“I’d like to know more about your Letterman databases, please,” said no one ever.
I. The Background
I had been compiling written logs for the first three-plus years of Late Night while recording most of the shows onto audio cassettes. Because I lacked a VCR until mid-February 1985, I’d scribble as much detail as I could so that it would hopefully remind myself what the various sight gags had been if and when I would ever convert those notes into something more permanent, like, say, a digital database. Here’s one page from the audio cassette logs, late December 1983:
I also lacked a computer, and so the search was on for what sort of database application would work best on an as-yet-to-be-determined machine. I had been working two jobs in the early ’80s as a graduate student in Historical Musicology at Columbia U.: The first was during the day in the Ethnomusicology Department, where I would dub and catalog rare reel-to-reel World Music audio tapes for the department’s archives; at night I’d be in the university’s computer lab, inputting Busnoy chansons in their original mensural notation, one coded note at a time, for a database project that would facilitate manuscript source comparisons.
It was the Busnoys project that suggested to me the potential of databases, and I wanted to see how such a project could lend itself towards organizing my Late Night logs. The Ethno. Dept. had bought its first non-networked multi-sectioned IBM behemoth in 1983 (256K RAM as I recall), and it intimidated the hell out of me with all of the time-consuming coding required to accomplish the most simple tasks. It wasn’t for me.
In 1984 I checked out my friend’s home Kaypro to see how it could handle databases. Its rigidity in creating fields and its search limitations left me unimpressed. I scratched that one off the list, too.
I found work as a music editor in an educational book-publishing company in midtown Manhattan in February 1985, a week after I had purchased my first VCR. The offices were networked with Wang computers — built for basic text editing, non-existent for graphics and databases. That summer, the computer manager had the company purchase a Macintosh; it was the earliest model at the time, a self-enclosed unit with a 9″ (diagonal) display, one internal 400K floppy drive, and RAM at a whopping 128K. He kept it on his mostly-abandoned office desk and invited anyone to test-drive it.
I was among the few who did. There was one “feature” that I found intriguing — the ability to sort the Finder window any way one wanted — by name, by file size, by date created. That flexibility was just what I was looking for, so any database that could do that was what I wanted.
At the time, there were several affordable database-focused applications for the Mac, all targeted for the business/finance world: Jazz, Excel, and OverVue. What attracted me to OverVue was that it was RAM-based, which meant processing speed would be much faster than disk-based, depending on the amount of RAM available (it would take Jazz five minutes just to load). Also, OverVue allowed for multi-paragraph-length text cells, a feature then missing in Excel and Jazz.
So in August 1985, I walked over to midtown’s 48th Street Photo and got a free demo of OverVue.
(Standing beside me at the counter was Will Lee. He was there shopping for a computer. I had never met him before, so we chatted for a few seconds, and after I picked up the demo disk and began walking out, I turned around and yelled back, “Get the Mac!”)
I launched OverVue into the office Mac. Because it was a demo, file-saving was disabled, but after inputting just a few fields and rows, I knew that this was just what I needed. It was time to finally buy a Mac for myself.
That happened a month later, in late September 1985. It was the most current model, a 512K RAM monster. The first applications I bought were Word and OverVue. I finally began to create my first Letterman database, transferring all of the hard-copy logs I had compiled since 1982 into the application. The best part was its flexibility; I could expand and create more fields whenever I wanted. It allowed for anticipating data not yet available as well as adding details that would become useful down the line.
It’d be ten years later when, in 1995, I finished inputting all of my Late Night data, then in a 1988-purchased Mac II, later upgraded to a IIfx. (There’d be subsequent desktop Macs bought — a PowerMac 7600 in November 1996; a G4 in April 2000, a MacPro in March 2007, and the MacPro trashcan model in August 2014.) By the early ’90s, OverVue had changed its name to Panorama, and it still thrives today. I use maybe 10% of its capabilities, but it fully serves my purposes.
II. The Databases
A. The Late Night Database (SSIDB)
1. The Source Material
The bulk of my Late Night database came from both my audio cassette and videotape logs. But it wasn’t complete. While I watched and logged in the shows, it was clear that there was a lot of information I didn’t have at the time — song titles, obscure guest spellings, shows taped on days other than when aired.
Much of that information was filled in when, in the mid-’90s, I received logs of the complete Late Night inventory from various sources. I had begun a phone correspondence with the producer of E!’s “Smothers Brothers” syndication in 1992, and when the network began airing Late Nights in December 1993, the Smothers Brothers producer connected me with the person handling the Late Night account.
