A. About The HistoryMakers The HistoryMakers is a 501(c)(3) not-for-profit corporation dedicated to preserving African American history through the creation of a national video archive of first person narratives of African American history makers, both well known and unsung. The HistoryMakers, the world’s largest video collection of African American oral history, is unique among other collections of African American heritage because of its massive scope, covering the stories of high-achieving African Americans from across a wide variety of fields. The HistoryMakers’ initial goal is to create an archive of five thousand interviews and make this important collection available to a wide audience. As of December, 2004, the archive’s holdings consist of approximately one thousand interviews.
1. Project goals With a grant from the Institute of Museum and Library Services, The HistoryMakers is collaborating with Carnegie Mellon University School of Computer Science's Informedia Digital Video Library to develop an archival search and retrieval system that will allow the display of digital video, audio, and text over local and wide area networks; and to create a database of four hundred digitized video interviews that are catalogued, indexed and searchable with the Informedia software.
2. Processing of interviews for HM/CMU project For each interview in this group, the Beta SP videos are encoded as MPEGs and transcribed. A team of noble and fearless cataloguers (AKA ‘Project Fellows’) then process these ‘raw materials’--proofreading the transcripts, adding data to clarify and aid in searches, dividing the encoded video into short clips (“Segments”) and indexing at the level of the individual Segments. As a corollary to the cataloguing process, we are collaborating with outside advisors to build a thesaurus for African American subject matter; this in itself will also be an important accomplishment.
Project Fellows will be expected to have a decent grounding in African American history whether or not they have a formal academic background in the subject. While we hope that everyone will continue to increase his or her knowledge in this area by study outside of work, this is absolutely essential if you are not already well informed. You are strongly encouraged to read some of the works on The History Makers’ bibliography of African American history and other books or writings recommended by your fellow cataloguers.
3. Cataloguing guidelines
Of the standards and procedures inherited from some earlier work on this project, we kept some, discontinued some, expanded upon most and also added many new ones. As we have begun processing these interviews, we have found that new questions frequently arise that require refining and adding to our rules. There will also occasionally be changes in the software and in related work at <em>The HistoryMakers</em> that will affect the work on this project. Therefore, these guidelines should be considered a work-in-progress that will be edited throughout the course of the project as we continue to develop our processes and standards. Every member of the team should always feel free to make suggestions, which should be discussed with the group and MUST be cleared before being put into practice.
Even with definite standards in place, the means of achieving them vary. Beginning cataloguers should talk to those with more experience about what works best for them; they are encouraged to come up with their own new ideas and ways of doing things in order to achieve the maximum in both quality and quantity of work.
4. Assignments and schedule The Cataloguing Schedule is an Excel file in the folder CATALOGUING PROJECTFELLOWS on the server (in Newserver\D\Staff Files\).
a. Assignments The Cataloguing Director will assign interviews to each cataloguer and list these on the schedule—if possible, a month in advance--but this is a cooperative process cataloguers themselves can have some input into which interviews they get. Some specific factors are:
i. Connection between HistoryMakers
In certain cases it will make sense that someone who has already processed an interview with one HistoryMaker should also process the interview of another with some relationship to the first; they may be relatives, colleagues or have been closely involved in other activities. This will avoid duplicating research on subjects that may be mentioned by both of the HistoryMakers. While we will get input from others who are familiar with many of the interviews, this is something that each cataloguer also needs to keep in mind, looking ahead at the list of upcoming interviews to be processed, and volunteering for those with some connection to previous interviews they have catalogued.
ii. Requesting Interviews
You may also request other HistoryMakers about whom you are particularly knowledgeable and interested. If two cataloguers both want the same interview, they should decide which of them is best qualified to catalogue that interview.
iii. Trading Interviews By the same token, two cataloguers can trade interviews already assigned to them in the schedule if they each feel better qualified for the interview assigned to the other; however, any such trades should be cleared with the Cataloguing Director who will then make the change on the schedule.
iv. Changing order of interviews
You can change interviews from one week to another within the month, but let the Cataloguing Director know. Remember to note the length of the interviews to see what you’re getting yourself into!
b. Time Schedule We need to complete 400 interviews by September 2005. The deadline is actually September 30, but we are aiming at a deadline a month earlier to give ourselves some time at the end to make any necessary changes (such as additional subject heading updates to earlier interviews--since Subject Headings will be created over the course of the project. There may also be unforeseen circumstances that cause delays. We do not want to be frantically rushing to finish in the final weeks, so it is very important that we not fall behind schedule.
By joining this team, each of us is committing to work together to ensure the accomplishment of the project. In pursuance of this goal, each Project Fellow will be expected to catalogue eight interviews per month. However, interview length will be taken into consideration when making assignments; for example, you may be assigned one six-hour interview rather than two three-hour ones.
A. Intro Here is a basic list of the tasks involved in our work, with notes pointing you toward more detailed instructions and information. This is included for the purpose of having a more convenient list to which you may refer—not as a substitute for the complete instructions.
You may end up finding a different order of steps that works better for you, which is perfectly fine as long as the work gets done correctly and in a timely fashion.
2. Prepare transcript for Segmentor Sec. IV, p. 11
3. Start Segmentor and open transcript and video Sec VI p.14
4. Decide boundaries of a Segment Sec IX.A. p. 19
a. Exceptions for Slating & Favorites – see
b. Choose a section of the interview to be a Segment.
i. You may wish to select all the Segment divisions in advance.
ii. Others may prefer to do this “on the fly” – deciding on your Segments one by one as you work with the video.
5. Select transcript text for the Segment
a. In the Movie Transcript view, highlight the text for your Segment.
b. Hit Ctrl +C to move this text into the Segment Transcript box (in Segment Text view).
6. Set Video TIME IN
Start video playing a bit before your intended Segment; set TIME IN Mark just before (c. one second if possible) start of speaking. (See Section)
Note: If you have previously decided on an “Out Point”, you may want to just “fast forward” using the video bars and go ahead and set it now too before proceeding to edit the text.
7. Edit Segment transcript
NOTE: Some Project Fellows prefer to do the tasks in this section all at once in a single pass through the Segment transcript; others prefer to go through it once for each type of editing (finding it actually more efficient to separate tasks which require different types of focus and concentration.) You may come up with yet another method that works best for you for both speed and quality.
a. Audit Compare video and transcript, comparing to make sure transcript matches what is actually said.
b. Proofread Check and edit Segment transcript for style, spelling etc.
c. Research and add info.
8. Set a TIME OUT mark Let video play as you audit it. (You may have already set both In and Out points at the beginning
9. Add Title Type a Segment Title In the Title field.
10. Save
a. Save Segment When you are done with a Segment, click on “Update” button under the video controls. The title of the Segment should appear in the box underneath the video controls. NOTE: You can also save Segment changes by clicking on a different Segment Title in the Title List; a prompt will then appear asking if you want to save changes to your Segment, and you naturally answer ‘Yes’.
b. Save Project In addition to saving Segments as you go, you need to also save Projects. Each time you get to the end of a video file you MUST save your project as an XML file, and then start a new Project (XML file) for your next video. (Each video MPEG file must have a corresponding XML File.) If you forget to do this and just continue adding Segments from a new video in the same Project you will have Segments with time code information relating to the wrong video.
The first time you save the XML file, name it in this format: Bond_Julian_01, and save it in the same folder as the Video files: First LD on Newserver:/Video/[Initial]/Last name, First name].
To save more work on a project you have previously saved, select Save As. You can save it with the same file name but only by “Saving As”. By default it will save in the same folder.
c. Auto Save.
XML files are automatically saved in C:\Program files\Carnegie Mellon University\Video Segmentor
11. Add Annotations (Annotations View) Date Annotations. (See Section
a. Subject Heading Annotations. (See Section
i. Add existing Subject Headings
ii. Request new Subject Headings and add
12. Movie Attributions View
a. Write Abstract.
b. Fill out other fields
c. Calculate SMPTE Offset
13. Complete Interview Evaluation Form See Section XVII
14. Finishing procedure See Section XVIII
a. Enter dates Segment & Annotations are done in Cataloguing Schedule.
b. Your work will be FTP’d to Melissa for Quality Control.
c. Print out copies of the XML files for Julieanna and leave them in her box.
d. Make any corrections from Melissa or others and re-send XML files.
e. Enter “Date Completed” in Cataloguing Schedule.
Whew! You’re done!
III. INITIAL RESEARCH When starting a new interview, first familiarize yourself with the HistoryMaker’s life.
A. Archival Box Each HistoryMaker has a box for paper materials in the Archives room, shelved in order of Accession Number. (You can look up this number in the Filemaker Pro database, or you can look through the Accessions lists in the filing cabinet or on the server.) You are required to look at the HistoryMaker’s Archival Box before beginning work on the transcript. taking the time to look over the information and get an idea of what is in there can save you research time later - -for example, a name you are looking for may have been mentioned in an article in the box.) Anything removed from the Archives MUST be signed out. You may keep the box at your desk while working on the interview. Below are some of the contents of the boxes:
1. Questionnaire If the HistoryMaker filled one out, it is a very valuable source of information about family, school, career, membership in organizations, awards, etc.
2. Biographical information folder This contains copies of magazine bio pieces, other press clippings, resumé, sometimes list of websites with information about the Historymaker, etc.
3. Paper copy of transcript If there is a transcript in the box, take a quick glance through it in case someone may have made notes on it that might be helpful later.
4. Photographs: These same photos will be in the video you are cataloguing and you can usually use the same captions.
This contains subfolders for various types of material and information about the HistoryMaker--Bios, Photos, etc.-- as well as Transcripts, in which you’ll find the interview transcripts you will need. It is located at (First LD on Newserver:// HistoryMakersfromsctnserver/Oserver_Mac/HM WebDev/Individual HistoryMakers/(Initial/Last Name, First Name)/ ) An exception to thiss: HistoryMakers who were included in the “Pioneers in the Struggle” project about Black Illinois legislators have their folders in the “Group HistoryMakers” directory; Filepath is same as above up to HM WebDev/ then Group HistoryMakers instead of Individual HistoryMakers
(First LD on Newserver:// HistoryMakersfromsctnserver/Oserver_Mac/ HM WebDev/Individual HistoryMakers/(Initial/Last Name, First Name)/ )
If there are several different versions of the transcript, use the most recent one. Most of the transcripts are in Word and will have to later be saved as text files.
B. Preparing transcript for Segmentor There are a number of things that must be done to the transcript to make it ready to be loaded into the Segmentor:
1. Strip extraneous material:
a. Speaker labels Example:Interviewer: First, I want you to state your name. Julian Bond: My name is Horace Julian Bond.
Remove these labels (in this case “Interviewer:” and “Julian Bond:”) from text (easily done using Find and Replace from the Edit menu/Replace All).
b. Timecodes
Some transcripts have time codes throughout which will need to be removed. We thought that they must be deleted one at a time--but Susan has found another way. Here are her detailed, step-by-step instructions!
Susan Perry's instructions for stripping time codes and preparing transcripts The Simmons-Lathan Media Group provided many of The HistoryMakers’ transcripts. These transcripts have printed time codes in varying increments and have multiple tapes within a single MS Word document. In addition, once the time codes are stripped, you will discover that some sentences are split in the middle rather than ending with a normal punctuation mark. A conventional time code entry looks like this:
1-4 02:03:12:22
{ISAAC HAYES cont}
Kids were there, watched that Lionel train, you know, going around. And the store, the grocery store, was owned by a Chinese guy, *Paul *Lee, and we sat up an account with *Paul *Lee at the grocery store. And we had our own, you know, funeral homes and we 02:03:38:15
[1-4] 02:03:38:16
{ISAAC HAYES cont}
had our own businesses, and those were corner mom-and-pop sundry stores, we used to call them. They were on different corners and things like that. It was all in our own neighborhood. But life changed. My grandfather was working at a tomato factory. 02:04:02:27
Before you start stripping time codes, FIRST check to see if the Double Space button has been selected on the document. The Double Space feature gives the appearance of hard returns in the document when in fact there aren’t. If this exists, you must create hard returns between the Interviewer and Interviewee dialogue. To do this you will need to:
Select All. (Ctrl A).
