The PAL Freeze Frame Machine

[Argus25]

This Dr. Who inspired freeze frame machine that I have recently completed may be of interest:

http://worldphaco.com/uploads/THE_PA...ME_MACHINE.pdf

[Graham G3ZVT]

It would be more telling to know when you started it.

Looking at the size of that prototype board, I am a little ashamed to say the one I was working on last night is only 9X9 holes, and contains a single ATTiny85 and not much else.

Graham.
G3ZVT

[Andrew2]

Saints preserve us, what a supreme effort that is. It looks as though it was started sometime in the early 80's and the builder has doggedly stuck to the design during the long and winding process of completing it.
I still do things the way I did 30 years ago, mainly because my component drawers are full of 1980's components.
Our previous telly had a freeze button....

[McMurdo]

Yes our old Ferguson FV26D had digital freeze frame.

[Argus25]

Originally Posted by rambo1152

It would be more telling to know when you started it.

In May this year, it took a bit longer to make than I had anticipated because I could only get to work on it outside working hours and some weekends.

[Terry_VK5TM]

Things change in such a short time.

When I started in television in the early 70's, the freeze frame unit we had was about 4' high, maybe 18" or so square and had a spinning disc on top similar to those of the old laser players.

It was also kept in what was euphemistically called a clean room.

Something like this would have been so much better.

Terry

[Graham G3ZVT]

Originally Posted by Terry_VK5TM

Things change in such a short time.

I remember contemplating how I could mix the video from a couple of B/W Vidicon cameras and perhaps a homemade FSS for captions. Everything would have to be genlocked and non-pro cameras and VTRs were not easily adapted.
I never did make much headway with that project.

Now, using an old PC and a few Ethernet cameras, a child could set up a colour HD studio and stream the output all over the world.

Perhaps that's what Angus intended to remind us, with his frame-store, that things, not so long ago, were much harder.

Graham.
G3ZVT

[TonyDuell]

There was a company in England called PPL (Process Peripherals Limited or something like that) who made what was effectively an analogue video hard disk. The disk rotated 50 times a second (or was it 25 times a second, anyway, locked to the vertical sync rate) and (monochrome) video was recorded one frame to a track using analogue FM modulation (a bit like a video cassette recorder). The heads did not move to different tracks like on a computer hard disk, rather there was a head per track and you could electronically switch between them, thus allowing you to record short sequences (perhaps 1 or 2 seconds) of video and replay it.

I am told they were used for action replays and the like.

There was a version (which I have) that was used as a computer video display. A rack of boards that linked to a DMA interface on a PDP11 (DEC's DR11-B interface). You could create a video frame in the PDP11's memory and then record it on a track of the video hard disk, and repeat. And then replay the frames yo had recorded to a monitor. There were 3 sets of video electronics on this unit, so you could record 3 frames and then play them back at the same time to the R,G,B channels of a colour monitor.

And then semiconductor memory became cheap enough to use instead...

[emeritus]

My Ferguson F01STB digibox has a "Pause" button on its remote that freezes the picture. Can't say I have ever found a use for it.

[suebutcher]

I remember that PAL freeze frame equipment in the 80's produced a still skipped-field image of 300 lines or so, to avoid flicker on the areas of moving image. I assume that the flicker problem was overcome?

[Argus25]

Originally Posted by emeritus

My Ferguson F01STB digibox has a "Pause" button on its remote that freezes the picture. Can't say I have ever found a use for it.

Yes, I'd have to agree. By the time that the Ferguson F01STB digibox had been invented the program material had deteriorated to the extent that there was probably nothing that interested you enough for you to want to pause it.

[red16v]

Sounds maybe like some of you are describing Ampex's HS100 machine?

http://www.cedmagic.com/history/inst...-100-deck.html

The BBC had one, as did LWT and one of the facility houses based in Soho.

[retroteck]

Hi Argus25

Reading your posting has inspired me for ideas on further design work for a stand alone
timebase corrector.

Some 30 years ago i embarked on a framestore project from an article published by an
electronics club in the uk.

The design was quite clever in certain respects. In the days before fast FIFO memory was
widely available the circuit used 16 x 200nS access time static ram chips, two for each of the
8 bit ADC data lines. Because of the slow access time of these rams it meant storing
8 consecutive samples of each of the 8 ADC bits serially into latches and then writing the
parallel output of these into memory every 500 nS or so.
Similarly for reading, each ram chip outputed data to an 8 to 1 data selector and was clocked
out serially to the approprate DAC input data line.
Using static ram obviously meant generating seperate read and write addresses for the
memory as well.

While the basic block design was sound enough I could never get this circuit to work properly,
(clock generator and numerous timing problems), due to some of the circuit techniques
employed in some of the stages, so i eventually abanded it.

However i recently resurected the project but quickly realised that i would like to design a new
circuit from scratch. Your posting has renewed my enthusiasm for this project and i am
thinking on the lines of developing a stand alone framestore\timebase corrector.
Also i always wanted something that was not tied to a pc.

Rod

[Terry_VK5TM]

Originally Posted by red16v

Sounds maybe like some of you are describing Ampex's HS100 machine?

Yep, looks like it, all our video gear was Ampex.

I guess (it was too long ago to be very accurate with the memory), that ours had what may have been a Genlock and or Timebase corrector under it as well.

Terry

[Argus25]

Retroteck,

Yes the AL422B lends itself very well to both standards converters and time base correctors because of its independent memory read/write functions. Since you won't want to digitise the incoming sync too, the AD8709 ADC IC would be a better choice than the AD8708 that I used.

[red16v]

Originally Posted by retroteck

The design was quite clever in certain respects. In the days before fast FIFO memory was widely available the circuit used 16 x 200nS access time static ram chips, two for each of the 8 bit ADC data lines. Because of the slow access time of these rams it meant storing 8 consecutive samples of each of the 8 ADC bits serially into latches and then writing the parallel output of these into memory every 500 nS or so.
Similarly for reading, each ram chip outputed data to an 8 to 1 data selector and was clocked out serially to the approprate DAC input data line.
Using static ram obviously meant generating seperate read and write addresses for the memory as well.

The technique you're describing was used in the Quantel DFS1751 frame synchroniser. It was the leading steady workhorse throughout the broadcast industry in the early/mid 80's. Memory chips in the day didn't have enough speed to be 'loaded' serially and so Quantel adopted a parallel load method - I think it was called 'barrell loading'? - sequentially pop the serial samples into a line of latches then 'barrell' load them all in the memory chips at the same time.

[snufkin]

I would love a frame buffer/store that would allow for some of the zany effects done in the very late 70s early 80s

Have you checked out http://synkie.net/blog/video-delay-now-in-colour/ or http://gieskes.nl/visual-equipment/?file=gvs1 both clever minimal designs

Also AMAZING WORK Argus25!