Interlaced Video

[Groucho2004]

Originally Posted by Katie Boundary

nothing about CRT technology inherently prevented us from making a TV standard with 100,000,000 scanlines or any other number we wanted.

I don't expect you to understand anything about electronics or manufacturing limits of cathode ray tubes but even common sense should be sufficient to realize the nonsense you're uttering.

[johnmeyer]

Original analog TV (NTSC, PAL, SECAM) was interlaced to provide more fluid motion within the constraints of the bandwidth allocated to each over the air channel (which was the only way most people could view video at home until the late 1970s). When home video recording equipment became available, that constraint actually got worse, because the "engineering tradeoffs" between cost and usability reduced that analog bandwidth even more. This was true regardless of the system: Beta, VHS, 8mm all had similar constraints.

So when TV went from analog to digital, why was interlacing still used, and in fact, used a LOT?

Same problem: engineering constraints. For those not familiar with that term, the more common way to say it is that you can't put ten pounds of stuff in a five pound bag.

The OTA frequency allocation is precisely what it was back in 1948. Digital did not change the laws of physics, and more frequency allocation is out of the question because the electromagnetic spectrum has a finite energy spectrum available for broadcasting.

Even when it comes to delivery by streaming, Blu-Ray, etc., similar constraints come into play. For instance, computer speeds stalled out over a decade ago, and we are not likely to see any faster CPUs for a long time. We are way down the learning curve on video compression (I was very much involved with this back in the early 1990s) and you aren't going to be able to get high quality video into smaller containers. While many systems can satisfactorily play 60 fps progressive material, not all satellite and cable channels can carry it.

The OP seems to think that we can just keep adding resolution and go to higher framerates, but in fact, with the technology we have today and the infrastructure that is in place to carry digital video and audio, even 4K is barely making it to most screens. There is no OTA 4K, and DirecTV and cable has almost no live 4K. Most 4K programming is a download then watch affair.

Finally, true interlacing, where the alternate lines are displayed 1/60 (or 1/50) of a second apart, actually works extraordinarily well and looks just fine. It is only when the display can't handle interlacing natively and only when the deinterlacing is botched that we have a problem. Most deinterlacing built into modern displays works amazingly well: I am very critical and have not once watched my 5-year-old LCD big screen and thought, "wow, I can see deinterlacing artifacts."

So, deinterlacing lets you get the fluidity of 50 or 60 events per second, but at 1/2 the bandwidth (more or less) that would be require to get 50 or 60 complete frames per second.

[Katie Boundary]

Originally Posted by Groucho2004

I don't expect you to understand anything about electronics or manufacturing limits of cathode ray tubes but even common sense should be sufficient to realize the nonsense you're uttering.

Keep in mind that I'm talking about what's physically possible, not what would be practical.

[StainlessS]

Originally Posted by Katie Boundary

Keep in mind that I'm talking about what's physically possible, not what would be practical.

Tut, tut, not physically possible (this century), nor practical.

EDIT: I seem to recall that sample(pixel) signal for PAL TV was something like 30MHz, so if new system has 100,000,000 scanlines
then might be about vertically (100,000,000 / ~625) 160,000 times greater, and making assumption that display is about 160,000 times
wider, then 160,000 * 160,000 = 2.56 *10^10 * 30MHz = 7.68 * 10^17 = 7,680,000,000,000,000,000 Hz.

Anybody know at what frequency X-Rays start ? (C/Wavelength, 300,000,000/Wavelength of XRAY)

XRays start at about 10 nano meters, so, 300,000,000/1.0*10^-8 = 30,000,000,000,000,000 Hz

So, would be well into XRays, or maybe even beyond that, (whats above XRAYS ?)

Numbers above are probably riddled with errors, but probably not that ridiculous.

Me guessin' not possible even beyond this century. (unless you watch from behind 6 inches of leaded glass).

EDIT: My brother is an X-RAY engineer, maybe I ask him (specializing in TV, ie monitors in MRI scanners etc)

[hello_hello]

Originally Posted by Katie Boundary

Keep in mind that I'm talking about what's physically possible, not what would be practical.

Keep in mind I provided a link where it was explained how the physical limitations of early CRT technology limited the number of scan lines used but you've chosen to ignore it.

