PAL (4.43) to PAL N

[practical]

Here in the US, I've come across a couple of ads for used TV sets originally sold in South America. These sets are so-called "trinorma" types, i.e. triple norm NTSC, PAL-M, and PAL-N. As these sets are usually CRT, they can be had for next to nothing.

I was wondering if one of these sets would be a good option for someone living in the US who wants to display PAL video tapes recorded in Europe (as opposed to Europe, multisystem capability is almost universally shunned by suppliers to the US market).

To be displayed, the European PAL signal would be transcoded into PAL-N, the Argentina-Uruguay-Paraguay 625/50 standard. Transcoding to PAL-N means that the chroma would be moved from its 4.43 MHz to a 3.58 MHz subcarrier. The scan rate would remain intact at 625/50, thus avoiding the need for scan conversion with its associated loss of resolution and motion artifacts.

My questions:

1) I know that as a broadcast standard, PAL-N has a smaller nominal video bandwidth (4.2 MHz) than PAL B/G or PAL I, which is dictated by the FCC-sized 6 MHz broadcast channels used in that part of the world. But as a baseband signal from a VCR playing a tape recorded in Europe or a so-called "PAL" DVD, the video bandwidth of the player should be the factor that in the end determines the (monochrome) video bandwidth. However, I am wondering if the location of the PAL-N chroma subcarrier at 3.58 MHz will confer any loss of quality compared to keeping everything in PAL (4.43) and displaying on an appropriate multisystem set. I am thinking that "Y on C" and "C on Y" may be a bigger problem in PAL-N. Is Y/C filtering more important in PAL-N than PAL B/G?

2) To make the project meaningful, one would need to use a PAL to PAL-N transcoder rather than a universal multisystem converter, as the latter seem to subject all signals to digital sampling and scaling, even if the input and output scans are identical (I could be wrong about this). Dedicated PAL to PAL-N transcoders can be had from Argentine on-line vendors very cheaply, but I am a bit concerned about the quality, maybe especially in terms of video bandwidth. Any leads for an inexpensive, good quality PAL to PAL-N transcoder?

3) A US company sells a NTSC 4.43 to NTSC 3.58 transcoder. I initially thought this would also do PAL 4.43 to PAL-N, but the company told me that it is not trivial that the NTSC subcarrier is 3.579545 MHz as opposed to 3.582056 for PAL N. This statement is confirmed by the fact that trinorma sets have three crystals for chroma detection (a third subcarrier frequency is used for PAL-M). In addition, they said the transcoder must be able to identify the signal as NTSC to work. I am wondering what aspect of the PAL might cause a PAL 4.43 signal to be incorrectly handled by an NTSC 4.43 to 3.58 transcoder. Wouldn't the most parsimonius design of the NTSC 4.43 to 3.58 transcoder be a device that just demodulates and remodulates the chroma than going through complete decoding and recoding? (based on my limited understanding, I am assuming that demodulation is a step that prededes decoding, and just involves separating the chroma signal from its subcarrier without converting it into its B-Y and R-Y components). Or maybe the low cost of NTSC decoding and encoding chips makes the use of them the most cost efficient design?

[ppppenguin]

Going from PAL-B to PAL-N by the techincally best route:

1: Separate chroma (C) from luma (Y), ideally by comb filter, otherwise by bandpass
2: Heterodyne the C down from 4.43MHz to PAL-N SC subcarrier
3: Add Y and C back together

No need to demodulate anything.

The SC frequency for PAL-N is slightly different to NTSC-M. A crystal cannot be pulled from one to the other.

I'd be tempted to modify the set. A crystal change and a few tuned circuits should sort it.

[Patrick Dixon]

Use a VCR with Y/C (S-VHS) out, and you won't even have to separate the Y and C ... which is the most difficult and degrading part of the process.

