HD NBTV?

[Steve Anderson]

At this stage without a sane enough storage system I'm going to have to go for 48-lines, perhaps if I lash out and buy a 'proper' soundcard, that might allow further progress.

For me, being somewhat of a CRT freak, a change of timebases/directions/aspect rations is all very simple, no new discs to drill.

As I mentioned before some of the Creative products look promising, the X-Fi Platinum in stereo mode has a bandwidth of 88kHz on each channel. This would allow for a good luminance channel on the left, and sound plus a colour sub-carrier on the right.

But I think I would need to upgrade my main PC to handle the data rates required, it is a few years old now....a bit like me really...

Steve A.

[DrZarkov]

I will go for 48 lines, too. I'm not sure which system I will choose, a mirror-drum or a lense-drum in combination with a rotating mirror. The advantage is that going for more or less lines is just a question of changing motor-speed and focus of the light spot. the disadvantage is much more work. Advantage of the mirror-drum are possibilities to make the monitor 3D with two light (and picture) sources, by using only one scanner and no shutter.

Now it depends on the software, or I need my own camera...

[Stephen]

48 lines...neatly fits into the bandwidth of 48kHz wave files and even an audio CD. But although Volkers pictures show what an improvement it is, just a 50% increase doesn't quite satisfy me.

60 lines...now we're getting serious, but unless using 96kHz sampling on a sound card most simple storage methods are non-starters.

With an overlapping scanning frame sequence, wherein the even frame lines overlap the odd frame lines by 1/2 line, a 48 line system may equal or exceed the resolution of the displayed 60 pictures. Assuming that the Kell factor is the ordinary value of 0.7, the 60 line picture would have an effective vertical resolution of 42 pixels. Overlapping scanning frames should push up the Kell factor to about 0.9, giving an effective vertical resolution of about 43 pixels, better than the 60 line picture with less bandwidth.

[Steve Anderson]

After what must be almost two weeks of debating with myself on what system to go with, I've come up with the following initial format specs....

48 lines, 12.5Hz frame rate, 4:3 aspect ratio (slightly landscape).

Line scan, left-to-right, frame scan, top-to-bottom.

Line rate = 600Hz, 1.667ms/line.

At a sample rate of 48kHz this gives rise to 80 pixels/line, of which 64 are active video and 16 are used for blanking/black-level clamping.

This will be a sync-less system, no sync pulses, in todays world, they're not needed.


..the last thing I'm working on is frame ID and perhaps for the future, colour. This is an interim spec as I would like to go to 60/80 lines once I can get a suitable storage medium, which for me is the bugbear of 60/80 lines at the moment.

Steve A.

[Steve Anderson]

On the 48-line format front mentioned above I've made some progress in that I've got some video out of the PC soundcard. Attached is a grab of same showing how (at the moment) I intend to do frame sync. This burst of 24kHz will vanish in time once I get the display up and running, or at least will be very shortened.

This is real video and it can be seen the the entire gamat of the 256 8-bit levels for the video is being used with black level reference between lines. This is akin to JLB who didn't use any sync pulses either.

Steve A.

[Stephen]

This is real video and it can be seen the the entire gamat of the 256 8-bit levels for the video is being used with black level reference between lines. This is akin to JLB who didn't use any sync pulses either.

This is very good, Steve. I agree: We don't need no stinkin' sync pulses!

[Steve Anderson]

We don't need no stinkin' sync pulses!

...so we agree to meet next Wednesday at 3:30pm...we both have moderately accurate timepieces, so there's no excuse short of some other factor that we will not be there.

With the 'conventional' 32-line system, a line is 2.5ms, I know how long that is, why do I need you to tell me the time? Once a year we reset our clocks and calenders on the first of January. I don't need a phone call from you to tell me it's Wednesday at 3:15pm and remind me.

You get my drift?

I've done a small modification to the waveform as the clamping circuitry was getting a little upset by the 24kHz burst....it's getting there, but then again, so is British Rail.

Steve A.

[Klaas Robers]

However the time burst occupies one complete line. One out of 32, that remains only 31 lines.

In the beginning there was also a complete line or half a line sync pulse to indicate that a new frame starts. This was the same as in the 819 line French TV-system. However that occupied also a complete line of video.

This was less a problem for CRT (oscilloscope tube) monitors, as they still needed some time to sweep back and discharge the rather large capacitor. But for Nipkow discs this line was spoiled.

And the sync pulses were any way ment for this CRT monitors. They were introduces by Freek Kerkhof, a Dutch guy that transmitted 30-line Baird standard TV every week at the 80 meter amateur band. He started about when the BBC stopped their transmissions and continued until 1939 when the Germans conquered the Netherlands. He worked with a small oscilloscope tube to display 30-line TV.

Nowadays we need no sync pulses any way. If you synchronise your display to an atomic clock then it is kept in sync for hours. Even the blanking is not needed. why not to start a new frame at the exact start of a second? GPS provides that information. So I don't see the need of the 24 kHz burst.