Select the Single Space button.
Highlight {ISSAC HAYES cont}
Open Find and Replace (Ctrl H) and Paste it into the upper field (Ctrl V).
Select the More button to expand on the Find and Replace features.
Tab down into the lower field and select Special and Paragraph Mark (^p).
Paste {ISSAC HAYES cont} immediately following the “^p” command
Select Replace All. You will notice that it places a space directly above the text.
Repeat steps for {ISSAC HAYES}, {INTERVIEWER}, etc.
To strip the time codes and other non-essential text, you will need to:
Highlight any time code entry “02:03:12:22,” including any spacing that follows the number.
Copy (Ctrl C) - In this example, we are working with Tape 02.
Open Find and Replace (Ctrl H) and Paste it into the upper field (Ctrl V).
Select the More button to expand on the Find and Replace features.
Select Use Wildcards and replace the remaining six digits with question marks. For this example, it would be “02:??:??:??.”
Leave bottom field blank. The time codes are not being replaced with any other numbers.
Select Replace All.
After hard returns are created and time codes stripped, repeat steps 1-4, 6 and 7 to remove the non-essential text such as {ISSAC HAYES}, {INTERVIEWER}, and {ISSAC HAYES cont}. Remember to deselect the Use Wildcards button before stripping these items. Copy and Paste each portion of the document before stripping the next set of time codes into a separate MS Word document so you can easily locate where the time codes of the next tape begin. In all cases, create a Text Only file for each tape, giving them File Names as follows: Hayes_Issac_01.txt, Hayes_Issac_02.txt, etc.
Note: Due to some formatting issues, stripping the bracketed numbers “[1-4]” seem to be problematic. Although they cannot be stripped using MS Word, however, they can be stripped in the Text Only file. Open the Text document, “Hayes_Issac_01.txt,” and follow steps 3, 6 and 7. After stripping the bracketed numbers, scroll through the document and shift the text to form complete paragraphs, i.e.: “funeral homes and we had our own businesses, and those were corner mom-and-pop sundry stores, we used to call them.”
c. Other material Any other printed material (for example, the name of the transcription company) besides the words spoken during the interview.
d. “Uhs” and “Ums” You may also use Find/Replace at this time to delete meaningless vocalizations such as ‘uh’, ‘um’, ‘er’, etc. (It is a very different experience reading these than it is hearing them: it can be distracting for the reader and may detract from what the HistoryMaker is trying to communicate.)
2. Separate transcript by original tapes and save as text files
The original transcripts, usually in a Microsoft Word format, must be saved as plain text to go into the Segmentor. The transcripts will usually have the transcription from multiple tapes combined in one document, but will indicate where each tape ends. We need to save the transcription for each tape individually so that each video file will have its own transcript file. File name format is: Last name_First name_number.Ex:Bond_Julian_01.txt (= first tape in Mr. Bond’s interview).
3. Punctuation:
Some punctuation symbols created in a Word document prevent the Segmentor from functioning.
a. Types of punctuation affected:
i. “Smart quotes” (quotes turned inward as seen here)
ii. Apostrophes
iii. Ampersands (&)
iv. Em dashes: replace one long dash with two short dashes.
b. Changing punctuation marks to be accepted by Segmentor.
To correct this, it is not enough just to save the document as a text file—even though after having saved the Word document as a text file, the problem punctuation is now indistinguishable from how it would look if it had been typed in the text file- -for example, quotation marks no longer appear as smart quotes. BUT they still retain some Word encoding that the Segmentor will not accept, so the punctuation must also be replaced with new punctuation that was actually created in the plain text format. This is easily done using Find/Replace. Example: How to change ‘smart quotes’ to regular text ‘Segmentor-friendly’ quotes:
i. After you have saved the Word documents as text files, select Replace on the Edit menu.
ii. For Find: copy and paste an opening/left quotation mark from a quote; for Replace: type a quote mark.
iii. Select Replace All. You will see quotes being replaced. (Apparently by identical quotes.)
iv. Repeat the process to replace end quote.
4. Find/Replace mistakes in transcript If you have a transcript with many repetitions of the same mistakes in spelling or punctuation, you may want to save time by fixing them all at once in the text file before opening it in the Segmentor. Of course, be careful about including in this mass change something that should not be changed!
5. The transcripts are now ready for the Segmentor. Note: Editing the transcript (which will be detailed in the following sections) could be done prior to loading it into the Segmentor. However, since you will also be using the application for other parts of the process, we have found that it takes less time to combine the proofreading and editing with these other tasks. Therefore these guidelines describe the editing as being done in the Segmentor.
For details on specific buttons or fields, see section on using Segmentor
A. Informedia Digital Video Library = Database system allowing full content search and retrieval of digitized video, created by Carnegie Mellon University’s School of Computer Science, who are providing software and assistance with it to <em>The HistoryMakers</em>
B. Metadata
= “Data about data” --often used for information about electronic resources; in this case the data it’s “about” is the digitized video; the metadata is the transcript plus the other data we are creating (subject and date annotations, titles, abstracts, etc.), saved as XML files.
C. Project
= the metadata (see above) relating to a digital video file; this data is saved as an XML file. In our case each Movie is and from each thirty minute tape; the Segmented transcript and metadata is.
D. Movie = digitized video, consisting of several Segments (see below), and the associated metadata (see above). In our case a digitized “Movie” always corresponds to one thirty-minute Betacam videotape, which is the format used for The HistoryMakers’ interviews. For each original tape there is a JPEG video file and an XML file.
E. Segment = a shorter portion of a Movie (see above); a Segment is a searchable / retrievable /watchable unit in the Informedia (See above) system. We are dividing each Movie into Segments and indexing at this Segment level.
F. Segmentor = an application developed by CMU to manually divide Movie into Segments and enter metadata
i. Open from Start menu, Desktop shortcut or on hard drive: C:/Program Files/Carnegie Mellon University/Video Segmentor/Segmentor.
ii. Choose Action box appears. (See left.) Select Open Existing Project and folder for most recent project opens.
iii. b. Segmentor already running: From File menu select Open Project. (See top right.) If most recent project is what you want skip to 3.
2. To find other MPEG & XML files
MPEG and XML files for all the interviews are on the server, organized in subfolders (by initial of last name, then individual HistoryMaker folders) (See rt mid & below.)
Filepath for video and XML files for surnames A-L is: FirstLD on ‘Newserver’(F:)\Video\[Initial]\[Last Name_First Name]
Ex: FirstLD on Newserver’(F:)\Video\ D \ Davis_Tyrone
Filepath for video and XML files for surnames M-Z is: FirstLD on ‘Newserver’(F:)\Video\[Initial]\[Last Name_First Name]
3. Select Project file from folder
From the HistoryMaker’s folder, select the correct XML file; for ex., for Project related to first video of Tyrone Davis interview, select Davis_Tyrone_01 (See right.) Note: If you don’t see any of the Project files make sure ‘XML’ is selected in the ‘Files of type’ box .
4. The Segmentor will open the project you selected.
However, there are times when, even with a previously saved project, you will still have to ‘Open Video’ and/or ‘Open Transcript’ from the File menu--see next section.
5. Saving Project. Each time you have finished processing a video file you MUST save your project as an XML file, with name matching the video file name. The first time you save the XML file, name it in this format: Bond_Julian_01, and save it in the same folder as the Video files: First LD on Newserver:/Video/[Initial]/Last name, First name].
When you start work on a new video file, you MUST start a new Project file (XML (Each video MPEG file must have a corresponding XML File.) If you open a second video file but forget to start another project, and just continue adding Segments in the old Project you will be create Segments with wrong time codes.
See Section VIII for instructions on watching the Segment.
To begin working on a new interview video in the Segmentor you need to start a new Project and load the encoded video and its corresponding transcript into the Segmentor; this is done simply by using the Segmentor to open these files. It does not matter which one you open first.
1. To start new project:
a. If you are just now opening Segmentor, click “Start New Project” in small box. (See top left.)
b. If already in Segmentor, select “New Project” in File Menu
2. Open transcript in Segmentor:
a. Saved as text files You will have already saved the transcript as text files, one for each thirty-minute tape, as detailed in Section III.B.
b. Click on File/Open Transcript
(See picture of File menu at left.) Folder most recently accessed will open.
c. Find transcript file
i. Either select a transcript file from the folder that first appears -or-
ii. Browse for transcript of another HistoryMaker. Here is the file path:
FirstLD on ‘Newserver’(F:)/The HistoryMakers from sctnserver/Oserver_Mac/HMWebSite_Dev/ Individual HistoryMakers/Initial/Last Name, First Name/Transcript /Last Name_First Name_[*].txt]
ex: F:/Historymakers from sctn server/Oserver_Mac/HM WebSite_Dev/Individual HistoryMakers/D/Davis, Tyrone/Transcript/Davis_Tyrone_A.
d. Double click on transcript file name to open transcript in Segmentor. To see where it is, select the “Movie Transcript” tab.
Note: The Movie Transcript Window’s only purpose is as a ‘holder’ for the transcript while you copy and paste parts of it into the Segment You cannot edit in the Movie Transcript window. Editing can only be done in the Segment Text window. (See Section VI below.)
3. Open video in Segmentor
a. Select File/Open Video.
(See left.) The folder last accessed will open.
b. Find video file
Either find video from that folder -or- browse to find video file of some other HistoryMaker. MPEGs are in same folders as the person’s XML files
FirstLD on ‘Newserver’(F:)\Video\ Initial\Last Name_First Name Ex: FirstLD on ‘Newserver’ F:)\Video\D\Davis_Tyrone
c. Select video file. You should see a list of MPEG files. Ex:Davis_Tyrone_01 (See left.)
d. Video will be displayed in the Segmentor video screen.
1. General definition When a video is recorded, a Time Code(TC), either visible or invisible, can be recorded simultaneously, automatically assigning a number to every frame of video, so that a specific place in the video can be easily accessed later for editing or research. The digital video you will be Segmenting works in the same way- -a time code is permanently attached to that frame of video and does not change if you cut out part of the video before it. (This is unlike a VCR time code that is not associated with specific frames and just counts how much of a tape has played. starts at 00 when you start a tape playing.)
2. Parts of the Time Code In ‘04:04:13:29’ (a TC seen in picture below) the 4 pairs of digits mean:
Tape Number:
Minutes :
Seconds :
Frames
04:
04:
13:
29
a. The first pair of digits represents a tape number set by the videographer. Interview tapes should be assigned consecutive numbers, starting from 01: - - in the case illustrated here, this is the 4th tape recorded during this interview. In most cases that’s what you will see; unfortunately, in some interviews, especially the early ones, this procedure was not followed and the TCs may start with totally irrelevant numbers.)
b. The second pair of digits refer to the number of minutes from the beginning of the video.
c. The third pair of digits refer to how many seconds past the last minute.
d. The last pair of digits refer to frames of video (at 30 frames per second).
TIME IN Point: Starting time code for current Segment
ii. Running time code:
Current time code for Segment being viewed; changes as Segment plays
iii. Segment
TIME OUT Point:
Ending time code for current Segment
Functions of specific controls
b. Video navigation controls
Main video time bar Located directly under Time Code windows. Represents duration of an entire video from beginning (left) to end (right).