[practical]

Originally Posted by ppppenguin

1: Separate chroma (C) from luma (Y), ideally by comb filter, otherwise by bandpass

Apropos, if the PAL-N capable receiver does not have a comb filter, does that mean that the bandpass it uses to separate chroma from luma will take away a chunk of bandwidth from the luma signal that wouldn't have been touched if it were a 4.43 set receiving a 4.43 signal?

Originally Posted by ppppenguin

I'd be tempted to modify the set. A crystal change and a few tuned circuits should sort it.

Good point, one would lose PAL-N in favor of PAL 4.43 but keep NTSC and PAL-M...maybe even gain NTSC 4.43.

[practical]

Originally Posted by Patrick Dixon

Use a VCR with Y/C out, and you won't even have to separate the Y and C ... which is the most difficult and degrading part of the process.

Yes, but doesn't the benefit of using a Y/C output assume that Y and C have been kept separate from the beginning? I have heard that if Y/C signals had been mixed to composite just once in the life of the signal and not been completely re-separated before recording, then one would have a problem of luma on chroma that no amount of filtering could remedy. Since most VHS tapes are likely to have been recorded from composite sources whether this was done in private homes or professional duplicating facilities, isn't the harm done already?

(By the way, I have a JVC S-VHS video. I have no S-VHS tapes. When I play regular VHS tapes, I have been disappointed that there is little improvement by going from composite to S-video out. In addition to the above, could a reason be that the luma bandwidth of the VHS system is too small for luma to interfere with the chroma signal in the output?)

[Patrick Dixon]

Originally Posted by practical

Since most VHS tapes are likely to have been recorded from composite sources whether this was done in private homes or professional duplicating facilities, isn't the harm done already?

To some extent - but since the Y and C are recorded separately on a VHS tape, there's no point in putting through a couple more processes if you don't have to - it won't improve picture quality that's for sure.

[ben]

A few years back I found a '90s Toshiba set which was tri-standard American (NTSC, PAL N and PAL-M. I seem to recall I could pipe a PAL 4.43/50hz composite signal thru the AV input and get a result, details are hazy though...

[Patrick Dixon]

OK, since my quote has been edited, I'll expand the explanation a little.

Originally Posted by practical

Yes, but doesn't the benefit of using a Y/C output assume that Y and C have been kept separate from the beginning? I have heard that if Y/C signals had been mixed to composite just once in the life of the signal and not been completely re-separated before recording, then one would have a problem of luma on chroma that no amount of filtering could remedy.

Yes, once you encode Y and C to a composite signal, you will never perfectly separate them again and you will always have an element of cross-colour and cross-luminance in the decoded signal (even with comb-filter decoders). But depending on the actual filtering process used to encode and decode, you'll also have some actual information loss each time you go through the process. So even if you use Y/C down the chain, there are still benefits to be had by doing so, rather than going through the encoder/decoder loop once again.

[cheerfulcharlie]

Why dont you just get a VCR from the same place your buying the TV set?


The old Sony TV sets in the UK could display NTSC at 4.43 even though the recording had obviously been made at 3.58

[ENGLISH VICTOR]

If the hetrodyned chroma signals on tape are the same for PAL and PAL,N then Cheerful Charlie's idea would seem to have merit providing of course the VCR does not object to 60Hz. mains/line frequency. I doubt it would trouble it.
Sony made a device type CFC100 Color frequency converter (their spelling not mine) it converts PAL 4.43 down to PAL 3.58MHz. The results are somewhat bandwidth limited in both Y and C channels but if one could be located at a reasonable price could be the basis of something better. I paid about $15.00 for mine.
The device works both ways, just tried my unit, i.e. 3,5 in 4.4 out or 4.4 in 3.5 out.
The C bandwidth is not much worse than normal VHS systems.
Victor.

[practical]

Originally Posted by Patrick Dixon

Use a VCR with Y/C (S-VHS) out, and you won't even have to separate the Y and C ... which is the most difficult and degrading part of the process.