Even if your distinction was accurate, which it isn't, in the real world what's possible is usually limited to what's practical so they often amount to much the same thing.

[johnmeyer]

Originally Posted by Katie Boundary

... nothing about CRT technology inherently prevented us from making a TV standard with 100,000,000 scanlines or any other number we wanted. The real reason for interlacing was to maximize the screen refresh rate without increasing bandwidth usage.

The second sentence is absolutely true. The first sentence is wrong. StainlessS' response to this obviously incorrect assertion addresses one reason why 100,000 scan lines cannot be done. You also have to dramatically increase the voltages used to deflect the electron beam. As many of the older people in these forums will remember, when the first large color TVs started to appear in the mid-1960s, there was quite a bit of concern about the level of X-rays they emitted. This happened because of the high voltages needed to drive only 525 lines at 60 fields per second:

Ionizing radiation in CRTs

If you build a circuit to drive the scanning lines faster, the x-ray level will increase dramatically. I can't even begin to calculate how much power would be required. If any of you have ever driven by one of the big nuclear accelerators, like SLAC at Stanford or the big cyclotron outside of Chicago, you get some idea of how much energy it takes to deflect particles (electrons in this case).

You also have to build an electronic circuit capable of scanning that many lines in a fraction of a second. The quote Scotty from Star Trek: " We can't even do that in the twenty-third century."

[hello_hello]

A bit of trivia (at least for me).

Where I am, the ABC (Australian equivalent of the BBC) have used the same logo (or variations of it) since the 1960s. Until now I'd been completely oblivious as to it's origins. I stumbled on this while reading other broadcast related pages. You can learn something every day.....

https://en.wikipedia.org/wiki/Austra...ous_curve_logo
https://en.wikipedia.org/wiki/Lissajous_curve

[Groucho2004]

Originally Posted by StainlessS

whats above XRAYS ?

gamma radiation. The TV will cause hair loss.

[StainlessS]

Originally Posted by Groucho2004

gamma radiation. The TV will cause hair loss.

Arh well, no change there then.

[Sci-Fi-Fan]

Thank you all for a very informative discussion.

In my op I was specifically referring to Blu-ray technology which was designed / Intended for High Definition Video Content, And have been recently frustrated to find that 2 Blu-Ray boxsets of Tv Shows (I recently brought) produced for
The BBC in the UK contained some older seasons that are clearly upscaled ports of the standard definition DVD versions of the series while the latest seasons are in full HD 1080i/p.

Now HD Capable displays have been around for a good decade now, so I wanted to understand why some tv networks are lagging so far behind in production / filming technology at least for Blu-ray.

As far as ota broadcasting goes, HD content can always be down scaled to fit into broadcasting limitations, but as has been pointed out, the cost of upgrading / replacing legacy equipment for smaller tv networks is prohibitive.

[wonkey_monkey]

Can any of you guru's out there explain to me why a Tv Network here in the United Kingdom is still using Interlaced Video in 2017?

Everyone's still using it. It was, and still is, a very quick and very efficient way to reduce bandwidth. Just chuck out half the lines and let the human brain - or a deinterlacer - fill in the gaps.

Arguably the need for a deinterlacer is onerous, but they were needed and made, and they are quite good these days. When the UK started shifting to digital broadcasts, everyone still had CRTs would require interlaced signals, so no point changing things then. As flat screen TVs came out, they needed to support interlacing. HD specs are old, so interlacing was still part of it, and it's supported by new TVs anyway so there was no good reason to stop using it.

I Guess when they created the blu-ray spec they didn't take the PAL 25fps standard into consideration :/

They did take into consideration. They realised it wasn't necessarily, because it can be perfectly (well, almost perfectly; not so you'd ever see the difference) encapsulated in 1080i50, which is part of the spec. This same encapsulation doesn't work between 1080p24 and 1080i60, so they are both in the spec.

And have been recently frustrated to find that 2 Blu-Ray boxsets of Tv Shows (I recently brought) produced for The BBC in the UK contained some older seasons that are clearly upscaled ports of the standard definition DVD versions of the series while the latest seasons are in full HD 1080i/p.

If it's Doctor Who, then the older seasons were only shot in SD. They're not upscales of the DVDs, but upscales of the SD master tapes.