Pointer Along with the running Time Code, the Pointer indicates current point of video, moving toward right as video plays. Pointer is also a control: it can be dragged along the time bar to move video forward or backward to whatever point you want.
Dragging the pointer along the Main Video Time Bar works sort of like a hyper=speed fast forward/rewind. Considering that the Time Bar represents the entire video (in our case, thirty minutes), moving it just half an inch makes the video move several minutes back or forward . So you can instantly jump to any point in the video. But it’s hard to precisely control where the exact point will be. (This is why we have more precise tools; see below.)
Precision video time bar
This bar represents a shorter section of video than the main video time bar, and thus gives you more precise control. It operates rather like an analog Rewind /Fast Forward; as you drag and hold the pointer b to left “REW” or right (“FF”) you see the video moving at fast speed.
Frame advance buttons These give you even more precision in selecting an exact point in the video.
Reverse
Forward
back 10 frames
back 5 frames
back 1 frame
advance 10 frames
advance 5 frames
advance 1 frame
c. Video Playing buttons
Play Plays the video from any point in it
Pause Stops video; Play button changes to this symbol while video is playing
d. Segment setting & playing buttons
Set Segment TIME IN Point
Set Segment TIME OUT Point
Blue area within time bar Indicates location of Segment within Video.
Play Segment Plays only the part of the video in current Segment. Useful for checking IN and OUT points
Segment start Goes to beginning of current Segment and plays it.
A. Open the Project File/Open Project; select XML file. See Section V.A.—“Opening an existing Project”. Video ‘screen’ and Segment fields will be empty until you select a specific Segment.
B. Select Segment The titles for Segments in this Project, along with their opening time codes, are listed in a frame underneath the video window and controls. (See picture at right.) Click on a title to select Segment; the video for that Segment should now be visible, as well as the data in all Segment fields.
C. Play Segment Use video controls (detailed in previous section) to play and navigate within the Segment video. Either of the right pointing arrows will play the Segment video.
In order to make it easier for a researcher to retrieve and view only those portions of the interview they want, we are dividing the digitized video into shorter clips, or “Segments”.
1. Subject We are dividing the video logically, based on the subject matter discussed. Often this will mean that a Segment is one question / one answer, but sometimes a few short answers can be combined in one Segment or, conversely, a really long answer may have to be divided.
2. Exclusions There is no need to include every second of video in Segments. You may omit:
a. Pauses to make adjustments, re-fill interviewee’s water, etc.
b. Technical side conversations about camera, sound, lights etc.
c. Other side conversations (possibly omit)
Talk that is not part of the interview is particularly common at the beginning of tapes, when a conversation that started while the tape was being changed continues for a bit once the camera is rolling. Such ‘side talk’ may come from some unidentified person in the room or may be between the interviewer and interviewee themselves; if the interviewee is speaking, it is up to the cataloguer to determine the relevance in order to decide what should be included in a Segment.
d. Unaware of being filmed (possibly omit)
A related issue is that sometimes (such as at the beginning or end of a tape or after a camera break), the interviewee is speaking without realizing that the camera is on. While we do have releases from the interviewees that encompass everything recorded, this can present some ethical issues. We should be considerate and not include in Segments comments the person most likely would not have made had they been aware of being recorded - -for example disparaging remarks about another person. On the other hand, sometimes there is a significant comment or great footage recorded at such times- - like Katherine Dunham singing to a baby- -that adds to the overall portrait of the person. …. You will have to use your own discretion aided by that of your colleagues.
e. Inaudible interviewer
Sometimes in early interviews the interviewer was recorded at a very low volume- -in fact sometimes it is difficult to hear them at all. In such cases it may be better to just omit the question and start the Segment with the answer.
f. Long question
Occasionally an interviewer may talk quite a bit leading up to their question and- -if it would not significantly alter the meaning -it may be better for the purposes of the Segment to cut off some of the “lead up” and begin with the actual question. After all, it’s the HistoryMaker people want to hear. (This can require some very precise editing! See III.C below.
3. Length
a. Minimum
Ten seconds is the minimum length preferred by CMU; five seconds is the absolute minimum.
b. Maximum
There is no absolute maximum but try to keep under six minutes.
c. Norm
Generally Segments should be one to five minutes long. Exceptions:
i. Certain long answers
If the HistoryMaker gives a brilliant answer over five minutes with no logical dividing place, it is okay to create a longer Segment. But if the brilliant answer gets to seven minutes it will have to be “Brilliant answer pt. 1” and “Brilliant answer pt. 2”!
ii. Short answers from interviewee
Sometimes a HistoryMaker will give a short answer or a series of short answers on a subject not related to the preceding or following Segments. It should be a separate Segment despite its brevity.
iii. Short Segments that occur in each interview
In addition there are certain cataloguing conventions we have that require short Segments for certain parts of the interview: “Slating”, “Favorites” and photographs. But…
iv. Long photograph narrations
Photographs should each have their own Segment, so even if the description /conversation accompanying a photograph exceeds normal limits, you should not divide it unless it’s REALLY long- -like over ten minutes.
B. Starting Segments Click on “New” to create new Segment. Both the video and the transcript need to be divided into Segments; it doesn’t matter which you do first.
1. Go to Movie Transcript view As mentioned above (V.C.4.d.), when you open a transcript text file from within in the Segmentor, it appears in the “Movie Transcript” view which is strictly there as a ‘holder’ for the transcript from which you copy sections of it into individual Segment Text windows.
2. Select text for your Segment
To get a section of transcript text into a Segment text window, select the text you want from the Movie Transcript view.
1. Set Segment TIME IN Find the place where you want the Segment to begin - -the “In Point” - -ideally about a second before speaking begins. To do this you will usually start playing the video shortly before this point, and when it gets to the exact point you want to be the TIME IN, click on the Segment TIME IN button. . You will see a tiny triangle appear at that point below the Time Bar.
2. Set Segment TIME OUT [Whether you have let the video play as you audited it (see Section X – Auditing), or have “fast forwarded” it using the video bars] when it hits exactly the point you want (ideally a second or so following speaking, so that it will not seem abruptly cut off) click on the Segment TIME OUT button. Another tiny triangle should appear at that point below the Time Bar.
3. No overlapping time codes NEVER set up Segments to have “overlapping time codes” (do not set the TIME IN for one Segment any earlier than the TIME OUT of the preceding Segment.) This will prevent the video from playing in the Informedia player.
4. Use Precision Navigation Tools if needed. Be very careful to set TIME IN and TIME OUT points correctly. Make the Segment:
a. …not too short:
Avoid setting TIME IN too late- -missing the start of speaker’s words, or setting TIME OUT too early- -missing the end of their words.
b. …not too long:
Avoid setting TIME IN too early or TIME OUT too late- -accidentally including unwanted words or syllables at beginning or end of Segment .
c. …but ju-u-ust right:
For more exact editing use the smaller Time Bar and Frame Back/Forward Buttons - -which allows you to move as little as one frame at a time. Segmenting can be especially tricky where people are speaking very close together with no pauses between speakers. It sometimes will require finding a split second gap in which to set the IN or OUT. (See details above in the section on video controls.)
d. Checking TIME IN and TIME OUT points
You can use the “Play Segment” button to make sure your IN and OUT points are set correctly, since it will play only your Segment and you can easily hear if it’s cutting off words or including parts of extra words.
5. Make sure you really set both marks! Hint: if you do not see a blue area (like this ) in the Time bar for a Segment you have created, it means you have forgotten to set TIME IN and/or the TIME OUT point. The blue section of the bar shows where the Segment is in the video.
E. Save Segment: Some Important Info! Click on “Update.” Another way is to click on another Segment title and save when the prompt appears. You may want to wait until you are totally done with a Segment to save—it’s actually easier to avoid losing data that way. But at least it should be mentioned in this part. of In addition to saving Segments by Updating them when you are done, whenever you are going to close the Segmentor you need to save the Project. See Section XVIII.
X. AUDITING: COMPARING TRANSCRIPT TO VIDEO
Listen and read the transcript carefully as the video is playing to make sure the transcriber heard and typed all words correctly, did not add extra words and did a reasonable job of punctuation to catch the meaning and emphases of the speaker. Things in transcription to note:
1. Auditing
The transcriber will usually have noted some words or phrases as “unclear”, “unintelligible”, “inaudible” or “?” Listen carefully--you may find that you are able to understand it even if the transcriber did not. If you can’t understand what a HistoryMaker is saying, you may ask other cataloguers for help; in fact, you should do this if it is important for comprehension of meaning.
In some of the early interviews the interviewer was not miked and it is often hard to hear them. You should try to get it right, but clearly this is less important than the words of the interviewee, and it is not something you should spend a lot of time on.)
2. Noting
a. If you figure out what the transcriber couldn’t understand, add the words to the transcript.
b. If you think you know what a word or phrase is but are not totally sure put the word(s) followed by a question mark in parentheses. (word?)
c. If you can’t understand it either, mark it as “(unclear).”
a. Auditing
Less obvious than seeing ‘(unclear)’ in the transcript—and requiring more vigilant listening and reading--is when the transcriber has misunderstood and typed the wrong word(s). This can be easy to miss especially if the transcriber has tried to make a reasonable statement out of sounds (s)he doesn’t quite understand, and the resulting error makes enough sense that it is not obviously wrong.
Example: Transcript: I joined the Elks Club of Tacoma over fifty years ago. Tape: I joined the Elks Club a little over fifty years ago.
b. Noting
Make corrections in the transcript when you hear a word or phrase in the video that is different from how it reads in the transcript. (As above, if uncertain, put the word(s) followed by question mark in parentheses.
2. Words left out: Sometimes a transcript is missing some of the words spoken in the video. Often just a few words are left out and that’s more difficult to catch than when a longer section is left out such as this example from the Na’im Akbar interview, with 25 words omitted from transcript:
Transcript: I could not get over the fact that I was here on Florida State’s campus, by the way, because that’s where they held the test. So, you know, just to even be on the campus was a major breakthrough, you know, for me.
Tape: I could not get over the fact that I was here, on Florida State’s campus, by the way, because that’s where they held the test, on a Saturday, with all these white kids, just me, and all my life I had not even been permitted to go on the campus. So, you know, just to even be on the campus was a major breakthrough, you know, for me.
If you find this problem, obviously the thing to do is add the missing words.
3. Extra words: As you might guess, this is less common than words being left out. It sometimes occurs when the transcriber types a more ‘standard English’ version of what was said. Delete extra words.
4. Pauses: The transcriber may not have noted breaks in the flow of speech. If there is a long pause this should be indicated in parentheses. (Pause). For shorter breaks use commas or dashes. (See below under Editing – Punctuation.)
XI. PROOFREADING AND EDITING The goal of editing a transcript is to create a document that is readable, coherent and accessible to users via keyword searches. We also want the document to be as error-free as possible. However (this is rather obvious but we’ll say it anyway) by “error-free” we mean such things as spelling and punctuation; we do not mean that the transcript should be changed to conform to rules of grammar, that vernacular should be changed into Standard English or that unpleasant statements or words (swearing, etc.) should be deleted. This is not an academic paper, but a record of what someone has said; we want the document to preserve the interviewee’s style of speaking and original intent.
A. Paragraphing The only paragraph spacing should come between different speakers (interviewer and interviewee--do not divide the interviewee’s answer into separate paragraphs. This is necessary for some aspects of the Informedia software to work right.