Through that contact, I was sent E!’s Late Night log in 1994. It listed all show dates, comedy bits, guests, song titles. Here’s the first page, from 1982:
Around a year later, someone at the Late Show handed me a copy of NBC’s own log of the shows. It included more details, mostly pertaining to NBC concerns, such as repeats and contract restrictions. Here’s one page from that log, from 1985:
A short while after I had acquired the NBC log, I was given another log that had been prepared by then-Late Night staffers Barbara Gaines, Barbara Sheehan, and, for a while, Jay Johnson. The three were positioned behind the stage, keeping track of what was going on in front by way of a tv monitor. One version, sent to me from California in ’94, had been compiled in 1989. It was titled “Almost Eight Years.” The final version included the entire 11 1/2-year run of the show, and it was titled, appropriately, “Eleven Years”:
Two sample pages from “The Gaines Log” from 1987 and 1991:
The last log I received was the ‘Talent Payments” book, begun in mid-1984 and stopped in the Spring of ’93. It kept a record of who appeared on each show and how much they were paid. The payment data has been blackened out, but the log’s value was more in the inventory of names. One sample page from 1989 (the week in Chicago):
Both the Gaines and Talent Payments logs have since been scanned and converted into searchable PDFs.
In addition to the hard-copy logs, starting in the mid ’90s I began cultivating contacts with writers, producers, and musicians from Dave’s Late Night and Late Show years. Through them I gained back-stories to comedy material, guests, musical memories, and histories of the writers’ tenures. And a whole lot more. It all went into both the Late Night and NYC Musicians databases (more on the latter below).
2. The Contents
The SSIDB consists of several elements:
a. Basic information on show numbers, dates, and repeats. Here’s a sample image from mid-November 1985 to early January ’86:
b. Broadcast sources from my collection. A sample image from the same period.
c. Sources from others via trades and acquisitions. (Image not provided due to some requests for anonymity.)
d. The main course: Show and band details.
The show details. Each cell contains as much information as I currently have. Here are two screen captures for June 12, 1985, the day “The Late Night Anthem” premiered. Included are details of its origins (because it’s so long, I copied the contents into TextEdit; otherwise, it would have taken 10 pages):
The band details. Each cell consists of all band members who played throughout any particular show. From this data I can locate any participating musician and band (David Sanborn, Bruce Kapler, Charley Drayton, Tower of Power horns, etc.) and determine total counts for each. Two examples from February 25, 1982, and September 14, 1988:
Much of the band information is collated into the “Into the Index” database, detailed below.
e. Various miscellaneous fields: Top Ten lists, TV Guide listings, Viewer Mail running count (establishing that Mark Hamill’s totals were way off), my pre-VCR audio cassette catalog, Cold Open intros to repeats, and other stuff for future video compilation projects.
3. The Name
Why it’s lovingly called the SSIDB: It’s short for “Super-Secret-Informational-Database,” and it springs from the following alt.fan.letterman newsgroup thread from the week of July 2004. Based on a skeptic who suspected and therefore resented that I must have had a privileged entry into a secret online source for all things Dave. Despite hints from fellow posters, he refused to consider the possibility that the information I had been sharing in the newsgroup had all been derived from my own files. The thread, initiated by another fan of mine, was titled “Donz5 Sux Ass.” I couldn’t have been more proud.
B. Into the Index Database
In 1995, I unexpectedly received a package from one of the Late Show producers. It consisted of hundred-plus-page printouts of two databases that Late Night staffers had put together throughout the show’s years at NBC. One consisted of guests, the other of comedy segments. The printouts were sorted in alphabetical order. Here’s the first page of the Guest printout:
And the first page of the Comedy printout:
I scanned the printouts, then, after fixing the OCR errors, I combined them into one database. Rather than alphabetical, I arranged the data by show date/number. Here’s a sample screen capture from June 1983:
I could now search and sort anything I wished. Such as the band members, which I added from the SSIDB. Here’s a sample image showing all of the times bassist Neil Jason subbed for Will Lee during Late Night’s first two-and-a-half years and, in most cases, the reasons for Will’s absences:
The “Into the Index” database is far from finished, as I continue to add staff information from the Talent Payments log, such as those seen in the “Players” field from the first screen capture.
I had also separately received a printout log of Dave’s 1980 morning shows, so I incorporated all of that data into this same file.
Searching through the SSIDB, the Gaines log, the Talent Payments log, and the “Into the Index” log, I can find nearly everything I’m looking for.
