logo Sign In

PAL vs NTSC laserdiscs — Page 3

Author
Time
Post link
Originally posted by: rnranimal
[the German set] should provide the ultimate transfer, if done properly.

You'll have to define 'properly'. I don't think anyone's gotten further than I have (even though I haven't released jack), and even I am not convinced PAL will be better. It is just incredibly difficult to find a PAL player that will rival the best NTSC-only players. No such animal seems to exist, in fact. I haven't given up (I have a line on a prototype player that may just do the trick), but once again there is a huge gap between the theory of saying "PAL is better," and actually getting better results.

I don't know that any transfer from the NTSC set could [measure up], it just doesn't seem good enough. The PAL set has much more detail and is without the bad jaggies. Not to mention not having to mess with reversing telecine or the 7.5 IRE NTSC black level boost.


The X0 transfer will be the best NTSC can offer. You'll have to see then if it stacks up. The PAL set has somewhat more detail, and includes a fair amount more dust and crap. If you get the Japanese Collector's Set, the black level is at the NTSC-J standard 0 IRE.

Author
Time
Post link
Originally posted by: Karyudo
Originally posted by: rnranimal
[the German set] should provide the ultimate transfer, if done properly.

You'll have to define 'properly'. I don't think anyone's gotten further than I have (even though I haven't released jack), and even I am not convinced PAL will be better. It is just incredibly difficult to find a PAL player that will rival the best NTSC-only players. No such animal seems to exist, in fact. I haven't given up (I have a line on a prototype player that may just do the trick), but once again there is a huge gap between the theory of saying "PAL is better," and actually getting better results.

I don't know that any transfer from the NTSC set could [measure up], it just doesn't seem good enough. The PAL set has much more detail and is without the bad jaggies. Not to mention not having to mess with reversing telecine or the 7.5 IRE NTSC black level boost.


The X0 transfer will be the best NTSC can offer. You'll have to see then if it stacks up. The PAL set has somewhat more detail, and includes a fair amount more dust and crap. If you get the Japanese Collector's Set, the black level is at the NTSC-J standard 0 IRE.



I didn't mean there was just one proper way to do it (or that I knew what it was). I just meant that the way the transfer is done is quite important in my theory that the PAL discs are better and that just because the PAL discs are the source, doesn't mean it's better than any transfer from the NTSC discs. At this point, I find the Editdroid discs to have the best picture (though, it has it's problems and I'm very much looking forward to seeing the X0 & Cowclops transfers), but think it could be outdone by the right PAL transfer. I have seen screengrabs of the current X0 transfer and the PAL grabs still look better to me. I know saying they have much more detail is subjective, but I said that as detail is very important to me and I chose the Editdroid set over the TR47 set for the minor amount of extra detail it had on it. So I chose to say much more detail is found in the PAL discs, because to me there looks to be. You can actually see the stars on this transfer, the crawl and opening shot look so much better. As far as a good enough PAL player not being out there, I see nothing wrong with the sample clips I've seen. But you can talk from actual experience, since you have these things. You certainly are better qualified than myself. These are just the conclusions I've come to from the many, many (and yes, many) hours of research and messing around transferring my own set and dowloading others, which I admit has nothing on hands-on experience. I could be dead wrong and this is another thing that keeps me from springing for the materials needed in my idea of the "ultimate" transfer. You're right about the Japanese LDs having the correct black level and I have seen grabs from them that suggest they would be a slight step up from the US discs. The dust and crap doesn't really bother me and I can accept that and the cropping on the PAL discs, I feel it a worthwhile trade off for the better detail. On the subject, does anyone know if any of the prior PAL releases are worth checking out? I've read posts by a few people that actually prefer the 1989 WS NTSC discs to the NTSC DC/faces sets. I've never seen them myself and can't comment, but the posters claimed that while the '89 WS discs have more dust and noise, the picture is more detailed and not flat looking like the DC transfer. I could go for that if it's true (but either way, they'd look like crap on my player with the CLV smear, so I haven't bothered). I can live with film sourced flaws when trading for them with mastering flaws like heavy handed NR, poor color correction, jaggies and such.
Author
Time
Post link
Originally posted by: Karyudo
The X0 transfer will be the best NTSC can offer. You'll have to see then if it stacks up. The PAL set has somewhat more detail, and includes a fair amount more dust and crap. If you get the Japanese Collector's Set, the black level is at the NTSC-J standard 0 IRE.