1. ‘Crutch’ sounds An exception to the goal of preserving someone’s speech is that we delete ‘uh’, ‘um’, ‘er’, and similar meaningless vocalizations that can be quite distracting to the reader. This was mentioned above in Section III.B.1.d. and you should have deleted most of these via Find/Replace when preparing the transcript for the Segmentor; however there may be some that were not deleted.
2. Response sounds during speaking Also omit “mm-hm” & other sounds of response made by an interviewer or another person while the interviewee is talking. (As our interviewers became more practiced over time this has become less of an issue.) Exceptions:
a. Do not omit if response follows a clear pause for agreement.
b. Do not omit this if it is the interviewee responding to a question.
B. Stylistic Standards Refer to The Chicago Manual of Style for very specific information on these issues. Following are some examples:
1. Abbreviations
a. The general rule is to spell out words in full, including:
names of currency: “Euros”, “dollars”, “yen”, etc.;
“number” instead of “#” or “No.”
“Okay” instead of “OK”
“and” instead of ampersand
b. A few abbreviations can be used: forms of address such as Mr., Mrs., Dr., Rev.; academic titles such as M.S., B.A., PhD., M.D.
c. States
Do NOT abbreviate states in Segment Text field, whether the name of the state is said during the interview or you are adding the name of a state in brackets.
DO abbreviate state name in Subject Headings when adding a location after an institution, etc.
Example: In that case, use
AP state abbreviations, NOT Postal version (2 upper case letters). Example: “Calif.”, not “CA” The AP abbreviations are in a file in the CATALOGUING PROJECT FELLOWS folder.
2. Punctuation
a. --Dashes
i. > -- “em dash”(long/double dash) for remark within sentence. When a speaker adds a supplementary or clarifying remark in the middle of a sentence--a strong parenthetical digression--the remark is set off by dashes as shown in this sentence; the dashes should be flush with (without a space between) the preceding and following words.
ii. -“em dash” (short dash)for cut-off sentence.
When a speaker fails to complete a sentence, indicate this by using double dash, flush with the last word spoken and followed by some form of end punctuation (period, question mark, etc. Example: “But I wasn’t--.”
iii. - “en dash” if word is cut off in the middle
iv. -- ‘em dash’ if multiple words are repeated
v. Interviewee spelling out names (or other words)
Use dashes between upper-case letters to represent this:
Example: “A-K-B-A-R”
b. , commas: for single-word repetitions or very brief pause in flow of speech; to separate a qualifying clause, between items on lists; between a location and larger location (such as city,state), etc. See Chicago
c. … Ellipsis: Use only for words omitted, for ex. a sentence cut off at beginning or end of tape; otherwise use dashes. (See above.)
d. ’ Apostrophe: contractions; possessives; Note: Per Chicago Manual, put ’s > after proper names ending in ‘s’, even it looks weird. Ex: Wilkins’s
e. ( ) Parentheses
i. to indicate actions, expression of emotion, etc.
Examples: (nods head for yes) if there is no spoken answer.
(tears in eyes)
(laughs) But if interviewee laughs a lot, note sparingly.
(Pause) Only for a long, noticeable pause in speech.
(Sigh) Only if very noticeable
(Simultaneously) if 2 people speak at once
ii. to qualify how something is said
Example: (sarcastically) I was a model child.
f. ‘_ ’ Single quotes: for titles of books, films, songs, TV shows, etc.
Note: Since the transcript must be in a plain text file to be loaded into the Segmenter, we are unable to use the style of italics or underlining for book or film titles.
g. “_” Double quotes: When speaker quotes what either (s)he or someone else has said.
h. [ ] Square brackets: used for inserting additional information. (See details in following Chapter.)
3. Capitalization The Chicago Manual of Style prefers lowercase in most instances. It is impossible to list all examples here, so refer to the manual if there are questions. Examples:
a. Names
i. Capitalize a person's proper name.
ii. Govt bodies, groups, etc.
Generally capitalize federal govt. bodies; not states.. Capitalize the official name of a group, institution, department, company, etc. Usually, don't capitalize unofficial /adjectival forms of names of groups
Capitalize:
Do not capitalize:
Government
Fed. govt. bodies:
United States Congress; Congress
congressional
State Department; Department of State
The department
the Judiciary Committee
I didn't know which committee I would be on.
State
(Unlike federal, do not capitalize)
Illinois legislature
City
Chicago City Council
the city council
Education
Morehouse College
We organized students at the college.
the Department of History
the history department
Business
General Foods Corporation
The corporation
The New York Stock Exchange
The stock exchange
iii. Names of specific historical events
Usually capitalized. But same words used in more general sense are not.
the Civil Rights Movement (meaning the U.S. civil rights movement for African Americans)
There have been civil rights movements in various different nations.
the Civil War (if referring to the U.S. one--since we're in the U.S.)
the Spanish Civil War
In the past decades there have been civil wars in several African countries.
b. Titles
Chi. Manual: "Civil, military, religious and professional titles and titles of nobility are capitalized when they immediately precede a personal name, as part of the name..
ex: President Francois Mitterand.
BUT..."When such titles are used in apposition to a name they are not part of the name and so are lowercased." Ex: The French president, Francois Mitterand.
AND... "In text, titles following a personal name or used alone in place of a name (other than in direct address) are, with few exceptions, lower-cased."
Capitalize:
Do not capitalize:
Senator Barak Obama
the senator from Illinois
Mayor Washington
The mayor of Chicago
Reverend Floyd Flake
a minister
Professor Hayes
the professor of history
Executive Director Julieanna Richardson
She is the executive director of the organization.
c. Geographic entities and directions: countries, continents, states, cities; neighborhoods. Examples: Kenya, Asia, Louisiana, Boston, Harlem, the South Side, the West Side, West Coast, East Coast, etc., the South, the West - -but NOT “southern” “the south of France”, “the western United States.”
d. Groups of people: Examples:
i. Race/ethnicity/nationality:
(A) Capitalize words that derive from names of countries, continents etc. that are themselves capitalized: Asian, African American, Koreans, New Yorkers, etc.
(B) Capitalize certain others such as: Latino, Native American
(C) Do not capitalize the words black or white.
ii. Religion: Capitalize Baptist, Catholic, Muslim, Buddhist, etc.
iii. Do not capitalize most descriptors used to characterize people, such as “the poor”, “disabled”, “gay”
e. Seasons: use lowercase: spring, summer, etc.
See Chicago Manual of Style for many more details on Capitalization.
4. Numbers
a. Spell out whole numbers < 100Example: twenty-five
b. Spell out these same whole numbers followed by hundred, thousand, hundred thousand Examples: six hundred; fifty-two thousand
c. Use numerals followed by million, billion, etc. Example: 3 billion
d. If spelled-out numbers cluster too thickly in the paragraph, use numerals. Example:“My oldest sister is 56—her name is Debbie--then I have a brother, Robert, who’s 53, my sister Lisa is 51, and the youngest, Francine, is 46 now.”
e. Numbers in the same category should be treated alike within the same context; do not use numerals for some and spell out others. For example, a paragraph giving several different percentages: if the rules say to use numerals for one of the numbers, then for consistency’s sake, use numerals for all of them.
f. Spell out numbers beginning a sentence even if this creates inconsistencies.
g. Numbers in addresses: Don’t spell out; keep in usual style: 1900 South Michigan
h. Public school numbers, when used as the name of the school Ex:P.S. 29
i. Percentages: Use numerals and spell out percent. Example: 96 percent
j. Do not spell out numbers that are part of common phrases, brand names, etc. commonly seen in numerical form. Examples: "24/7", "20/20 vision" "7-Up"
k. Money: If number is spelled out, so is the currency. Example: ten dollars
If numeral is used, then use symbol for currency Example: $549
l. Age decades: spell out. Ex.: "I was in my thirties." (For specific ages, follow standard given above, spelling out numbers 99 and under.)
m.Dates: When a month and day is given in the interview, the month should be spelled out rather than abbreviated numerically. Example: "April 21st, 2005" or "the 21st of April" (depending on which is actually stated.)
Years should be numerical, whether a four-digit year (1965) or an abbreviated two-digit year ('65) is stated.
Decades should be written numerically with an apostrophe first and an 's' after: Example: "in the '50s."
1. Books
We have a small collection of African American reference books in the office- - Who’s Who, the wonderful big Africana one-volume encyclopedia, Who’s Who, etc. Bring in some of your own books that may be helpful, if you want.
2. Internet
We are in the process of assembling a collection of African American history and other useful research links, in a Favorites folder in the CATALOGUING PROJECT FELLOWS folder on the server in Staff Files.) Cataloguers should make each other aware of great links that will make research easier. As with any Internet research, be careful that you are getting reliable facts from trustworthy websites; if in doubt try to find a more ‘official’ site.
3. Archival Box
As noted in the beginning of this manual, an excellent and essential resource in <em>The HistoryMakers</em> office are the archival boxes for each individual HistoryMaker, which you should have already perused before starting to work on the transcript.
4. Colleagues
Use co-workers as a reference source; we all have different areas of special knowledge and we can help each other out. At the same time, be considerate: remember that every other cataloguer also has a lot of work and don’t bombard someone with lots of questions to which you could easily find the answers yourself.)
5. HistoryMaker
If you are having trouble finding information you need, another possibility is to directly contact the HistoryMaker and ask them. Some are more easily reached than others; check with Crystal Foster.
This is related to auditing, because the transcribers are often ignorant of the people, places, companies, groups, etc. being discussed, and if they don’t understand what the interviewee says, their guess may be wrong. Types of facts to check:
1. Dates
Check dates and years mentioned for important events, legislation, etc.
2. Names and spellings
Check names (and spellings) for people, locations, organizations, companies, etc. mentioned in the interview.
If an acronym is mentioned find out the entire name.
Don’t take things for granted! (For example, in the Tyrone Davis interview, the transcriber typed “Wonderful Records” as the name of a record company- -the actual name was “One-derful Records.”).
However, don’t spend hours trying to track down one obscure fact that in the long run is not that essential for a researcher to find the material.
As stated above, if there is, for example, a name you can’t understand and can’t find out who the person was, one possibility is directly asking the interviewee; check with Crystal Foster about whether and how you should contact the person.
a. To assist in searches
The main reason is to aid researchers in finding Segments with the information they need about events, people and topics discussed in these interviews.
i. Partial names given.
For example, an interviewee might mention another person only by their first or last name or a nickname. Or they might refer to the landmark Supreme Court decision on school desegregation merely as “Brown”. In any of these cases a keyword search for the complete name would not find these Segments. (And a keyword search for “Brown” alone would probably return an unmanageable number of results.)
ii. To repeat terms from another Segment
Also, since we are dividing the interview into Segments, there will be situations in which relevant information was not repeated in each Segment-- for example, if an interviewee discusses job-related discrimination at a particular company but only mentioned the name of the company in a previous Segment.
b. To provide clarification or elucidation
We are also adding information that might clarify what is said on the tape for a potential user. We are still working on setting standards for this. This should not be overdone--we don’t want to treat the user as if they’re totally ignorant, and we should assume they are capable of looking up information in reference sources. You do not need to insert a notation to inform the user that Cassius Clay is a boxer; however, if someone is talking about Cassius Clay, you should add "[later Muhammad Ali]" so that a user could find it by either name.
2. Format for added data
Additional information should be placed inside square brackets [ ] to clearly distinguish it from the actual words spoken.
3. Types of information to add:
a. Names
i. Persons mentioned only by a shortened name, nickname or as “Daddy”, “my mother”, etc.
Example: “That's where Daddy [Robert Booker Flake] lived.”