C. Closing Credits Database
Every so often there’d be enough time in the Late Night hour to include a full closing credit. I compiled nearly all of them into my “Closing Credits” database. A sample screen capture from July 7, 1982:
Again, now inputted, searching and sorting is a snap. Here’s a search for Barbara Gaines. You can see her gradually rise during the 1982-1989 period from Third Production Assistant to Production Coordinator. (By 2015 she’d be one of Late Show’s Executive Producers.)
There’s another field, not shown, with biographies of and notes from many staffers.
Also not yet finished; there are around 40 more Closing Credits to input.
D. NYC Clubs Database
I spent the mid-’90s summers in the Columbia U. library microfilm room, where I searched its microfilms of Village Voices from the early ’60s up to when I began collecting the hard-copy issues after I moved to NYC in the Fall of ’78. I was initially interested in tracking the Late Night band members’ careers — their recording sessions and public performances, mostly NYC-centric — but it soon expanded into something broader, tracing the career whereabouts of countless other selected musicians (such as members from Encyclopedia of Soul/Stuff; Steps Ahead/Steps; the Brecker Brothers, Larry Coryell, Joe Beck, Steve Marcus, Mike Mainieri, and many, many more).
Here’s one page of notes from January 1977:
And here’s the same data, plus all of February 1977, as input into the NYC Clubs database:
Here’s one of the “Musicians” cells opened up, from Mike Lawrence’s gig at Mikell’s on January 9, 1977:
And, finally, an opened cell from the “Comments” field from the same January 9th gig, an email from Hiram Bullock:
There are currently over 19,000 entries, and its far from complete. Much more to input.
E. Non-Late Night Database
This database consists of as many of Dave’s broadcast appearances as I could find before his career landed at NBC. It includes game shows, Tonight Show guest and guest-host dates (and his guests), and singular appearances. A sample image from 1976 to early ’79, with some air dates still to be determined:
F. Music/Bumpers Database
This one began this year. It documents every song Paul’s band performed during the commercial breaks, as well as all of the bumpers used within those breaks. It’s only up to September 1982 so far. A sample screen capture from June/July 1982:
G. ManWhos Database
Another ongoing project, inputting all of Bill Wendell’s show introductions, in two sections, “From New York…” and “And now, a man who…” A sample image from late March to late April 1982:
H. The Late Show Databases
I maintained my own bare-bones Late Show databases: One consisted of guests and performers throughout the show’s 22-year run; the other focused on the more basic data — dates taped, aired, and repeated and where they could be found in my videotape collection. It’s since expanded to include other show-specific details, such as band song titles as compiled by Micah While from 2002 until 2007.
I plan to incorporate other logs into the latter database: one put together by Late Show media producer Walter Kim from 1993 to September 1996, all of the Wahoo Gazettes that Production Coordinator Mike McIntee had published in Late Show’s online sites from late 1998 until the end in May 2015, and Supreme-Fan David Yoder’s logs that he had maintained throughout all of Late Show. Walter’s and Mike’s are in Word, while David’s is in PDF format.
A couple months after Late Show ended, I was sent the show’s own databases, split up, like their Late Night counterparts, into two categories: Guests, and Comedy Segments, but here as Excel files rather than printouts. Like all projects of this enormous size, it was not error-free: A few music acts would find themselves in the Comedy database; there’d be some misspellings, which would make searching more difficult; and some guests would be missing. All minor and fixable, but it meant that all of those end-of-show stats that were made available to the media were incorrect.
Still, it was an invaluable resource, and once I had converted and combined both Excel files into one Panorama database, organized it into a more accessible format (for me), and corrected errors when I found them, it ended up saving a ton of time when looking for nearly any detail.
So, as with the Late Night data, combined searches through Walter Kim’s and David Yoder’s logs, Mike McIntee’s Wahoo Gazettes, and the Late Show’s own databases, as well as my own, I can find nearly everything thrown at me.
I. You Tube Videos Database
Lastly, my You Tube Videos Database. It seemed wise to keep track of the 1,400-plus videos I’d so far uploaded onto my You Tube channel, and a database was the logical solution. It’s gotten to the point when I’ll get a video request and will have to check the database to see whether I’ve already put it up. And if I have, it’s a simple and quick matter of just providing the link.
There you have it. Drive safely.
P.S. A chronicle of my complete Late Night video collection and its conversion to digital is here: https://donzblog.home.blog/2019/01/16/the-journey-begins/