In the following post, I’m going to talk a little about the 8mm video format (Video 8, Hi8, and Digital 8) and what makes it a difficult format to work with, how XFR Collective hooked up our 8mm deck, and the results of some preliminary testing we did with the deck and our TBC. In trying to write this post, I had a lot of trouble finding resources about Hi8, aside from Wikipedia and random internet links. Most of the information in this post was gleaned from other people’s personal experiences, and our own testing. I would love to see more resources about how to work with Hi8, given its proliferation among media makers, in particular amateur, personal, and non-professional individuals, whose materials are often in the most danger of neglect and obsolescence. If you have links or insights, please share them in the comments!
8mm Video format
8mm video format is a term that encompasses three different generations of video formats: Video8, which is an entirely analog recording, Hi8, which is an analog video and audio recording but has capability for digital audio depending on the equipment used, and Digital8, which is a digital recording format and encodes video in DV. Sony was the original developer of the 8 mm video formats, releasing Video8 tapes and camcorders in 1985, although tapes were manufactured by different companies as time progressed. It was a popular consumer format, which means that it was accessible to a large audience and therefore a lot of content that was recorded on 8mm video formats was amateur, personal, or non-professional.
8mm video format has a reputation of being a difficult format for a few reasons. One problem is the size of the head drum and tape. As is well documented, magnetic media suffers from high rates of degradation. In particular, some 8mm video tapes are made with evaporated metal, not metal particle, which is even more fragile. One of these types of degradation is tape dropout, in which magnetic particles are eroded from the magnetic tape’s surface, essentially a loss of information akin to a scratch on a piece of film, or bit rot to a digital file. As the surface area of 8mm magnetic tape is smaller, a dropout has a much more damaging effect than it would on magnetic tape that was larger in size. Some hi8 tapes are known to curl up at the edges due to the nature of the evaporated metal, which also causes issues with playback.
Another issue with 8mm video is the relative scarcity of playback equipment, a problem that reaches across all video formats to varying degrees. While obtaining working playback equipment is difficult for most formats, 8mm video was primarily used as an amateur, consumer format, which means that the technologies for playback were aimed at the consumer, and probably did not take into account as high a quality of playback as other formats that were used in professional settings. A quick search on eBay shows that professional-grade decks can be expensive, and for an organization that doesn’t have many materials in 8mm video format, spending over 500.00 on a deck doesn’t seem worthwhile. Furthermore, not all decks play back each generation of 8mm video- multiple sellers note that the some decks, such as the Sony EVO-550H (which is a professional grade deck), don’t support the final generation of 8mm video- Digital8. While cameras are much more reasonably priced, they lack many of the important features of decks, and so content captured may not be as high of quality.
Recently, XFR Collective was gifted a Sony GV-D200 NTSC 8mm video format portable VCR Deck, with which we hoped to hook up to our existing digitization rack and begin offering the capability to digitize a few Hi8 tapes as part of our partnership with the puppeteering and theatre organization, Great Small Works. The deck we were given was a consumer deck, which means that it was intended for consumer playback, not professional use. We were able to plug the deck in and power it on, so we knew that it worked- a small victory! We then went about trying to integrate it into the rest of the deck’s setup. As is documented in xfr’s workflow documentation, we are working with the following equipment:
- AJA analog to digital converter
- time base corrector
- CRT monitor
- Mac Computer running Final Cut pro software
Both the AJA converter and the TBC have a limited amount of inputs and outputs, so we unplugged our VHS deck, and plugged in the Hi8 deck using an S Video cable. The video signal travels from the Hi8 deck to the TBC, and then the TBC is connected to the AJA. The AJA is connected to the computer and Final Cut Pro, and (oversimplification) turns the signal into a digital file that is captured according to specific parameters set in our Final Cut Pro capture options. For audio, we connected RCA cables, via an XLR cable, from the Hi8 deck to the AJA.
With this configuration we were in business! We were able to play back a test tape that Ethan borrowed from NYU, of XFR Collective friend and all-around great person Rick Prelinger no less, and capture the signal in Final Cut Pro.
Now that we knew the deck worked and we could capture material, we moved on to testing different settings for capturing content. In an initial test, we noticed that there were a lot of errors with the picture quality. We turned to the extremely helpful and informative AV Artifact Atlas to try and determine what playback errors we were seeing. We noticed the following errors when we captured without any sort of pre-capture adjustment with the TBC: Dropout, Ghosting, Quilting, and Chrominance Noise (pictured below)
We also realized that our Hi8 deck came with an internal TBC, so we decided to see how well this worked- if it did work well, this meant that we might be able to take our Hi8 on the go by connecting the deck directly to a computer.
We ended up performing 6 different tests with various modifications:
Our first test used the external TBC, using the levels on the TBC that were from a previous capture. We experienced some blur, halos, and the image was pretty pixelated.
For our second test, we decided to bypass the external TBC and connect the Hi8 directly to the AJA. We experienced some drop out of the signal, and we noticed that the whites were brighter.
For the third test, just for fun, we decided to see what the quality looked like without running the signal through any type of TBC. Spoiler alert, it looked bad.
For the fourth test, we returned to the external TBC, and reset the levels. We saw ghosting, chrominance noise in background screen, and quilting on diagonals, as well as visual distortion on the edges of the whites (shirt).
Our fifth test tested the Hi8 deck with the internal TBC, and connected the deck directly to the computer. This means that we can’t really do any type of monitoring and adjustments, and we got warnings that our audio sample rate didn’t match and could possible go out of sync. The visual results of this test weren’t great, and although we did successfully capture the signal, it was not of a good quality. However, it did work.
Our final test, which was the best quality transfer we did, was using the external TBC, adjusting the video, black, chroma, and hue levels. These settings made the skin less pink, the blacks had more information, and the picture quality errors, though still present, were much less obvious than in all the prior tests.
This entire process was exciting and valuable for numerous reasons. Working together to do any sort of troubleshooting is instructive- you learn from one another (horizontal membership) and the feeling that you receive from solving a problem, or making significant progress towards solving a problem, is deeply rewarding. Doing this testing also helped us refine our workflow for transferring Hi8 tapes going forward, which is incredibly important for our partners who have valuable materials on Hi8 tapes. We’re looking forward to incorporating our setup into our workflow documentation, and hopeful that we’ll continue learning more about the format. If you have resources or personal anecdotes about working with Hi8, we’d love to hear about them!
AMIA-L archives, Hi8 Preservation Project: http://lsv.uky.edu/scripts/wa.exe?A1=ind1006&L=amia-l#73