“As [William] Bill Clay says….”
ii. Nicknames /common names of institutions, places etc. that have or had at the time discussed a different official name. Examples: Ole Miss [University of Mississippi, Oxford, Miss.] The Sorbonne [University of Paris, Paris, France]
iii. Partial names of persons, institutions etc: Examples:
Martin [Luther King, Jr.]
[Adam Clayton] Powell
Harriet [Beecher] Stowe
Columbia [University]
When adding part of a title, think in terms of retrievability via a phrase search.
Example: Wrong: “Florida A and M [Agricultural and Mechanical] University”
(Nobody would do a search for that exact phrase.) RIght: “Florida A and M [Florida Agricultural and Mechanical] University”
Note: re colleges, see also ‘Locations’ later in this section.
iv. Former/later names of persons or entities: Examples:
Tuskegee Institute [later Tuskegee University]
Kwame Ture [formerly Stokely Carmichael]
v. Acronyms or abbreviations:
When an acronym is given, follow it with the complete words spelled out in brackets. This need only be done for the first time it appears within each Segment. Example: SNCC [Student Nonviolent Coordinating Committee]
For an entity often known by its acronym, if the full name only is given, add the acronym in brackets. Example: Equal Employment Opportunity Commission [EEOC] Exception: Academic titles such as M.B.A., PhD, M.A., M.D.
b. Dates:
For search purposes, if interviewee mentions:
i. a year in abbreviated form, follow this with a bracketed insertion of the entire year. Example: “… in ‘65” [1965]…”.
ii. …a decade in abbreviated form, follow with entire decade. Example: “…the seventies [1970s]…”
iii. …an act or famous event without mentioning date Example: “Brown v. Board of Education [1954]”
c. Locations
i. If a city is named, put the state (or foreign country) in brackets following it. Examples: Paris [Texas]
Paris [France]
ii. If an institution is named, follow with location (city, state or city,foreign country). Example:
Howard University [Washington, D.C.]
iii. Unnamed location of described events : if a subject discusses an event or situation at a specific place but does not identify the location by name--for example, (s)he may refer to it as “here” or “there”--follow with the location information in brackets. Example: “I had fights with professors here [Harvard University, Cambridge, Massachussets], I still do, about things that I think are not right…”
Note: The editor must make a decision whether the event being discussed is important enough to warrant adding the location.
d. Corrected information:
If a speaker makes a clear factual error, (for example gives the wrong location for an institution), follow it with the correct information. This should be only for undisputed facts and especially if it would aid in retrieval; you should not insert arguments against statements you disagree with; that is not our role.)
4. How often to add the same bracketed information?
Add this bracketed information once per Segment. For example, if there is a section where the NAACP is mentioned frequently, it is not necessary to add “[National Association for Advancement of Colored People]” after the acronym every single time. More frequent bracketing may be distracting to a reader while not serving any purpose—the reader does not need to be informed twice, and only one instance is needed for a record to be retrieved in a database search.
There are two Segments usually at the start of an interview that are exceptions to some of our rules and have their own sets of rules. (Another exception is photo segements: see following section.)
A. First Segment of interview: slating The interviewer asks the HistoryMaker to state his/her full name and to spell it out, to state interview date and location, birth date, maybe occupation.
1. Highlight the transcript text. In the Movie Transcript view, highlight the text at the beginning of the interview with these questions.
2. Hit Ctrl +C to move this text into the Segment Transcript box (in Segment Text view).
3. Set Video TIME IN before first question
a. NO COLOR BARS! If the beginning of the digitized video retains some “Color Bars” from the videotape, set the Segment TIME IN mark AFTER the ‘color bars’.
b. Occasionally the first question was not recorded and you will have to start with the answer.
c. As explained above, if there is any unrelated talk about video etc., set the TIME IN mark after this side talk.
4. ‘Audit’: make sure Segment transcript matches Segment video If there were any unrelated talk you cut out of the beginning of the video, also delete it from the text. (This does not apply to example above.)
5. Proofread for spelling, punctuation NOTE - Spelling of name: upper case letters with dashes between. Ex.: G-O-R-D-0-N.
6. Set a TIME OUT mark after these questions and answers.
7. Title In Title field, type : Slating of [name of HistoryMaker] interview.
8. Save Click on “Update” button under the video controls. The title of the Segment should appear in the box underneath the video controls.
B. Second segment: “Favorites” After the Slating Segment there is another Segment which contains a few short questions asked for fun about the HistoryMaker’s favorite food, color, vacation destination and saying. Below is continuation of the Gordon Parks interview transcript, showing “Favorites” questions.
1. Click on “New”.
2. Select text in “Movie Transcript” window and copy to Segment Transcript.
3. Set TIME IN just before first question.
4. Audit to make sure text matches video.
5. Set TIME OUT mark right after end of last “favorites” answer.
6. Title this one [Name of HistoryMaker]’s favorites
XIV. Photo section of interview At the end of most interviews is a section of photos or other items the HistoryMaker has selected; usually accompanied by The HistoryMaker or another person talking about the photos.
Project Fellows are not only doing the usual segmenting and proofreading of the transcript, but are also writing a description/caption of the photo to be used for The HistoryMakers Photo Archive. Since we are already going through the videos and segmenting this photo section anyway, we do this to avoid duplication of effort.
1. Each item (photo, etc.) should be listed in separate Segment.
2. Determining boundaries of Segment This depends on whether and/or how much the photos or other items are being described aloud. Sometimes the photos stir up memories for the HistoryMaker and can lead to more stories and comments. However, on other occasions, the interviewee is tired or needs to leave and it ends up being a family member, friend or even our own videographer who describes the photos aloud for the camera. In a few other cases there is no audio at all.
a. Segments in which they are describing photos aloud:
The general rule is to include all of the video while the photo is shown. However, you can leave some of it out, such as the part where photo is being put into place (unless the HistoryMaker is saying something relevant right then).
b. The audio is to be considered more important than the video.
i. If a HistoryMaker’s description carries over into the next photo do not cut off the spoken description when the photo changes.
ii. If a HistoryMaker is talking about something unrelated to the photo, but what they are saying IS historically significant, then DO include all of the audio.
c. Segments with no relevant description or no audio at all:
If they are talking about technical things, side conversations unrelated to the history and issues we are preserving in these interviews, or else there is NO audio at all, then cut it down to the best quality 10 seconds. The absolute minimum is 5 seconds.
B. photo segment title and description standards Before creating new photo segments, check the existing photo captions on the server; if these meet the following standards, there is no need to re-do the work!
The "Segment Text" section for a photo record in <em>The HistoryMakers</em> Digital Video Library contains three parts: 1. Title field 2. Transcript and 3.Full Caption (both of the latter are in "Segment Transcript" field)
1. Title field
In a set of search results this is the text that will pop up along with each thumbnail. The Title should be a concise one-line description of photo. There will not always be room in the Title field to include all the information about the content of the photo; the full information shall be in the Full Caption.
*The title should be in the present tense.
*The title should be expressed as simply as possible.
*The title does not need to be a sentence.
*Do not follow with period.
*Only proper nouns should be capitalized.
*Do not use the word "poses." (Instead of " William Clay poses with campaign supporters", simply say " William Clay with campaign supporters" plus the location and year.)
*Should contain (in this order) name(s), event, geographic location and year; location and year should be set apart by commas. Format: Photo – Name, Event, Location, Year
a. Names:
Title should always include the name (both first name and surname) of the HistoryMaker if present in the photo. The name should be preceded by title such as "Dr.", "Rev", or "Bishop" if commonly used as a form of address for this HistoryMaker.
Other persons' names (first and last) should be given if possible. Use names given by the H
If the photo is a relative of the HistoryMaker, use the possessive form of The HistoryMaker's name, plus the name of the relative.
Example: Photo – Dr. Horace Smith’s grandmother, Alberta Pryor
(Note: When several people are in a photo, there may not be room to list all their names in the Title field; this information will, however be included in the Full Caption.)
b. Events: Title should give a basic idea of what the occasion is, if known.
c. Geographic location:
Name of city, state or foreign country. States should be spelled out, not abbreviated. Major cities such as Chicago, New York, Los Angeles, Atlanta, etc. need not be followed by state in this Title field.
d. Year:
Preceded by comma.
Approximate dates: circa (ca.) If a year is not specifically named, you may be able to estimate the year or range of years. When estimating, type "ca." before the year ("ca. 1972"), decade ("ca. 1950s") or another range of years ("ca. 1961-1964")
Example: The HistoryMaker says "This was from about 1968." Type "ca. 1968."
Note: In the case of a really old photo (1800s) with exact year unknown, you may expand the range of possible years even as much as "ca. 1865-1899" or "ca. late 1800s" (meaning 1870-1890s) so that users looking for nineteenth century images will be able to find it.
Ideas on estimating years:
If the HistoryMaker says "This is when I was eight years old.", just do the math from their year of birth to get the year of the photo (use "ca." if you don't know if it's before or after their birthday).
Similarly, if a HistoryMaker born in 1935 identifies a photo as being from their high school years you can estimate a probable four-year range from ages 14-18, or 1949-1953. (That may not be absolutely exact if they graduated early or were held back a grade but 'circa' doesn't mean exactitude.)
For HistoryMakers' parents, hopefully they've given a birth year, in which case you can add on years to that as in the examples above. Or else a HM may have mentioned a parent's age at married or at the birth of the HM.
If the photo is of a person holding a particular office, you can find out the years during which they held that office. If the HistoryMaker is pictured with a U.S. President and you know it was while that president was in office, use "ca." with the range of years the President served.
Visual clues! Sometimes there may even be a clue to the exact year in the photo itself--for example a prom photo with the year as part of the decorations (which is the case with a photo from Na'im Akbar).
2. "Segment Transcript" field
a. At top of field: Transcript
This is a record of what is spoken during the photo segment. Edit/audit as you would any segment. Fact-checking should be restricted only to names and places that are subjects of the photo. Subject headings should be limited to people, places, and dates in the photo unless there is a broad discussion about another topic. Exception: General "Autobiographical" subject headings such as Marriage, Education, Work and Career, etc.
If there is no audio, type [No audio.]
b. Below Transcript section: Full caption:
Following the transcript there should be a complete description of the photo, including ALL the information you may have put in the Title field, PLUS information related by The HistoryMaker in the Transcript. This section may have more detailed information about the photo that does not fit into the Title field, such as:
i. Names and positions of people in a group. Position information: After initial phrase explaining the subject, follow with semicolon, then in parentheses, a designation like "(L. to R.:) " or "(clockwise from left:)". In the case of more than one row, list from front to back. (Front row, L. to R.:) , then (Back row, L. to R.:)
Example: [Marion Barry with fellow Washington, D.C. City Council members; L. to R.: Harry Thomas, Marion Barry, Frank Smith, Kevin Chavis, and Senator Pendleton, Washington, D.C., January, 1993. From the collection of Marion Barry.]
(This position information is not necessary if the positions/identities of the people are clear from the rest of the description--for example, if your description identifies a HistoryMaker's parents, it should be clear which is which.)
Unidentified people: When there are several people identified in a photo, and one or a few are unknown, list them as usual but use the term "unidentified man" or "unidentified woman" for the unknown person. A person may also be referred to by their job or similar identity.
Example: [Civil Rights Movement leaders being served at the Davis Brothers Cafeteria in Selma, Alabama, spring,1965; seated at table, clockwise from bottom left: Reverend A.D. Williams King (brother of Dr. King), unidentified man, Dr. Martin Luther King and Reverend Ralph David Abernathy; standing behind, L to R: unidentified waitress distributing menus to the table, James Bond (brother of Julian Bond) and Julian Bond. From the collection of Julian Bond.]
