What's so cool about "analog"?

[tooki]

In digital terms, NTSC content was produced in 640x480, and PAL content in 768x576.

No. Digital NTSC video was 720x480.

No. Digital video has always been produced in many different resolutions, including 640x480 (NTSC), 720x480 (NTSC), 720x576 (PAL) and 768x576 (PAL) being some of the most typical. There is no single resolution.

The purpose of my comment, which you took outside of context, was to explain the bandwidth difference in digital terms. PAL has higher video bandwidth which also translates into slightly better horizontal resolution, so using more pixels for it makes sense. 640x480 (NTSC) and 768x576 (PAL) happen to use square pixels for 4:3 material, which makes them less confusing. Their horizontal pixel count difference also roughly captures the analog performance difference.

Towards end of the SD video era, most formats and processed converged to 720x480 or 720x576. Neither have square pixels.

Sorry, I know what I’m talking about, and you’re also wrong about the history of the resolutions.

Don’t confuse the square-pixel display resolutions on computers with the actual resolutions used in NTSC and PAL digital video formats. FYI, the “display” pixel ratios are called the “display aspect ratio” and the ones actually stored are the “storage aspect ratio”.

The first digital video format ever on the market was D1 videotape (1986), which used 720x486 for NTSC and 720x576 for PAL, in both cases a storage aspect ratio of 5:4. Only much later, when digital video distribution on computers became common, did any square-pixel formats become commonplace. But in production, the actual native storage aspect ratios are used.

768x576 is the display resolution when scaling the non-square storage pixels of 720x576 up to 4:3 aspect ratio using square pixels and keeping the 576px dimension untouched. Many computer video playback apps don’t report the storage resolution, instead displaying the already-scaled display resolution, which is why people mistakenly think that 640x480 NTSC and 768x576 PAL are a thing. It didn’t help that when converting to most compressed video formats, the compressor often scales to square pixels and then compresses that. But again, those were never the production resolutions.

Note that the resolutions are defined in broadcast standards, like SMPTE 259M. “There is no single resolution” is fundamentally incorrect.

If you calculate the horizontal x vertical resolution x frame rate to get a pixel rate per second, you’ll find that 720x486x29.97 and 720x576x25 are very, very similar. 768x576x25 would require about 25% more.

(Many later digital video standards did adopt 720x480 for NTSC, dropping 6 lines to get a number cleanly divisible by 8. 576 is already divisible by 8 so it was left alone.)

[Siwastaja]

Yet I did own a fairly expensive A/D D/A digital video interface card, for digital video production, that used 768x576 with hardware MJPEG codec, somewhere mid-1990's. Definitely not SMPTE 259M, but why does this matter? Is there a law which defines what "digital video production" means? I have done a lot of digital video production in whatever resolutions. Now for interfacing with others, the choice will be more limited, but I said production, not distribution. So you are just plain wrong.

There are as many official standards, and industry standards, as one wishes.

But I understand your issue. I originally was unclear, with the intention of discussing the effect of BW put in digital terms - much like one says "C cassette is 6 bits" without it actually having bits -, but I was not clear enough that I meant this. Then you aggressively reply, and posting in a hurry, make a factual mistake in process, which I correct just for fun, but then you are too proud about your experience that you forget that maybe others also know what they are talking about.

But don't worry. It doesn't matter. I know very well you know what you are talking about. I know it because I know what I'm talking about, too. And thanks for good additional information.

Edit:

768x576x25 would require about 25% more.

And, surprise surprise, PAL has 5.0MHz of video BW, whereas NTSC has 4.2MHz; PAL has 19% more. This is exactly the point I was making with the resolutions. 640x480 vs. 768x576 reflects the difference in amount of pixels actually needed to represent the analog BW, and they are also resolutions that were actually used in video production, albeit not the most common ones. Heck, I used them in video production. One example is enough to prove they were used.

