I’ve been interested in Super NTSC since reading about it here in the LDDB forums. Disclord (RIP) had the most to say about it and is the muse for my further investigation of its use on Laserdisc titles.

More people are also stumbling across this info here and have more questions and are interested in seeing the difference.

So, after watching Deep Impact yesterday I’ve decided to do a write up and see if we can convince people with the technical knowledge to come forward and share what they know. Please, if you know more that what is being covered here we’d love to hear it!

Background on the Technical Details
As noted by Disclord
Super NTSC on LD has to be from a component source and be encoded to Super NTSC as opposed to normal NTSC. The products used were likely Faroudja brand since they invented the technology and hold the patents. The source tape was either Sony’s D1 Digital component or Panasonic’s D5 HD.

D1 boasted uncompressed digital Rec.601 compliant (interlaced) 4:2:2 video in YCbCr color space. While this format was released in 1986 it was primarily used for broadcast since at that time there was no home video format that was in the component space. Sony released D2 a similar composite tape format in 1988. When you see “Digitally Mastered” on a LD it is probably on a disc released after 1988.

Once DVD was being developed the studios had to make new component masters from new film scans for newer movies that were just making it to home video. Think about movies that came out theatrically in 1996. Those were making their first home video appearance on VHS, LD, and DVD. The masters would likely be in component format to take advantage of the component nature of DVD and then be easily converted for other formats. I highly doubt they used Standard Definition D1 as a mastering format for DVD. If they were smart the studios would scan new films at a higher resolution to future proof them and save time/money. By that time some of the LD mastering facilities had new technology from Faroudja called Super NTSC. Kurray in Japan being one of, if not the first to use it.

It wasn’t until DVD came along late in 1996 that there was a need to produce component masters for duplication. I’m not positive, but I speculate that DVD master tapes were Panasonic D5 HD format (1994). D5 HD allows 1080p@23.98. If you rip any major studio DVD movie (film) you’ll see that the VOB/MPEG file on the disc is usually 23.98 progressive. 1080p/24 down-res’d to 480p/24 is easy to do and you have a Blu-Ray ready HD master at the same time. Makes sense to me but Disclord mentions that Kurray used D1 component tapes converted to D2 for LD mastering despite those being interlaced and hence locked at Standard Definition and not future proof.


What is Super NTSC anyway?
Super NTSC was an encoding/decoding process very much like NTSC or PAL are. You input component and the output is NTSC. The difference is all the things Disclord mentions; higher resolution due to pre-combing luma/chroma and extended color bandwidth. Also much less video noise. Its a very clean stable signal that looks like a soft or analog DVD.

I have PDF files by Yves Faroudja describing the technology but can't upload them here because PDF is not allowed. I'd hate to have to convert them to jpg...I see all the manuals are in PDF format so maybe there is a way.

People have been asking for a list of titles and some have been mentioned; Titanic, Fifth Element, Deep Impact, Star Wars (Faces), etc.

How to tell the difference?
All I can add is that if you have any of the above titles and a Faroudja decoder you can tell what is Super NTSC quite easily. In fact, you don’t even need a Faroudja decoder as any NTSC decoder will still produce a better image with a Super NTSC encoded disc. Look for video noise, especially in dark scenes. If there is none its probably SN. Look for places where you’d normally see artifacts on fine details like car grills, space ships, AC vents, clothing patterns, wide moving shots of nature, etc. If there are few to none its probably SN. Look at the overall picture and check out the color saturation and bleeding. If it looks solid and doesn’t bleed its probably SN. At the other end of the spectrum pop in any disc from say 1994 or earlier. It should be pretty obvious that those discs are not SN encoded.

I watched Goldeneye the other day and it looked stunning. Then I put on Conan and it looked noticeably not as good. Then I put on Deep Impact, again a very noticeable improvement. Then put on Memoirs of an Invisible Man which was back to looking more like Conan. What I’m saying is that unless you A/B it can be hard to distinguish but it becomes pretty obvious when you see Super NTSC vs non-Super NTSC on the same system.

As mentioned previously, if you’re looking for Super NTSC encoded LDs good candidates are movies theatrically released ca.1996 or later and mastered by Kurray. They will all have new film scans made and be stored on D1/D5 HD.

Note that ALL DTS titles were released after 1997 and mastered at Kurray however if there was a catalog title that already had a LD release the older master tape may have be reused and DTS sound added in mastering. I have no idea but I’m saying that not all DTS titles pressed by Kurray are defacto SN encoded just that they are good candidates. I'd almost bet that all DTS LD releases are the same (video) masters as their DVD counterparts. It would be easy to reduce the bitrate of the DTS stream to fit on the DVD as required by the bit budget.