Large group with only a few identified: Sometimes the HistoryMaker only names one or two people out of many in a group photo (which may be necessary when a few of the other photo subjects are famous people and are therefore pointed out by the HistoryMaker as he/she describes the photo) identify these people, following their name with parenthetical info on their position in the photo--such as "(Front row, 3rd from left)" and follow with "and other members of"/"employees of"/"directors of" etc. if they are all part of the same organization, or use "and other" plus a noun for a type of person, such as "and other African American business leaders" . If they are not from a specific group or category of people just use "and others"
Example: [Paul Adams (center left), U.S. President Ronald Reagan (center), W. Clement Stone (third from left) and others at Providence-St. Mel School, Chicago, Illinois, 1983. From the collection of Paul Adams.]
i. Details about the occasion that do not fit in the Title field can go here, such as the name of an organization, an award, etc.
ii. Location:
same as in Title field but more specific if known. (However, NEVER give someone's home addresses.)
iii. Date:
Same rules and suggestions as in Title field, but in this field, be as specific as possible--so if you know the month or the date, include it. If the date is absolutely unknown and cannot be estimated even within a span of ten or fifteen years, then use "n.d." (which stands for "no date"; per Chicago Manual of Style)
iv. Extra info Other information of interest about the photo can be included--just don't write a book--one or two sentences at most.
Example: Photo - Renowned African-American entertainer and activist Paul Robeson following a concert at Lincoln University, Pennsylvania, ca. 1950, with children of the President and Dean; L. to R.: Jane Marguerite Bond, Sylvia Hill, Paul Robeson, Horace Julian Bond. Robeson's father, a former slave, was a graduate of Lincoln University. From the collection of Julian Bond.
v. Any information known about photographer and publication. Example: [Georgia State Representative Julian Bond tours the poverty-stricken Vine City area of his legislative district, Atlanta, Georgia, 1968. Photograph by Vernon Merritt, originally published in 'Life' magazine. From the collection of Julian Bond.]
vi. Sentence identifying the photo as from the HistoryMaker's collection. Example: [Horace Smith and his wife, Susan Davenport Smith, at his ordination as pastor of the Apostolic Faith Church, Chicago, Illinois, 1980. From the collection of Horace Smith.]
3. Annotations for photos
a. Dates (same principals as in estimating dates for description)
b. Subject Headings
Add headings here as with other Segments, but sparingly:
i. People: as usual, only if well-known people.
ii. Organizations:
Do add if it’s a group photo or some event in which the members can be seen participating. In the Floyd Flake example above, add “Congressional Black Caucus.”
c. Locations:
DO add:
DON’T add:
if the location can be seen in the photo…
for example, if you can see it’s on the steps of the Capitol, add Washington, D.C.
-if the people are participating in some important event associated with the location…for example, a “Freedom School” in Mississippi.
if it’s in some generic sort of place—some place that “could be anywhere” - - like somebody’s living room
if it’s a close-up and you can’t see anything about where it is
Other subjects: Add sparingly. But in the example above at right, it would make sense to add Freedom Schools as well as Mississippi.
A. DATES Date annotations will help researchers to narrow down search criteria and results and will enable visualization of search results along a timeline.
1. Annotate most Segments with date info:
a. Years or decades mentioned in reference to events discussed.
b. Specific dates of notable events - for ex, assassination of MLK.
c. Approximate years/date ranges of events discussed
Try to include some date annotation, whether or not the interviewee names the specific year, whenever (s)he talks about temporal events (as opposed to general & timeless ideas). Usually you can estimate a year or a range of years based on the HistoryMaker's age or on other events mentioned in the interview. (Note:there is no way to designate “circa.”)
2. Do not annotate: a date that is not historically important (wedding anniversary, the party for their seventh birthday, etc.) However such cases may have an approximate decade added.
3. How to add date annotations
a. "New"
If another date is in the box hit "New" to clear it. JUST TRYING TO TYPE OVER A DATE WILL CAUSE AN ERROR.
b. Specific date
Type in the date in the left box.
mm/dd/yyyy
Click on Add; date should appear in box at right. (See picture below.)
c. Year (w/o month/date): Enter zeros for the month and the date then enter year, for ex.: 00/00/1947 +Add
d. Decades or other date ranges: Enter beginning and ending years (As when adding single years, use zeros for month /date)
(Do not list each year in a range separately; just put a range from the first year mentioned in the Segment to the last year mentioned in the Segment.)
1. Use of “Subject Headings” in this project Subject Headings will be assigned at the “Segment” level. As mentioned previously, our cataloguing team is collaborating with outside advisors to build a thesaurus for African American subject matter as a corollary to the Library of Congress Subject Headings which alone are insufficient for our purposes. Indexing at the Segment level allows us to achieve a level of specificity impossible with indexing at the interview level--which has been the most common practice for oral history cataloguing.
For the researcher, the use of a controlled vocabulary allows them to have more control over their searches in a database that will include the transcripts from c. 1200 hours of interviews. Keyword searches may sometimes return too many undesired results. Using our thesaurus and limiting their search to the Subject Headings field, researchers will be able to target their searches more narrowly. And the anticipated hierarchical search will allow them to explore the collection by browsing through records in the subject they want to research, moving down a level to a more specific aspect of that subject, or moving up a level to broaden their search to include related subjects.
2. How to add Subject Headings to Segments In Segment Annotation view, you can scroll through the list of Subject Headings. which are generally organized in categories of Mainheadings and Subheadings. Subheadings are expressed with the higher level headings on the beginning, like: Law enforcement::Race discrimination in law enforcement::Driving while Black.
Let’s say you have a Segment in which the HistoryMaker is talking about Martin Luther King, Jr. and the Selma-Montgomery Rights March.
Scroll down the list to the Persons section, and find Dr. King’s name. (See left.) Select the name by clicking on it, then on the Add button at right. The name appears in the “Annotations” window at right. (See below.)
Now find a heading for the event itself. Something that famous should have its own heading, so check. Click on it and add it. (See right.)
Always add location if the person is discussing events that happened in a certain place. So scroll down to Locations, arranged from general to specific: US, then Alabama, then Selma. Click & add.(See left.)
Now you have three headings listed in your Annotations box (see above). Note: this is not There would be others- -for example always add the name of the interviewee.
3. Content Analysis: Guidelines for assigning Subject Headings NOTE: For adding new headings see Section 4, below.
In all cases, think in terms of the user: if someone doing a Browse search by subject heading finds a Segment based on a name, topic etc. that you chose to assign as a heading, will that user likely be satisfied or disappointed with the result?
Types of terms that should be annotated:
a. Name of HistoryMaker themselves should be indexed in each Segment.
b. Names: of other significant persons who are discussed.
i. “significant”: Not everyone who is talked about in the interview should have their own subject heading. If we assigned a separate heading for every family member mentioned the list would really get out of control.
ii. “discussed” (not just mentioned) Example of names not to include: A user searching for stories of people’s memories of Charlie Parker would not be thrilled to be directed to a Segment in which Parker is just one name in a list of an interviewee’s favorite musicians; in such a case, do not index for all the names on the list. (Even of persons discussed, you should exercise some discretion not 100% will be appropriate Subject Headings - - for example the names of interviewees’ relatives who are themselves not well known)
c. Locations: of significant event related by interviewee. Again, not if just mentioned. Example of locations to include as heading:
William Clay discusses in detail his campaign for alderman but he does not specifically mention the name of the city. In this case, St. Louis should be a heading because the location of the events he is relating is significant.
Sonia Sanchez talks about attending an integrated public school in New York, and then compares the curriculum, assemblies, etc. to the segregated school she had attended in Birmingham, Alabama. In this case, include headings for both cities since she has really described her experiences in both places.
Example of location not to include as heading: If Sanchez had said “School in New York was really different from Birmingham.” but said nothing else about Alabama, do not add heading for Birmingham.
d. Other proper names: events, institutions, organizations, etc.
Of the terms spoken, again, only suggest headings for those that are actually discussed. Example: if interviewee says “I thought about attending Howard, Fisk and Hampton, and eventually I decided on Howard”, and then goes on to discuss the experience at the university, Howard University would be indexed for this Segment but not Fisk or Hampton.
i. Main subject headings under which to list proper names. Examples:
Banks
Businesses (most) are in Business::Business enterprises –
Hospitals are in Health and Medicine
Radio stations in Media
Television stations in Media
Record companies are under Music trade
Particular theaters are list under Theaters.
Churches, mosques, etc. are listed under Religion
Universities and colleges are under Education::Universities and colleges. Other schools are under Education::Schools, then organized by grade level.
Most organizations are just listed under Organizations except civil rights orgs which are listed under that topic Under Organizations there are sub-categories for youth groups, Greek letter societies, and professional organizations.
e. Topics
This is used to mean everything except the above listed categories of proper names: this encompasses the concrete and the abstract.
f. "Autobiographical" subject headings
There is a special “Autobiographical” section we have created to use for certain ways the HistoryMaker talks about his/her life - - for discussion of family members, their own personal philosophy, what they want their legacy to be, etc.
Using the Autobiographical section shouldn’t be an excuse to ignore the rest of the list which is much richer and more detailed in terms.
4. Adding new terms to Subject Headings list Many of the new headings to be added to the subject list will be proper names: people, organizations, schools, locations, etc. Because of the great number of these and in order to expedite your work, Project Fellows themselves may add Subject Headings for proper names- -specific persons, locations, or corporate bodies – businesses, institutions, organizations. However be very careful when doing this to follow correct standards. To add headings, simply add them to the text file in N/Staff Files/CATALOGUING PROJECT FELLOWS/subjectlist
Put this form into our Locations template: country::state::city::neighborhood
If not listed there, create the heading using our standard form.
b. Names of Persons, corporations, institutions, etc.
i. Check Library of Congress Authorities! http://authorities.loc.gov, then search for Name of person, institution, etc. to find the form of the name used. Be careful: NOT EVERYTHING LISTED IS AN AUTHORIZED HEADING!! Look at the red buttons in the left column: you want to find ones that say "Authorized Heading". There are others that say "References" --these are for alternate versions of the name or subject heading; if you click on one of these, the record will tell you to "SEE…[ whatever the Authorized Heading is]”
Example of a search for a person’s name: Angela Davis. In LC Authorities, select Name Search, type "Davis, Angela" and hit Begin Search. It jumps to that point in the alphabetical authorities list; there are several Angela Davises listed. Let's say you already know her birth year is 1944. There is a listing that says "Davis, Angela, 1944-" BUT note in the left column a red button that says "References" This means that this listing is NOT the authorized version; it's just there to refer you to the right one.
Click on “References” You'll get a reference telling you: "See Davis, Angela Y. (Angela Yvonne), 1944-" Click on that version of her name,
It returns to the list, at the point of Davis, Angela Y. (Angela Yvonne). You can see another red button at the left of this name, which says "Authorized Heading" If you click on the button you'll get a page, with the name listed as a hyperlink, which will take you to the "Authority Record".
Authority Record will look like it is in code because it is—a standard library code called MARC. But you can click on the tab that says "Labelled Display" to see a plain English version. The Authority record tells where they got the version of the name that they have decided is the “Authorized” one—usually a book by or about the person. It may also give info such as birth & death dates, a list of alternate, non-preferred names, and sometimes there is even a brief bio. So it's a handy clue as to whether you've got the right person. (Example: If you had not known Angela Davis’s birth year or middle initial, if you clicked on Angela A. Davis you see she has written books about wetland restoration, so that's probably not the right Angela.)
