Between 1984 and 1993 I worked professionally in Broadcasting on the Engineering side of the house, mostly at TV stations. Right around 1990 there was a new low cost video tape format on the market, Hi-8, that was supposed to be better than the bulky 3/4″ U-matic. The TV Station I worked at decided to try it and bought three combination camera / recorders. They were given to three of the News Videographers who were told to shoot with them and see what they thought.
Keep in mind that adding a new format took more than just putting gear in the field. We had to equip some of our 5 edit bays with Hi-8 playback decks (the final edits were still cut to U-matic so they could be played on air). So the station had a pretty hefty investment of both time and money in this new format.
So when the more discerning Videographers started complaining that the Hi-8 material did not “look as good” as the footage shot on U-matic, we in the Engineering Department had to figure out why.
My boss, the Chief Engineer, came to me and said that we needed to put together a comparison of all the various portable video formats at the time and grade them from best to worst in terms of “how good they looked”. I also had to explain why each looked the way it did.
To summarize, the portable formats of the day were, in no specific order, with the tape width and whether each was a professional or consumer format:
- 3/4″ U-matic (pro)
- 3/4″ U-matic SP (pro)
- 1/2″ BetaCam (pro)
- 1/2″ BeatCam SP (pro)
- 8mm Hi-8 (consumer)
- 8mm 8mm (consumer, yes the format name and size are the same)
- 1/2″ Super-VHS (consumer)
- 1/2″ VHS-HiFi / VHS (consumer)
- 1/2″ Super-BetaMax (consumer)
- 1/2″ BetaMax-HiFi / BetaMax (consumer)
VHS-HiFi and VHS were grouped together as they had the same video performance, same for BetaMax-HiFi and BetaMax. Note that BetaMax is a consumer format while BetaCam is a professional; so you really can’t say just “Beta”, even though the tape cassettes are the same, they are loaded with different types of tape. You can use base BetaCam tapes in a BetaMax (I did), but you could not go the other way, nor could you use BetaCam SP tapes. The U-matic SP and BetaCam SP formats are different enough from their non-SP cousins that we looked at then separately, and it shows in the results. I did not include SMPTE 1″ Type C in the above list even though we had a “portable” VPR-20, it was not really portable enough for News Gathering, which was the end goal here.
EDIT: Someone mentioned the Japanese industrial formats of the 1970’s used in the Sony EV-200, EV-3xx series, AV-34xx series (portable), and AV-36xx series. These were 1″ (EV) or 1/2″ (AV) and generally B&W, but there was a color adapter available for at least some of the EV-3xx series. I have always seen them described as “EIAJ 1″ for the 1” and “EIAJ 1/2″ for the 1/2”. I have also seen them referred to as JIS format. They were not included above as we did not test them because they were not suitable for broadcast. Then again, some might argue that VHS and BetaMax were not suitable for broadcast either… See my longer port on these formats here.
We already had a bunch of these in house, the ones we did not, we managed to borrow from a local dealer.
I sat down and talked to the Videographers who thought the Hi-8 did not look as good as the U-matic and tried to understand what about the images they did not like. The technical specifications of Hi-8 beat out U-matic, so I could not just repeat the measurements the manufacturers did. I got a lot of very subjective language, but that lead me in a certain direction. While the sharpness was better in the Hi-8 images, the colors did not look right. That directed my testing and our subjective rankings. Since the technical explanation of both how video recording works and the testing we did gets pretty deep, here are the results first.
The Portable Video Formats of 1990 in Order from Best to Worst (Subjectively):
- 1/2″ BetaCam SP
- 1/2″ Super-VHS
- 3/4″ U-matic SP
- 1/2″ BetaCam
- 1/2″ Super-BetaMax
- 3/4″ U-matic
- 8mm Hi-8
- 1/2″ BetaMax-HiFi / BetaMax
- 8mm 8mm
- 1/2″ VHS-HiFi / VHS
Note that the Professional formats did not all come out on top. To say that we, in the Engineering shop, were surprised that Super-VHS came in second is an understatement. We all owned BetaMax decks at home. But, Super-VHS was a marketing term for one (or more) of six separate features that improved the look of the video. The deck we tested probably included all six, but you could put the label on the machine if it included just one.