What About Star Wars?
This is my speculation with Star Wars faces set from 1995. According to lddb user Mitch the 2nd run was in 1998 at Kurray.The Definitive Collection, mastered by Mitsubishi, came out pre-DVD era (1993) and so was not likely on D1 unless they were thinking ahead (possible). Then again, since DVNR was applied it seems likely a new master was made and thus could be on D1 component. Legend here says the faces set is just the DC in CLV format (same master). I’m going to check out my faces discs today since they’re all the Kurray pressings. Being that they're later releases its possible that either Kurray had the time to implement SN encoding by 1998 if the master was D1 or that if the master was D1 in 1993/1995 for the DC and the first run of the Faces that Mitsibushi/Pioneer may not have been up to speed on SN encoding. However, the SE Trilogy Set from 1997 made by Pioneer does have Super NTSC encoding according to Disclord's comment. Everyone loves that version right?

So Pioneer ca.1997 has SN encoding capabilites....


The List
Titanic
The Fifth Element
Deep Impact
Meet Joe Black
Rounders
Wizard of Oz THX CLV 1996
From Dusk Til Dawn
The Man Who Knew Too Little
Good Will Hunting
Great Expectations
Nothing To Lose
Wild Wild West
Goldeneye AC3 (not Kurray)
Mission Impossible (not Kurray)
Top Gun AC3 (not Kurray)
Star Wars Trilogy (Faces 2nd press at Kurray) *maybe*
Star Wars Special Edition Trilogy 1997 (Pioneer)

If you have both Goldeneye AC3 and DTS you can do an A/B comparison since the DTS was manufactured at Kurray and the AC3 at Pioneer. Probably hard to tell since the AC3 is already impressive.

As mentioned by Disclord, at a certain point probably around 1997/1998 almost all mastering facilities were using Super NTSC encoding so that is probably why late releases look so good, blame DVD.

This was long winded but if you have anything to add please do. There is not enough reliable information on Super NTSC encoded Laserdiscs. I'd love to know more about the post-production video workflow using D1 sources to make composite products like LD. Please add movies to the list if you think they're SN encoded. There are probably a lot more than we previously thought.

According to wikipedia: "Later "SP" and "OS" models ran Off-SPeed, making them technically friendly for 24-frame telecine film transfers to D1 tape - and allowing a single tape to provide both NTSC (525 vertical lines) and PAL (625 lines) masters at one time."

So D1 could do 24 fps as well.

Some of the attributes you associate with SuperNTSC could just be from improved mastering techniques. My Abyss, The: Extended Version (1989) [0896885] SuperNTSC LDs are noisy in dark scenes, for instance. Conversely, I have a superntsc decoder, CFD-SN (2 actually). I recently did some A/B comparisons with Kurraray Faces vs. FOX technidisc pressing vs. JSC. The technidisc and Faces were about the same for cross color artifacts, but both were way better than the JSC. This tells me that there are other factors that contribute to the cross color artifacts, because there is no way the Star Wars: A New Hope (FOX) (1977) [1130-85] technidisc pressing is SuperNTSC.

My Criterion Armageddon: Special Edition #384 (1998) (Uncut) [CC1548L] looks horrible. Really bad cross color artifacts. It is the last criterion disc. Pressed in 1999 at Kuraray. So just because it checks all the boxes, doesn't mean it is SuperNTSC, sadly. Was very disappointed in that LD. It has other issues too, bad smearing.

I wish there was a truly definitively way to tell, like there is with PAL+ (extra lines of resolution in the blanking area), but i haven't found any.

I’m not very technically up on this so forgive me here...but I’ve been under the impression that SuperNTSC’s main thing is that it has full chroma resolution. In that case the real world benefit is going to vary greatly by disc. There for sure are ways to check that with a scope. I don’t know exactly how to do it but for sure it’s not hard those Domesday guys will know and I’d expect Substance as well.

^^^ Emphasis mine. ^^^
I'm pretty sure the full color bandwidth is only "after"; "bandwidth expansion" by a SuperNTSC decoder, so you wouldn't easily detect it from the RF the domesday decoder gets (also, if you didn't see in the ld-decode thread, i actually used a LimeSDR to capture and decode the LD RF). The RF bandwidth on the LD is finite, that extra color bandwidth isn't in the raw signal, but "encoded" in the base signal. Similar to how AC3 is "encoded" into one analog channel, but you need an AC3 decoder to get it. I have that book that he quoted, and read it about 5 years ago. I won't state it as fact, but that is best i can recall.
[edit]I found the reference: see post below for the link to the paper where this reference is from.[/edit]


This is referring to cross color/raindbowing and jagged diagonal lines, which to me, is the big benefit.

