A little bit about my current workflow. Likely to be updated in the future (with pictures, samples, and prior implementations) when I am more focused on video tapes again.
I am of the mind that even a digitized copy of a tape from a late 90s mono VCR is better than no transfer of a tape at all. Although I would, of course, like every tape to be copied in as nice a quality as possible, I do not fret too much about if I am using the most high-end options available when I can achieve a quality I am satisfied with. I use a variety of VCRs for playback, with my main VHS machine currently being a Panasonic NV-DHE10 D-VHS machine (previously I used an industrial Sony S-VHS deck) for VHS and its variants and a Sony EDV-9000 ED Betamax deck for Beta and its variants, as well as a variety of machines for other formats and VHS and Beta tapes that do not want to play nicely in the Panasonic or the big Sony. Originally I used a Mitsubishi VHS hifi and a Sanyo-made Sears Betavision, but I have upgraded several times in the years since.
For capturing I currently use a number of Dazzle DVC-100s connected to primarily Windows 10 computers, where the capture cards still work with the right drivers. I use VirtualDub or VirtualDub2 to capture and cut older video where I do not plan on making complex edits. I capture over S-video cables and run everything through a Sony DVD recorder to take advantage of its time-base corrector. I would like to use the FireWire port on the D-VHS machine to capture MPEG2 files both from D-VHS recordings and also regular VHS and other sources passed into the NV-DHE10 but the CapDVHS software to capture the not-DV MPEG stream over FireWire seems to have some issues under Windows 10. I have gotten it to work successfully a few times, but I already have a good candidate for a Windows XP or Vista machine that should be able to handle it better in mind. After capturing to AVI I deinterlace the video and usually have to boost the audio of non-hifi recordings and then save them to MP4. This is a lossy conversion but I do not think I capture in high-enough quality nor anything of important enough subject matter to keep the AVI files around (perhaps this would be worth it if I had a more expensive higher quality capture device and more disk space) and uploading gigabyte MP4s is easier than hundred-gigabyte AVIs on my internet.
At first I would simply cut the commercials out of recordings but since my second period of mass tape digitization I retained both full-length recordings and clips as well as ordered lists of the commercial contents in an accompanying text file.
A variety of different decks were used throughout my last major tape digitization phase. A Sony EDV-9000 will forever remain in my VCR suite and for a time I was running two of them. Now I usually use a later slimmer digital picture Japanese Hi-Band Beta hifi to have some variety in my secondary machine. I finally acquired some S-VHS decks during this period (later 2022 and earlier 2023) and mostly used an industrial Sony deck until acquiring a Panasonic D-VHS deck. With the most recent round of video tapes I began to catalog more information such as recording speed and the audio systems present on the tape as well as what "collection" they originated from. When I next digitize tapes en masse I plan to try to retain the timestamps of commercials and have enough information to be able to take the timestamped lists and be able to re-cut the same commercials from higher-quality transfers taken in the future. I also would not mind uploading a version of my recording database for others to peruse, complete with links to where the content can be found on the Internet Archive or YouTube were applicable. Another future goal is to be able to batch-upload videos of individual commercials to YouTube using a CSV file for metadata in a similar manor to how I currently handle uploads to the FRC Archive.
March 2024
I have successfully uploaded a number of videos to YouTube using the YouTube API and Python however this number is relatively small because I am limited by Google's daily data quota to about five videos a day. As of writing I have not fully uploaded a single block of mid 1990s ABC commercials and have many, many hours worth of recordings I hope to eventually upload. Without a way around this quota limit it is probably not in my best interests to fully automate the process - I automated almost all of the process in Excel fairly quickly, with the output being the lines to paste into the terminal to upload the video from input parameters.
I will continue to try to use the script I have written but I am not sure if it is really that much better than just using the YouTube Studio batch editing functions. Another option may be to upload the videos ahead of time with no metadata and try to use the API to edit the titles and descriptions based on a CSV.
The Internet Archive also has a method for uploading from Python. I may look into getting this set up for the future as well. I am not primarily focusing on tapes right now. I think the last huge batch I did burnt me out on commercials and random TV for a while.
Another workflow optimization that I would be interested in pursuing is some assistance in cataloging commercials. I think my ideal form for this would be some kind of software analysis of the video file's audio transitions, scene/frame changes, vertical blanking interval captions, etc. to ideally form a timestamped list of the detected commercials with a suspected title for the commercial from the caption data that I can then check through and tweak.
I had planned for a time to upgrade to some sort of FireWire capture device, and will probably still do so to capture footage of things that are not VCRs, but I am also interested in the developments of vhs-decode and the domesday duplicator projects and, perhaps when I get around to (and set aside money for) building a new computer I will look into what the then-current state of suggested hardware looks like. I would be very interested in modifying both a Beta and VHS deck to try both formats out.