayal as “a poor unknowledgeable girl” has both a disparaging and a seductive element to it.

In both cases, the use of the word “professor” is quantitatively increased when the setting (context) and the dialogue (register) are staged to be in contrast with each other, creating a distinct tension in the scene. The film emphasizes the tension by recalling to the audience the ordinary context and hierarchy associated with the word when the following scene shows the original framework of the students and the professor.

Of course I am only being playful here and not trying to construct a serious argument. However, it is interesting to experiment how distant reading can work here. More interesting, however, than what I did here, would be to do the same with actual plays and their adaptations such as Die Dreigroschenoper and Multiple Language Versions of films; on a very basic language level one could trace some of the changes and adaptations and interpret them.

The limitation for using Voyant for such work is that it focuses too much on word frequency and not so much on computationally comparing two works. There are most probably some features of Voyant that would allow some work into that direction. Nevertheless, it may be useful to just keep an eye out for a tool that would be more effective for that kind of comparative work.

A potential benefit of this kind of software is to use it in order to find keywords, words that can then lead to themes, hence subject headings for the database. Of course, all this remains on the surface of a text (as I hope my superficial interpretation above will have shown). Nothing, in my opinion really replaces close-reading.


Mapping Movements of the Cast and Crew

1/22/2017 \\ Development

Copyright: U.S. Holocaust Memorial Museum. No infingement intended.

The idea of using information gathered in the “People” section in order to create a map of the presence of artists active during the Weimar sound cinema era (1929-1933) had been on my mind from the very beginning. However, I did not wish to limit myself to the emigration of Jews forced into exile in the period following the Rise of the Nazis (as it is more generally detailed for all Jews in the map above). That was intended to be only a subsection of the movement I wanted to depict.

It is crucial to be aware, when assembling the information, what kind of end result one intends. In my case, this has only cristalized when going over the way in which a map for Black Travel Memoirs was developed by Dr. Stephanie Y. Evans and Dr. Moya Bailey. I wanted to visualize the artists various movements from their birth till 1945 in an animated map, in effect creating a timeline on which you would see the concentration of Weimar talkies artists in a particular location and their movements across the globe before, after and during the Weimar era.

Now that I am aware of what I want to visualize, this implies looking through the biographies of the artists to carefully enter a location or multiple locations for each year between their birth and 1945.