Personally I'd prefer a slightly grainier/dustier image over one that looks like this:
http://www.haku.co.uk/pics/StarWarsLDjaggy2.jpg
I mean, c'mon, that's atrocious! I'm seriously surprised the matter of what appears to be 'half the picture missing in removed lines' subject hasn't been brought up before about the definitive collection, nobody has answered wether the 95 Faces set suffer the same picture detail loss and if the Japanese discs do too.

Besides, a grainy image is what was originally experienced in the cinemas all those years ago
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***
Author
Time
Post link
Yeah, that frame sucks. But we haven't seen that same frame from the X0. Might be different; I'm not sure.

I have the DC, Faces, and Japanese Collector's Set. Making some comparison shots is on my lengthy list of things to do...

The PAL transfer has burns on some frames -- you remember that from the theatre?? Nah, me neither.

So far, I like PAL better except for two biggish problems:

1. crap in the picture that will need to be rotoscoped; and
2. dot crawl, even from a CLD-D925.
Author
Time
Post link
I'll post some later, but if you've got Moth3r's ANH, you can find them there. Look in the first couple of scenes with the droids on Tantive IV. Two frames, each with very obvious burn marks.

I've roto'ed them out during my own experimentation, but it's a PITA.
Author
Time
Post link
I don't think anyone's gotten further than I have (even though I haven't released jack),


Karyudo, I have been consistently impressed by your interesting posts and technical knowledge. You are obviously questing for the "ultimate" transfer yourself. I'm just wondering if you are intending to release when you're done. If so, how about starting a thread so we'll be able to follow your progress?
Author
Time
Post link
Citizen, are you sure that the jaggedness is not caused by capturing to DV format?

The "burns" on the PAL discs, if I'm correct, seem to occur regularly on the frame prior to a scene change. However, I think the DC set exhibits the same defects, so maybe Karyudo is talking about something else. I don't have the ability or patience to rotoscope out such things, and I don't find dirt/scratches too objectionable (I feel they add a little authenticity and "retro" feel to the transfer ).

The dot crawl is a bit disappointing, and it's exacerbated by the sharpening. I realise it's because the 3-line 2D comb filter in the D925 isn't the best. Karyudo, do you think that capturing from the composite output of a 2950 with a SweetSpot card (contains a more advanced 3D comb filter) would give a visible improvement?

Guidelines for post content and general behaviour: read announcement here

Max. allowable image sizes in signatures: reminder here

Author
Time
Post link
The DV format has it's issues, but does not cause such jaggedness. It's only real problem is blockiness in solid color areas and that's basically just for NTSC. It's not from the DV format, for sure. I would bet money that not only does that shot not improve (in the way of jaggies) with the X0, but that the faces and Japanese DC all look the same. That's what I was saying in my original post, it's just the NTSC disc's transfer. Still, it's greatly appreciated whenever someone takes the time to do some screengrabs and turns the guessing into fact. Would be glad to help out, if I had any of these discs. Lucas has always been ahead of things and it almost looks as if he did a higher resolution transfer (maybe even HD, which means only more reason why we could've had a nice OT release) and then scaled down to 480i. That could certainly cause this. Of course, HD to 480i scaling now can look a lot better, but maybe it didn't do so well in '93, even with what Lucas was using. Whatever the cause, it's from the transfer. It does look a lot like crappy de-interlacing where they just remove every other line, so maybe somewhere along the line in the film>video transfer, this major boner was made. Wouldn't be the 1st or the last time in the case of SW releases. Oh, wait, maybe it was a creative decision!

The US DC set doesn't have the burns, I know I would've noticed those. You're talking of the burns signaling a reel change, correct? I also am not bothered by these. I actually have seen them on DVD releases before (cheapo ones, mind you) on movie channels (70's movies) and in the theater as recently as Harry Potter 3.

Dot crawl does bother me (the main problem with the otherwise great Editdroid set) and usually can be reduced or done away with when using composite and a good capture card, unless this is one of those players that simply outputs the s-video treated signal from the composite. Then there's really nothing that can be done.
Author
Time
Post link
In most of what I read, it says there's no quality difference between CAV & CLV and I believe in most cases that is probably true.


I always felt CAV looked a lot better. I seem to remember that blacks (i.e., the black bars around a widescreen movie) were much blacker with CAV, whereas CLV would introduce some blue speckling. I assume it's because CAV allows more bits (about double, in total, though it varies from the beginning to the end of a CAV side) for digitising the composite video signal.

DE
Author
Time
Post link
On burns:

What I'm talking about is not the discolouration at scene/reel changes. That looks more like glue than a burn. The two frames I'm thinking specifically of are in the middle of a scene. PAL is a step up in terms of resolution, but it's far from being a panacea.