KEEP TRACK OF NAMES ADDED FROM LC AUTHORITIES TO OUR LIST. After finding and adding a proper name of a person, institution, etc., to our headings list, copy the MARC record (the code one) from LC Authorities into a Word or text file; include the name of the what interview it came from.
ii.Adding names not listed in LC Authorities.
If no existing Authority Record can be found and there is ample reason this name should have a heading, you may create a new one. KEEP TRACK OF NON-LC AUTHORITY NAMES ADDED Put list of new names at the end of your list of MARC authority records, and with each name include the source interview and any other reliable source you checked to research spelling and dates. Send this list with all names added to our list to the Cataloguing Director.
STYLES FOR NEW NAMES ADDED: Persons: list the name as Last Name, First Name, Middle Initial., YYYY- or YYYY-YYYY if the person is deceased.
follow name with birth (and death if applicable) dates.
Institutions, corporations, etc.: if they are a localized entity, follow name with location- –city, state (Use AP abbreviations). For national/multinational bodies you don’t need to list location (unless it is given in Library of Congress Name Authority File). Radio and television stations.: Call letters should be followed by “radio/or television station”+ location” Example: WGCI (radio station, Chicago, Ill..)
c. Topics:
For topical terms you should NOT just add your own heading ideas. If there is a subject heading you think should be added, suggest this to the Cataloguing Director; send an e-mail with your heading suggestions and tell what interview they came from. It is helpful if you can check Library of Congress Subject Headings to see if there is an existing heading that would apply – be sure to mention it if you have done this. If no existing heading Often we discuss ideas for new terms as a group. Keep track of which Segments you wanted to annotate so that you can do that as soon as you hear that the heading has been added.
The Cat Dir. will send out regular e-mails letting everyone know what new headings have been added.
We are avoiding geographical subdivisions of topics. The Location headings and Informedia’s dynamic map feature should be able to help people narrow down topic searches by location without us having to subdivide each topic by all the different locations mentioned.
A. MOVIE NAME
Give the Movie a name in the format Last name_First name_number of video. This should match the names of the video file, transcript file and XML file. If you have video files previously given a different type of name (like Jeff_Donaldson_A) re-name it in the correct format.
B. TITLE For the title, use format: Jeff Donaldson Interview, Tape 1.
C. Abstract For each 30-minute videotape, an abstract should be written summarizing the content.
DO: Write the abstract as a single paragraph and keep it no more than half a page long. DO: Briefly summarize the important subjects discussed during that half hour. DO: Write in the form of complete sentences. DO: Write in simple, direct language. DO: Write for a reader who is intelligent but not necessarily an Af-Am Studies scholar. DON’T: Try to describe each Segment or cram in everything that was mentioned. DON’T: Write it as a list of subjects. DON’T: Write in wordy academic prose. DON’T: Assume that the reader is an African American history expert who will understand any reference you might make to events, groups, etc. DON’T: Mix up tenses! This is a common error in writing these summaries In this correct example below the cataloguer used the present tense to refer to the time of the interview and the past tense to refer to….the past! However, sometimes people slip back and forth between this style and using the present tense for the past.
D. INTERVIEWER The interviewer’s name should be on the transcript (but check spelling). It may be mentioned during the interview. It definitely should be on the Production Report. Check in FileMaker Pro database and if it is not there, ask Crystal Foster. Here is a list of current and former HM interviewers.
Julieanna Richardson
Sandra Johnson
Racine Tucker-Hamilton
Larry Crowe
Nefertiti Makenta
Lorretta Henry
Adele Hodge
Reginnia Williams
Clarissa Myrick-Harris
Jodi Merriday
E. INTERVIEWEE
Enter same format as used in adding the name to the subject headings.(Library of Congress)
F. DATE OF INTERVIEW Given during “Slating” at start of interview. (If not, check HM’s Filemaker Pro record or Production Report.) Express date as YYYY-MM-DD (ex.: 2005-01-03).
G. LOCATION Given during Slating (or check Filemaker database or Production Report).
1. Time code discrepancy There may be a slight difference between the time codes of the original videotape and those of the MPEG digital video you see in the Segmentor. The reason for this stems from the fact that some of the “color bars” (recorded at the start of each videotape) are often digitized along with the actual interview footage. The MPEG time code begins at zero immediately at the start of whatever video was digitized (thus including whatever length of color bars was digitized - -see left). But the Betacam video time code starts immediately AFTER the color bars (also see left). However many seconds of the color bars got digitized is the amount you need to “offset”, so that the Segmentor (and later the end-user version) will show the correct time codes.
2. How to fix it
a. Find a point on the digital video in the Segmentor that is easy to remember when it’s coming up - - like a certain word spoken. Write down the time code (TC) for this word. (For example, in diagram above, let’s say the point you choose is at 00:04:26:11.) Repeat this with each video file for that interview.
b. In the Archive, check out the Beta dubs (copies) & take them to a tape deck. On the first tape, find the exact point (probably a word) you chose on the first MPEG video file. Write down that TC. (In diagram above, 00:02:10:00.) Repeat with each video for the interview.
c. For each tape, subtract Beta TC from MPEG TC; enter result in SMPTE Offset field. (In the example above it would be 00:02:16:11.) NOTE: Remember that you are not dealing with two eight-digit numbers in a decimal system! The first two digits are the number the videographer gave to the tape. Next is minutes (which will never be over 30) then seconds – (since there are 60 sec. in a min., this pair of digits goes up to 59 (not 99). The last two digits are frames – - there are 30 per second. So they go up to 29 (not 99).
I. MEDIA LENGTH This refers to the exact time length of the actual video of the interview. To get it, find the first frame in which the interviewee appears after the color bars; note the time code. Move time bar to end of video and find the last frame of video; note that time code. Subtract the first from the last to get the Media Length. Remember that this little bit of arithmetic is not in a decimal system--see previous explanation for SMPTE Offset.
Each Video (MPEG file) must have a corresponding “Project” (saved as an XML file). Every time you are shutting down the program you must save the XML file for the Project on which you are working. And, each time you get to the end of a video you MUST save your project as an XML file, and then start a new Project (XML file) for your next video. If you forget to do this and just continue adding Segments from a new video in the same Project you will have Segments with time code information relating to the wrong video.
The first time you save the XML file, name it in this format: Bond_Julian_01, and save it in the same folder as the Video files: First LD on Newserver:/Video/[Initial]/Last name, First name].
To save more work on a project you have previously saved: you must select Save As instead of Save—or else it won’t! You can save it with the same file name but only by “Saving As”. By default it will save in the same folder.
XVIII. Standards for HM Evaluation Forms In addition to a HistoryMakers’ (HM) interview being processed in the Segmentor, a completed Evaluation Form (EF) is required to provide a brief, written synopsis of what is contained in the interview. Also information will need to be culled from the FileMaker Pro database and/or the written Production Reports that may be enclosed in HM’s archival files. The purpose of these standards is to provide conformity to the actual Form itself.
In all cases, the form should have a Font of Times with a size of 12 pt. MS Word will provide you with those settings. In addition, the Evaluation Form’s questions must be in Bold, while the data can remain Normal. A template of this form is on the server for you to copy to your desktop for frequent use. Do not type in HM information on the server’s copy of the Evaluation Form. When saving the document, use the following format: Davis_Ossie_Evaluation Form. Within that HM’s file in Individual HistoryMakers on the server (First LD on Newserver:// HistoryMakersfromsctnserver/Oserver_Mac/HM WebDev/Individual HistoryMakers/(Initial/Last Name, First Name) create a New Folder name it Evaluation Form. Move the completed form into the Evaluation Form folder.
1. Dates – Always spell out full dates for the Date you are processing the EF, the Date of Interview, and the Date of Previous Interview – i.e.: January 26, 2005.
2. Name of HM – Write out the entire name in the order as follows: First Name/ Middle Name/Last Name. In the case of males, follow the last name by “Jr.,” “Sr.,” “III,” “IV,” if applicable. In the case of married women, followed the last name by “née Maiden Name” in parentheses – i.e.: Elnora D. Daniel (née Belle). If one is commonly known by his nickname, follow the First Name by his Birth Name in parentheses – i.e.: Ossie (Raiford Chatman) Davis or Scatman (Benjamin Sherman) Crothers – NOTE: It is not necessary to include nicknames if they are not popularly known. Also include “Rev.” and “Dr.” for clergy, physicians and those with doctorates, providing they are routinely referred to as such. Check the questionnaire or biographical folders in the HM archives box, just in case.
3. Videographer and Interviewer – Check FileMaker Pro database or Production Report in the archival box. If unavailable, at times both are identified in the slating of the interview. Also consult the list of Videographers’ Names in the CMU file on the server for proper spellings.
4. Completed Questionnaire and Release Form Check archival box and simply mark these spaces with an upper case “X.”
5. Interview Properly Slated The interviewer is responsible for slating the interview by having the HM state their name, date of birth, place of birth, occupation and the date and location of the interview. If it is not properly slated, note what was missing or out of place in “Other Comments.” Mark the appropriate space with an upper case “X.”
6. Audio and Video Levels It’s assumed that both have been checked prior to the start of the interview; simply mark these spaces with an upper case “X.” If problems exist with either, indicate them in the “Comments about audio/video levels” query immediately following.
7. Location of interview This protocol should not only be used on this Evaluation Form, but in the Movie Attributions field on the Segmentor. These “variations on a theme,” are based on how the HistoryMaker specifies the location of the interview during slating.
For privacy purposes, interviews conducted in his/her home should be written as Residence of HM/City/State. If the HM gives only an address, just write the City and State. It shouldn’t be assumed that the interview is being conducted in their home. In the case where interviews are conducted in a place of business or house of worship, list the full name of the Business and/or Church/City/State. Include the street address of the business only if it is given in conjunction during the slating. In all cases, spell out the full name of the state – i.e.: Residence of William Greaves, New York, New York or Travis Realty, 840 E. 87th St., Chicago, Illinois. Lastly, in if the interview was conducted at any of The HistoryMakers’ offices, include the geographic location of the office – i.e.: The HistoryMakers, Washington, D.C.
8. Interview of HM This is dependent on whether the HM has conducted their interview in more than one sitting. Mark the appropriate space with an upper case “X.”
9. Date of Previous Interview Check FileMaker Pro if previous dates exist and always spell out full date. If no previous interview exists, mark it with an upper case “N/A” (Not Applicable).
10. Interview Complete Mark the appropriate space with an upper case “X.” If not, make note of why not, for instance, whether an interview abruptly cuts off on the last tape.
11. Transcript Complete Mark the appropriate space with an upper case “X.” Hopefully this is the case, however, if the transcript is problematic, include comments in with item #12.
12. Transcript Proofread Mark the appropriate space with an upper case “X.” If not, make qualitative notes as to why.
1. Standard Questions Asked? To note whether questions about Family Background, Earliest Memory, Sights/Smells/Sounds, Remembrance and Legacy were asked, mark the appropriate space with an upper case “X.” Note that the interviewer may not always ask these questions in the exact same words or the HistoryMaker may give the information without having been asked; if this happens check "Yes".
2. Other Interview Quality Questions (Philosophical Questions, Done Well, Need Improvement, Follow Through and Other Comments) Keep answers brief but include whatever you think should be noted.
a. Philosophical Questions
Note if interviewer asks more than plain factual questions; this is not limited to questions containing the word "philosophy" or about the study of philosophy.
b. Done Well
List positives about the interview.
c. Need Improvement
List areas in which interview could have been improved.
d. Follow Through
Assess how interviewer has done at asking for more details or explanations of subjects mentioned. It is quite possible that the HM will be interviewed again at a later date, so missed follow-through questions may get addressed then.
e. Other Comments
Note issues not covered under other questions. Note when there are noticeable skips in the video, ‘off’ coloring or lighting, superfluous background chatter, etc. In some cases, re-encoding or re-dubbing the video may remedy problems. All in all, noting this will assist in producing a better, more professional interview.