It was no surprise that BetaCam SP came out on top, it was the most sophisticated of the formats and cost almost twice as much as any other option in the list. The dividing line between U-matic SP and BetaCam was a fine one and on a different day they might have swapped places on the list (it is a subjective list). Those were the only two that were that close.
All of the formats listed, with the exception of the two BetaCam formats, use a technique called “color under” to record color video. The BetaCam formats separate the video into three components (luminance or Y, and two color axis R-Y and B-Y) and record them separately, hence they are considered “component video” formats.
NTSC video, the legacy analog color video specification in the US, is a composite signal consisting of both luminance (brightness) and chromanance (color or chroma). The luminance is the base signal, based on the original US video signal. The color element is carried on a 3.58 MHz carrier that is both amplitude (for saturation) and phase (for hue) modulated. The overall signal is bandwidth limited to just over 4 MHz to avoid interfering with the sound signal (at 4.5 MHZ) in the television RF transmitter. So to record this color video signal you need about 4 MHz of bandwidth. The bandwidth you can record on tape is limited, to use a modern term, by the information density you can write and read from the tape. So the faster you move the tape, the higher bandwidth you can record.
The very first practical video tape recorder developed by Ampex (Ray Dolby of later audio fame was on that engineering team) used 2″ wide tape and moved the heads perpendicular to the tape motion. There were four heads, and as one head was leaving the tape the next was coming into contact. Because the tape was moving linearly, the heads traced a diagonal line across the tape. This format was known as 2″ Quadralplex or quad for short. The video signal was directly recorded on the tape. There were various variations of 2″ quad for black & white and later color (in both high and low quality). These machines were huge and real marvels of mechanical design.
In order to make video reorders smaller, a way to use smaller tape (moving slower) had to be devised. To accomplish this the bandwidth necessary needed to be reduced. By stripping the 3.58 MHz color sub-carrier from the composite video you could do a good enough job of recording the luminance part with small, relatively slow tape (1″, 3/4″, even 1/2″). My High School had a number of 1″ and 1/2″ industrial video recorders that handled black & white (and the 1″ looked much better than the 1/2″, yes, I was a video geek even back then). The video heads still rotate past the tape (to increase the head-to-tape speed), but at a much more shallow angle than the almost 90 degrees of 2″ quad. This class of tape transport was known as “helical scan” because of the path the (typically 2) video heads took past the tape.
Someone then had the brilliant idea to use the superheterodyne principal (see a write up here) to convert the 3.58 MHz chroma signal down to a manageable frequency.
Wait a minute, you say. By down converting the chroma signal aren’t you loosing information? Yes, and that is the crux of this whole story. You loose the ability to discern small color details and rapid changes in color. Each color under system makes a slightly different compromise in terms of how much you loose, but they all loose something.
The reason the images from the Hi-8 systems did not look as good as the U-matic is that the Hi-8 systems had a noticeably lower chroma bandwidth. While the Hi-8 system could resolve a thin vertical pole better than a U-matic system, with the U-matic system you could tell that the pole was red, while the Hi-8 system showed a very well defined grey pole. This is what the Videographers were seeing and not liking. The color details were getting lost. Color bandwidth was not a parameter that any video recorder reported, they all just reported luminance “lines of resolution” and the Hi-8 system was designed to optimize that at the cost of the color.
The other real technical challenge in color under systems is making the color signal line up correctly with the luminance after it has been down converted and then converted back up to 3.58 MHz and recombined with the luminance. And remember, this was all, for the most part, being done with purely analog circuitry. The cost for the converters to digitize video was just getting low enough to see the start of widespread adoption of digital techniques in 1990.