The database called TapeStore is largely completed, built up from hundreds of audio and video recordings and comprising just under a terabyte of raw information. I’ve worked a bit on getting everything organized, but in keeping with the Retrochallenge Theme I need to be able to locate and recall that information in a manner usable on a vintage computer system. Sending streaming video to a Commodore VIC-20 over a serial link is probably impractical, given the low data rate and tiny memory, but can actually be accomplished if we treat the TapeStore server as a peripheral and transmit our program content using standard video and audio cabling to a second monitor connected directly to the TapeStore System.
What I envision at this point is sitting at the console of my VIC-20 (or any other 8-bit antique) and requesting a particular audio or television program from the TapeStore server. TapeStore checks its table of contents, and if the program is available begins playback over a standard video and audio link. It takes an extra monitor and some ancillary equipment but what we have is a version of YouTube circa 1984.
Right now, the plan is to move the TapeStore drives to a Linux system that supports serial terminals and allows shell access via a terminal program running on the vintage PC. Server-side programming accepts the program request and begins playback using dedicated video and audio links to a second monitor located adjacent to the 8-bit console. If I can figure out how to generate a plain-old NTSC video signal on TapeStore, the server software should be a small matter of programming. In the end, all we have is a high-capacity and overly complicated VCR that runs under programmed control, but it’s a neat hack and worthy of the Retrochallenge.
I have nine days to implement this. See you soon!
The number of different media types here is overwhelming… My project started with the Digital-8 Videotapes because of the ease in converting the digital video stream on the tape into a digital stream on disk. Likewise, the 9-track tapes were easily copied from tape to disk-images using the Linux DD command-line utility. Making sense of the data might take a bit longer in some cases, but most of the 9-tracks contained TARballs from earlier incarnations of the Unix Operating System, so decoding them won’t present much of a challenge.
I’ve now moved back into conventional analog media — in this case, audio cassette recordings. I’m using Audacity to record them to disk and generate MP3 files, but the sheer number of audio cassettes (and a few reel-to-reel tapes that I’ll convert with the same setup is pretty daunting. I’ll start with this “small batch” of 100 tapes or so, and work my way through the library given time. I probably won’t get through everything in time for the retrochallenge deadline.
The TapeStore project is moving along slowly. I’ve managed to digitize and upload the first 22 8mm tapes from the Sony Digital Handycam and fill about half of my first 500 GB Hard Drive. While uploading the tapes, I’ve been taking notes and plan on starting my index database soon.
The next phase of the project will be to dig out the Ampex VPR-80 and Time-Base Corrector so that I can digitize and upload the dozen or so 1″ C-Format tapes in the collection. Some of these are edited copies of stuff that was sourced on the Digital-8, so you’ll see a certain degree of duplication. Moving several hundred pounds of vintage video gear is a project in-and-of itself and I plan on filming the operation. Stay tuned for more video updates from Paleoferrosaurus.
Scanning the first few tapes is going well, but it seems I’ve vastly under-rated the amount of online storage I’m going to need to pull off the entire TapeStore Database. I started with a 500 GB media drive on the video acquisition system in addition to the “native” 140 GB Windows volume. This is on a “vintage” PC from 2007 (at 10 years, it actually meets the criteria for the Retrochallenge!) running Windows VISTA. While scanning the 8mm videotapes, I’ve been looking at other media; the VHS tapes are relatively low-res so I’m not worried about the filesize, but even with compression I’m going to be hard-pressed to store all the Ampex videotapes on this system. I haven’t even gotten as far as the 9-track stuff yet.
This screen-shot shows the directory listing for the first 16 tape cassettes. Yes, I can count. Tape 13 broke while playing, necessitating another file (AV8MM0014) to store the remainder of the tape. I have another broken tape that’s going to need some TLC before I can scan it, so the day isn’t over yet!
This version of Windows doesn’t give you much of a preview while importing video, so an outboard monitor makes it easier to see what’s actually going onto the hard drive. I’m also taking notes on the content of each tape while it’s uploading for the purpose of building an index to the content of each tape while building the TapeStore database.
Each file created from an individual tape gets a simple, systematic “name” describing the type of media stored therein. This series is named as follows:
- AV — For “Audiovisual” media
- 8mm — for the source format
- A unique sequence number
Thus, the first file in this series was AV8MM0001 and so on, leading up to the current tape, AV8MM0017. Each tape has a general title, such as “Aurora Band Concert”, a series of named scenes, and notes based on content keywords. I’m not sure yet how I’ll organize the rest of the database, but I’m hoping to come up with something that can link related documents in the grand scheme of things — I might want to find both the original “un-cut” footage and edited copies of a tape, for instance.
It might seem redundant when viewed in the context of a relational database, but I hope to keep much of the meta-data “built-in” to the directory structure.