Analog video limitations and video synthesizers

roxics

roxics

Veteran
I got some video 8 footage transferred to DVD yesterday and I was looking at it. Got me thinking, I switched to digital video back in 1999 and haven't looked back since. The footage I got back off those Video 8 tapes was far from the best we see today (or even back then). Interlaced, low resolution, low dynamic range, etc. But how far could analog video really go today with modern sensors and technolgies? Could we have composite or component analog inputs/outputs that are 4K and look really nice? What would those connections look like? I assume an old yellow RCA composite cable wouldn't be good enough since we required an S-Video connection for only a slightly higher resolution.

The biggest reason I started thinking about this is because I recently watched that documentary on Netlfix called I dream of wires about the resurgence of analog audio synthesizers and the less expensive Eurorack modules people buy to create interesting sounds. Got me wondering if we'd ever go back and experiment with analog video in the same kind of way, patching various video feeds though different modules to get different effects. If we could potentially do that with higher resolution footage through some kind of 4K analog connection.

There is something kind of fascinating about that idea to me. Which is probably the reason the audio guys do it. I'm not saying we'd come up with anything better than weird psychedelic videos only really usable for dance tracks, but it's still an interesting idea in an age where so much is just done with software.
 
Analog video was constrained first and foremost by the requirement of fitting the audio and video signals into the constraints of a 6 MHz wide broadcast TV channel, and the equipment never really grew beyond that limitation, even as VCRs and DVDs gradually started to erode over-the-air delivery as the dominant source of programming for many viewers. This limitation gave us such horrors as interlacing, the chroma subcarrier and its resulting low bandwidth assigned to the color content of the signal, and all the other sins associated with what we consider to be the "analog" video look.

But even capturing and recording a high quality 525-line analog signal took some pretty fancy and expensive equipment. Broadcast quality cameras were typically $40,000 and up, and an analog 2" quadruplex or 1" Type C videotape recorder could easily cost $100,000. The industrial and consumer formats like U-Matic and VHS imposed even more terrible compromises on the quality of the signal.

There were a few competing proposals for analog high definition broadcasting floating around in the days when the FCC was first considering creating HD standards for broadcasters, including some that used double-width (12 MHz) over-the-air channels, but these were swept away when they decided to simultaneously require broadcasters to adopt digital transmission via the infamous 2009 transition. I doubt much equipment for these old experimental analog HD systems was ever produced, and if it had been, it would have been a long time before prices became competitive with SD systems for consumer or low-end professional use.

The information content of a basic 1080i60 broadcast HD signal is about six times that of an analog SD transmission; it fits (with some compromises) in one of the old 6 MHz channels only because of digital compression, which is much more effective and transparent (although by no means perfectly so) than anything that's possible in the analog domain alone. Double that information content again to make 1080p60, quadruple it again to get 2160p60, and the equivalent analog bandwidth of a 4K video stream is about 48 times greater than an analog SD signal -- roughly 288 MHz by my reckoning, if you really want to retain something resembling 4K quality. This could be off by a factor of two or four if you want to compromise on frame rates or come up with some clever encoding of the color information as was done in NTSC and PAL TV, but any way you slice it, the bandwidth is huge for an analog HD signal.

Problem is, I don't think there exists any analog recorder using any medium that can do justice to recording a low-noise, high-bandwidth signal of a couple of hundred megahertz or more. There has been no justification for developing anything like that since digital recording has become the norm. Your desktop computer can fairly easily process a digital bit stream of that tremendous rate, but the only analog equipment I know of that can handle it is designed for instrumentation work in areas like radio astronomy -- very specialized, staggeringly expensive and probably not adaptable to video as we know it without a tremendous amount of re-engineering.

The other problem is that you'd be stuck with designing most of the specifications and equipment for such a niche system by yourself. There are no standards, and no companies like Sony, RCA, Ampex and the others who pioneered analog video equipment in the "old days" to lend you their teams of engineers to help out. The present and future are digital, and they are focused on the products that will make money in that domain.

Another way to look at it, though, is that a lot of old SD analog equipment, even some pretty high end stuff, is now being consigned to dumpsters as the industry switches to digital, HD and beyond. There's probably never been a better time to pick up old cameras, switchers, VCRs and such cheap or for free if you want to experiment with it, take it apart, rebuild it and create your own little world of old-school production. (Essentially, this is what's happening with the audiophile gear you mentioned.) As long as you don't try to fundamentally push the standard beyond the original resolution and frame rate limits, which seems a monumental task, there's a lot of room for an individual to fool around and create new and exciting art.

My "Second Law of Technology Acquisition" has always been: "Any high-tech equipment that you covet but cannot afford will be available for free in a dumpster somewhere in 20 years."

(And if you're wondering, the First Law is: "Always buy the bigger battery.") :)

- Greg
 
Last edited:
Batutta said:
Film is analog.
Click to expand...

Technically it's not analog. It's chemical. In order for something to be analog it needs to have an analog signal and be electronic. At least that is my understanding.
Although today we use the word analog as a catchphrase to encompass anything that is physical and not digital. Such as analog as used in the e-cigarette community to represent traditional tobacco cigarettes compared to electronic versions, most of which are not digital either.
 