In my posts in the screencaps thread with "episode one" and "first contact", those shots look good, with minimal jaggies and cross color, because i used my SuperNTSC decoder. I don't have a lumagen or super fancy anything like that. If i play the JSC through that setup, every star in the space scenes is a different color. I had intended to post caps from my Armageddon criterion, assuming it was SuperNTSC by the stats it has, but when i started looking at them, it was bad. My fairly recently purchased 4k TV looks good, but i can't capture that.

From IEEE: NTSC and beyond by Y.C. Faroudja

You may not be able to download that, unless you have access. My affiliation with the University grants me access, so i have downloaded a copy here:
NTSC and beyond by Y.C. Faroudja If you want a copy, download it and save it, i can't promise how long it will be there.

This a paper where Y.C. Faroudja talks about NTSC, and how Cross Color is the #1 thing to solve, cross luminance the #2 thing to solve, and limited chroma the #3 thing solve, with SuperNTSC.

Took some time to read that full PDF again today:



So as @sonicboom said, the "darker" scenes should have less noise. Maybe "The Abyss" is just an exception?

I didn't have time before and couldn't remember from memory but searched it up and it is super ntsc.

Gone with the Wind (1939) [ML107199]
Not an easy to find copy and it truly looks stunning.

What players have the ability to take advantage of this enhanced signal?

? Sorry I don't understand, all players will play these LDs its how they were mastered.

I'll have to try out my Faroudja LD-100's SuperNTSC decoding one of these days. Plug it in to one of my old VGA monitors or something - though a CRT pj would be perfect to recreate the 90s high end ht experience. Someone paid $15k for it new.

In addition to my CFD-SN, i also have a Faroudja LD200 but unfortunately, it reduces perceptible resolution. I wouldn't be surprised if the LD100 does the same.

LDs from D2 Digital tape have crisp resolution but sometimes inaccurate skin tones (towards green).
For my taste best are LDs mastered from Type C/Betacam SP pro-analog videotapes

Máster tapes for DVD as far as i know have been always Digibeta tapes with HDcam and another formats being punctual exceptions

Interesting thread. In my pursuit for the best laserdisc setup I came to the realisation that Super NTSC discs really make the format shine. One of the best has to be the DTS release of Con Air. It certainly is a different transfer from the non DTS version and a simple A/B comparison will prove it. The non DTS release is full of aliasing artifacts and is noticeabley more noisy. Apparently any 3D comb filter is capable of processing the Super NTSC signal.

According to the blam website, any Dolby digital disc after 1996 was pressed and mastered by Kuraray themselves. Whereas in 1995 they were pressing masters done by Pioneer Japan. Leads me to believe they were doing their own masters for their DTS discs too.

Great responses from everyone so far.

I've been watching more movies with an eye for the mastering. I would have to agree that a major part of why later LDs look good is the component sources and higher resolution scanning of the film as compared to pre-DVD era LD mastering, even digital mastering. Anything from a D1/D2 source was always a 720x486 sample so that is baked into the image. When you scan a film at 2K then reduce that 2K to NTSC it will always look better than scanning the same film at D1 resolution. Especially now with large fixed resolution screens and the tendency to zoom in. Maybe it didn't matter as much with CRTs.

Also, consider the source generation. A movie released in 1982 and made available on LD in 1995 for the 4th time might be using a safety copy or other multiple generation copy away from the actual theatrical print. Criterion, and some other titles, specifically list what the film source was (negative, inter-positive, etc). So when all the stars line up and you get an old film that gets a new component HD scan (for DVD) from a new/better source and a SN mastered LD you win big. I'm assuming this is what Rein-O is talking about with GWTW. I have the snapper DVD that likely is the same transfer and it does look great. I think the THX Oz from the box or the single disc mastered by Kurray are similar situations.

Moving on, I found it interesting that while Kurray is touted for being the plant that used Super NTSC the most it seems that Pioneer USA has more titles that are noted by Disclord as being SN (Fifth Element, Titanic, Deep Impact, etc). How does he know anyway? Was it mentioned in an article/review back then? Did he have contacts? I never quite understood how he had all that info. I think he was a projectionist but I have no idea. He sounds like he could be a sales rep...

Another small point to be made regarding color is that D1/D2 uses 4:2:2 sampling while DVD compliant MPEG-2 uses 4:2:0. That means that technically speaking there is more color information on a LD versus a DVD. Combine that with Super NTSC and it is possible to have a much more vibrant color on LD but for things like chroma noise, etc.