[Steve Anderson]

However the time burst occupies one complete line. One out of 32, that remains only 31 lines.

Nowadays we need no sync pulses any way. If you synchronise your display to an atomic clock then it is kept in sync for hours. Even the blanking is not needed. why not to start a new frame at the exact start of a second? GPS provides that information. So I don't see the need of the 24 kHz burst.

Klaas, I cannot agree more, you're abosultely correct. As I mentioned this is an interim method I'm using. I'm just keeping those interested updated on progress....at least I hope there are at least a few that are interested....if not, I'll shut up...

Next thing is I'm using 48 lines, not 32, in a 4:3 aspect ratio which in effect is a 64 x 48 pixel picture, landscape. Using (as an interim as I have said) one line out of 48 for frame sync is somewhat less that the standard 525/626 system that uses some 8% of frame/field time for vertical blanking and sync.

OK, that 'dead time' has come in useful for Teletext, ITS, VITC and other uses, but that was simply a result of the technology available at the time of the development of the system. This is not different conceptually.

As I mentioned, I intend to get rid of the 24kHz burst, again this is all interim and I'm not asking anyone to follow my example.

For those interested in a reasonably priced GPS frequency and time reference here's where I have purchased two:-

http://www.jrmiller.demon.co.uk/project ... frqstd.htm

...you'll never have an excuse to be late again...my Nixie clocks are so unforgiving in their display of real time, I have no excuse any more...

Steve A.

[Klaas Robers]

Oh Steve, a Nixie Clock. Why are people alway doing the same things? However, my Nixieclock is 60 times less forgiving than yours.

What you see on the foreground is a VLF receiver that receives the time encoded in second ticks from a transmitter on 77,5 kHz, somewhere in Volkers country. A microcontroller of the type 8051 decodes and encodes everything. I made a description for the students, however it's in Dutch....

[Steve Anderson]

Oh Steve, a Nixie Clock. Why are people alway doing the same things? What you see on the foreground is a VLF receiver that receives the time encoded......

Yes, it's amusing and interesting at the same time. I have a commercially made clock that can use the MSF signal from the UK, others in Europe and the one used in Japan. It was given to me as a gift when I was last in the UK.

Sadly it's useless in Asia as none of these signals are transmitted. Oh well...

But once in a while it does aquire the signal from Japan and sets the time, wonderful, but it's two hours ahead of our local time. So I have three clocks beside the bed, don't ask me why, we just have. One on local time, one on GMT and the other on Japan time...bizzare.

As ever, your build standards are far better than mine...very neat, very tidy.

Steve A.

P.S. I did add a seconds display the the Nixie clock in time.....sorry that was intentional. Anyway, let's return to the original theme of this thread....

[Steve Anderson]

Nowadays we need no sync pulses any way.

This will be a sync-less system, no sync pulses, in todays world, they're not needed.

We don't need no stinkin' sync pulses!

...it seems the sync pulse is doomed...long live the sync pulse.

Steve A.

[Steve Anderson]

So I'm looking for a method of frame identification, currently I'm using a burst of full-amplitude 24kHz which as Klaas points out uses up a complete line, just over 2% in a 48-line system. The burst could be shortened reclaiming a large percentage of the line, but if you don't have a complete line as a result, there's not much point.

I need some brainstorming here to try and eliminate that overhead.

I want my 48th line back.

I also intend to 'close the gap' and at least reduce the percentage of time devoted to black-level clamping.

I will confess that I have considered using the other channel in a stereo system, but I feel that would be cheating.

I've also considered a form of 'preamble', wherein at the start of playback a time reference is given, i.e. "Follows is frame one, if we both keep time there should be no problem." By using a GPS derived reference a re-sync once an hour would be more than enough. Which would be no more annoying than a commercial or a promo.

Steve A.

[Viewmaster]

We don't need no stinkin' sync pulses!

I for one would not wish to see part of TV history omitted, especially on NBTV.
Why the ol' sync pulse is as much of historic TV as eggs are to
"Egg, sausage and chips" here in UK.
So long live my historic sync pulse, warts and all.
Albert.

[Steve Anderson]

I for one would not wish to see part of TV history omitted, especially on NBTV. Why the ol' sync pulse is as much of historic TV.....So long live my historic sync pulse, warts and all. Albert.

I guess it all depends on how 'authentic' one wishes to be. For the true purist it would be 30 lines with that unequal outer-edge spacing that JLB used. As far as I'm aware no-one is currently using that standard. But I'm sure I'll be proven wrong.

The same goes for our lowly friend, the sync pulse which JLB also didn't bother with. A part of television history, yep, it sure is, and today it's surplanted with audio data, a development of the old sound-in-syncs system. The sync pulse needs to be acknowledged but its days are over.

Steve A.

...as much of historic TV as eggs are to "Egg, sausage and chips" here in UK.

P.S. I'm a vegetable...oops, I mean vegetarian. Or maybe I was right in the first place.