Greg Smith said:
Analog video was constrained first and foremost by the requirement of fitting the audio and video signals into the constraints of a 6 MHz wide broadcast TV channel, and the equipment never really grew beyond that limitation, even as VCRs and DVDs gradually started to erode over-the-air delivery as the dominant source of programming for many viewers. This limitation gave us such horrors as interlacing, the chroma subcarrier and its resulting low bandwidth assigned to the color content of the signal, and all the other sins associated with what we consider to be the "analog" video look.

But even capturing and recording a high quality 525-line analog signal took some pretty fancy and expensive equipment. Broadcast quality cameras were typically $40,000 and up, and an analog 2" quadruplex or 1" Type C videotape recorder could easily cost $100,000. The industrial and consumer formats like U-Matic and VHS imposed even more terrible compromises on the quality of the signal.

There were a few competing proposals for analog high definition broadcasting floating around in the days when the FCC was first considering creating HD standards for broadcasters, including some that used double-width (12 MHz) over-the-air channels, but these were swept away when they decided to simultaneously require broadcasters to adopt digital transmission via the infamous 2009 transition. I doubt much equipment for these old experimental analog HD systems was ever produced, and if it had been, it would have been a long time before prices became competitive with SD systems for consumer or low-end professional use.

The information content of a basic 1080i60 broadcast HD signal is about six times that of an analog SD transmission; it fits (with some compromises) in one of the old 6 MHz channels only because of digital compression, which is much more effective and transparent (although by no means perfectly so) than anything that's possible in the analog domain alone. Double that information content again to make 1080p60, quadruple it again to get 2160p60, and the equivalent analog bandwidth of a 4K video stream is about 48 times greater than an analog SD signal -- roughly 288 MHz by my reckoning, if you really want to retain something resembling 4K quality. This could be off by a factor of two or four if you want to compromise on frame rates or come up with some clever encoding of the color information as was done in NTSC and PAL TV, but any way you slice it, the bandwidth is huge for an analog HD signal.

Problem is, I don't think there exists any analog recorder using any medium that can do justice to recording a low-noise, high-bandwidth signal of a couple of hundred megahertz or more. There has been no justification for developing anything like that since digital recording has become the norm. Your desktop computer can fairly easily process a digital bit stream of that tremendous rate, but the only analog equipment I know of that can handle it is designed for instrumentation work in areas like radio astronomy -- very specialized, staggeringly expensive and probably not adaptable to video as we know it without a tremendous amount of re-engineering.

The other problem is that you'd be stuck with designing most of the specifications and equipment for such a niche system by yourself. There are no standards, and no companies like Sony, RCA, Ampex and the others who pioneered analog video equipment in the "old days" to lend you their teams of engineers to help out. The present and future are digital, and they are focused on the products that will make money in that domain.

Another way to look at it, though, is that a lot of old SD analog equipment, even some pretty high end stuff, is now being consigned to dumpsters as the industry switches to digital, HD and beyond. There's probably never been a better time to pick up old cameras, switchers, VCRs and such cheap or for free if you want to experiment with it, take it apart, rebuild it and create your own little world of old-school production. (Essentially, this is what's happening with the audiophile gear you mentioned.) As long as you don't try to fundamentally push the standard beyond the original resolution and frame rate limits, which seems a monumental task, there's a lot of room for an individual to fool around and create new and exciting art.

My "Second Law of Technology Acquisition" has always been: "Any high-tech equipment that you covet but cannot afford will be available for free in a dumpster somewhere in 20 years."

(And if you're wondering, the First Law is: "Always buy the bigger battery.") :)

- Greg
Click to expand...

Awesome explaination, thank you so much!
So technically it is possible to have a great 4K analog signal, just way too complex and expensive to be worthwhile. Which I understand. Although if we skipped the recorders and went from digital source through analog processors to play with the signal and then back to digital recorders, I wonder if that would be at all cheaper. Even just a little bit.
Although I did find this, which is interesting. Takes a digital video or in this case a gif and turns it into an audio signal you can then feed through a mixer and back into a digital video signal when recording.

 
roxics said:
Technically it's not analog. It's chemical. In order for something to be analog it needs to have an analog signal and be electronic. At least that is my understanding.
Although today we use the word analog as a catchphrase to encompass anything that is physical and not digital. Such as analog as used in the e-cigarette community to represent traditional tobacco cigarettes compared to electronic versions, most of which are not digital either.
Click to expand...

I'm not sure that's true. I believe Batutta may be correct. After all, analog clocks aren't necessarily electronic.

Webster's definition of 'analog' is "of, relating to, or being a mechanism in which data is represented by continuously variable physical quantities." That certainly sounds like film to me. I suppose it could be argued that light is a 'signal', which the density and coloration of the film modifies.
 
I would very much argue that film is analog, along with mechanical clocks, slide rules, (most) automobile fuel gauges, bathroom scales, rulers and a lot of other devices that are not mostly electronic.

- Greg
 
I wish I could remember the article I read years ago back during our transition to digital and HD, and they said film is actually a digital medium instead of an an analog one.
 
You must log in or register to reply here.
Back
Top