@9954Tony you're totally with me on this and I love it. I have the same CFD-SN but I'm using the FLI2300 in the Cyrstalio II. Thanks for sharing the PDF. I think that you're correct in that the benefits are encoded into the signal and need to be decoded properly to work as intended. However, a lot of the Faroudja patented stuff is imitated and implemented under different names by other manufactures so its possible that they too decode SN in some way. Look at Dolby Stereo you can decode it with Pro Logic, PLII, PLIIx, or even DTS NEO, etc. And, again I agree that ALL of this info is speculative so from title to title and plant to plant there will be variances. I had that Abyss and sold it since I have both box sets and the DVD. Film does have grain and when you're shooting in a tank in the dark there will be grain no matter what. Underwater film grain has to be hard to encode/decode. Check out Sphere and see what you think. Pioneer USA press with 59-xxxxx mint marks. Similar visuals. Armageddon sounds like it may have been delivered to Kurray on D2 not D1...baked in NTSC. In fact, all Armageddon US LDs were Kurray pressed and probably all look exactly the same.

@signofzeta yes, I think you're correct that one could observe the difference on a scope. Though if SuperNTSC has to be compatible with NTSC (and also decoded) then it may look like PONTSC (Plain old NTSC) on a scope. I would guess that somewhere in the VITC you could find some info about it. I recall that some THX discs use the VITC for something like that.

@ertolli you can master a DVD from anything but for major Hollywood movies I'm hoping DigiBeta was not used as it records 2.34:1 DCT-compressed digital component video signal that then needs to get MPEG-2 encoded. I'm guessing not. But you're right there are some really amazing looking 100% analog LDs mastered from 1" C.

@odotb3 Good note on Con-Air. I have the AC3 that I should re-visit but remember it being "meh". Big action movies (especially 2.35:1) with lots of moving wide shots and a very detailed set are difficult for NTSC (and MPEG-2).

So maybe the Domesday crowd can help us out and see if there is any way to identify Super NTSC encoded LDs. Without that or a definitive list from a definitive source we're all just chasing at shadows in the dark.

I may have found the primer....

So investigating some of my suspected titles and Disclord's "confirmed" titles(*) I've come across an interesting pattern in the mint marks on Pioneer USA mastered discs:

59-546 Fifth Element*
59-290 Deep Impact*
59-288 Titanic AC3*
59-643 Meet Joe Black
59-492 Trainspotting CC
59-212 Goldeneye AC3
59-230 Mission: Impossible AC3
59-260 Jurassic Park AC3
59-266 JP: Lost World AC3

These are arguably some of the best looking LDs and I think we know why now. I've personally watched all of these and confirm they do look fantastic. Also notice that these are all newer movies getting their first home video release on DVD so component masters were made and the films were likely scanned at 2K HD. JP was likely a new master for DVD since there has been several LD/VHS releases at that point but none with AC3.

THANKS DVD!

I'm no expert in mint marks (or anything really) so it could just be a coincidence that has to do with the something else altogether.

Kurray, however, has no such pattern that I've been able to decode. It seems Kurray just uses the catalog number for the mint marks from the LDDB submissions.

8763-85 Star Wars Faces 2nd press
0896885 The Abyss SE AC-3*
15625 U.S. Marshals* <--- This does look spectacular
12881AS Con Air DTS

I've also watched a few DTS movies recently and discovered that they are definitely not all Super NTSC encoded. So just to re-iterate and paraphrase what Disclord said, according to his anonymous source:

If Kurray received a component master, then it was SN encoded.

The catch is that not all the movies supplied to Kurray were in a component format and so thus NOT SuperNTSC encoded.

I think that far too much emphasis and speculation regarding Super NTSC has been put on Kurray as a result of the Star Wars effect. I will go out on a ledge and say that based on what I've seen and documented that Pioneer USA in Carson, CA had the most prolific Super NTSC mastering run.

The mystery deepens.....

I don't have a foolproof way of determining SuperNTSC titles, but they're less likely to have noticeable artifacts when run through a good 2D comb filter.

(Also, I'm more interested in relative oddities - most of the late disks mentioned have decent DVD's, if not decent Blu-Rays that not even an X0+ADV7842 has a chance of beating.)

That's actually a very good point and an easy way to tell.


I agree that the LDs are inferior to the DVD/BR versions of the same masters - obviously.

except "the abyss"

Another thing of note, Disclord made a comment on this disc that it was SuperNTSC:
Abyss, The: Extended Version (1989) [0896885]

which i own, released in 1996. but i also have this disc, released in 1993:

Abyss, The: Special Edition (1989) (Uncut) [1988-85]

I did captures of portions of both discs, using several different hardware chains, and the 2 discs "look" identical when processed through the same hardware chain.

I've often wondered about this. Thanks for making this thread. I think indeed you pretty much just have to go by sight and feel of the overall transfer. I just use the onboard comb filter in my XBR960 and indeed most of the titles above look stunning-but what makes it extra difficult is most titles post 1995 look stunning regardless and pretty much all discs 1997-1999 look great across the board.

The only thing I run into is the occasional aliasing or shimmering effect in a spot or two at times which Disclord used to say was a lack of low pass filtering which makes artifacts like that show up on CRTs. On First Contact the transfer is great save for a little bit of this popping up a few times.