On CAV/CLV:

I don't believe the signal is different at all between CAV and CLV. LDs are not digital, so the concept of bitrate is a red herring. I suspect that any improvements seen are improvements made because of purely mechanical reasons: optical gain, crosstalk, flutter, wow -- all that sort of good analogue junk. It will perhaps manifest itself in something more electronic-looking (e.g. a better player will maybe do a better job on CAV than CLV, maybe giving a better picture), but I just don't buy that there is anything fundamentally different about the information recorded on each kind of disc.


On dot crawl:

I have yet to try a SweetSpot card. But I'm already using a card with a Philips SAA-7133 chip, which has a 10-bit ADC and an adaptive 3-line comb filter. I believe that's just about what the SweetSpot card has, too, isn't it? So except for build quality, and maybe some goofing around with the Philips reference design, I don't see how things would improve by leaps and bounds -- at least, not on the capture side.

I am also using a CLD-D925, which is apparently one of the two PAL-spec players that is generally regarded as being the best there ever was. I've looked at the schematics, and as best I can understand (I'm no electrical engineer), the composite signal is separated into Y/C early on, processed with a now-obsolete Sony comb filter, tweaked however Pioneer sees fit, and then recombined. And then I think separated again for the S-video, but memory fades. So even at the pointy end, things are thoroughly sub-optimal. Too bad there's no such animal as a PAL-spec X0 player. I'm hoping there's something out there that might come close, but that's but one item on my list of things to do.

To answer Moth3r's direct question, I'd love to get a 2950 and a SS card together, and see what transpires. Problem is, all this stuff takes gobs of time, and costs money. I don't know how many of you have spouses/significant others, but even just having the money doesn't mean you can spend it without repercussions... Still, my wife's cool enough to have tolerated all my fiddling thus far, so I'm not ruling out being able to get all the parts in one place at one time, and see once and for all whether things can be improved substantially from what I'm looking at now.

This is fun stuff, ain't it?


On projects:

I'm not sure I'm ready to start my own thread. For now, you can imagine a swan: looks pretty calm and boring on the surface, but underneath the legs and feet are pumping away! Lots of behind-the-scenes work going on in which I'm heavily involved. If/when I start to get serious about letting the general public know what I'm up to, I assure you that people here on OT.com will be the first to know. For now, I like the freedom of being able to read, lurk, make the occasional comment or observation, and not have to feel too much pressure to keep saying, "I'm still here! And still working!"
Author
Time
Post link
I don't believe the signal is different at all between CAV and CLV. LDs are not digital, so the concept of bitrate is a red herring.

I'd forgotten that LD doesn't store a digitised signal (I think it encodes by the [analogue] length of pits and bumps). But VHS tapes aren't digital either, and if you record in Long Play you get a noticeably worse picture. From a Laserdisc FAQ:

CAV also provides constantly improving signal-to-noise ratio as the program proceeds toward the outer edge, but this is typically not that noticeable on properly manufactured discs.


I'd question the second part, unless all the CLV+CAV discs I've ever had were improperly manufactured.

There's more technical stuff here: http://www.modeemi.fi/~leopold/AV/LdFaq/Introduction.html#MediumClvCavQuality

DE
Author
Time
Post link
Ok, thought you meant the cigarette burns at the reel changes, as I've heard them called. So you actually mean bulb burn like when you're watching your old home movies and it gets left on a frame too long? That's pretty bad, but still wouldn't bother me as much as the NTSC disc's problems.

CAV/CLV issue I know has become quite the arguement in the past (on other boards, don't know about this one) and a lot of people say they're the same quality-wise. However, I've seen it with my own eyes when comparing the faces & DC sets on my V8000. Wish I still had the faces set and I could post some screenshot comparisons. I can't tell you what the difference is within these formats, but one holds half what the other does and looks better on my player, so there is something going on there. So much that originally I was going to use my faces discs for my transfer (I had a stand-alone recorder at the time and wanted the least side changes) and picked up a cheap DC box for the extras. For the hell of it I checked out the movie to see if there was any difference. Right away, I thought it looked better (sharper, no smear), so I pulled out the faces set for a compare and sure enough the CAV set was enough of an improvement that I sold the faces discs, sold the DC set and scoured the universe for a late pressing DC box with the flaws fixed (not just the Leia disc replaced). Actually, I do have the CAV Jaws & ET boxsets which both contain long docs in CLV. I could grab a frame from the movie that is also shown in the docs. Will do that when I get home Monday. Like I said, I don't doubt that some players can make the 2 formats look the same quality, but I also know for a fact some players have CAV looking better. Noise is also better on CAV on my player (for most of the disc). CAV disc's s/n ratio keeps getting better until the end of the disc. When comparing the 2 sets, the CLV had the same amount of noise that the very beginning of the CAV sides have. After a few minutes into the CAV discs, the noise was already improved over the CLV (which has the same noise throughout). Let me put it this way, I WISH my player played CLV as good as it does CAV, I don't see a difference because I want to.
Author
Time
Post link
Originally posted by: rnranimal
I can't tell you what the difference is within these formats, but one holds half what the other does and looks better on my player, so there is something going on there.