The BBC had only just started experimenting with HD when Doctor Who restarted. They shot Torchwood in HD partly as an experiment, and as a result some of the first season episodes are a dark blurry mess because they didn't know exactly what they were doing. Effects for Doctor Who would also have been prohibitively expensive in HD back then - Torchwood's were more limited.

[hello_hello]

Originally Posted by davidhorman

If it's Doctor Who, then the older seasons were only shot in SD. They're not upscales of the DVDs, but upscales of the SD master tapes.

My perception of time must be pretty bad. I would have assumed the BBC had been shooting everything in HD before the new series of Doctor Who (which is dead to me now) started.
Obviously not. Seems the first four season were SD. Do they put disclaimers on the Bluray packaging?

After reading your post I consulted Google and did some more reading:

https://en.wikipedia.org/wiki/List_o...leases#Blu-ray
The 1996 telemovie received a Blu-ray release on September 19, 2016. Curiously, despite being shot entirely on film, the Blu-ray release is simply an HD upscale of the 480p broadcast copy

How pissed would you be if you'd paid good money for that?

[sneaker_ger]

Originally Posted by davidhorman

Everyone's still using it.

There are channels that use 720p50 or 1080p50. On DVB-T2 in Germany I think not a single channel uses interlacing. HEVC interlacing is inefficient.
If you look through some older posts on this forum you may also find such result with x264. You still save bandwidth in the encoding chain and power on encode/decode but in terms of DVB bandwidth interlacing is not good anymore.
Think about it: throwing away lines is a very stupid way of compression (as is chroma sub-sampling). It shouldn't be surprising for AVC, HEVC or other modern lossy compression schemes to surpass it.

[SeeMoreDigital]

Too bad that the french 819-line system never took off in the analogue TV world

[Sci-Fi-Fan]

Originally Posted by hello_hello

How pissed would you be if you'd paid good money for that?

I payed over £100 to import "Merlin Complete Series Blu-Ray" from Amazon.com (as it was not availlable in UK) only to find that the first 3 Seasons are SD upscales of questionable quality.

So yeah I was pissed.

And no they don't tell you in the pakaging or description that they are upscales.

[kolak]

Yep- common practice, specially these days when they expect you to do Blu-ray almost for "free". They use any existing master even if it was done ages ago and it's not good enough to use as Blu-ray source.
No one cares anymore, all what counts is sell numbers.
Spoken with my friend who is in BD/UHD BD industry and UHD is doing terribly bad. After initial not bad start now almost no one (except few big places in the whole world) doing UHD projects. When he told me what studios offer for authoring UHD+BD+DVD I'm not surprised- it just doesn't make sense from authoring facility business point. Now it's all in hands of big places which are factories- only numbers count, so don't expect any attention to details or polished transfers. 10 years ago there were many smaller authoring houses in London, today there is none.
My old company left BD+DVD business few year ago and even then it was already barely profitable.
Shame as Netflix with its sub 20Mbit UHD encodes is far from delivering high-end quality experience at home.

[Sharc]

Originally Posted by kolak

..... It's just a "hack" as 25p has been not included in BD spec for whatever reason!

It is included in the specs as 720x576 25p (or 25i), Level 3.2, for Secondary Video.

[kolak]

We are talking about HD.
Yes, you can have 576 25p as secondary video, so why not have HD 25p as primary in the spec?
Decoding power needed for 25p compared to 24p is about the same.

[wonkey_monkey]

Originally Posted by Sharc

It is included in the specs as 720x576 25p (or 25i), Level 3.2, for Secondary Video.

But are they separate, or is 25p available simply as a by-product of 50i, as it is at HD resolution?

Originally Posted by kolak

so why not have HD 25p as primary in the spec?

There's no need for it. 1920x1080x50i covers it.

[poisondeathray]

Originally Posted by davidhorman

There's no need for it. 1920x1080x50i covers it.

"Covered", but not as "ideal" as native progressive .

1) Progressive > MBAFF > PAFF in terms of encoding efficiency or quality at a given bitrate .

2) It's not really an issue with BD playback, and even less so with "50Hz" area equipment, but there is potential for deinterlacing progressive content depending on the playback setup. It's been known to happen with "29.97p" encoded as fields with software (MBAFF or PAFF triggers deinterlace instead of content based detection for weave or 2:2 pulldown removal)