[Terry Bites]

The notions of logic and digital are not necessarily interchangeable terms. Logic is philosophical idea with a mathematical expression, while digital refers to a technology made of real stuff. So just because something is logical doesn’t imply its digital in nature. The universe has logic in the "laws" we have contrived to understand it, but that doesn't make it digital. Analog clocks are not illogical. Logical bits are an abstraction- they don't have a mass a charge a location in spacetime or any other physical property. Freaky.

On its way to our screens data has only ever existed as an analogue in stuff.  Which is stuff made of atoms made of atom gizzards made of fluctuations in fields, all real stuff. Analog technology uses analogues of real stuff with other real stuff. Then the stuff performs mathematical operations. It doesn't use numerical methods to approximate a result. I'm likely prejudiced against of digital systems because being more than a little dyslexic, I can’t code for shit. Still horses for courses.

This is the immersive experience in the analog Matrix right here, right now.

Just sayin'

[vk6zgo]

And try mentioning MOSFETs to the youngsters of today, & they think you are referring to big, clunky power switching devices. Most of my old ham stuff has a multitude of dual gate MOSFETs  in the RF/IF.

Oh please. I hate hearing people talk down about young people. There’s no reason to make that assumption. I think it’s reasonable to assume that anyone learning EE now is going to learn that the CMOS technology practically all modern ICs are made with are based on the MOSFET.

It was a bit "tongue in cheek", harking back to "The Three Yorkshiremen", but I was referring to discrete MOSFETs.

Over the time I have been on this forum, I have not seen any reference to MOSFETs in that context here, (apart from when I have referred to them), other than the power devices.

Well, honestly, isn’t that completely sensible? Where else would you use a discrete MOSFET these days? Even the smallest modern MOSFETs can switch surprisingly large currents, and for signals, you’re unlikely to use a discrete transistor (of any kind). Ham radio is a vanishingly small part of electronics these days. Heck, outside of digital wireless protocols (Wi-Fi, Bluetooth, cellular, lorawan, etc), all of which we implement using off-the-shelf modules, radio is of practically no importance in modern electronics.

So you have become "module jockeys?"

Free to air broadcast radio & TV still exists, & Transmitters have long lifetimes, so for people involved in that industry, there is quite a lot of "legacy" stuff, still using some discretes.

Signals can be surprisingly high power, & RF power amps still in use from 2000 to 2014 or so definitely used discrete LDMOS devices, usually as  two separate devices in the same package.
The two amps are connected in parallel, then the outputs of several of those pairs are in turn paralleled & so on.

I had occasion to work on some 2kW transmitters which were used in a SIMS application. I had to individually sweep each discrete amplifier, & match the frequency response to within 1dB, or they would kill the combiner unbalance loads-----AAAARRRRGGGGHHH!

[tooki]

Yet I did own a fairly expensive A/D D/A digital video interface card, for digital video production, that used 768x576 with hardware MJPEG codec, somewhere mid-1990's. Definitely not SMPTE 259M, but why does this matter? Is there a law which defines what "digital video production" means? I have done a lot of digital video production in whatever resolutions. Now for interfacing with others, the choice will be more limited, but I said production, not distribution. So you are just plain wrong.

There are as many official standards, and industry standards, as one wishes.

But I understand your issue. I originally was unclear, with the intention of discussing the effect of BW put in digital terms - much like one says "C cassette is 6 bits" without it actually having bits -, but I was not clear enough that I meant this. Then you aggressively reply, and posting in a hurry, make a factual mistake in process, which I correct just for fun, but then you are too proud about your experience that you forget that maybe others also know what they are talking about.

But don't worry. It doesn't matter. I know very well you know what you are talking about. I know it because I know what I'm talking about, too. And thanks for good additional information.

Edit:

768x576x25 would require about 25% more.

And, surprise surprise, PAL has 5.0MHz of video BW, whereas NTSC has 4.2MHz; PAL has 19% more. This is exactly the point I was making with the resolutions. 640x480 vs. 768x576 reflects the difference in amount of pixels actually needed to represent the analog BW, and they are also resolutions that were actually used in video production, albeit not the most common ones. Heck, I used them in video production. One example is enough to prove they were used.