I don't think CLV 'holds half' what CAV does. I suspect they actually hold the same amount of signal information, just distributed differently. However, like I wrote before, I suspect purely mechanical considerations come into play. The laser spot size is finite for both kinds of discs, so it stands to reason if the data are further spread out (i.e. in CAV) then it will be easier to read accurately.

Imagine you're re-typing a block of text. If that block of text came in two versions, one on two pages, double-spaced, in a large readable font and the other on one page, single-spaced, in 8-point Arial Narrow, which do you think you'd have an easier time re-typing? Same information, different packaging. I don't know for a fact, but I strongly suspect that this is akin to how CAV and CLV differ.

Author
Time
Post link
Citizen, thanks for taking the time to do that comparison. I guess that proves that this is an issue with the DC disc.
Originally posted by: Karyudo
I have yet to try a SweetSpot card. But I'm already using a card with a Philips SAA-7133 chip, which has a 10-bit ADC and an adaptive 3-line comb filter. I believe that's just about what the SweetSpot card has, too, isn't it? So except for build quality, and maybe some goofing around with the Philips reference design, I don't see how things would improve by leaps and bounds -- at least, not on the capture side. "Adaptive" = "Motion Adaptive" = "3D", correct?
If you check the SweetSpot FAQ, it mentions the 3 notional price bands for capture cards:
1. under £50
2. under £200
3. under £1000
The card I'm currently using is a BT878-based card in the first bracket. In fact I think it was about £20, as there's no tuner on it, just a composite and s-video input. The SweetSpot falls into price bracket 2. Would your card be comparable then?

Originally posted by: Karyudo
I am also using a CLD-D925, which is apparently one of the two PAL-spec players that is generally regarded as being the best there ever was. I've looked at the schematics, and as best I can understand (I'm no electrical engineer), the composite signal is separated into Y/C early on, processed with a now-obsolete Sony comb filter, ... I'm not an electrical (or electronic?) engineer either. I understand the process to be as you described - the composite video from the disc always goes through the 2D comb filter, so will always pick up the dot crawl introduced at that stage. I believe that the composite output is simply the s-video combined. No one seems to know if the 2950 is similar, or if the comb filter is bypassed when the composite output is used on that model (and thus, using a 3D comb filter in such a capture card would benefit the quality of transfer). Where's Laserman when you need him? Also, I have heard some opinions suggesting that the 2950's playback is less noisy than the 925.

Originally posted by: Karyudo
To answer Moth3r's direct question, I'd love to get a 2950 and a SS card together, and see what transpires.
I really want to know what the outcome of such a combination would be. I read that Laserman was going to try this, but don't know if this ever happened. Rest assured, if you or Laserman don't beat me to it, I will definitely be doing some tests of my own in the future.

Originally posted by: Karyudo
Problem is, all this stuff takes gobs of time, and costs money. I don't know how many of you have spouses/significant others, but even just having the money doesn't mean you can spend it without repercussions...
Yeah, I know. I've already called in my birthday present early - the PAL equivalent of the JVC SR-S365U (a rather nice VCR). But my wife's an expert at last minute bidding on eBay, and some shrewd purchases (tip: when searching for the PAL discs, try using krieg or guerre as keywords - sets without "Star Wars" in the auction title usually don't go as high) means that I've not spent anywhere near as much as some of the other preservationists. I think I've proved that you don't need to spend a huge amount of money to get a fairly decent transfer of your own.

Guidelines for post content and general behaviour: read announcement here

Max. allowable image sizes in signatures: reminder here

Author
Time
Post link
I suspect they actually hold the same amount of signal information, just distributed differently.


Might be getting a bit philosophical here, but is it still the same amount of information if we can't extract the same amount of information? Any system of measurement is going to get a more accurate reading from a CAV disc than a CLV disc because the SNR is lower, like the FAQ said - even if it's the best pressed laserdisc in the world and all the noise is coming from the player's circuitry.

Build a better laserdisc player and I think it would still get more "information" from a CAV than a CLV.