It is a little bit irritating that VHS VCRs never had Y/C ports unless they were S-VHS. Not even professional machines had them. JVC invented the VHS system and presumably continued to have some say over its use including, for example, the criteria for being able to label a VHS machine "HQ". I wonder whether JVC had licensing rules prohibiting Y/C outputs on non S-VHS machines, to avoid eating into the S-VHS market share, which was lucrative to them.

[Patrick Dixon]

I think some SCART connectors support Y/C but I don't know whether any VCRs are likely to have it. You could also look at modifying the VCR output to get Y and C before they are combined in the decoder.

[David Looser]

Only S-VHS VCRs ever had Y/C on the SCART connector.

[neon indicator]

The colour under subcarrier on tape is AFAIK the same.

A Pal N VHS will play regular Pal 4.433 as Pal N. Who knows what it does with secam!. It will probably play NTSC tapes in some non-standard fashion.

Barco in Europe in late 1970s had AV TV sets that could display nearly anything.

Generally only S-VHS machines have Y/C on outside world, though it's internal on all Colour VCRs

[barcomarc]

Dear friends,

What is de decoder chip used in the receiver? TDAxxxx type from Philips
If it use a multidecoder chip than it is very easy to replace the Xtal of PAL N/M
to PAL BG ( 4.43 or 8.86 )
No need for standard conversion or difficult electronics just change the Xtal en adjust the zero beet
Please look at datasheet of TDA4650 as an example

[geofy]

Originally Posted by practical

(By the way, I have a JVC S-VHS video. I have no S-VHS tapes. When I play regular VHS tapes, I have been disappointed that there is little improvement by going from composite to S-video out. In addition to the above, could a reason be that the luma bandwidth of the VHS system is too small for luma to interfere with the chroma signal in the output?)

Not sure if by S-Video output you mean using an S-Video input lead to another recorder, but as far as I understand this, it is Separate Video and not to be confused with Super Video (more scan lines) as recorded on Super Video tapes. Your S-JVC recorder might have the S-VHS-ET function which allows Super VHS quality recordings to be made onto ordinary VHS tapes, which still can't be as good as using S-VHS tapes.

[geofy]

The S-Video is usually an input only, alongside the component input and audio phonos. There is not an S-video output.

[neon indicator]

often SCART on S-VHS has S-video out
see pinouts.ru

Regular vhs rarely has s-video in or out, but should not be hard to fit if you have sechamtic

[geofy]

Originally Posted by neon indicator

often SCART on S-VHS has S-video out
see pinouts.ru

That's a good point, I have found just the composite video gives the best results, when I have tried separate outputs the colours have never seemed right.

[terrykc]

I've only just encountered this thread, so it may be a little bit late in the day, but here goes, anyway.

The thread seems to have started with with an error based on the assumption that ...

Originally Posted by practical

... To be displayed, the European PAL signal would be transcoded into PAL-N, the Argentina-Uruguay-Paraguay 625/50 standard ...

and the error has trickled through the thread (in fact, it's in the thread title!) with various suggestions including rebuilding the decoder for European PAL - I assume these sets have SCART inputs as converting the IF strip for 5/5.5 MHz vision bandwidth and 5.5/6MHz sound (depending on which part of Europe the VCR comes from) would be a major project on its own ...!

The confusion comes about from the assumption that the video and transmission standards are inseparable - but they are not!

What is recorded on the tape is 625 line PAL, pure and simple. Not system B, G, H, I, N, .... Y or Z. (You're allowed to make up your own transmission standard for this exercise - so long as it handles 625 line signals, it makes no difference!)

The transmission standard is determined by the part of the VCR which is customised for the region in which it is sold. Late in the day, the truth emerges:

Originally Posted by neon indicator

... A Pal N VHS will play regular Pal 4.433 as Pal N ...

Well ... almost. Because it isn't "Pal 4.433", it is just PAL, nothing more.

So, if the OP can find a VCR with PAL N capability, it should be a simple "Plug'n'Play" solution ...