I wasn’t aggressive, I was simply direct. (Something you do all the time, and it’s fine.) You made a statement that NTSC was produced in ONE resolution which is NOT the resolution within the digital standards for that analog broadcast standard. That’s what I am disputing. You didn’t say “was often” or anything like that. You made a 1:1 association, and that’s incorrect. Certainly I am not “just plain wrong” because you then had to backtrack to a mushier definition of what you meant.

As for the issue of resolvable image detain within a frame (what you apparently meant to say, but wasn’t what you actually said), of course you’re absolutely right: PAL has more than NTSC.

[tooki]

So you have become "module jockeys?"

Free to air broadcast radio & TV still exists, & Transmitters have long lifetimes, so for people involved in that industry, there is quite a lot of "legacy" stuff, still using some discretes.

Signals can be surprisingly high power, & RF power amps still in use from 2000 to 2014 or so definitely used discrete LDMOS devices, usually as  two separate devices in the same package.
The two amps are connected in parallel, then the outputs of several of those pairs are in turn paralleled & so on.

I had occasion to work on some 2kW transmitters which were used in a SIMS application. I had to individually sweep each discrete amplifier, & match the frequency response to within 1dB, or they would kill the combiner unbalance loads-----AAAARRRRGGGGHHH!

So… you’re complaining that young people think of discrete MOSFETs for power applications, and now you’re explaining that one still encounters modern and legacy discrete MOSFETs in… power applications? What am I missing here? What’s bothering you, if young people are only talking about discrete MOSFETs in the situations where one encounters them?!? 

P.S. Why are you accusing me of becoming a module jockey? I’m not a kid, far from it.

[vk6zgo]

So you have become "module jockeys?"

Free to air broadcast radio & TV still exists, & Transmitters have long lifetimes, so for people involved in that industry, there is quite a lot of "legacy" stuff, still using some discretes.

Signals can be surprisingly high power, & RF power amps still in use from 2000 to 2014 or so definitely used discrete LDMOS devices, usually as  two separate devices in the same package.
The two amps are connected in parallel, then the outputs of several of those pairs are in turn paralleled & so on.

I had occasion to work on some 2kW transmitters which were used in a SIMS application. I had to individually sweep each discrete amplifier, & match the frequency response to within 1dB, or they would kill the combiner unbalance loads-----AAAARRRRGGGGHHH!

So… you’re complaining that young people think of discrete MOSFETs for power applications, and now you’re explaining that one still encounters modern and legacy discrete MOSFETs in… power applications? What am I missing here? What’s bothering you, if young people are only talking about discrete MOSFETs in the situations where one encounters them?!? 

P.S. Why are you accusing me of becoming a module jockey? I’m not a kid, far from it.

Yes, you are right---I should have qualified that by saying "Power switching" situations.

Except for those talking about legacy audio, (& even that is rare, many of the analog "audiophiles" seem to be tube enthusiasts) I haven't seen anyone mention MOSFETS in any analog capacity, as distinct from switching, on this forum for years. 

"Module jockey?"-------well, you did use the term "we", so I assumed you were one of the group of people who use modules, "kid" or not!
In any case, I am a very "Old Fart", so you would probably seem "young-ish" to me!

[tooki]

Yes, you are right---I should have qualified that by saying "Power switching" situations.

Except for those talking about legacy audio, (& even that is rare, many of the analog "audiophiles" seem to be tube enthusiasts) I haven't seen anyone mention MOSFETS in any analog capacity, as distinct from switching, on this forum for years. 

"Module jockey?"-------well, you did use the term "we", so I assumed you were one of the group of people who use modules, "kid" or not!
In any case, I am a very "Old Fart", so you would probably seem "young-ish" to me!

Oops, my mistake! When I said “we”, I was thinking (but didn’t state) “we” the electronics industry today, as a whole. I should have been clearer!

Well, according to others I do look young for my age, so thanks! :p (I’m 42 now, but around 3 years ago, on my last real travels, some drunk girls thought I was 25! Oh, what beer goggles can do! 🤣 But I do get routinely pegged at early 30s.)