Edit: I've just remembered. When I made a DVD from my SE LDs, I did an experiment. I captured the same scene multiple times and averaged it, to see if I could reduce noise. It barely made any difference, which suggests to me that "noise" is mostly already encoded on the disc by the manufacturing process and its limitations. The slight difference I saw would have been from averaging out the lower noise introduced during reading.

DE
Author
Time
Post link
I don't think CLV 'holds half' what CAV does

I meant that CAV holds half the amount of video than CLV, timewise.

I suspect they actually hold the same amount of signal information, just distributed differently. However, like I wrote before, I suspect purely mechanical considerations come into play. The laser spot size is finite for both kinds of discs, so it stands to reason if the data are further spread out (i.e. in CAV) then it will be easier to read accurately.

Imagine you're re-typing a block of text. If that block of text came in two versions, one on two pages, double-spaced, in a large readable font and the other on one page, single-spaced, in 8-point Arial Narrow, which do you think you'd have an easier time re-typing? Same information, different packaging. I don't know for a fact, but I strongly suspect that this is akin to how CAV and CLV differ.



That's actually kinda what I was trying to say, but you said it better. I'm not trying to say that I think the CAV discs actually have more or even better information than CLV (and I'm not saying they don't, I don't know). I just mean that my player (and surely some others) do play CAV & CLV with different results, whatever the reason. Which causes me to prefer the CAV format, regardless of wheather the CLV have the same info or not. As said previously, if it can't be extracted, then it doesn't do me any good if it's there or not. If my player played them both at the same quality, I wouldn't care which format I had.
Author
Time
Post link
My German THX LD set arrived this morning I'd already got bored so started work on a DVD transfer of my Deep Red (1994) LD, not the Italian horror film but a not so known sci-fi starring Michael Bein (Terminator, Aliens) & John de Lancie (Q - Star Trek TNG) so I'll finish that and then I'll set to on making my SW DVDs.

But first a comparison, cropped non-scaled caps of the definitive collection vs the German THX collection:
http://www.haku.co.uk/pics/StarWarsLDjaggy2.jpg
http://www.haku.co.uk/pics/StarWarsLDjaggy4.jpg
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***
Author
Time
Post link
First impression is that the PAL disc actually looks softer than the NTSC. The horizontal distinction between the white lights at the top right is more defined in the NTSC version. Makes me think that the awful jaggies in the DC could be caused by some sort of sharpening filter?

Guidelines for post content and general behaviour: read announcement here

Max. allowable image sizes in signatures: reminder here

Author
Time
Post link
Yeah the difference between the grid of white dots on the right (some control panel?) is quite noticable in terms of sharpness, but the PAL set doesn't suffer the jaggies which is much, much more preferable to me as they really show up when projected onto a big screen.

For my transfer I'm going to have to use some video clips from the definitive set, I can't use the German set's video completely because the intro text is in German as are the subtitles (they burnt it into the picture instead of the black border ), hopefully the difference when swapping scenes between the German & definitive set should be unnoticable.
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***
Author
Time
Post link
Thanks for the screencaps. I have to agree, the NTSC looks sharper. Does anyone have this frame from the French LDs? Maybe it's a different transfer. Looks like the German set is a no-go as the ultimate source. I didn't even think about the crawl and sub scenes. How is the crawl and sub situation on the French set with english audio?
Author
Time
Post link
Originally posted by: rnranimal
Does anyone have this frame from the French LDs? The image I posted on page one is from my DVD, which is sourced from the French THX laserdisc.
Originally posted by: rnranimal
I didn't even think about the crawl and sub scenes. How is the crawl and sub situation on the French set with english audio?
The French set has the English crawls, and the subtitles are outside the picture in the lower black bar. But they are unfortunately missing those frames from ANH at the end of side 1/start of side 2.

Guidelines for post content and general behaviour: read announcement here

Max. allowable image sizes in signatures: reminder here

Author
Time
Post link
The same frame from the French LD.
http://img246.echo.cx/img246/8868/french4ka.jpg

A new release of Star Wars is like a box of chocolates, you never know what you're gonna get.

Author
Time
Post link
The definitive set does look sharper but at a significant cost, those jagged horizontal edges/lines, which look horrible when the footage is smoothly scaled up to 1024x436 and projected to 5 foot wide, the PAL set whilst looking softer looks a whole lot better at 5 foot wide.

Now all we need it the same screencap from the faces set and the Japanese set, anyone?
http://www.haku.co.uk/pics/LukeCruise.gif http://www.haku.co.uk/pics/dontcare.gif
***Citizen's NTSC DVD/PAL DVD/XviD Info and Feedback Thread***