3. Historic Relevance
Keep the answer simple and limit it to about a paragraph. You are briefly stating what makes the HM unique – i.e.: Charles Burrell is one of the early black musicians to break through the color barrier and enter the predominantly white field of classical symphonic music. If the HM makes references during the interview to other documentaries/interviews that he/she was the subject of, cite them here – i.e.: Two documentaries were also produced about Howard Armstrong: “Louie Bluie,” by Terry Zwigoff (1985) and “Sweet Old Song,” by Leah Mahan for PBS (2002). Note: If the HM is a filmmaker, it’s not necessary to list his/her films on the EF.
4. Historically Important Issues Covered
Write topics here rather than full sentences, using a loose form of the CMU Subject Headings format. List about five to six major topics covered over the course of the interview, and separate each topic with a semicolon – i.e.: African American actors; African Americans in the performing arts; Rose McClendon Players; Howard University in the 1930s; The U.S. House Committee on Un-American Activities (McCarthyism and anti-communist blacklisting).
(Note: This step was discontinued as of June 2005.) After following the instructions for identifying and segmenting any photos in an interview, export the transcript of these photos into a text document and store it in the Evaluation Form folder. Notify the appropriate Project Fellow (currently Susan Perry) responsible via e-mail so he/she may make the necessary changes. This list will serve as the archival list of photo captions.
Photos – To export the list of photographs from the Segmentor, go into the appropriate interview tape. Pull down the File menu and select Export Transcript. Name it simply “HM photos Tape 1, HM photos Tape 2, etc.” Since this will be reformatted as Archival Photo Captions, Save it in the Evaluation Form folder for the time being. You will then receive a prompt asking if you wish to include the SMPTE Offset time codes. Select Yes. Remember, you’re saving most of the bracketed information detailing the HistoryMaker, photo subjects, additional sitters, geographic locations and dates as specified in the CMU Photo Standards Suggestions document. Once ‘cleaned’, Number each photo as 1), 2), 3), with a single space separating parenthesis and the bracket preceding the time code, Copy and Paste the list into the EF – i.e.: 12) [07:24:59:06-07:25:19:13] Marion Barry with fellow Washington, D.C. City Council members; (L. to R.): Harry Thomas, Marion Barry, Frank Smith, Kevin Chavis, and Senator Pendleton, Washington, D.C., January, 1993.
A. Information to be collected for FileMaker database. Part of the work the cataloguer is asked to do may include entering information about an interviewee in another database. The information is extracted from the HistoryMaker’s interview and the questionnaire contained in his/her archival files. NOTE: As of summer, 2005, Project Fellows are collecting information from the interview and/or HistoryMaker Questionaire, and passing it to another staff member (currently Frederick Adams) who is doing the actual data entry. The information to be collected for each HistoryMaker is:
HistoryMaker's name HistoryMaker's occupation Interview Date Interview Location Interviewer Videographer Accession number HistoryMaker's birth date HistoryMaker's birthplace Family members' names Educational institutions attended and dates Career information
These are guidelines for data entry. Some information already exists in FileMaker so search before creating new records. If there are any errors or formatting issues in the existing data, please make the necessary corrections.
1. To get to the FileMaker database on the server, you must:
a. Open FileMaker Pro.
b. Select “Open an Existing File.” Select Okay.
c. Select “Remote”; wait for hourglass icon to find server files.
d. Select Newserver, then HM Interviewee Recovered.
e. Choose Search Mode to find HistoryMaker
Select the magnifying glass icon in the gray margin at the left. This is to FIND your specific HistoryMaker. Remember to ALWAYS do this before creating any records (including all "pages" listed below). It will prevent accidentally writing over existing information and considerably cut down on duplicate records in the database.
2. Interviewee Page Check HistoryMaker’s name, date of birth and description (occupation) for accuracy.
a. Names
Note that there is no field for nicknames, so if the HistoryMaker is commonly known by his/her nickname, included it in the Middle Name field – i.e.: Ted “Double Duty” Radcliffe; Holmes “Daddy-O” Daylie; Lenora “Roscoe” Dellums. Remember to put quotes around the nickname to distinguish it from their christened middle name. If a HistoryMaker is commonly addressed by his/her initials, spell out their name (if given) in the appropriate fields rather than treating it as a nickname.
b. Birthdate/Death Date of HistoryMaker
Use the following format when entering any dates in FileMaker Pro: mm/dd/yyyy. Single digit months and days should be indicated by 01/, 02/, etc. and the entire year must be called out since there may still be some HistoryMakers born in the 19th or early 20th century—it creates less confusion.
c. Birthplace of HistoryMaker
Give the city and spell out the full name of the state. If the HM gives the town AND county, enclose the county in parentheses – i.e.: Dayton, Tennessee (Rhea County); Arlington Heights, Illinois, etc. Sometimes the HM will give the county name AND name a nearby town if they did not officially reside in the town. If this is the case, put ‘near’ and the town’s name and state in parentheses – i.e.: Jo Daviess County (near Galena, Illinois). If a nearby town isn’t provided, the county’s name will suffice.
d. Description
In FileMaker it’s the occupation of the HistoryMaker. Keep the descriptions simple – Housewife, Minister, Teacher, Accountant, etc. Some occupation descriptions may also exist in a pull-down menu in the Family Members field in FileMaker.
Also scan for Gender, Marital Status and Spouse’s Name if it is revealed during in the interview or on the questionnaire, please fill in these fields or make the necessary corrections.
3. Status Page On the Interviewee page, each HistoryMaker is given a Unique ID #. Select the green Status button on the Interviewee Page. Select Find, and type in the Unique ID #. Select Interview. In this portion of FileMaker you will find information pertaining to the production of the interview. Much of this information contained here needs to be either added to or expanded upon, especially if an interview takes place on different dates, in different locations, or is conducted by multiple interviewers. The information should come from the Production Report or gleaned from the interview’s slating if no Production Report is available. DO NOT CREATE NEW STATUS PAGES. Fill in the fields or make the necessary corrections. Multiple dates should be separated by a comma – i.e.: 03/02/2003, 05/15/2003. Multiple Videographers should be written using their initials and the portion of the interview they participated in – For example, MH-1, SS-2 means that Matthew Hickey did the first portion of the interview while Scott Stearns did the second.
4. Family Members Page Select the green Family Members button on the Interviewee Page. Select Find, and type in the Unique ID #. If no records exist, select Records, and then New Record. Type in the Unique ID # from the Interviewee page. When the appropriate HistoryMakers’ name appears, you may now begin to enter records. You are given 18 spaces with which to add family members’ names—just select the tab for each new name. If more than 18 spaces are needed, create another New Record and you will be given 18 additional spaces.
Family Members – Write out the entire name in the order as follows: First Name/ Middle Name/Last Name. In the case of males, follow the last name by “Jr.,” “Sr.,” “III,” “IV,” if applicable. In the case of married women, followed the last name by “née Maiden Name” in parentheses – i.e.: Minnie Bethany Bishop (née Slade). For siblings (and step-siblings) of the HistoryMaker, give full middle names or middle initials if known. In the case of married female siblings, write the married name excluding the maiden name unless it is entirely different from their last or maiden name. For other relatives, list them in the order of the immediate family and work outwards: Parents/ Grandparents/Great-Grandparents/Spouse/Children/Uncles/Aunts/Cousins/In-laws/Ex-spouses (only if they’ve produced children). If the HM indicates which sets of grandparents belong to his/her mother or father, try to cluster the family members together and precede the list of names by using “Maternal” or “Paternal”. Also, if a HistoryMaker mentions a cousin whose name is unconventional, it may be necessary to indicate the gender by following the first name – i.e.: Beverly (male)Bishop, or Hasna (female) Muhammad. If the HM continues to give the full names of extended family members (i.e. cousins) and you’ve run out of room in the initial FileMaker record, create a New Record and reuse the assigned Unique ID. However, if they only give the first name of a relative, don’t bother guessing last names, just omit them.
5. Education Page Select green Education button on the Interviewee Page. Select Find, and type in the Unique ID #. If no records exist, select Records, and then New Record. Type in the Unique ID # from the Interviewee page. When the appropriate HistoryMakers’ name appears, you may now begin to enter records. Unlike Family Members, you will need to create a New Record for each school entered.
Schools – Always spell out the full official name of the school, city and state. Use the term Grammar when referring to Elementary (K-6) or Catholic/Parochial (K-8) schools; if known, remember to note the religious denomination when listing parochial schools – i.e.: St. Patrick’s Episcopal School or Faith Lutheran School. Use the term Intermediate in FileMaker when referring to Junior High, Middle School or Intermediate School (commonly grades 7-9), and use Secondary when referring to High School, College Prep or Technical/Vocational Academies. For Education Type, indicate “Public,” “Private” or “Private religious” school.
A. CMU QC Project Fellow responsible for FTP connection between THM and CMU (currently Tyler Barnett) will ftp the XML and MPEG files to Carnegie Mellon. Melissa Keaton there will go over them for Quality Control and send back reports about errors via e-mail.
B. Corrections in QC report Tyler will make the corrections noted in the QC report and will write summaries of what the problems were in each. He will send a copy to Cheri. In making the corrections, for cases in which Melissa suggests something that does not coincide with The HistoryMakers’ standards as laid out in these guidelines, our own standards will take precedence. If you disagree with or are unsure about a correction / suggestion Melissa has made and there is no set standard here yet, check with the Cataloguing Director (currently Cheri Pugh). We will come to a decision on the issue, and you should then make any required change. (Note: the Cataloguing Director will maintain contact with Melissa to keep her informed about our standards as they develop.)
C. FTP Final Version He will send corrected XML files back to CMU and type the date sent in the Master List ("Transcript Holders Extra Columns" Excel File on N drive, in the column "Final FTP") (Our Digital Video Library is currently being maintained at CMU; this process will change when we get our own data server.)
Home
i. Open from Start menu, Desktop shortcut or on hard drive: C:/Program Files/Carnegie 
Mellon University/Video Segmentor/Segmentor. 
a. Select File/Open Video.
folders as the person’s XML files
c. Select video file. You should see a list of MPEG files.
A. Open the Project
. You will see a tiny triangle appear at that point below the Time Bar. 
Another tiny triangle should appear at that point below the Time Bar. 
to make sure your IN and OUT points are set correctly, since it will play only your Segment and you can easily hear if it’s cutting off words or including parts of extra words. 
) in the Time bar for a Segment you have created, it means you have forgotten to set TIME IN and/or the TIME OUT point. The blue section of the bar shows where the Segment is in the video.
The interviewer asks the HistoryMaker to state his/her full name and to spell it out, to state interview date and location, birth date, maybe occupation. 
enter year, for ex.: 00/00/1947 +Add
Let’s say you have a Segment in which the HistoryMaker is talking about Martin Luther King, Jr. and the Selma-Montgomery Rights March. 
There may be a slight difference between the time codes of the original videotape and those of the MPEG digital video you see in the Segmentor. The reason for this stems from the fact that some of the “color bars” (recorded at the start of each videotape) are often digitized along with the actual interview footage. The MPEG time code begins at zero immediately at the start of whatever video was digitized (thus including whatever length of color bars was digitized - -see left). But the Betacam video time code starts immediately AFTER the color bars (also see left). However many seconds of the color bars got digitized is the amount you need to “offset”, so that the Segmentor (and later the end-user version) will show the correct time codes.