As for MOSFETs, I don’t think there’s any doubt that the vast majority of them today are optimized for switching applications. Both as a cause and consequence of that, as we do more things with switching (e.g. how both power supplies and audio amps have gone largely switching), MOSFETs operated in the ohmic region become rarer and rarer. (Ironically, the 20 year old Sony (linear) power amp right next to me is MOSFET based.)

I’m partway through a formal electronics program right now (one year and change to go) and it surprises me how much focus is placed on BJT calculations (both linear and switching applications), while in practice we encounter (switching) MOSFETs much more often. (We also did linear JFET calculations.)

I suppose that with the efficiency gains from switching operation and the relative cheapness of digital circuitry to control it, linear is going to get used even less in the future, though it certainly won’t ever disappear. I mean, who would have predicted 50 years ago that it would make sense at all, never mind be cheaper, to perform radio reception in software?

[magic]

I haven't seen anyone mention MOSFETS in any analog capacity, as distinct from switching, on this forum for years.

Power MOSFET oscillators or attempted dummy loads appear every few months

[AndyBeez]

I never understood why people seem so fascinated with "analog".

I've been pondering all week how to reply to the OP?

I could mention the analog/analogue joy of superheterodyne receivers or,  the harmonic enigma of the lissajous figure drawn on a pure analogue oscilloscope. But there is a thing that analogue does that digital will never do better.
This...
...the moving iron VU Meter. Stick it on any front panel and it says class.

VU Meter: https://en.wikipedia.org/wiki/VU_meter

Superheterodyne Receiver: https://en.wikipedia.org/wiki/Superheterodyne_receiver

The Lissajous thing: https://en.wikipedia.org/wiki/Lissajous_curve

[Tomorokoshi]

marantz VU meter.

[rsjsouza]

Not as great of a picture (photogram of a short clip of mine), but a parallel association of four in a reel-to-reel quadraphonic deck is quite an analog beauty. Looking at a Reel-ro-reel in operation is also quite the sight - I am pretty sure you can find various Techmoan videos (this short clip of mine shows the aforementioned Akai in operation).

[coppice]

Not as great of a picture (photogram of a short clip of mine), but a parallel association of four in a reel-to-reel quadraphonic deck is quite an analog beauty. Looking at a Reel-ro-reel in operation is also quite the sight - I am pretty sure you can find various Techmoan videos (this short clip of mine shows the aforementioned Akai in operation).

Its strange how many of those quadraphonic reel to reel machines were made without the simul-sync switches that Teac used to make their recorders great for semi-pro music recording. Most quadraphonic machines sold in tiny numbers, but the Teacs had that ace up their sleeve that gave them access to a much bigger market.

[rsjsouza]

Not as great of a picture (photogram of a short clip of mine), but a parallel association of four in a reel-to-reel quadraphonic deck is quite an analog beauty. Looking at a Reel-ro-reel in operation is also quite the sight - I am pretty sure you can find various Techmoan videos (this short clip of mine shows the aforementioned Akai in operation).

Its strange how many of those quadraphonic reel to reel machines were made without the simul-sync switches that Teac used to make their recorders great for semi-pro music recording. Most quadraphonic machines sold in tiny numbers, but the Teacs had that ace up their sleeve that gave them access to a much bigger market.

Coppice, thanks for the mention of Simul-sync - I had to look it up.

The GX-270D-SS has a similar feature but named Quadra-sync, which I recall let us do quite some interesting mixing, and even letting us play tracks in reverse to find out about those subliminal messages in popular music.

I recall that my dad had some ancient (for the time) mono tapes with four sides. Good times.

[Bud]

What's so cool about "analog"?

[tooki]

The fact that that’s only slightly harder than for software and embedded makes me think it’s not actually that hard compared to the MCU-based digital stuff the analog graybeards like to complain about so much.

[tggzzz]

The fact that that’s only slightly harder than for software and embedded makes me think it’s not actually that hard compared to the MCU-based digital stuff the analog graybeards like to complain about so much.

It is, of course, also helpful to know the absolute number of vacancies as well as the relative number.